Thursday, November 13, 2008

Thoughts on Melendez-Diaz: chain of custody, products of a machine, who must testify, etc.

While Melendez-Diaz is pending, I am going to write about various significant issues beaing on it. Here I will write about the chain of custody, products of a machine, who must testify, and other related topics.

Chain of custody is not a technical requirement that should be belittled as a mere technicality. The chain is part of the requirement of authentication, which is that the proponent demonstrate that a piece of evidence is what its proponent claim it to be. Absent authentication, the evidence has no substantial probative value, because its connection to the case has not been demonstrated. Authentication is a fundamental part of the whole story told by the prosecution, which is how a course of events that included the crime charged wound up producing the evidence that the prosecution has presented in court. Some items are treated as self-authenticating because from the face of the item it is readily inferrable, on the basis of information of the type that jurors are allowed to bring with them to court, that the item is what it appears to be; no evidence need be produced that an item that appears to be the New York Times of November 14, 2008 is exactly that. But assuming, as is usually the case, that the item is not self-authenticating, then, like any other part of that story that the jurors cannot infer based on the information they are entitled to bring to the job, authentication has to be proved by ordinary admissible evidence.

So let’s say a breathalyzer spits out a red splotch on paper if the subject’s blood alcohol level is excessive. Plainly, the prosecution cannot prove its case simply by introducing a red splotch on paper of the type generated by the breathalyzer. Absent evidence that this splotch was generated by a test performed on the accused’s breath, the splotch is essentially irrelevant; its connection to the story we are concerned about has not been shown.

Suppose now that a lab technician who performs the test writes on the margin of the paper, "Jack Sparrow, Nov. 14, 2008." Taken in context, that is an assertion that the red splotch is the result of a test done on Jack Sparrow on the date recited. That assertion was plainly meant for evidentiary purposes, and it is testimonial. The red splotch itself is not testimonial, of course – as has been noted on the states' side in this case, and by some courts, that is just the product of a machine. But without a testimonial statement as to just what that product is, it is of no help to the adjudication. And if the paper with the writing is offered by the prosecution, then a testimonial statement has been offered without the witness having been subjected to cross-examination. That is a Confrontation Clause violation.

So what can the state do, assuming the defendant does not waive the confrontation right? Obviously, the way it can most clearly solve the problem is to present as a witness at trial, the technician, who will testify how the splotch was generated. I will discuss here five other possibilities. Two are clearly permissible. Another should be permitted, in my view, though it is controversial. The final ones I believe are clearly unconstitutional.

1. A properly drawn notice-and-demand statute is constitutional. Petitioner has acknowledged that a "bland" notice-and-demand statute is constitutional. That is, so long as all the defendant has to do to get confrontation is make a timely demand, there should be no problem. Of course, a statute could make an unreasonable demand on the defendant, and that could create a problem. And I think some leeway ought to be allowed if some reason arises belatedly giving the defendant reason to demand confrontation. But in general, a well-drafted notice-and-demand statute is constitutional, and that relieves much of the problem; most defendants have no desire to have the technician tesify live.

2. The constitutional requirement is not that the person who performed the test or wrote the report testify subject to confrontation, but that someone testify from first-hand knowledge about the process leading to the reported results. The critical factual propositions to which the technician would testify at trial, if she does appear, are:

a. The person or bsubstance tested was the one material to the case.

b. The test was performed as described by the witness.

c. (1) The results of the test were as testified by the witness, or (2) A
given tangible item is the product of the test.

Testimony subject to confrontation as to these propositions must be presented by a witness with personal knowledge of them. That witness could, of course, be the technician who performed the test, but it need not be. Suppose that at every step along the way the technician is accompanied by a trainee who touches nothing and writes nothing but observes everything the technician does. (Some jurisdictions routinely have multiple physicians present at an autopsy.) Then that trainee could testify at trial as to these propositions. For that matter, suppose the cop who brings the suspect, or the sample, to the lab stays and watches the entire test being performed. The cop could then testify from first-hand knowledge as to these propositions as well.

Now note that if the technician testifies she will often testify as to a fourth proposition:
d. The proper interpretation of the test results is as testified by the
But even assuming that interpretive evidence is necessary, there is no requirement that the witness providing the interpretation be the one who testifies as to the three factual predicates identified above; as long as a witness with first-hand knowledge is able to testify to those predicates, then any witness with sufficient expertise to interpret the test results may do so.

3. The state may provide a pre-trial opportunity for confrontation, and if it does the courts should be generous in holding that the witness is unavailable at the time of trial. Much of the inefficiency that Massachusetts and its supporting amici complain about would be avoided if the state provides a full pre-trial opportunity for confrontation. The deposition can be scheduled ahead of time, and does not have to depend on what happens at trial. A witness coming from a distant city could schedule several depositions on one trip. In most cases involving lab reports, there is no mystery early on what the lab report would be used to prove and what the defendant would like to achieve on cross. Early cross-examination offers the possibility that the witness will have a clearer memory of the test than at trial. And the deposition can be videotaped.

The sticking point is that under Crawford a pre-trial opportunity for cross suffices only if the witness is unavailable to testify at trial. I think it would be wise to be rather generous in deeming a witness who has been subjected to cross-examination to be unavailable at the time of trial. Unavailability is often a matter of degree, and given the prior cross and the rather small probability that a second cross would add much, I think the trial court should be allowed to characterize as unavailable a lab technician who would have to travel a great distance. Also, if at the time of trial the witness has virtually no memory of the particular incident -- and that should be subject to demonstration without live testimony from the witness -- then that is a basis for holding her unavailable.

Pretrial testimony is a second-best solution. But in this context, I think it should be constitutionally acceptable.

4. A surrogate should not be allowed to testify to the factual predicates necessary for proving the test results. I stated above the three predicates necessary to admit lab results. Notwithstanding the position taken by several other law professors in an amicus brief, I think it is plainly inadequate for a witness who did not observe the conduct of the test or the handling of the substance tested to report someone else's assertions as to the chain of custody of the substance, the test performed on it, and the results of that test. ("I didn't see the stuff being handled, I didn't see the test being performed, and I didn't see the results of the test. But I'll tell you what my colleague wrote on these points.") There is simply no justification for presenting a surrogate witness to report the testimonial statement made by another witness who has not been subjected to confrontation and whom the state cannot or would rather not produce.

5. The state should not be allowed to shift the burden of presenting the witness to the defendant. I have said that a simple notice-and-demand statute should be constitutionally acceptable. But the confrontation right is not satisfied by providing that the defendant may call the technician as his own witness. The passive language of the Confrontation Clause suggests that such a burden-shifting statute is unacceptable. So does the fact that, given the Compulsory Process Clause, the Confrontation Clause would be rendered a virtual nullity if such a statute were valid. But to my mind, the most significant factor is that practical considerations make clear that telling the defendant, "You may call this technician whose report has been entered against you" is simply not the equivalent of telling the defendant, "You may cross-examine the prosecution witness who has just testified against you." My pending petition in Briscoe v. Virginia explores the reasons why this is so.

Monday, November 10, 2008

The Melendez-Diaz argument

I have not been able to work much on the blog lately, because of the press of other work. But I attended the argument of Melendez-Diaz v. Massachusetts in the Supreme Court yesterday and I want to report on that. (The transcript is available on the Court's website; you can see it by clicking here.)

For those who have not been tracking the case, it concerns the question whether the Confrontation Clause covers a certificate of a lab report that a given substance cocaine. I believe the answer is clearly affirmative, and I am hopeful that the Court will reach this result; I would not be amazed if it were unanimous, though that is of course hard to predict.

There was no substantial doubt that the certificate was prepared in contemplation of use in prosecution; the statute makes it clear on its face that these certificates are to be prepared for usei n law enforcement, and the Commonwealth (represented at argument by Attorney General Martha Coakley) has not denied the point.

Much of the discussion concerned the practical impact of holding that lab reports are testimonial. Jeff Fisher, arguing for the petitioner, explained effectively that in most case the author of the report would not have to come to trial because usually defense counsel would be willing to do without confrontation. Defense counsel might make this clear by stipulating to admission of a certificate. Also, a state can validly create a simple “notice and demand” system, under which the state gives notice of intent to introduce the certificate and it is admissible unless the accused demands that the state produce the author of the report. But it is the defense’s choice whether to demand confrontation, not one that can be made by a court. (Jeff made the same point in response to questions about whether cross-examination of the analyst would be futile given the analyst's presumed lack of memory; he might have also added that the Supreme Court has said, in upholding admissibility of a witness's prior statement despite the witness's inability to respond to many questions on cross, that showing the witness's loss of memory is a benefit to the defense.) Jeff also pointed out that numerous states – including California and other large ones – currently have procedures that are constitutionally satisfactory, and they seem to be able to operate effectively. At one point, Ms. Coakley said that California had joined the amicus brief of Attorneys General – but the Chief Justice soon pointed out that she was mistaken.

Lisa Schertler, arguing for the United States as amicus, contended that the certificate should be admitted because it merely reported the output of a machine. I don't think she made much headway with the argument, and for good reason. Justice Scalia presented an interesting hypotheical -- the time of murder is significant, and a witness says the clock was striking twelve. Obviously, that would be testimonial. The output of the machine is of course not itself testimonial, but the certificate reflects much more. At the outset of the argument, I believe there may have been some confusion on this score, but by the end -- there was a good exchange between Jeff and Justice Souter at the very end -- I think the matter was clear: The state must present a witness who can testify from personal knowledge about the chain of custody from the time the material enters the lab to the time the test is performed, and about what test was performed. This, I believe, must include testimony establishing that the particular test output is the result of the test on the relevant substance and not on another. But once this is done, any qualified expert can give an interpretation of the output.

Another issue that came up a couple of times was whether the Confrontation Clause was satisfied because the accused has a right to call the author of the report himself. Jeff answered that this argument could lead to trial by affidavits of any witness the prosecution chose, and that it ignores the passive nature of the confrontation right -- that is, the accused has a right to be confronted with the witness, to demand that the witness be presented to him. Also, it would essentially render the Confrontation Clause a nullity, given the Compulsory Process Clause. He did not address the practical reasons why the defendant's chance to call the analyst is not the equivalent of the right to cross-examine the analyst; his reply brief pointed out that I raised these practical problems in my pending petition for certiorari in Briscoe v. Virginia. If the Court decides to treat this issue outside the question presented by the Melendez-Diaz petition and instead to address it in Briscoe, I would not be unhappy. But it may be that the Court will simply hold here that the right to call a witness does not satisfy the confrontation right.

Justice Breyer in particular pushed one issue: Suppose there is a non-testimonial document, not prepared in contemplation of litigation, and the prosecution offers a certificate, prepared in contemplation of being used in prosecution, that the underlying document is authentic. Must the author of that certificate testify in court, and if not what is the distinction between that certificate and the one at issue here? Jeff responded that a certificate of this sort is not evidence as such, but is akin to a seal, which can be used as foundation for the document. In contrast, the certificate here is being used to establish an element of the crime.

I think this is at least the kernel of the answer. This is a tricky matter that could not be adequately addressed in oral argument -- and that, I believe, warrants further historical research. The United States took the view that "human assertions that merely establish the foundation for admitting nontestimonial evidence do not themselves trigger Confrontation Clause rights," but that cannot be right. Suppose a non-testimonial letter allegedly by the victim is critical in a murder case; the prosecution could not authenticate it by presenting the certificate of a handwriting analyst that the letter was in fact written by the victim. But when the prosecution wishes to produce a copy of an official document, I do not believe there is a constitutional requirement that there be evidence of proof that the document is what it appears to be; there are some documents that the law treats as self-authenticating, and this could be one. Traditionally, documents under seal were regarded as self-authenticated, and certificates of copies are a modern counterpart. They amount virtually to a form of judicial notice of authenticity. It is clear that documents under seal were not thought of in the founding era as an alternative form of testimony (as dying declarations were). They were categorically different from testimony, and they do not resemble the certificates in this case, which report on a test performed in contemplation of prosecution.