When is a Witness Unavailable?
We’ve written before about hearsay and its exceptions. As you learn in evidence class, hearsay exceptions come in essentially three flavors.
The first is “not hearsay.” These are out of court statements or writings that sound an awful lot like hearsay (e.g. “he told me that…”) but don’t technically qualify as hearsay under the rules. This is generally because the statement has some evidentiary value that’s independent of whether it’s true or not. A good example is a bank teller testifying that a defendant told her to empty the cash drawer, or he’d shoot. The point of the testimony isn’t whether the defendant was factually correct in his statement. Maybe he would shoot or maybe he’d lose his nerve. Rather the point is that he made the threat, regardless of whether the statement is correct.
The second category of hearsay exceptions are statements that definitely qualify as hearsay, but that courts allow into evidence anyway. This is theoretically because, as a policy matter, courts or legislatures have decided that the statements are reliable. In reality, this is often an exercise in amateur psychology and received “folk wisdom” with some tradition to back it up. A good example is the “excited utterance,” a statement made when somebody is really amped up about something. It’s allowed into evidence on the phrenology-adjacent theory that really emotional people are physically incapable of deception. Ancient judges actually believed this, at least a little, and the drafters of the 1972 Rules of Evidence were too polite to contradict them. In the Federal system, these exceptions are mostly set forth in Fed. R. Evid. Rule 803.
The third category of hearsay exceptions are the “well, if we’ve got to,” exceptions. These are statements that are definitely hearsay and not permitted into evidence unless the person who made the statement is “unavailable” to testify. Under the Federal Rules of Evidence, these exceptions include former testimony, statements against interest, dying declarations, and statements of personal or family history. If whomever said them can testify, all of these things are out, but if the witness can’t, they can be admitted under Fed. R. Evid. 804.
You might ask what distinguishes category three from category two. In fact, there are some major similarities between exceptions in the two groups. For example, most folks are fairly emotional about their own impending death. As such, a dying declaration looks an awful lot like a subspecies of excited utterance. This is even more obvious when you walk through hypotheticals. If you’re too excited to lie because you saw someone get hit by a car, imagine your condition if you were actually hit by the car and lay dying in the street. Despite this, the later category requires unavailability while the former does not.
If you’re hoping for a solid analytical justification for the difference, you’re probably studying the wrong area of law. The Advisory Committee notes to Rule 804 are fairly open about this and admit that the “unavailability” requirement was cobbled together out of a tangle of similar common law doctrines that were built into the caselaw surrounding specific hearsay exceptions.
But what does it even mean for a witness to be unavailable? And how many types of “unavailable” are there?
The first thing to straighten out is that there are at least two different sets of evidentiary rules about unavailability that apply similar, but very much not identical definitions. Fed. R. Civ. P. 32(a)(4), which governs the use of depositions at trial, defines a witness as “unavailable” such that their deposition can be used as follows:
(4) Unavailable Witness. A party may use for any purpose the deposition of a witness, whether or not a party, if the court finds:
(A) that the witness is dead;
(B) that the witness is more than 100 miles from the place of hearing or trial or is outside the United States, unless it appears that the witness's absence was procured by the party offering the deposition;
(C) that the witness cannot attend or testify because of age, illness, infirmity, or imprisonment;
(D) that the party offering the deposition could not procure the witness's attendance by subpoena; or
(E) on motion and notice, that exceptional circumstances make it desirable—in the interest of justice and with due regard to the importance of live testimony in open court—to permit the deposition to be used.
In contrast, a witness is “unavailable” for the purposes of hearsay exceptions under Fed. R. Evid. 804(a) in the following scenarios:
(a) Criteria for Being Unavailable. A declarant is considered to be unavailable as a witness if the declarant:
(1) is exempted from testifying about the subject matter of the declarant’s statement because the court rules that a privilege applies;
(2) refuses to testify about the subject matter despite a court order to do so;
(3) testifies to not remembering the subject matter;
(4) cannot be present or testify at the trial or hearing because of death or a then-existing infirmity, physical illness, or mental illness; or
(5) is absent from the trial or hearing and the statement’s proponent has not been able, by process or other reasonable means, to procure:
(A) the declarant’s attendance, in the case of a hearsay exception under Rule 804(b)(1) or (6) ; or
(B) the declarant’s attendance or testimony, in the case of a hearsay exception under Rule 804(b)(2) , (3) , or (4).
While there are obvious similarities, the fact that they’re different at all is stranger than you might think.
Imagine a witness who lives more than 100 miles from the courthouse. If you have deposition testimony from the witness – even in another case between the parties – it’s almost certainly going to be admitted and used for any purpose. But if you have prior trial testimony – even prior trial testimony in the same case – you need to affirmatively show that you tried and failed to get the witness to appear. This is true even if the witness is much much further than 100 miles away, even if the witness lives abroad or is otherwise unambiguously beyond the subpoena power of the court.
One version of this scenario occurred in United States v. Yida. [1] In that case the government prosecuted an alleged drug smuggler named Yida. One of Yida’s co-conspirators had previously pled guilty and sought to be deported to his native Isreal. The government detained the co-conspirator pursuant to a material witness warrant and made him testify at Yida’s trial.
The jury deadlocked and the court declared a mistrial. The co-conspirator asked the government if he could go back to Isreal, and the government agreed, so long as he promised to return for the retrial. Obviously he didn’t, so the government sought to introduce his prior trial testimony at the second trial. The court rejected this and the Court of Appeals for the Ninth Circuit affirmed, holding that by allowing the co-conspirator to leave, even after getting a promise that he’d return and offering to pay his way back, the government had not exercised “reasonable efforts” to ensure the witness’s attendance.
Yida is an extreme example, but it illustrates the large potential difference between “anyone more than 100 miles away” and “anyone you couldn’t reasonably get to come.” Where does this difference between depositions and trial testimony come from? It’s a bit of historical accident arising out of the separation of law and equity courts at common law. The deposition was actually invented in Chancery Court, where they were treated as equivalent to trial testimony. [2] Until the 1800s, there simply was no equivalent in courts of law. [3] And given the crude and unreliable nature of historical court reporting, transcripts of prior in-court testimony in courts of law were viewed with suspicion and treated as little more than ordinary hearsay. [4] When the law and equity courts merged both kinds of evidence became available, but they never really shed their historical baggage.
Beyond literal physical absence, what else qualifies a witness as unavailable? For Rule 804 purposes, a witness can be unavailable even in the courtroom if they refuse to testify when ordered, assert a valid privilege, or testify that they cannot remember the subject matter. But even these are fraught with complications.
For example, how little does a witness need to remember about something to qualify as “unavailable” such that their past testimony can be used? Some courts take a very hard line, suggesting that something close to complete amnesia – “no memory of the events to which his hearsay statements relate.” [5] But other courts have suggested that any lack of recollection will do, a standard that arguably applies to every witness. [6]
Refusal to testify is also potentially problematic. As written it requires not merely a refusal, but one that persists even after an order from the court. Read literally, it would require a party to call a witness to the stand so that they can refuse a direct order, a tricky and potentially wasteful exercise.
So what can we glean from all of this? First, while depositions and prior testimony may look similar, they are not treated the same. You may want to take a deposition of a potentially troublesome witness even if you have good prior testimony on a particular point. Second, if you can characterize a dying declaration as an “excited utterance” or a statement against interest as a “party admission” or a “business record,” you should do it and save yourself the trouble of proving unavailability. And finally, if you are stuck relying on Fed. R. Evid. 804, make sure that you document your efforts to get the witness to court.
[1] 498 F. Ed. 945 (9th Cir. 2007).
[2] John Henry Wigmore, A Treatise on the System of Evidence in Trials at Common Law Volume II § 1331 (1904).
[3] Id.
[4] John Henry Wigmore, A Treatise on the System of Evidence in Trials at Common Law Volume II § 1330 (1904).
[5] Gilmore v. Palestinian Authority, 53 F. Supp. 3d 191, 207 (D.D.C. 2014).
[6] Crump, David, What Does Unavailable Mean?, 55 Cumb. L. Rev. 1, 12-14 (2025).
