When is a statement presented for purposes of the Confrontation Clause?

I had virtually completed the post below when Bullcoming was decided, and then I had to put it aside for a while. As it happens, the two issues discussed here are both involved in Williams v. Illinois. I hope to write later on Williams, but for now I'll post this without further reference to that case.

In 2008, I wrote a post titled No confrontation issue if the statement isn’t even offered. I believe my analysis of the case discussed there was correct, but the title was overstated. Prosecutors occasionally try to evade the Confrontation Clause by offering, rather than the actual words of a statement, evidence from which the substance of the statement can be inferred. This attempt should not be countenanced. So we have what actually can be a tricky problem, determining when a statement is sufficiently presented to invoke the Confrontation Clause.

Courts are sometimes but not always alert to the problem, which sometimes is tied to another problem that I have labeled the "not-for-truth end run".

Jeff Fisher has brought to my attention a particularly crude prosecutorial attempt at evasion, one that fooled a district judge but was corrected by the First Circuit. United States v. Meises, 2011 WL 1817855 (1st Cir. May 13, 2011).

After being arrested, one Rubis gave an interview to DEA agents. The prosecutor recognized that statements made by Rubis during this interview were testimonial, and so decided not to offer the “actual statements”. Instead, the prosecutor established through one of the agents that Rubis had agreed to cooperate, and that he was interviewed alone. This exchange followed:

Q: After this interview, did the targets of your investigation at this point change?

A: Yes, sir.

Q: Okay. After this interview, what did you decide to do with Defendant [Reyes–Guerrero]?

The agent then testified that Reyes-Guerrero and Meises were processed and taken to a federal detention facility.

(As an aside, the only objection made at trial explicitly referred only to the hearsay rule, and the district judge held that the statement was admissible because it was by a co-conspirator. This is incorrect as a matter of hearsay law, as the First Circuit noted, because Rubis had already been arrested. It also does not address the Confrontation Clause issue, which the First Circuit held was adequately raised by the hearsay objection.)

On the merits, the First Circuit held “that a reasonable jury could only have understood [the testifying agent] to have communicated that Rubis had identified appellants as participants in the drug deal. It makes no difference that the government took care not to introduce Rubis's ‘actual statements.’

And, the court added, “any other conclusion would permit the government to evade the limitations of the Sixth Amendment and the Rules of Evidence by weaving an unavailable declarant's statements into another witness's testimony by implication.”

I think the First Circuit got it basically right, though I would put the governing standard somewhat differently. I think the key point is not quite whether “a reasonable jury could only have understood” that the in-court witness was communicating the substance of an out-ofhttp://www.blogger.com/img/blank.gif-court statement; the evidence might be ambiguous and still create a confrontation problem. I think I might instead pose this question:

Is the prosecution, in offering the evidence on a given ground, effectively asking the jury to, or is there an excessive risk that the jury will

(a) infer that the in-court witness is effectively communicating some or all of the substance of an out-of-court testimonial statement, and

(b) use that statement as proof of the truth of an assertion it makes?

See Ocampo v.http://www.blogger.com/img/blank.gif Vail, 2011 WL 2275798 (9th Cir. June 9, 2011) (concluding that before Crawford it was clearly established that "testimony from which one could determine the critical content of the out-of-court statement was sufficient to trigger Confrontation Clause concerns" and that Crawford did not alter this result; finding a violation in part on the basis of an in-court witness's statement that "did convey some critical substance" of the out-of-court testimonial statement); State v. Swaney, 787 N.W.2d 541 (Minn. 2010)(trial court "violates the Confrontation Clause when it admits testimony that inescapably implies a nontestifying witness's testimonial hearsay statement").

In Meises, I think the answer is affirmative. Apparently, the prosecution offered the testimony on the ground that it explained the subsequent conduct of the officers. But how would it do that? The most natural, if not the only, plausible inference is that Rubis made a statement suggesting guilt on the part of the defendants, and that the officers believed it. Why does that matter? The most likely use that the jury will make of the evidence is to conclude that the officers got it right, and had good reason to arrest the defendants.

And this is where the "not for truth" end run comes in. The prosecution says that it was only offering the evidence to explain the conduct of the police. The First Circuit properly rejected this argument, elaborating on “the limitations on so-called background or context evidence” in a footnote that included this passage:

We take it to be common ground that the government may not have an agent testify, “X told us that the defendant was involved in the crime.” Quoting X's out-of-court accusation remains impermissible if the agent's testimony is changed to say, “We began to investigate the defendant because X told us that the defendant was involved in the crime,” and the government seeks to justify it by arguing that X's out-of-court statement was offered not for its truth but only to explain why the agent focused on (or arrested) the defendant. Nor does the result change if, instead of quoting the out-of-court statement, the government communicates its content to the jury by implication.

The simple fact is that in a case of this sort the jury doesn't need to know why the police arrested the defendants -- apart from knowing the evidence properly presented to them. But there is a very real danger that the jury will conclude that the out-of-court speaker made a truthful accusation to the police. And so the First Circuit properly decided that whatever valid probative value the evidence might have had was outweighed by the prejudicial danger of the evidence. That sounds like a discretionary ruling under Fed. R. Evid. 403 or a state counterpart. But it goes beyond that. Because the probative value for setting context is so minimal, and because there is a substantial probability that the jury will use the evidence to infer that a testimonial statement was made and is true, thus greatly aiding the prosecution, in this context the balance indicates that in effect the prosecution has presented a testimonial statement for the truth of an assertion it made. And so this is a judgment used to determine whether the Confrontation Clause is invoked -- it is not merely a discretionary matter of weighing on which an appellate court should defer to a trial court.

My comments here are of a piece with the standard I tentatively offered in a 2007 post, Conflict deepens on expert evidence, on the general question of judging when a testimonial statement supposedly not offered for the truth raises a Confrontation Clause problem:

It seems that a court . . . must first ask whether the testimonial http://www.blogger.com/img/blank.gifstatement has substantial value in proving the proposition for which it is supposedly offered without respect to whether the statement is true or not. If the answer is negative . . . then that proposition does not offer a basis for admitting the statement. If the answer is affirmative, then the court must assess whether that value warrants the risk that the jury will, notwithstanding an instruction to the contrary, use the statement to prove the trhttp://www.blogger.com/img/blank.gifuth of what it asserts.

As for that 2008 case, United States v. Tucker, 533 F.3d 711 (8th Cir. Jul. 17, 2008), here for ease of reference is what I wrote about ti then:

Tucker assisted her boyfriend Robson in a get-away from a bank robbery, and the critical issue was whether when she did so she knew that he had robbed the bank. So, as the 8th Circuit indicated, evidence that Tucker knew that Robson had committed other bank robberies was admissible to show her state of mind on this occasion, and it was also admissible to impeach the credibility of her testimony that she didn’t know Robson had robbed the bank. In cross-examining Tucker, the prosecutor asked whether on three specified prior occasions Robson had robbed a bank and either he had told Tucker about it shortly afterwards or she had helped him get away. Tucker denied each. No evidence of the other robberies was admitted.

If the prosecutor had no good-faith basis for asking questions of this sort, they would be highly prejudicial and improper. But if she did, then they were fair game – even if she had no other admissible evidence of the other robberies, she was entitled to hope that Tucker would tell the truth (as the prosecutor believed the truth is or may be, based on the information that gave her the good faith belief), and that would give her the evidence she wanted. And in this case, she did have a good-faith basis – a statement Robson had made to the FBI. The statement was clearly testimonial, and couldn’t be admitted against Tucker, given that she had not had a chance to cross-examine him. It wasn’t offered, and the jury never even learned of its existence, though they might well have inferred that the prosecutor was basing her questions on something of the sort. But the statement could, without violating the Confrontation Clause, form the good-faith basis for the prosecutor to ask her questions, and that is what the 8th Circuit properly held.

So clearly in Tucker the prosecutor did to ask the jury to infer the truth of a statement that could not itself be admitted; the logic of the offer did not require that. There may have been some risk that the jury would infer such a statement, but (a) the jury would not necessarily know what basis of information the prosecutor had for asking the question, and (b)the risk of such an inference is made much more tolerable by the fact that the defendant opened the door to impeachment by denying a proposition that the prosecution had good reason to believe was true. So neither aspect of the test I have suggested was violated.