A useful opinion on statements not reported explicitly at trial

On the first of this month, the Michigan Court of Appeals issued a useful decision, People v. Washington, concerning the situation in which a prosecution witness does not explicitly report the content of an out-of-court testimonial statement but nevertheless conveys its substance.  A 2-1 majority of the panel properly held that such testimony violates the Confrontation Clause. (I didn't want to post on the case when it first came out, because I used it for a series of exam questions.)  And I'm glad to say that the brief for Washington was drafted by two students in the University of Michigan Law School's Criminal Appellate Advocacy clinic, Greg Momjian and Erica Arensman, under the supervision of Doug Baker; it was argued by Kierston Nunn.

The defendant, Washington, was convicted of possession or use of body armor by a person who had previously been convicted of a violent felony.  A toll-both operator on the Canadian side of the Blue Water Bridge, which connects Port Huron, Michigan, and Sarnia, Ontario, testified that she had seen a Ford pickup truck blow through her lane, following close on another car.  She immediately notified her supervisor, and several officers gave chase.    

Paul Stockwell, a supervisory officer with U.S. Customs and Border Protection, testified that on the American side of the bridge he met Matt Lavers, one of a crew of Canadian officers, and that they had Washington in a cruiser.  Neither Lavers nor any other member of the crew testified at trial.  Stockwell's testimony included this exchange:

[The Prosecutor]: At some point did Officer Lavers from the Canadian services hand you any other evidence?

Officer Stockwell: Yes, he did.

[The Prosecutor]: What did he hand you?

* * *

Officer Stockwell: A body armor.

* * *

[The Prosecutor]: Now, without saying anything about what was said, the only question I have for you is were there communications between you and Officer Lavers?

Officer Stockwell: There were.

[The Prosecutor]: Okay. And ․ based on those communications you took custody of [defendant]?

Officer Stockwell: Yes, we did.

[The Prosecutor]: And you took possession of the body armor that was turned over at the same time?

Officer Stockwell: Yes, sir.

The testimony took that form because, in response to an in limine motion to exclude evidence of the armor (a bulletproof vest) the trial court had ruled that Stockwell could testify as to what was turned over to him so long as he did not testify as to what was said.  Washington was convicted and appealed.  The court of appeals, in an opinion by Judge Elizabeth Gleicher (Judge Sima Patel concurring), reversed.

First, and most significantly, note that no statement by Lavers was actually introduced.  But, the majority pointed out, there would have been a clear violation of the Confrontation Clause had the prosecution asked Stockwell, "What did Officer Lavers tell you about the body armor?" And, the court noted, Stockwell's actual testimony "conveyed precisely the same information: that Washington had been wearing the body armor when caught by the Canadian authorities."

This has to be right, I think: What would otherwise be a Confrontation Clause violation cannot be excused by the fact that the in-court witness does not purport to report the actual statement made, if the effect of the testimony is to convey the substance of that statement to the trier of fact.  It seems plainly obvious that it cannot be that the Confrontation Clause is violated only when the in-court witness purports to be repeating the statement verbatim; if that were true, the Clause would be rendered a virtual nullity with respect to unrecorded oral statements, because the in-court witness would not usually remember the statement verbatim anyway.  And similarly it cannot be that the Clause can be evaded by asking the in-court witness, "Please summarize in your own words what [the absent witness] said."  Now in Washington, and in similar cases, the prosecution did not explicitly ask the in-court witness to communicate the substance of what the out-of-court witness said.  But in some circumstances, as in Washington, that is, as the majority said, precisely the effect of a series of questions that leads to an inference as to what was said.  Here, the prosecutor asked Stockwell whether Lavers made a statement and what Stockwell did as a result.  What could the jury infer in the circumstances but that Lavers said he had found Washington wearing the vest?  Indeed, if that were not the inference to be drawn then the case should never have gone to the jury, because there was not other evidence that would support the inference that Washington was found wearing the vest.

Judge Jane Markey, in dissent, argued that implied assertions are not hearsay and should not be deemed to create a confrontation problem.  It seems to me that this is a rather clear category error.  What are commonly called implied assertions (and perhaps should more accurately be called inferred assertions) are a type of conduct that is not intended to assert the proposition at issue but that indicates the actor's belief in the truth of that proposition.  So, most notably, in the famous case of Wright v. Tatham (H.L. 1838), letters written to the decedent Marsden assertedly indicated the writers' belief that Marsden was of sound mind, but none of the letters asserted that proposition.  The Wright majority deemed these to be hearsay, but under modern law, they would not be (see FRE 801(a) ("intended . . . as an assertion").  And there would not be a confrontation problem. (Notice that conduct of this sort is almost inevitably not testimonial; witnesses who testify intend to make assertions.)  But this is an altogether different matter from what is at stake in the Washington case and others like it:  In Wright-type cases, the out-of-court actor did not assert the proposition at issue, and the evidence of the actor's conduct is thought to imply his or her belief in the proposition.  In Washington-type cases, the out-of-court witness presumably did assert the proposition at issue ("Washington was wearing this vest"), and that fact is inferred from the evidence actually presented at court.  These are altogether different situations.

Judge Markey raised other issues, standard ones in confrontation cases, but I don't think either has merit, and both were rejected by the majority.  First, she contended that the evidence of Lavers's communication to Stockwell was not offered for the truth of the presumed assertion but rather only to explain why Stockwell took Washington into custody.  This is the "why the officer did that" end run that prosecutors often offer and on which I have commented before.  In most cases, including this one, there is no reason why the jury needs to know that.  And besides, as the majority pointed out, "Stockwell took Washington into custody because Washington was in possession of body armor." 

Second, Judge Markey contended that the presumed statement of Lavers was not testimonial because "the purpose of the communication was simply to provide Officer Stockwell with the information necessary for Officer Stockwell to formally take defendant into custody from Canadian authorities and proceed with the charging process under Michigan law."  But, as the majority said, "The 'explanation' for Stockwell's conduct is relevant only because it inculpated Washington." Under a proper view of what is testimonial – I am tempted to say under any reasonable view – if one officer passes on information to another on the understanding that the other will use the information in prosecution of a crime, and presumably will transmit the first officer's report of that information to the trier of fact, that conduct is testimonial.  If not, we have created a system in which a witness can testify by passing information to an officer, without even an intention that the witness will show up at trial.  That's a clear violation of the confrontation right.

Shields v. Kentucky: A cert petition to watch on preliminary-hearing testimony

Jonathan Ellis and Gregory DuBoff of the McGuire Woods firm have filed an excellent cert petition in Shields v. Kentucky, No. 22-450.  The question is whether, or in what circumstances, testimony of a prosecution witness at a preliminary hearing may be used at trial if the witness is then unavailable.  This is an issue on which the Supreme Court has not given any guidance since Crawford, and the petition ably shows that the lower courts are in clear dispute.

Shields was charged with murdering his uncle, and the prosecution presented the testimony of Shields's elderly aunt, the decedent's widow, at the preliminary hearing.  It did so without advance notice, or any indication that it was seeking to preserve the aunt's testimony for trial, and without making disclosures of exculpatory evidence that would have been required by trial.  Cross-examination was perfunctory.  The Commonwealth never attempted to take a deposition. The aunt died before the time scheduled for trial.  Shields entered a conditional plea of guilty, allowing him to appeal the trial court's ruling that the preliminary-hearing testimony was admissible.  The Kentucky Supreme Court affirmed.

My own view is that preliminary-hearing testimony of a prosecution witness who has become unavailable by the time of trial should be allowed if and only if certain conditions (apart from the obvious ones that the accused has a right to be present, to be represented by counsel, and to have the witness testify under oath) are satisfied.  The purposes of a preliminary hearing and of trial, and defense counsel's goals, are very different -- it is usually, as it was here, a foregone conclusion that the defendant would be bound over for trial.  Accordingly, preliminary-hearing testimony cannot be a substitute for trial testimony unless (1) the prosecution has given the accused notice that it intends to use the testimony, if necessary, for preservation; (2) the notice is given in sufficient time to allow adequate preparation; (3) the prosecution has made relevant disclosures that would be required by the time of trial, at least those that would be constitutionally required; and (4) there were no limitations beyond those that would apply at trial.  In addition, I think that probably, if the accused is able to show that developments subsequent to the preliminary hearing would have opened up significant lines of cross-examination, that would justify a holding that the earlier testimony is not adequate.

I've expressed these views in an amicus brief in support of the petition.  You can find the papers filed so far in the case -- including the petition, my amicus brief, and two other amicus briefs filed in support of the petition, by clicking here.

Samia v. United States: Cert granted in a Bruton-type case

 Earlier this month, the Supreme Court granted cert in Samia v. United States, in which it may resolve one of the issues left open by Bruton v. United States, 391 U.S. 123 (1968), Richardson v. Marsh, 481 U.S. 200 (1987), and Gray v. Maryland, 523 U.S. 185 (1998).

Samia was tried along with two other defendants for murder for hire and related crimes.  One of the co-defendants had made a confession that named Samia.  The court instructed the jury not to consider it against Samia, but under Bruton, that alone would not have been sufficient.  Nor, under Gray, would it have sufficed if the confession had been redacted with Samia's name replaced by some indication such as "[name deleted]."  But under Richardson, if it had been redacted to remove all references to Samia, that would have been acceptable.  Samia's case falls between Gray and Richardson; the trial court allowed redactions so that the confession as reported to the jury included references in forms such "somebody else" and "the other person he was with." Given other evidence in the case, it could be readily inferred that the references were to Samia.

I'm going to guess that the cert grant suggests that the Supreme Court will come down on Samia's side.  We should know before the end of the term.

A good decision in a fresh-accusation case

I use the term fresh accusations to cover statements made shortly after an event asserting that someone committed a crime.  Often, though not inevitably, they are made to a police officer or 911 operator.  They are especially common in domestic-violence cases.  I believe that Davis v. Washington left the law in this area in a very unsatisfactory state.  If a court resolves that the statement was made primarily to resolve an "ongoing emergency," then it will call the statement non-testimonial.  And courts will very often seize on any facts that allow them to characterize a statement as having been made in response to an emergency.  

So State v. Tapper, a unanimous decision by the Minnesota Court of Appeals, is a welcome development in this respect, though it should be utterly unsurprising.  This was a domestic-violence case, and the statements in question were made to a responding officer after a 911 call.  But the court recognized that this particular incident had ended and that the complainant was not in immediate danger; she was outside her apartment with the officer and her ex-husband, the alleged assailant, was apparently asleep inside.  Her children were inside and she said that she would not leave them, but she did not "express fear for their safety or indicate they [had] been harmed by Tapper in the past."  She had suffered injuries – she was covered with blood and reported that her head was "messed up," but most of the conversation concerned what Tapper had done on the incident in question an on prior ones.  She was sniffling at the beginning of the conversation, wiping her eyes and nose and breathing rapidly, but soon she was able to speak calmly, and she answered the officer's questions coherently.  Some courts might have used these factors – the presence of the kids in the apartment with the alleged assailant, the complainant's injuries, and her demeanor at the beginning of the conversation – to justify a conclusion that the statement wasn't testimonial.  But it clearly was, and the Minnesota appellate court, like the trial court, recognized this.  Too bad it designated the decision as non-precedential!   

A good decision from the Fifth Circuit on the "why we acted the way we did" end run

I have written before -- most recently here -- about how too often a court allows a prosecutor to introduce a testimonial statements on the ground that it is not offered for the truth, when the prosecution has no real need to prove the proposition for which the evidence supposedly is offered. In particular, courts often allow prosecutors to use such statements to prove the reason for authorities’ conduct, especially why they investigated the accused.

This is a particularly difficult problem, because a testimonial statement can be introduced for a legitimate purpose that the prosecution has, other than to prove the statement’s truth, and one cannot say that explaining officials’ conduct is utterly immaterial.  So some degree of balancing tends to come in; if the prosecution’s need is flimsy, usually the court should recognize the very substantial probability that the jury will use the statement for its truth and so the statement should not be admitted.  But courts tend not to do a good job of balancing when the confrontation right is at stake; the temptation to admit probative evidence is often too great, and that is one reason the categorical rule of Crawford was necessary.

Some courts get it, though.   Eric Freedman, who has been keeping me informed for more than four decades, has called to my attention the fine opinion Judge Jerry Smith for a unanimous panel of the Fifth Circuit in United States v. Hamann, which is not only keenly aware of the problem but emphasizes a categorical rule that, if followed, would solve a good deal of it.

Hamann was charged with conspiracy to distribute meth.  An investigator, Stanley, testified to a statement by a confidential source to the effect that “Cali,” understood to be Hamann, was “moving multiple ounces” of meth and to also to statements by local law enforcement that an unknown declarant had said Hamann “was selling narcotics.”  Stanley also testified to details concerning a controlled purchase that was used to establish probable cause for a search warrant – but this testimony was second- or third-hand, because Stanley did not observe the transaction.  The panel had little difficulty that Stanley was thus able to testify to out-of-court testimonial statements.  The fact that he did not recount an “identifiable statement” concerning the controlled purchase did not matter, given that he described the purchase in detail despite not having observed it, making it inferrable what he purported to have been told.  All good.

The more serious question was whether the statements were used for the truth of what they asserted.  The panel noted that prosecutors often try to introduce evidence highly probative of guilt “[u]nder the guise of explaining why police began their investigation or conducted it a certain way.” But even if the evidence is relevant for this purpose, the panel said, quoting the court’s earlier decisions, the trial court must be “circumspect in its use”; the evidence would be deemed admitted for the truth of what it asserted if it “specifically links a defendant to the crime” or, put another way, “points directly at the defendant and his guilt.”  Thus, “[t]he government must advance a specific reason why it needs to provide inculpatory ‘context’ for its investigation.” The defendant might open the door by challenging the adequacy of the investigation, but “[o]therwise, there is no reason why it cannot begin its account by explaining that it got a search warrant or that ‘a tip prompted’ it to begin investigating a suspect.” 

Application of these principles here was straightforward.  Hamann had never contended that the investigation was inadequate.  And so the panel said, “We perceive no reason why the government could not have begun its case-in-chief by explaining that officers arrived at the motel to execute a search warrant and found Hamann and Davis together in the parking lot holding distributable amounts of meth.”  The prosecution’s “inculpatory prequel” was far from circumspect, and highly prejudicial.  Hamann hadn’t had an opportunity to cross-examine the confidential source or the non-testifying source (and neither was shown to be unavailable).  So there was a clear Confrontation Clause violation, which the panel concluded was not harmless.

The panel made clear that it was simply reaffirming principles it had often stated, but the vigor with which it stated them is notable; it said that “ the government has repeatedly failed to take the lesson” and it “remind[ed] prosecutors to take note” – it could and probably should have have included trial courts in these statements.  And the decision appears to state not only a stern and skeptical attitude but a very useful per se rule: If a testimonial statement specifically links an accused to a crime, and the accused does not challenge the adequacy of an investigation, then the statement may not be introduced on the basis that it explains the official conduct.  That does not solve the whole problem, but it does take care of a good large part of it.

An encouraging cert denial in an autopsy case

It's always a mistake to read too much into a denial of certiorari, but I find encouraging the Supreme Court's denial this morning in Lee v. Garlick.  This is a homicide case, and at trial the prosecution was allowed to introduce and rely on an autopsy report that indicated that stab wounds were the cause of death; that was crucial evidence, because it was clear that the accused, Garlick, had stabbed the victim and that another person had hit him repeatedly in the head.  But the report was introduced through a medical examiner who had played no role in the autopsy or in preparation of the report.  Garlick was convicted, the New York courts upheld the judgment, and the Supreme Court denied cert; in a prior post, I referred briefly to the case and posted the petition.  (That one included a link to a useful research memo by Cody Reaves, a recent Michigan Law alum, on the topic.  Among my other prior postings on autopsy reports is this one.) But Garlick sought federal habeas relief.  The district judge, acting over the recommendation of a magistrate judge (who thought the point was not sufficiently well established for habeas), granted the petition.  And a panel of the Second Circuit unanimously affirmed.  The decision basically says – correctly, in my view – that the New York courts have been clearly wrong in treating autopsy reports in murder cases as non-testimonial.  Michael Hutter of Albany Law School, who called my attention to this phase of the case, has pointed out to me that one of the judges on the panel, Richard Wesley, is a former judge of the New York Court of Appeals, though he had left that court before it began issuing decisions on this question.

It was this Second Circuit decision that the Supreme Court declined to review today.  Although I've thought for fifteen years or so that the Court ought to resolve definitively (and affirmatively!) the question of whether autopsy reports in homicide cases are testimonial (note my failed cert petition in Craig v. Ohio, I think it's probably better if the issue comes up cleanly, not enmeshed in the rigorous standards for granting habeas.

Beyond that, I think the cert denial is also good news because the Supreme Court evidently didn't turn a hair at a federal appellate decision holding that it is clearly established that autopsy reports are testimonial.  What is more, the decision was a forthright, boundary-pushing one, in the right direction.  The panel waved at the "primary purpose" test but did not rely at all on it; instead, it seems to have regarded the case, quite appropriately, as a simple, straightforward application of Melendez-Diaz v. Massachusetts, and the key sentence applies what I regard as the appropriate standard:  "The circumstances under which the autopsy report was created would lead any objective witness to 'believe that the [report] would be available for use at a later trial.'" And this is from the same court that years ago had issued an atrocious decision, after Crawford but before Melendez-Diaz, saying that autopsy reports should be admitted as business records; the Garlick court didn't even feel the need to cite that case.  So all in all, I think it's great that this decision stands.

By the way, it now appears that when the Supreme Court resolves this issue, and all other Confrontation Clause issues, it will be with Justice Breyer's successor.  For some years, he has been one of the Court's steadiest skeptics about Crawford.  Justice Kennedy was another.  I'm hoping that the changes in membership will lead to better, sturdier decisions.  And one of them should be a clear affirmation that when an autopsy report points to homicide as the cause of death, it's clearly testimonial, whatever other purposes it might serve.

The "why I acted the way I did" end-run

 Prosecutors often try to get in testimonial evidence from absent witnesses by arguing that it is not offered for the truth of what it asserts but explains why a testifying officer acted the way he or she did.  This points to one of the great vulnerabilities of the confrontation right:  A prosecutor has an incentive to come up with some ground on which the evidence is relevant, other than for its truth, and given the low bar of relevance it is not hard to do so.  The courts therefore have to be closely vigilant to ensure that the right is not eviscerated in this way, state categorical rules as to when an asserted ground of not-for-the-truth relevance does not support admissibility of the evidence.  One frequently used such ground is the supposed need of the investigating officer to explain why the officer acted in the way he or she did.  But in most cases, there really is no need for such an explanation.  Tell us what you did, officer, and what happened after that.  Presumably you detected criminal activity, and the jury will know all it needs to about why you acted the way you did.  If the defendant raises some doubts about why you acted in a given way, well that may be a different story.  But short of that, there's no need to get into it.  

Some courts have realized this.  The Mississippi Court of Appeals issued a decision yesterday, Lowe v. State, that collects some of the cases.  In Lowe, a confidential informant told the officer that she could buy drugs from the defendant and had done so in the past.  The officer testified at trial but the CI did not.  The trial court allowed the statements in to explain the officer's conduct.  The appellate court seems not to have had any difficulty with the prediction -- it should have -- but it recognized that the statement of past conduct was testimonial and there was no need for it; its admission was a violation of the Confrontation Clause, though the court held the error harmless.        

Hemphill v. New York: the right call

 

Apologies for being silent so long.

The Supreme Court decided Hemphill v. New York this week.  The Court held that the Confrontation Clause does not tolerate the admissibility of an out-of-court testimonial statement on the basis of a judicial determination that doing so is necessary to correct what the court deems to be a misleading contention by the defense.  On this point, the Court stood 8-0.  Justice Thomas dissented alone, and on other grounds; he believed that the defense had not properly preserved the issue.

Briefly, Hemphill was accused of murder, but there was substantial evidence that another man, Morris, had committed the crime.  Indeed, Morris had originally been charged with the murder, but then was allowed to plead guilty to a much lower-level weapons charge.  Hemphill’s attorney referred to the fact that police had found cartridges from a 9 mm. handgun, the type of weapon used in the killing, by Morris’s bedside.  The trial court then allowed the prosecution to prove that at his plea hearing Morris had said he brought another type of weapon, a .357 magnum revolver, to the scene of the crime; Morris was unavailable at the time of Hemphill’s trial and Hemphill had never had an opportunity to be confronted by him, but the court thought this was necessary to dispel a misleading inference created by Hemphill’s contention.  The intermediate appellate court and the Court of Appeals agreed, but the U.S. Supreme Court did not.

The case should have been regarded as an easy one on the merits, and the Court treated it as such.  Justice Sotomayor’s opinion for the Court declared:

For Confrontation Clause purposes, it was not for the judge to determine whether Hemphill’s theory that Morris was the shooter was unreliable, incredible or otherwise misleading in light of the State’s proffered, unconfronted plea evidence.  Nor, under the Clause, was it the judge’s role to decide that this evidence was reasonably necessary to correct that misleading impression.  Such inquiries are antithetical to the Confrontation Clause.

Exactly right.  On the whole, Hemphill is a welcome development, because of the result, because of the lack of any dissent on the confrontation point, and because it is an unqualified endorsement and application of Crawford.  In contrast, say, to Justice Sotomayor’s majority opinion in Michigan v. Bryant, there is nothing in her opinion for the Court here that suggests buyer’s remorse about Crawford. On the contrary, it speaks approvingly of “Crawford’s emphatic rejection of the reliability-based approach of Ohio v. Roberts,” and underlines the point by saying:

If Crawford stands for anything, it is that the history, text, and purpose of the Confrontation Clause bar judges from substituting their own determinations of reliability for the method the Constitution guarantees.

(I wish the Court would stop talking about reliability in this context; even cross-examined eyewitness testimony is not reliable.  Speaking of accuracy would be an improvement.  But oh well, old habits die hard.)

The Court made clear that it was not holding invalid the traditional rule of completeness as applied to portions of testimonial statements that would otherwise be inadmissible under the Confrontation Clause.  Rather, it reserved the issue.  In my view, such an application poses no problem under the Clause; it would be absurd if an accused were allowed to introduce the portion of a statement up to but not including “But . . . “ and then object to the latter part on the basis that he did not have an opportunity to be confronted by the maker of the statement.  Justice Alito, joined by Justice Kavanaugh, wrote a separate concurred to endorse this position, but they joined the majority opinion as well.

One aspect of the case is somewhere between rankling and troubling.    Although the statement at issue was made in a plea hearing, the State contended below that it was not testimonial because it did not “incriminate or point a finger at all against Mr. Hemphill.”  Before the Supreme Court, the State did not deny that the statement was testimonial, and in a footnote Justice Sotomayor wrote that the Court “expresses no view on the matter.”  Justice Alito’s concurrence makes clear in its first sentence that its conclusion that admission of the statement violated the Confrontation Clause is drawn “assuming Morris’s statement was testimonial.” So this point-a-finger idea is a clear reference to Justice Alito’s suggestion in Williams v. Illinois that a statement is not testimonial if it is not aimed at a “targeted individual.”  Though Justice Alito’s Williams opinion spoke for four justices and was on the prevailing side, that point was explicitly rejected by five justices, so I am disappointed to see that he is keeping it alive and that the Court evidently felt that to achieve virtual unanimity it was necessary to keep the point alive.  Apparently, Justice Alito wants to be able to contend in the future that if in a similar case Morris was sworn before a grand jury investigating the incident and asserted that he was at the scene with a .357 and not a 9 mm gun, that would not be testimonial, and neither would a sworn statement by another observer describing the scene of the incident and saying that the shooter appeared to take deliberate aim and was a young male wearing a blue sweater.  (By the way, I think the lab report was in fact targeted at a given individual – the person who had the DNA profile described; there is presumably only one person with that profile, but many, many young men may wear blue sweaters.)  I hope that if the issue is again presented, a majority of the Court will again slap the theory down.  It’s unfortunate that it still has breath.

But on the whole, Hemphill is good news.