Confusion over confrontation.

In a combined case, Davis v. Washington and Hammon v. Indiana, the Supreme Court answered a tricky question about the Sixth Amendment's Confrontation Clause that had bedeviled the lower courts.' The issue was whether a 911 call (in Davis) and a police interview with a witness (in Hammon) were "testimonial." If so, then they couldn't be used at the defendants' trials without the witnesses testifying.

The Court, in an 8-1 decision written by Justice Antonin Scalia, concluded that the 911 call was not testimonial but the police interview was. (2) This ruling only furthers the confusion over the "testimonial/non-testimonial" dichotomy the Court created in its decision in Crawford v. Washington two years ago. (3)

In Crawford, a defendant's conviction for assault was based partly on a statement his wife made to police officers. She did not testify at trial. In striking down Crawford's conviction, the Court overruled Ohio v. Roberts, (4) which had found that hearsay statements did not violate the Confrontation Clause if they fell under a "firmly rooted hear-say exception" or had other "adequate indicia of reliability." (5) Applying Roberts, the state courts had admitted the wife's statement on the grounds that it appeared reliable because it corresponded closely to what the defendant had said. (6)

But in Crawford, the Supreme Court said that "reliability" was no longer the appropriate test; instead, courts must determine whether a statement is "testimonial."

Finding the framers' intent

The Confrontation Clause was intended to curb the use of evidence allegedly given to investigating magistrates by witnesses who did not appear at trial. (7) But the historical law the Court cites in Crawford supports only the proposition that no hearsay evidence could be used against a defendant. (8) This proposition runs counter to historical fact.

Indeed, the Crawford Court boldly declared--without any historical support--that "not all hearsay implicates the Sixth kanendment's core concerns.

An off-hand, overheard remark might not be unreliable evidence...." (9) Why not? Well, the Court reasoned circularly, because it is not testimonial, and only testimonial evidence concerned the framers.

Dissenting in Crawford, Chief Justice William Rehnquist, joined by Justice Sandra Day O'Connor, pointed out that "the Court's distinction between testimonial and nontestimonial statements ... is no better rooted in history than our current doctrine." (10) There was apparently no consistent rule, he wrote, and hearsay was "often heard by the jury." (11) In fact, the most clearly testimonial statements--those taken by magistrates under oath-were the most likely to be admitted. (12) Because of this, Rehnquist concluded that the old rule was correct, and the statement by Crawford's wife was admissible.

Other than statements made under oath, it is unclear which statements are testimonial and which are not. Contrary to the historical evidence Rehnquist presented, only nontestimonial statements are now admissible in the absence of the declarant. Crawford held that a statement given to police officers after they issued Miranda warnings was sufficiently formal to count as testimonial. But what about the 911 call in Davis, and the non-Mirandized statements in Hammon?

Before discussing the Court's answer, it is helpful to consider an alternate approach. Given that the history on this question is fuzzy, (13) isn't "reliability" a better standard? While often inconsistent and confusing, reliability at least presents a sensible target. Instead, courts must now define "nontestimonial"--a target that is equally hard to hit and that makes no sense, historically or otherwise.

Take 911 calls. They are inherently unreliable. People who make them are usually upset and prone to the type of overstatement that calls out for cross-examination: "Now, when you told the operator, 'He's trying to kill me,' you didn't really think he was trying to kill you, did you?" Yet Davis/Hammon concludes that such nontestimonial statements may be used against a defendant without the declarant testifying.

Oddly, under the old hearsay rules, which Rehnquist advanced, such a statement was also admissible as a spontaneous utterance. The theory was that spontaneous utterances are less likely to be calculated lies--which is true, but they're also more likely to be exaggerated or mistaken, and even more likely than calculated lies to be repudiated on cross. It's enough to make any lawyer or judge cry, "A pox on both their houses!"

And what about the non-Mirandized statement? In Davis/Hammon, the Court found that it is testimonial--and therefore can't be used to obtain a conviction. It is not clear whether such a statement would have been admissible under the old rules (but Rehnquist probably would have said it was).

An elusive definition

So this is the basic question before the courts: How do we know whether evidence is testimonial (inadmissible) or not? The answer depends on whether "the primary purpose of the interrogation is to establish or prove past events potentially relevant to later criminal prosecution." (14)

Now let me see if I have this straight. The definition depends on the police officers' purpose when they took a statement. But, as Justice Clarence Thomas noted in his dissent, the Court has repeatedly rejected police purpose in other cases as being "indeterminable and often mixed." (15) Thomas would distinguish between "formal" and "informal" statements. He would consider a Mirandized statement formal (testimonial and inadmissible) but would find the statements in both Davis and Hammon informal and admissible.

This rationale seems closer to historical intent and easier to apply. (16) But it has a serious flaw: Any test that allows an un-cross-examined 911 call as evidence to convict is simply wrong.

It remains unclear to what extent such calls will be admissible. The majority opinion said that if a 911 operator's purpose shifts from "an interrogation to determine the need for emergency assistance" to "questions designed solely to elicit testimonial evidence from a suspect," (17) then statements made during the call become inadmissible. Good luck to any trial courts trying to determine when that point is reached.

As with 911 interchanges, police questioning of victims and witnesses at the scene of a crime will likely have several functions: finding out what the emergency is and what needs to be done now, as well as discovering generally what happened and gathering evidence to use in court. At one point in its decision, the Court said that only "interrogations solely directed at establishing the facts of a past crime" produce testimonial statements. (18) If so, most statements will be admissible, since police purpose is usually mixed.

But at another point in the opinion, the officer's primary purpose seems to be the test. (19) And Davis itself excludes responses to police questioning at the scene.

Take, for example, "Describe the man who hit you," requested at the crime scene or during a 911 call. In general, once the police have arrived, the emergency is over. The same is true if a suspect flees before the victim can call 911. In these situations, the investigators' questions seem designed to elicit testimonial (nonadmissible) answers. But when testifying in court, police officers can argue that they needed to get information in order to find and arrest a dangerous, often armed, person, which renders the answer non-testimonial--because it is designed to dispel an ongoing emergency--and admissible.

A sensible standard

Finally, what does all of this have to do with the Sixth Amendment? As a practical matter, the right to confrontation cannot be absolute. As the Court conceded in Crawford, certain exceptions to the clause (such as business or government records) have always been recognized? (20) But that's because such records are likely to be reliable and cross-examination would not help. And although it is imprecise and requires a case-by-case analysis, that consideration should be the test.

Forget about "traditional hearsay exceptions." They rarely make sense. The trial judge should simply ask whether or not cross-examination would be useful. (21) If the declarant had reason to lie, was drunk, or was upset, then the statement should not be admitted unless the person who made it can show up to be cross-examined.

Eyewitness statements should never be considered nontestimonial, because cross-examination is always useful to the defense. The testimonial/nontestimonial confusion created by Crawford and reinforced in Davis/Hammon only takes courts further away from this common-sense approach to confrontation.

But as long as we must abide by the Court's testimonial/nontestimonial dichotomy, there is another way of reading the Davis/Hammon decision--one that does not depend on the elusive concept of police purpose. The Court drew a distinction between statements that describe events "as they were actually happening" (non-testimonial) and those that describe past events (testimonial). (22) The majority noted that the testimonial statement in Hammon was "neither a cry for help nor the provision of information enabling officers to end a threatening situation." (23)

This distinction is clearer. Under this reasoning, it doesn't matter what the police purpose is. Also, the standard could apply to statements given to a third party--for instance, a rape victim's description of the attack to an examining nurse would be testimonial, regardless of the nurse's intent. (24) I suspect that this is where the Court will ultimately settle.

The Supreme Court should be congratulated for moving away from the often irrational "traditional exceptions to hearsay" approach of Ohio v. Roberts, but this decision still doesn't get to the heart of what the Confrontation Clause should be about.

Don't miss the news

The ATLA L@w News Digest is a weekly summary of news and cases of interest to plaintiff lawyers, delivered to members via e-mail. To have the digest sent to your computer, send your name, member ID number, telephone and fax numbers, and e-mail address to [e-mail], or fax the information to [phone].

Notes

(1.) 126 S. Ct. 2266 (2006).

(2.) Justice Clarence Thomas concurred only in the result in Davis and dissented in Hammon.

(3.) 541 U.S. 36 (2004).

(4.) 448 U.S. 56 (1980).

(5.) Crawford, 541 U.S. at 40.

(6.) Id. at 41-42.

(7.) Id. at 44-46 (citing, as example, the unjust conviction of Sir Walter Raleigh).

(8.) "The court said that one of the 'indispensable conditions' implicitly guaranteed by the state constitution was that 'prosecutions be carried on to the conviction of the accused, by witnesses confronted by him, and subjected to his personal examination.'" Id. at 49-50 (citing an 1844 decision by the South Carolina Supreme Court).

(9.) Id. at 51. In Davis, the Court further declared that a slavish adherence to "the precise forms against which [the Sixth Amendment] was originally directed is a recipe for its extinction." (Davis/Hammon, 126 S. Ct. at 2278 n. 5.) Yet such slavish adherence is what motivated Crawford.

(10.) 541 U.S. at 69 (Rehnquist, C.J., dissenting).

(11.) Id. at 69 n. 1.

(12.) Id. at 70.

(13.) Thomas Y. Davies, What Did the Framers Know, and When Did They Know It?, 71 Brooklyn L. Rev. 105, 189-206 (2005). Davies says the history is not fuzzy and the Court is simply wrong.

(14.) 126 S. Ct. at 2274.

(15.) Id. at 2283 (citing New York v. Quarles, 467 U.S. 649 (1984) (Thomas, J., dissenting)).

(16.) However, as the majority was quick to point out, Thomas then waffles, making an exception to his "nontestimonial" formula for the state's use of "out-of-court statements as a means of circumventing the literal right of confrontation," whatever this might mean. Davis/Hammon, 126S. Ct. at 2278 n. 5.

(17.) Id. at 2277.

(18.) Id. at 2276 (emphasis added).

(19.) Id. at 2278.

(20.) 541 U.S. at 56.

(21.) I wish to thank Richard Friedman and Robert Mosteller for pushing me away from the "reliability" approach, which was so unsuccessful in the Roberts regime, and toward a similar (but unfreighted with the past) "cross-examination likely to be useful" approach. See Richard D. Friedman, We Really (for the Most Part) Mean It!, Mich. L. Rev. First Impressions 1 (2006), http://students.law.umich.edu/mlr/firstimpressions/vol105/friedman.pdf (last accessed Aug. 24, 2006), and Robert D. Mosteller, Beating Expectations, Mich. L. Rev. First Impressions 1 (2006), http://students.law.umich.edu/mlr/firstimpressions/vol105/mosteller.pdf (last accessed Aug. 24, 2006).

(22.) 126 S. Ct. at 2274, 2276.

(23.) Id. at 2279.

(24.) Presumably this is what the Court meant by its otherwise cryptic pronouncement that "police conduct [cannot] govern the Confrontation Clause; testimonial statements are what they are." Id. at 2279 n. 6.

Don't miss the news

Craig M. Bradley is the James Louis Calamaras Professor of Law at Indiana University School of Law in Bloomington. He can be reached by e-mail at [e-mail].