Responses to Requests for Production (RFPs) vary widely between extreme candor and obfuscation. Between those poles are the responses that look something like the following: "Responding Party will produce relevant, responsive documents within its possession, custody or control." That almost looks like a commitment to produce documents within the scope of the request at issue. Hooray! But they say the devil is in the details, and the devilish detail lurking in this particular response is that one, innocuous little word: Relevant.
This response might not even make the top five on most attorneys' lists of annoying discovery games. But it's a sneaky little trick, partly because it seems innocuous. Including a qualifier about producing only "relevant" documents in a discovery response rests on the half-truth that a party has no obligation to produce "irrelevant" documents. This has a ring of common sense. After all, the concept of relevance is central to civil discovery because the scope of discovery itself, in both state and federal court, is limited to material relevant to the issues in the case. See Fed. R. Civ. P. 26(b)(1); Cal. Civ. Proc. Code § 2017.010. Thus, by operation of law, anything that is not relevant is per se not discoverable, and a commitment to produce only relevant documents amounts to little more than giving the requesting party what it deserves. Makes sense, right?
Not so fast. This argument begs an important question: Who decides what's relevant, and how? Should it be a responding party? A requesting party? The court? The wrong answer makes for mischief.
A commitment to produce only relevant documents implies the producing party will decide whether an individual document is or is not relevant and therefore subject to production. The problem is obvious if you stop to think about it for more than a moment. Giving one lawyer unsupervised discretion to interpret the scope of another lawyer's discovery request is an invitation to sharp tactics. Attorney ethical guidelines and the requirement of signing and certifying discovery responses establish some guardrails that could curb the worst potential abuses of such liberty-taking; but the vast outer boundaries of what is ethically permissible leave ample room to frustrate an adverse party's discovery goals.
Many lawyers will argue our system is based on trust, and an attorney seeking discovery needs to trust opposing counsel to produce all the relevant documents. Some judges will even repeat this reasoning. See, e.g., UAB "Planner5D" v. Meta Platforms, Inc., 746 F. Supp. 3d 806, 808 (N.D. Cal. 2024). This line of thinking, however charitable about the scruples of litigators, is false. But not because lawyers are untrustworthy. It is false because our system does not run on trust. It works because we have a set of procedural rules that apply equally to all parties, and those rules yield trustworthy results (even if we don't like them) because they are the product of a transparent, objective, and verifiable process. Trust has place, to be sure, but trust is always moderated with "but verify," and there's no way to verify an opaque process.
It is not hard to set up a transparent process for securing discovery of only relevant documents. All questions of relevance must be resolved at the document request level, either by agreement of the parties or by court order. The operative question is, "Does this request seek relevant information?" If it does, producing parties only need to worry about whether a particular document is responsive to the request. Document-by-document determinations of relevance have essentially no place in discovery.
The distinction between "relevant requests" and "responsive documents" is key to ensuring the discovery process functions transparently. Reserving relevance determinations for the court as a threshold inquiry ensures both parties operate under the same rules. Relevant will mean whatever the court, in the exercise of its sound discretion, determines its means. Once that rule is in place, executing the command to locate and produce responsive documents is fine to leave to producing party's discretion because, unlike determining relevance, locating responsive documents requires making an objective determination. A document is either responsive to a request or it is not. If a party cannot tell whether a document is responsive, then there is a vagueness problem with the request itself that must be corrected.
By putting relevance and responsiveness in their proper places, the process of producing discovery is fully transparent because all parties operate under the same rule for what is discoverable. The party receiving discovery can be assured they're receiving what they deserve, nothing more, and nothing less. That assurance simply doesn't exist when one party takes it upon itself to make a blind determination of what is relevant to the case.
Although there is little case law addressing this exact issue, there is ample case law addressing the analogous issue of unilaterally redacting documents to scrub them of purportedly irrelevant information. The general rule in those cases is that parties must produce responsive documents in their entirety even if those documents contain some information that might ultimately be irrelevant under the discovery rules. See, e.g., Coe v. Cross-Lines Ret. Ctr., Inc., 342 F.R.D. 539, 548 (D. Kan. 2022). The logic of the redaction cases applies to relevance-qualified responses--courts determine relevancy at the request level, not the document level, and if the request seeks relevant information, the producing party has no discretion to pick and choose what gets produced on relevancy grounds or for any other reason. Transparent process, trustworthy results, and a lot less anxiety over whether the other side is secretly withholding discovery.