Litigation hold: 7 things Rule 37(e) expects of your company

The demand letter is on your desk. Maybe it came from a former partner's lawyer, maybe from a customer who says you breached. Before anyone decides whether to negotiate or fight, a quieter job has already begun: the litigation hold, which is the practice of stopping the ordinary deletion of information that may matter to the dispute.

In federal court, the rule that polices this is Federal Rule of Civil Procedure 37(e). It governs what a judge may do when electronically stored information is lost. Its text and the advisory committee's note to its 2015 amendment say a good deal about what's expected of a company long before a judge is involved.

A word on scope first. The federal rules govern civil actions in the United States district courts, and Oklahoma is divided by statute into three federal judicial districts. Oklahoma's state courts apply Oklahoma's own discovery rules and case law, and this article doesn't describe that state standard.

If you're still weighing which forum the dispute belongs in, our view is that preservation shouldn't wait for that decision. Here are seven things the federal rule expects, as we read it.

1. A litigation hold starts when a dispute is foreseeable

The committee note puts the duty this way: potential litigants must preserve relevant information "when litigation is reasonably foreseeable." Not when the complaint is filed. Not when you're served.

Rule 37(e) didn't invent that duty. The note says the rule is based on the existing common-law duty to preserve and "does not attempt to create a new duty to preserve."

You won't find the words "litigation hold" in the rule's text. The phrase is practitioner shorthand. An earlier committee note to the same rule uses it once, describing intervention in the routine operation of an information system as one aspect of what is "often called a 'litigation hold.'"

What makes a dispute foreseeable? In our reading, a lawyer's demand letter usually does, and so can a pointed threat in an email or a text. So can your own decision to sue: the note speaks of "potential litigants" and of "a party," never of defendants alone, so we read the duty as running to the company planning a claim as much as to the one expecting one.

There's a fair counterweight in the same note. Courts are told to consider how far a party was on notice that litigation was likely and that the information would be relevant, and the note acknowledges that early warnings often carry "only limited information" about the dispute to come. It cautions courts against judging those first choices with hindsight drawn from the case as it was eventually filed.

2. The rule reaches electronic information, and only that

Rule 37(e) applies to electronically stored information, usually shortened to ESI: email, texts, chat messages, accounting files, shared drives, anything that lives as data instead of on paper. The note is direct about the limit. The rule "applies only to electronically stored information."

Paper files and physical evidence aren't free to disappear. They're judged under other law, which this article doesn't cover.

For a closely held company, the practical question is where the ESI actually sits, and it's rarely on one tidy server. It's on the owner's phone, in a group text among managers, in a personal email account someone used over a weekend, on the bookkeeper's laptop, in a chat app set to erase messages after a few days, and in the account of the employee who gave notice last week.

We see this most sharply in ownership fights. When a business divorce is brewing, the people holding the most relevant messages are often the people on opposite sides of it, each with company data on a personal device.

Those files often do double duty. An accounting file that matters to a contract claim usually matters to the tax return as well, which is one reason we treat preservation as a business question and not only a litigation chore.

3. Reasonable steps are the standard, not perfection

Rule 37(e) asks whether information was lost because a party "failed to take reasonable steps to preserve it." The note explains that wording: reasonable steps suffice, and the rule "does not call for perfection."

That concession is deliberate. The note recognizes that perfect preservation of all relevant ESI "is often impossible," and it tells courts to be sensitive to a party's resources and to its sophistication about litigation. As we read that, a family-owned distributor's efforts aren't measured as if it had a public company's records department.

Cost counts too. Proportionality is named as a factor in judging preservation efforts, and the note says a party may act reasonably by choosing a less costly form of preservation if it's substantially as effective as the more costly ones.

The drafters were candid about why. Federal circuits had applied significantly different standards, and the note says those differences led litigants to "expend excessive effort and money on preservation" to stay clear of the most serious sanctions. Keeping everything forever isn't what the rule asks, and it carries its own expense.

Still, reasonable is not the same as nothing. In our reading, the concession helps a company that made sensible choices and can show them, and does much less for one that never made a choice at all.

4. Routine deletion has to be interrupted on purpose

Most data loss in a small company isn't sinister. It's a retention setting doing its job, or a phone traded in at the carrier's store.

The note treats that routine fairly. The "routine, good-faith operation" of an information system is a relevant factor when a court evaluates whether a party failed to take reasonable steps. The same passage adds that the prospect of litigation may call for reasonable steps to preserve information by intervening in that routine.

Here's a pattern we see. The controller who handled the disputed account resigns, and the IT vendor works through its usual offboarding checklist. Her laptop is wiped and reissued before anyone connects the departure to the dispute.

So the first working step of a hold is usually technical, not legal. Switch off auto-delete where it runs, pause the offboarding wipe for the people involved, hold back the phone upgrade, and keep retired laptops in a drawer instead of sending them out for recycling.

Events outside your control are a different matter. The note recognizes that even when a party takes reasonable steps, preserved information can be destroyed by events it doesn't control, and it gives a flooded computer room and a failed cloud service as examples. Courts may still ask how far the party knew of such risks and protected against them.

Your own retention schedule is part of the picture as well. Independent obligations to keep records can come from statutes, administrative regulations, an order in another case or a company's own protocols. As the note describes it, though, having such an obligation doesn't necessarily mean a party had a duty with respect to the litigation, and failing to observe one doesn't by itself show that preservation efforts were unreasonable.

5. A hold is a list of people and systems, not a memo

Nothing in Rule 37(e) prescribes a form of notice. What follows is practice, not law.

In practice, a hold starts with two lists. One is people: everyone likely to hold information about the dispute, including the owner and including people who've already left. The other is systems: every place those people's information lives, whether the company owns it or not.

The written notice comes after the lists. A typical one describes the dispute in a sentence or two, names the kinds of information to keep, tells recipients not to delete or alter anything, says who to call with questions, and makes clear that it stays in force until someone lifts it in writing.

Think of a hold as closing the drain, not filling the tub. The aim on day one is to stop the loss; collecting and reviewing what was kept can come later, on a schedule that fits the dispute.

Then comes the part companies skip, which is follow-up. Someone has to confirm that the settings were actually changed and add new names as the dispute widens.

Keep a dated record of all of it. Preservation questions tend to arrive long after the fact, in discovery, when a witness is asked under oath whether she texts about work and what became of last year's phone. The company that can show what it did, and when, is arguing about reasonableness from records instead of from memory.

6. Scope is something to raise early with the other side

Preservation isn't meant to be a guessing game played alone. Under Rule 26(f), the parties to a federal case must confer, and when they do they must "discuss any issues about preserving discoverable information." Their discovery plan has to state their views on ESI issues, preservation among them.

That conference is where the cost argument belongs. Discovery in federal court reaches matters relevant to a claim or defense and "proportional to the needs of the case," measured by factors that include the amount in controversy and whether the burden or expense of the discovery outweighs its likely benefit.

Rule 26 also recognizes that some ESI is hard to reach. A party need not provide discovery from sources it identifies as "not reasonably accessible because of undue burden or cost," although a court may still order it on a showing of good cause. That provision is about producing information, not about keeping it, so we don't read it as permission to let hard-to-reach data go.

A proposal made early is, in our experience, easier to defend than a decision explained late. Something along the lines of "we'll keep these named people's email and phones plus the accounting file, and we won't copy every workstation" gives opposing counsel a concrete position to respond to. If the two sides can't agree once a case is underway, the Rule 37(e) note itself says that promptly seeking the court's guidance about the extent of reasonable preservation may be important.

What it will cost to keep the data, and later to produce it, belongs in the same conversation as whether to settle or fight.

7. Lost data is judged in two tiers, and intent separates them

Suppose something is gone anyway. The rule comes into play only for ESI that should have been preserved in the anticipation or conduct of litigation and was lost because a party failed to take reasonable steps. Even then there's another gate: the information must be something that "cannot be restored or replaced through additional discovery."

ESI tends to exist in several places, and the note points out that loss from one source "may often be harmless when substitute information can be found elsewhere." The email deleted from one mailbox may still sit in the recipient's. If the information is restored or replaced, the note says no further measures should be taken.

Tier one: curing the prejudice

If the information can't be recovered, the court asks whether another party was prejudiced by its loss. On that finding, subdivision (e)(1) allows measures "no greater than necessary to cure the prejudice."

The note describes the available range as broad. Its examples include forbidding the party that failed to preserve the information from putting on certain evidence, and permitting the parties to present evidence and argument to the jury about the loss. The rule places the burden of proving or disproving prejudice on neither side and leaves the judge discretion over how best to assess it.

Notice what tier one doesn't ask about: motive. As we read it, "we didn't mean to" is no answer to a request for curative measures, because (e)(1) turns on prejudice from the loss and not on why the loss happened. An honest mistake can still end with your opponent telling the jury about the missing texts.

Tier two: the severe measures

Subdivision (e)(2) is different in kind. Only on a finding that the party acted "with the intent to deprive another party of the information's use in the litigation" may a court presume the lost information was unfavorable to that party, instruct a jury that it may or must presume so, or dismiss the action or enter a default judgment.

Carelessness doesn't get there. The note rejects decisions that allowed adverse-inference instructions, meaning instructions that let jurors assume the missing information would have hurt the party that lost it, on a finding of negligence or gross negligence. Its reasoning is that such behavior "does not logically support that inference," since information lost through negligence may have been favorable to either party.

Two details of this tier are easy to miss. No separate finding of prejudice is required, because the intent finding itself can support an inference that the other side was harmed. And the judge may make the intent finding when ruling on a pretrial motion.

Even a finding of intent doesn't require a court to impose any of the listed measures. The note says the severe ones shouldn't be used when the lost information was relatively unimportant.

The note also limits where a judge can look for the power to impose these measures. Courts speak of inherent authority, meaning a general power to manage cases apart from any specific rule, but the note says Rule 37(e) "forecloses reliance on inherent authority or state law" to determine when these measures should be used. The stated aim is a uniform standard in federal court.

In our reading, the intent requirement protects a company that was disorganized but not dishonest. It protects your opponent the same way, so if the other side carelessly lost the messages you needed, the relief open to you may be limited to curing the prejudice.

If a demand letter or your own plan to file suit has you wondering what your company should be keeping, that's a sensible moment to talk with a lawyer, ideally before the next phone upgrade and not after it. You can reach us through the contact page or at our Oklahoma City office. A conversation about preservation at the start usually costs far less than explaining a gap in the record later.

This article is general information about Oklahoma, Texas, and federal law, not legal advice, and it does not create an attorney-client relationship. Facts matter; talk to a lawyer about yours.