Oral argument — United States v. Castro-Caicedo

United States v. Gustavo Castro-Caicedo Mr. Wood, good morning. Good morning. May it please the Court, Chauncey Wood for the Appellant Gustavo Castro-Caicedo. I'd like to reserve two minutes for rebuttal, if I may. You may. The Appellant asks this Court to incorporate the knowledge gained from 30 years of scientific research into its analysis of the reliability of eyewitness identification. Scientists have now reached a broad consensus about the numerous factors that can affect eyewitness memory and lead to mistaken identifications. This consensus was reflected just last week in a publication from the National Academy of Sciences entitled Assessing Eyewitness Identification. The Second Circuit and several State Supreme Courts have already incorporated the consensus. It's also consistent with Supreme Court precedent. Most importantly, it will minimize the impact of mistaken identifications generally regarded as the single greatest cause of wrongful convictions in the American criminal justice system. What's the effect on this case that no argument similar to this was made before the District Court? I think that it's, in fact, properly reviewed under the Court's plenary. The Court's review is plenary. First of all, the government concedes that a District Court's denial of motion to suppress identification evidence is reviewed de novo. The government also concedes that Castro-Caicedo argued below that the identification procedure was impermissibly suggestive and the identification, more specifically, was not reliable. More specifically than that, the Defense Counsel, in his motion to suppress, argued that the array was impermissibly suggestive because the defendant was the only older person in the array. But that's not even an issue on appeal. The District Court found that it was impermissibly suggestive. What you're attacking is the other part of the District Court's ruling, the Court's reliability finding. to upset the District Court's reliability finding based on an argument that was never made or even hinted at in the District Court, and I'm troubled by that framework. Yes, Your Honor makes an important point. I think initially I would say that counsel did, in fact, challenge the reliability finding directly. But you never challenged it on the ground that you're challenging it on appeal. There was no mention of scientific studies, no mention of anything of the sort. That's right. So I think that the place to start is the fact that at least three, as I've laid out in my briefs, at least three courts have not merely cited but taken judicial notice as legislative fact of the science of memory. Actually, you know, the thought that Mark Wolfe was totally ignorant of the scientific debate about eyewitness testimony is belied by the fact that we extensively discussed such evidence in the Jones case, where the issue came up in terms of whether you should have expert testimony on this or whether a judge's instructions would be fine. This was in the air for a long time, even before our Jones decision, and the specific argument is not made to the district judge. But the notion that he's ignorant of all of this, they do read our opinions. They may regret it, but they read our opinions. And so this, and despite that, months later when this case goes to trial, there's nary a mention of any of this. So it's worse than there's no mention. One would think that the district judge had all of this in mind when he made the reliability finding, because he could read the case law. Well, Your Honor, I would begin with the obvious, that the judge made no mention of that. Why would he? Well, I think that it's crucial for an appropriate analysis of the reliability factors under Neal v. Biggers to specifically incorporate the 30 years of scientific research into one's analysis. There's absolutely no evidence in the record from Judge Wolf that he incorporated those findings. And I think that what I'm really asking, I think the appropriate model here is Young v. Conway, Second Circuit case, a federal habeas case arising out of a state court conviction. The Second Circuit in that case made clear that they were relying on this 30, they incorporated this 30 years of scientific research, which has achieved a broad consensus, into their analysis of the state court's review of the Neal v. Biggers factors. And they made clear that this scientific research was not presented to the state court below. And of course, you've got the issues of comedy in a federal review of a state court case, which you don't have here in a direct review of a federal case. And so you've got, I think, procedurally an even stronger case for the defense here. But the point is, the judge made no mention of this in the state court. And your evidence that the judge did not consider any of this is merely that he didn't mention it after defense counsel didn't even raise the topic. Well, at the end of the day, I don't actually think it matters to my argument whether the district court judge considered it or didn't consider it. Let's, for the sake of argument, agree, I admit that I don't know whether he considered it, and I agree that Judge Wolf is a judge of great experience, and so it's a fair inference that he was aware of this research. I think the ultimate issue is how this court is going to approach its review of his decision on the reliability of what he admits was an impermissibly suggestive identification. Okay, so let's get down to brass tacks. Taking this scientific knowledge, which of his findings do you think are vulnerable? I think that virtually all of them are vulnerable. I would go through them very quickly. The first being opportunity. I guess I need to preface that by pointing out that there's a fundamental flaw, which Judge Wolf didn't acknowledge, in the Neil B. Biggers analysis, which scholars and all the courts that have looked at this have acknowledged, that there's this problem of self-reporting. Three of the five Biggers factors, opportunity to view, attention paid, and certainty, all require self-reporting by the eyewitness. And a suggestive ID procedure can contaminate a witness's memory of the events, increasing the danger that these self-reported factors will be enhanced by the defendant. So the New Jersey Supreme Court described this this way. The irony of the current test is that the more suggestive the procedure, the greater the chance eyewitnesses will seem confident and report better viewing conditions. Courts, in turn, are encouraged to admit identifications based on criteria that have been tainted by the very suggestive practice the test aims to deter. So the point here is that if the court doesn't acknowledge that three of these five Biggers factors are a product of self-reporting, and doesn't incorporate the suggestiveness, the impermissible suggestiveness, which this district court admitted existed, then there's a danger that the impermissible suggestiveness will, in turn, create the reliability of the self-reported factors. I'm just trying to figure out what we're supposed to do with this point. So if you go back to the Second Circuit case you talked about, was that a case in which the identification was after a substantial claimed meeting with the person who he was identifying? Or was that just an eyewitness to a pleading event by a stranger? The facts of the case are not, I don't immediately remember them. The issue there was slightly different. It wasn't identification. It was whether there was an independent source for the witnesses. Let's say that, in general, the eyewitness evidence has some reliability issues because of the scientific data. And maybe that should factor into any case in which there's a stranger identification. But what are we supposed to do with a case in which they say they met for an hour and a half two times? We just say that's inherently unreliable? Well, OK, that's the first point. I'll go through these quickly. Here's how I would go through that. Just that point. Could you just answer that? Sure. The opportunity to view. That's the first Biggers factor. The court characterized this as they met for, well, first of all, Rosette said that they had two meetings. The first was less than an hour. The second was less than 30 minutes. So this was characterized as an hour and a half. In fact, what you have is a self-report from Rosette following an impermissibly suggestive identification procedure and thus prone to exaggeration. He gave this self-report four and a half years later after failing to mention So you would say because of the skepticism that science has raised about eyewitness testimony, that a person who says I met with someone for about an hour in total, we have to just treat that as inherently unreliable? I would say that that's one factor. And the scientific research would say that, this is not me, the scientific research would say that the problem with that by itself is that it's Suppose they had been together for three days. Would we have to disregard that? I think that these are, it's a totality of the circumstances. And did the scientific evidence tell us at what point we're supposed to decide? No. Well, then what are we supposed to do with it on appeal? What you're supposed to do with it on appeal, I think, is to look at the totality of the circumstances. So in this case, that is certainly a factor you can consider, his estimate, but you have to recognize that it's a self-report prone to exaggeration. Also, I guess just one last sentence. Suppose I recognize all that. What am I supposed to do after I have now recognized it? Are you saying that it's inherently unreliable? If you're not, I don't quite understand what we're supposed to do on appeal. It's not a, I think what your Honor is getting at is, is this a per se test? I'm not, it's a totality of the circumstances review. And the court's review of the motion to suppress is plenary. The review of the factual findings is for clear error. So the district court, I think, said they met for about an hour and a half. So this specific factual finding is, I'm not saying that that is clear error, but that's not the dispositive point. The dispositive point is, ultimately, was this identification, notwithstanding it was impermissibly suggestive, based on the five factors in the LV Biggers and the totality of the circumstances, sufficiently unreliable that it should have been excluded. And I would argue that if you incorporate the scientific research here and you properly analyze the Biggers factors in light of the impermissibly suggestive identification procedure that we all concede happened, that you would conclude that it was insufficiently reliable to present it to the jury, as under all of the circumstances, that that one factor that he met with him for, according to his own self-report, after an impermissibly suggestive identification procedure, after he had been given immunity, was not sufficient to establish conclusively that this is sufficiently reliable to go to a jury. No, but Judge Wolf resolved four of the five factors contrary to your client. And which of the five factors, which of those four factors, did he resolve in a manner that was clearly erroneous and thus could be disturbed on appeal? The first, the most glaring is the certainty. He said that the witness, he stated once, and this is a quote, transcript, volume one, page 295, once he saw that photograph, Rosette was quickly certain that that was the owner. In fact, there was no evidence of certainty. In fact, DEA agent Grella merely testified, quote, there was no hesitation in his ID. And you don't think quickly certain is a fair paraphrase of there was no hesitation in his ID? You think that's clear error to paraphrase it in that manner? I think that there was no hesitation is clearly a different concept than expressing a level of certainty. That's clearly different. And more importantly, scientific research, and this is the amicus brief, page 16, citing authority, demonstrates that rapid identifications, which are different than expressions of certainty, are not good indicators of accuracy. But more important than that, even if we say that this is a synonym for certainty, that this court has actually spoken most clearly on that factor that the witness is, I'm quoting Jones here, the witness's lack of confidence is certainly a reliable warning sign, while the presence of confidence is probably closer to a neutral factor. In other words, while a lack of certainty undermines reliability, expressions of certainty do not support reliability. That doesn't really answer my question as to which of these four factors that he resolved against you, can we say he resolved in a manner that was clear error? I think that I, respectfully, I think that just answered, from my perspective. It didn't answer it from mine, but what about any of the other three? Sure. I believe that all of them, again, witnesses, quickly, I'm out of time, but I believe all of them, and I've briefed them in my papers. Your Honors, may it please the Court, Randall Crum on behalf of the government. I think I'll start where Judge Saliu finished up. I think where we come down on this is, in the end, the questions of standard of review don't matter as much as the fact that, in the end, as Judge Barron also asked, it doesn't really make any difference on the factual record that we have that we're all bound to the decisions the district court made with respect to individual factors were clearly supportable. And what you have left over is generally questions that the defense feels could have been asked. Was he under stress? Was he intoxicated? Did he overestimate time periods? But nothing that actually alters the record we do have, which when he said when he was offered a beer, he took a coffee, that he didn't appear to be under stress, and that he met with the man for a certain amount of time. And based on those factors, even including the social science research, doesn't really alter the conclusion that those factors, four out of five, weigh in favor of reliability, and the fifth is not so overwhelming as to outweigh those four. And so, in the end, I think our view is while we argue, obviously, standard of review, and we think it's important to note that this Court's not in a position, as the amicus appears to suggest, to simply read certain bigger factors out by suggesting they're so unreliable that they shouldn't be considered. We're stuck with that test as well, but applying even a certain amount of skepticism with respect to the factors, the factual record we have supports a finding of reliability, and that that shouldn't be disturbed on appeal. You know, the thing that troubles me about the case is perhaps less the issues that have been argued than the fact that the government should have known that its photo array lineup was impermissibly suggestive, and it went ahead anyway. The mere fact that most of the lineup happens to be photos of the conspirators, it's not a big deal to get photographs emailed to you if you need additional photographs. I understand the circumstances under which they finally do the array, but this notion of the government agents almost willfully being blind to what they're doing is very disturbing. Well, I would respond. I think I don't – I understand that the way the array was set up of including multiple conspirators is something that this Court and other courts have raised questions about, although I don't know that it's been viewed as automatically disqualifying. But in this circumstance, I think the government took quite a number of steps to try to inoculate the procedure, again, as you know, under circumstances in which they weren't anticipating doing a photo array until the man gave such a detailed description of the person prior to the photos even being brought up that they felt that, well, he might just be able to identify this person that we have not had anybody able to identify yet. So they put up a series of pictures. But they were clear in stating, and I think they did what they could under the circumstances to say, you might not recognize anybody. It's okay if you don't. There might be one or more than one. That made this array permissive? No, Your Honor. Again, I'm not suggesting that this was an ideal array by any means, but simply that the circumstances were not all on the side of suggesting that there was no attempt to render it less suggestive. I think they made this, again, the pictures were the pictures they had. I understand Your Honor's point that it could have been delayed and done differently. But there was an effort made, and that's why the government did argue below that it wasn't impermissibly suggestive, although, again, the inclusion of multiple co-conspirators was a factor against it. Okay. His basic claim is this should never have gone to the jury. The defendant had another shot at convincing the jury that this was not a good identification. As a factual matter, at the trial, did the defense try to persuade the jury that this was not a good identification? It was unreliable for all of the reasons, the scientific research. No, Your Honor, not with respect to the scientific research. The witness was definitely cross-examined on his credibility and his opportunity, what he was doing, and that kind of thing, but not with respect to the scientific research. So although there was an opportunity potentially to bring that up, it was not re-raised in the cross-examination of this witness. Okay. Now, with respect to the other arguments on the case, we've briefed them. Does the Court have any questions with respect to them? Could you just speak a little bit about, I think you make a harmless error argument in your brief, or do you not? No, Your Honor, we don't make the argument that it would be harmless beyond a reasonable doubt of the preserved standard or that if he had not been allowed to testify at all. But we do make an argument and a footnote that because there's a suggestion that the government would have had insufficient evidence without this testimony, we seek to rebut that by simply saying there was a large circumstantial case. And, in fact, that was the case the government was prepared to go to trial on until this man proved able to identify Kessler-Caicedo. But we're not arguing that it would be harmless if it had never come in at all. Your Honor? Thank you. Thank you very much, Your Honor. I'd like to use my two minutes just to focus on Judge Celia's question about the factors and focus on just one of them, one of the vigorous factors, accuracy of the description. And there was this claim that he gave such a – the reason they had this idea was he gave such a detailed description. We would dispute that, and it's crucial, I think, to the third vigorous factor, the accuracy of the prior description. Rosette's description of the owner, an older man of dark complexion and medium height with a limp, was not helpful. The ID procedure was impermissibly suggested because, one, it depicted only five men with dark complexions. At least four of these were linked to the conspiracy, and the defendant was the only one who was older. The bottom line here is that the defendant was the only person in the array who fit Rosette's general description. In other words, there's a danger that Rosette picked the defendant's photo because the defendant was the most similar to the description he gave. So this is – and this is the point that academics and other courts have repeatedly emphasized with the vigorous factors, that when Rosette's description of the owner is, quote, weighed against the corrupting effect of the suggestive identification itself – that's Manson – the description added nothing to the reliability of the identification. On top of that, there's the cross-racial aspect of the identification, which, notwithstanding the government's claim that there was nothing in the record to support this, I submit clearly was established by the record because Mr. Rosette, who's Filipino, testified in the hearing in front of Judge Wolf, and it's clear that Mr. Castro-Caicedo was of African descent. So this clearly was a cross-racial identification. Judge Wolf made no mention of the fact that it was a cross-racial identification, which undermines reliability according to the academic research. And so I think that this illustrates that when you apply the scientific research here and you recognize the danger of the flaw in the vigorous test that these reports – the flaw in the vigorous test is that you need to recognize the problem of suggestiveness, and you have to weigh that against the evidence of reliability. Just out of curiosity, going back to Judge Barron's comments about – this is an identification of a business partner. This is not an eyewitness who was on the street corner when he saw the two gangs mix it up and the cops come waiting in. This was not a momentary, unexpected encounter. Does the scientific research sort through those sorts of issues? The scientific research says on the opportunity for a view, there's no definitive answer. No, that doesn't entirely capture the difference I'm asking about. It's not opportunity to view. It's sort of an intentionality of the interaction. I think – well, I guess another issue – I think what you're getting at is it's not as stressful as a robber pointing a gun at somebody in a surprise situation. No, that's not quite what you – how would you do a study that would get at whether people tend to misidentify their business partners? I can understand how you do randomized experimental studies that generate the research when you just say somebody walks into a room, they leave the room. Then you describe the person. People misdescribe it. But this is different. This is trying to figure out somebody claims I worked with this person for a substantial enterprise and we had two meetings for an hour and a half. Are there studies that show in that context people tend to misremember? I am not aware of a study that – So if you're not, how can we then say that there's scientific evidence that goes to the reliability in this case? There are scientific studies on – I'm trying to answer the question. There's – I'm not aware of a study that deals with business partners. I would dispute the premise of the question that they were business partners. They had two meetings. And the important points that I would emphasize is there's definitely studies on the length of delay between the meeting and the formal identification. Four and a half years is absolutely a significant factor in academic research. Also, the notion that this was not a stressful encounter because it was business partners, I vigorously dispute. This is a guy who for the first time ever is meeting with Colombian drug dealers in their house. And he's agreeing to, for the first time in his life, as far as we know, move large amounts of cocaine for them on a ship. That is, we argue, a very stressful encounter. But it might lead you to take a pretty good look at him. Academic research, that's an important point. The academic research, that's a common argument that prosecutors have made for years. In a stressful situation, you focus on this person and it makes it more reliable. That's exactly why the research is so important. But in fact, research demonstrates that the more stressful the situation, the less reliable the identification becomes. That's precisely why, among many reasons, one among many reasons, it's crucial that we incorporate scientific research and it's counterintuitive. That's the most important point. It's counterintuitive. So I would submit that unless we, this situation where it's not merely a meeting with your business partner for an hour and a half, it's rather a very stressful encounter with a stranger and an identification four and a half years after the fact. The research does speak to both of those factors. Thank you. Thank you, Mr. Wood.