Saturday, December 17, 2011

A (mostly) bad decision from the Fourth Circuit

Several readers have pointed me to a decision made on Thursday by the U.S. Court of Appeals for the Fourth Circuit, U.S. v. Summers, 2011 WL 6276085. Here are a summary and some thoughts.

Summers was charged with drug-related crimes. The stuff was allegedly found in a black North Face jacket that he left behind in a chase. The prosecution authenticated the jacket through the testimony of several officers who testified that a particular jacket, Government Exhibit No. 1, was the one that Summers wore on the night in question. The defense introduced an internal log of the FBI lab, which purported to show the chain of custody of the jacket, but the signers of the log did not testify at trial.

There was no Confrontation Clause issue in this respect, and the court got this part of the case right. Given that the prosecution wanted to introduce the jacket as an exhibit (I'm not sure that it had to, but let's put that issue aside), its burden of authentication, as usually stated, was merely to present sufficient evidence for the trier of fact to conclude that the particular jacket was what the prosecution claimed it to be. (I actually think that the burden ought to be lower – see my article Conditional Probative Value: Neoclassicism Without Myth, 93 Mich. L. Rev. 439 (1994) – but let's put that issue aside as well.) The prosecution satisfied this burden by the direct evidence of witnesses who testified that the exhibit was the very jacket that Summers wore. (How they would remember this particular jacket among all others, and why they didn't tag it to make the identification surer – again, let's put those issues aside.) So the prosecution didn't feel any need to provide the log, perhaps because the jacket itself wasn't so important to its case. The defense introduced the log, but it couldn't pry its way into having a Confrontation Clause issue by presenting evidence on which the prosecution might have relied – had it chosen to and brought the witnesses in – but decided to forgo.

There has been loose talk in connection with Williams to the effect that there can't be a confrontation problem with anything that comes out on cross or as part of the defense case. I hope to write on that soon, because I don't think that's categorically true – if the defense flushes out that the in-court witness has testified without personal knowledge and instead has acted in effect as a conduit for an out-of-court witness who has not been subjected to confrontation, then I think there is a problem under the Confrontation Clause. But that's not what happened here – the prosecution presented the testimony that it deemed satisfactory with respect to the jacket, and that testimony was from personal knowledge. The defense could introduce the log if it wanted to, but that did not create an issue under the Clause.

As to the other part of the case, however, I think the majority of the Summers court erred pretty badly. This part involved proof that the predominant DNA found on the jacket was that of Summers. An FBI analyst, Shea, testified at trial, but he did not do any testing. One judge on the panel thought the court should not have reached the issue, because it found that any error was harmless, and I think that this is right: If, as the court concluded, it really believed that the case was lock solid without the DNA evidence, then it should not have reached the issue, especially given that Williams is pending. (The court never took note of Williams.)

The court, drawing on its precedents, said,
We perceive little difficulty with the admission of Shea’s testimony, given the predominance therein of his independent, subjective opinion and judgment relative to the lesser emphasis accorded the objective raw data generated by the analysts.
And then it cited and quoted Fed. R. Evid. 703, as if a late-20th century evidentiary rule can answer a constitutional question.

This mirrors an argument that has been made in the Williams case, and it is as wrong-headed here as it is there. Sure, Shea added his subjective opinion in evaluating the data – but the fact that the prosecution has an extra step to present (evaluation) cannot relieve its burden of proving the underlying facts in a constitutionally acceptable manner. Two critical points must be borne in mind. First, the underlying data were conveyed to Shea in a testimonial statement, a report by his colleagues. No confrontation issue arises unless this is true. Second, though Shea's testimony may not have emphasized that underlying data, his opinion was based on them; the court acknowledged that the data were "crucial" to the opinion; if the data were inaccurate, Shea had no basis for his opinion.

And in fact, though Shea might not have emphasized the data in his testimony, the report itself was introduced, with the data. This gave the majority "pause," but they overcame their doubts. Based on a prior decision in the Circuit, the majority thought that there was no problem with proving "the numerical identifiers of the DNA allele here, insofar as they are nothing more than raw data produced by a machine."

Let's be careful here. What was actually introduced, as I understand it, was an allele chart, similar to the one in the Cellmark report in Williams. Sure, such a chart reports data generated by a machine, but it in itself is produced by humans. (Moreover – though this should not be crucial – it requires judgment to produce the chart; this is not as simple as recording a number off a screen.) A machine makes a streetlight green, but a witness who observes that phenomenon and reports it for use at trial is making a testimonial statement, as Bullcoming made clear. It is no different here.

Indeed, it is somewhat remarkable that the Summers majority got to the result they did with respect to the report in the face of Melendez-Diaz and Bullcoming. And the way they did so shows just how subject to manipulation the law will be if the Supreme Court were to hold in Williams that there was no problem because the in-court expert presented her own opinion in evaluating the data. The Summers majority said,
The notarized certificates of analysis at issue in Melendez-Diaz revealed considerably more than raw data; they concluded that the substance attributed to the defendant’s possession "was found to contain: Cocaine."
And further:
Melendez-Diaz and Bullcoming each involved one or more absent expert’s "certification" with respect to the meaning of the underlying raw data, and no such certification is at issue here.
So in other words, in Melendez-Diaz, if the report had just avoided the very last word, it would have been acceptable to introduce the certificate without live testimony from anyone who observed performance of the test. I'm not sure just what "certification" in Bullcoming the Summers majority believes was fatal, but evidently there, too, the change of a few words in the report would have enabled it to be admitted without testimony of anyone who performed the test. In my amicus in Williams, I warned of precisely this manipulation: The report that gives everything but the bottom line.

The Summers court seems to have some qualms about this bad result, and it expresses some hope that if an accused makes a timely demand the prosecution will do the right thing and produce a lab witness. But if the courts don't compel the prosecution to do s, I'm not sure why a request from the defense will persuade it to.

The Summers Court takes note of but disagrees with Derr v. State, 2011 WL 4483937 (Md. Sept. 29, 2011), which is discussed in Williams's reply brief, and on which I have written a prior post, A nice decision in a Williams-like case. Derr got it right when it said:
In light of Bullcoming and Melendez, it is inescapable that the testing procedures and method employed, the DNA profile created, and the conclusion that there is a match are testimonial in nature, and therefore the analyst who performed the DNA testing or the supervisor who observed the analyst perform the DNA testing must testify in order to satisfy the Confrontation Clause, unless the witness is unavailable and the defense had a prior opportunity to cross-examine the witness.


pv said...

The Court in Williams should follow Summers (especially footnote 3) and hold:

Numerical identifiers of DNA alleles & electropherograms are not "statements" implicating the Confrontation Clause. Thus, even had the prosecution introduced any such raw data, via Lambatos' expert testimony, the Confrontation Clause was not violated.

Richard D. Friedman said...

But this isn't quite right. An electropherogram, produced by machine, is not itself a statement, though the human-input indication of what sample the electropherogram has tested is. But a chart presenting a listed set of alleles as the male profile found in a given sample most certainly is a statement; that doesn't seem to me to be a close call.

If the Supreme Court is going to follow lower courts, they should look to the Maryland Court of Appeals in Derr,, 422 Md. 211,, especially footnote 14.

pv said...

In Williams, neither the electropherogram or the allele chart (contained in the Cellmark report & reviewed by Lambatos) were introduced into evidence by the prosecution either directly (by marking them as exhibits and introducing them) or indirectly (because they were examined and relied upon by Lambatos in forming her ultimate opinion).

Indeed, because neither the electropherogram or allele chart, on-their-face, contained a human statement that they were generated from a particular source, i.e., the semen on the swab from the victim, the prosecution could have marked them as exhibits and introduced them into evidence without violating the CC.

Finally, I do agree with Prof. Friedman to this extent:

If the electropherogram and/or allele chart, in addition to being machine-generated raw data, actually contained a human statement (or, in Prof. Friedman's lingo, a "human-input indication") that the sample that resulted in the data came from a particular source, that statement would come within the ambit of the CC, assuming it was introduced for truth and was testimonial (under the "primary purpose"/"formality" test).

TASA said...

I agree.....bad decision.