« UPDATED: Jurisdiction and Venue Clarification Act | Main | When Easy Cases Make Bad Law »

Friday, December 02, 2011

Supremes to Sort Out Harmless Error (Maybe), But Does It Matter?

I'm glad to be back for another stint on Prawfs.  Thanks to Dan for inviting me.  At the risk of not being invited back, I'll start with a post on harmless error -- a terminally unsexy topic that nonetheless is much on my mind as I prepare to teach Post-Conviction Remedies next semester.

Earlier this week, the Supreme Court agreed to hear a case that offers an opportunity to clarify a longstanding ambiguity in harmless error law.  As many scholars have observed, the Court has at different times articulated the harmless error standard in two different ways, without ever clearly indicating whether the two formulations are substantively different and, if so, which one is preferred.

In the new case, Vasquez v. United States (No. 11-199), the defendant’s cert. petition focused squarely on this ambiguity, arguing that the majority opinion below (635 F.3d 889 (7th Cir. 2011)) rested on one formulation, while the dissenting opinion rested on the other.  In Vasquez’s view, the choice of harmless error standard is more-or-less dispositive in his case, thus making the case an appropriate platform for deciding which standard is the right one.  In its response, however, the government disputes that there is any substantive difference between the standards.

Here are the (allegedly) competing standards.  On the one hand, there is the standard from United States v. Chapman, 386 U.S. 16, 22 (1967): “[t]he question is whether there is a reasonable possibility that the evidence complained of might have contributed to the conviction.”  On the other, there is the “overwhelming evidence of guilt” standard from Schneble v. Florida, 405 U.S. 427, 430 (1972): “[i]n some cases the properly admitted evidence of guilt is so overwhelming, and the prejudicial effect of the [erroneously admitted evidence] is so insignificant by comparison, that it is clear beyond a reasonable doubt that the improper use of the [evidence] was harmless error.”

Admittedly, it requires some effort to see a difference between the standards, but I think the formulations do indeed suggest two distinct ways of performing the analysis, at least at a conceptual level.  The first standard invites an assessment of what actually happened at trial and what the jury’s actual views of the case were.  The second standard invites more of an independent assessment by the appellate court of the strength of the government’s case.

In practice, though, I suspect that the distinction between the two standards collapses in most cases.  That's because the appellate court will rarely have any reliable basis for determining what really mattered to a particular jury.  Juries normally render bare-bones general verdicts — a simple declaration of guilty or not guilty — and formal examination of jurors about what happened during deliberations is almost always forbidden.  In light of these limitations, an appellate panel really has no alternative in most cases but to try to imagine what it would have done if it had been the jury.  This no doubt helps to explain why the courts have not felt it necessary to resolve the longstanding ambiguity.

It is possible, though, that there are some cases in which the formulation really does matter, and Vasquez just may be such a case.  Here’s what happened.  Vasquez drove to a Denny’s parking lot to meet Perez, who was attempting to complete a purchase of a kilogram of cocaine.  It turned out, however, that the seller was working with the police, and Vasquez soon found the car surrounded by cops.  Rather than surrendering, Vasquez attempted a getaway, crashing into a couple of police cars in the process.  After a chase by car and on foot, Perez and Vasquez were apprehended and prosecuted on drug charges in federal court.  Perez pled guilty, but Vasquez went to trial.

At trial, Vasquez’s defense was that he was simply in the wrong place at the wrong time.  Perez’s wife was his key witness. She testified that she had asked Vasquez to pick up Perez at the parking lot, apparently indicating that he did it as a favor for her, not as a knowing participant in a drug deal.

In order to undermine the testimony of Perez’s wife, the government played for the jury recordings of phone conversations between Perez and his wife while Perez was in jail following his arrest.  Through those recordings, the jury learned that Vasquez’s lawyer had advised his client that he would likely lose at trial and should plead guilty.  When Vasquez’s lawyer then stood up to deliver his closing argument a little while later, one imagines that he had very little credibility left with the jury.

On appeal to the Seventh Circuit, all three members of the panel agreed that the trial judge had committed an error in admitting the sensitive portions of the recordings, at least without an appropriate limiting instruction for the jury.  However, the majority concluded that the error was harmless, reasoning that the evidence of Vasquez’s guilt was “overwhelming.”  The majority particularly emphasized the importance of Vasquez’s getaway attempt as evidence of guilt.  The majority also cited some other items of evidence, including testimony indicating that Vasquez knew the car contained $23,000 in cash.

For the majority, the harmless error determination was quick and easy, requiring only a brief recounting of the evidence against Vasquez, with no attention to the question of what the jury actually thought about the evidence.  The dissenting judge provided much more extensive analysis, carefully parsing the trial record to determine what impact the evidence actually had on the jury.  Here’s a taste:

The [Chapman] standard is not easy to satisfy, and four factors here lead me to conclude this error was not harmless: the modest strength of the rest of the government’s case against Vasquez, the prejudicial character of the evidence that was admitted erroneously, the fact that the jury acquitted Vasquez of one of two charges, and the importance that the government itself attributed to its flawed rebuttal evidence.

. . .

Without the flight evidence and the MCC tapes erroneously admitted for their truth, then, the government had evidence that was legally sufficient to convict Vasquez, but the case was far from a slam-dunk. The dramatic evidence of the dangerous flight strengthened the case substantially and makes it easier for my colleagues to describe the district court’s error as harmless. But the flight evidence cannot carry that much weight, in my view. The Supreme Court and we have repeatedly cautioned against too much reliance on flight as evidence of guilt for the crime charged because there are so many links in the chain of inferences . . . .

We must also consider the prejudicial effect of the improper evidence. The evidence from the MCC tapes, admitted here erroneously for their truth and with no true probative value, was just about as prejudicial as one could expect to encounter in a trial. The jury heard that Vasquez’s lawyer — the man who would soon make a closing argument asking them to find reasonable doubt in the government’s case — had told Vasquez that he should plead guilty and had said that if he and his codefendants went to trial, “everyone is going to lose.” A juror who heard and believed that evidence would surely discount anything she heard from that lawyer. In terms of prejudice, these harpoons are comparable to evidence of a defendant’s own withdrawn guilty plea. Such a plea is virtually never admissible because of its powerful force.

We also have strong indications from both the jury and the government itself that the erroneous admission of the MCC tapes was not harmless. Even with the prejudicial and erroneous evidence, the jury still found Vasquez not guilty on the charge of attempted possession with intent to distribute. That verdict is hard to reconcile with the jury’s conviction on the conspiracy charge, and the split verdict certainly has the whiff of a compromise verdict in a close case. Such verdicts are permissible in criminal cases, of course, but when determining whether, beyond a reasonable doubt, a conceded error was harmless, we should not ignore that strong signal that the jury viewed the case as a close one, even with the evidence of flight and the improper rebuttal evidence.

The government also showed how important it believed the improper rebuttal evidence was by its extraordinary efforts to obtain its admission. The trial seemed nearly over when the government filed its emergency Sunday motion for a continuance to enable it to prepare this rebuttal case. The events of the next several days, including especially the government’s emphasis on the improper evidence in its closing argument, showed that the government believed that Mrs. Perez had seriously weakened its case and that the improper rebuttal evidence strengthened its case considerably.

In light of the considerations highlighted by the dissenting judge, I think Vasquez does have a plausible argument that the choice of which harmless error test to use was critical in his case.

On the other hand, it’s not entirely clear to me that Vasquez must necessarily lose under the “overwhelming evidence” test.  It really comes down to how much weight you want to give the flight evidence: does this really demonstrate in an “overwhelming” way that Vasquez was conspiring to deal drugs?  I’m skeptical, although I’d really like to know more about the circumstances of the flight.  Minority men (based on his surname, I’m assuming that Vasquez is Hispanic) have plenty of reason to fear encounters with the police even if they are not doing anything illegal.  Although trying to bull his way through a cordon of police cars was stupid and dangerous on many levels, Vasquez seems to have made a snap decision in a high-stress setting that may not really say much at all about what he was thinking when he drove to the parking lot.

So, the Court may be able to dodge sorting out the harmless error standard yet again.

If it does resolve the ambiguity, which way will it go?  Given the restrictions on finding out what a jury was actually thinking, I’m dubious that the sort of approach exemplified by the dissenting judge in Vasquez can really work on its own terms.  Applying the test will almost always seem a highly speculative exercise.  And it may be one that actually works against defendants in some cases.  For instance, if a jury convicts on all counts with only a very short period of deliberation, that would seem rather compelling evidence that the jury did not find the case at all close and would help to support a finding of harmlessness.  (This would be the flipside of the compromise verdict in Vasquez.)

Our normal approach is to handle jury verdicts in a highly formalistic way: a guilty verdict is a guilty verdict is a guilty verdict, and we rarely recognize the possibility that deliberations may fall short of the rational, conscientious ideal.  (The prejudice test from Strickland is a good example, expressly removing from consideration the possibility of an idiosyncratic jury.)  The “overwhelming evidence” test seems most consistent with this jurisprudential tradition; it avoids any exploration of the “sausage-making” that takes place in the jury room, and treats juries as so uniform and predictable in their operation that an appellate panel can readily stand in for a jury without any need to consider evidence of the inclinations of the actual jury.  To be sure, though, as Vasquez points out, it is not easy to reconcile this approach with the Sixth Amendment guarantee of a jury trial.

Cross posted at Life Sentences.

Posted by Michael O'Hear on December 2, 2011 at 05:49 PM in Criminal Law | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Supremes to Sort Out Harmless Error (Maybe), But Does It Matter?:


Doesn't Sullivan v. Louisiana, 508 U.S. 275, 279 (1993), provide some clarity on this question?

"Consistent with the jury-trial guarantee, the question [Chapman] instructs the reviewing court to consider is not what effect the constitutional error might generally be expected to have upon a reasonable jury, but rather what effect it had upon the guilty verdict in the case at hand. See Chapman, supra, at 24 (analyzing effect of error on 'verdict obtained'). Harmless-error review looks, we have said, to the basis on which 'the jury actually rested its verdict.' Yates v. Evatt, 500 U. S. 391, 404 (1991) (emphasis added). The inquiry, in other words, is not whether, in a trial that occurred without the error, a guilty verdict would surely have been rendered, but whether the guilty verdict actually rendered in this trial was surely unattributable to the error. That must be so, because to hypothesize a guilty verdict that was never in fact rendered—no matter how inescapable the findings to support that verdict might be—would violate the jury-trial guarantee."

Posted by: Patrick | Dec 14, 2011 12:32:54 AM

I can see kind of a difference, but after thinking about it more I think the effect comes from rhetorical framing rather than substance. You call the Schneble standard the "overwhelming evidence of guilt" standard. But it could just be as accurately called the "insignificant prejudicial effect" standard. And the two framings differ. If one were asked whether the evidence of guilt was overwhelming, one might readily agree with the majority. If one were asked whether the tape was "insignificantly prejudicial," one would be inclined to agree with the dissent. In that sense I am seeing less of a conflict across the cases than a difference of rhetorical emphasis.

Posted by: TJ | Dec 3, 2011 3:47:38 AM

The comments to this entry are closed.