Oral argument — United States v. Haggerty

Next case is number 23-2084 United States v. Robert Haggerty. Ms. Stern. Good morning and may it please the court. Samantha Stern on behalf of Robert Haggerty. I'd like to request three minutes for rebuttal. Granted. This appeal presents the question of their commentary to the number of images enhancement in Guideline 2G2.2 is entitled to deference after Kaiser v. Wilkie. Starting with the text, in common usage, the term image encompasses still and moving images. Therein lies the mystery of this case, to me anyway, and that is how does ordinary usage and plain text require us, compel us, to interpret the image, whether it's a still photograph or whether it's video as one image, as you argue. Let me just put my cards on the table. If it isn't an oxymoron, it comes awfully close to one, too. So perhaps you can enlighten me. As this court recognized in Mercado and Adair, in conducting Kaiser step one, you arrive at contextually appropriate definitions considering ordinary meaning. No rabbit in the hat. I understand all of that. Well, I think what I'm asking you and why I use the word oxymoron is how can one have, on the one hand, an image with what that not simply connotes, but what it means as a matter of textualism, with a moving video or a moving image, more to the point? Well, we know images do move. That's what we see when we watch the videos. We know frames move. Do we know that an image moves? I mean, I'm not sure I can agree. The frames don't move. The frame is a still image. When spliced together, we see a moving image, and we see it as continuous moving image. So that's the ordinary meaning that I think is consistent with this text. But appreciating Your Honor's question, yes, there is a place where this use of the term image somewhat departs from ordinary meaning. And the answer is it's in a particular statutory context here, a statutory That statutory context strongly indicates that Congress wanted to punish videos more harshly than stills. Did it not? I disagree. There's no indication. There's no indication that Congress adjusts videos to be more harmful. You could certainly intuit that, but there's no indication from this text. Not only the text, we look at the structure. How can you see the structure, which has gradations based on amounts, as seeing anything other than a clear congressional intent to punish quantities imperatively, increasing. I agree with that. But the sort of assumption that Congress is measuring quantity by parsing videos assumes the very interpretive question that this court needs to decide. They didn't parse. That's why they use the I look at them as plug numbers. They're just invented out of thin air. They're estimations. Right. And that's why you've got a strong argument under Kaiser. Right. It doesn't help you necessarily. It doesn't get you to the finish line to win the case because we still interpret text, even if we don't give any deference to the statutory context. I think what's really helpful for this court's analysis is reference to 2256 5 and 2256 8. Those are the provisions that define visual depiction in this context. Visual depiction means photograph, film, video. That's right. Why does that help us in any way deal with the question we have here, which is one of quantification and the words that are used to support that quantification? Because it supports the conclusion that when Congress used the term image here, it was actually referring for the most part to visual depiction. And one video equals one visual depiction equals one image in this all videos. Yes. All right. So what is a term of art? What about a what about a video file? One video file that contains multiple scenes with different perpetrators and different victims filled at different times that your argument is that is one image. It's one depiction. It's one image for purposes of counting. Yes, I think it's compelled by 2256 5 and which I'll note it's it's somewhat odd. This is the part of the depiction definition that defines it to include things like undeveloped. So now now you're asking us to write an opinion that would result in similar punishment or a similar guidelines range for perpetrator who has one photograph of child sexual abuse material and a different perpetrator who has two hours of child sexual abuse material with 100 victims and 100 perpetrators. That's what you're asking us to do. That reading is compelled by this text. And I mean, how can you argue that that's reasonable? That's not the question. I mean, that's not exactly the question for this court. The question is, what's the best reading of this text? And I mean, there's not going to be a parade of horribles. You want us to take an actual approach to this is what you're arguing. Absolutely. You want us to look at this and say that one video is one image. Okay, if we do that, don't we end up in the position where Judge Larson was in her concurrence, Mary Phillips case is that one video isn't one image. It's a collection of frames and the frames within that video are so numerous that they're way above the 600 number to require the five point enhancement. Is that the risk you're running by arguing this textual approach? I don't think so. I mean, I think this court has to analyze the text, but the court has to keep in mind the statutory context, the structure of the guideline and the purpose. And that's where Judge Larson's approach and the frames approach go off the D-band. Wasn't Congress purpose, at least in huge measure, a concern about the increased harm that came from how many depictions there were aggregated in a particular offense or in a series of offenses that gave rise to increased harm? I don't disagree. One way to adjudge how many depictions are in a case, the best way to adjudge that is by counting the number of photos, videos, films. If we want to address the additional harm of video... If we're going to look at the text, we have to count the number of images, right? That's right. And I would follow up Judge Fischer's question to simply be careful what you wish for, because if we were to interpret, as you suggest, this particular guideline, it would probably be greeted by Congress with such disapproval that they would soon amend the guideline, correct our mistake, and then you might not have that 600 cap any longer. You might be looking at an increased amount. We're stuck with the text and the structure and the purpose, their purpose. I think we need to be careful how much we read into the purpose. There's one line about the purpose here, and it's to increase punishment based on the amount of child pornography involved in an offense. The only term that Congress used to define that was image. And image does not, certainly unambiguously, refer to frames in a video. And the reason why is because even if that's somehow a reasonable reading of the term image in ordinary usage, it doesn't comport with the incremental structure of this guideline. You would agree, believe it or not, that such an authority, and I think it's an authority, is the Motion Picture Association, which had input during the notice and comment period, disagrees with you and defined video as 24 frames per second and also said each frame is equivalent to one still image. That was their statement during the notice and comment period. If anybody has insight and expertise into this area, it would be the Motion Picture Association of America. I think that was actually referenced in the Department of Justice's letter to the Sentencing Commission. Whether it was or was not, did they not provide that to the comments about the standard 24 frame rate for films? I'm not sure where that gets us. And nobody's taken, no, that's not actually what I sought emphasis to. I'm sorry if I misled you. They, yes, did point to the 24 images per second and went on to say that each frame is equivalent to one still image. That was the quotation that I saw. The Department of Justice letter to which you're referring, which references the frame rates, is interesting because it actually in some ways supports my view that image here, when Congress used the term image, it meant visual depiction as defined in statute. And so it does make sense to count photos, videos, and not parse videos further. But the oddity of that Department of Justice letter was that it then went on to urge the commission to adopt some rule that would penalize videos more because of the greater harm of videos and films, which sure, we can all intuit and understand. The key here is that if the commission wants to address that harm, which Congress didn't even speak to wanting to address, it needs to do so in the text and not the commentary. So far from a parade of horribles that will ensue, I mean, even under this court's interpretation, if you agree with me about the term image here being a term of art, a district courts, even with the law as it would be in this circuit, are not without tools to consider the additional harm of videos as the district court noted in this case, you know, he found there to be sort of a greater harm. Judges can consider that they could depart upward from the guidelines, they can consider it in the 3553 a mix. They could depart down. They can depart down and in this area, they routinely do due to the major flaws of this guideline where Congress really took the unprecedented step, telling the commission what to do in this area. Maybe, maybe I'm misrecollecting this recalling wasn't there a downward variance or defied pressure in this case, there was, but not on the basis of disagreement with to understand. I understand I'm over time now. So I'm happy to save my time for a bottle. Thank you very much. We'll hear you on your bottle. Mr. Hello. Good morning, your honors and may it please the court. Adam hello for the United States under section 2g 2.2 of the guidelines, a video ambiguously contains more than one image. That is enough to affirm the five level enhancement in Hagerty's case and this court could stop there. If it goes further, this court should defer to the guideline commentary here because based on section 2g 2.2 is text structure history and purpose. The guideline is genuinely ambiguous as to what the number of images in a video is the number of frames or some smaller number, because Hagerty qualified for the enhancement and because there was no abuse of discretion. Arguing that it's not ambiguous and that it's ambiguous, which is I'm arguing that the zone of ambiguity, your honor, includes the interpretation that the commission reached in commentary, but does not include a one to one reading that Hagerty proposes between a video and an image. And that's based on the text, primarily because the guideline does refer to images. And as this court has said in Heatherly, a video is a set of images. That's how many images is how many images in this case, your honor, the commission reasonably determined 75 images. That's that for us to defer to the commission. If it's not ambiguous, how many images is it? It's ambiguous between a larger number than one image. You just said you said a minute ago, a video is a series of images, correct? How do we determine how many images any given video has? Yes, your honor, you would look you do what the commission did here. And you'd look and invent 75 or 150, 322. I'm asking you a 10 second video. If we were put to the test of answering the question, how many images are in this 10 second video? How would we go about doing? Well, you do that for the purpose of this guideline. How would we go about doing that as a factual matter? I'm not asking you about the guideline. Yes. As a factual matter, you'd look to you'd understand that as the Sixth Circuit said in a human, I can't perceive every frame. And so you might think that this guideline referred to less than the number of frames. So the district court would need to make a record, right? If it was, if it might require expert testimony, if it was, if it was interpreting outside of, if it was trying to determine on its own, how many images were in a particular video, are we talking about whether or not there is a discernible difference discernible to the human eye between frame number one and frame number 24 of those 24 that seems most people or there is a consensus that there are that many frames in one second? Yes. Yes, you're. So assuming the, the typical 24 frame per second rate, there may, there may be a very imperceptible difference between frame one and frame two. So doesn't that lead us then to the question at what point does that discernible difference render frame number one and frame number 25 images? Correct, Your Honor. It's a, it's a very, it's a very specific inquiry. It depends on the facts of actual. Isn't that the inquiry we end up with as implied by Judge Hardiman's question. It's the inquiry that every district court would end up with if the commission hadn't offered its own data-driven expertise to look at the range of pardon me for this, but directly go to the position taken by your friend on the other side. And then that is to see if we can find some basis for agreement on a predicate here. That is whether a video is a set of images. What force should we give to the fact that this court has already said in Heatherly that a video is just a set of images. I think that resolves whether Haggerty's reading... It's not, it wasn't a holding as I understand it. We may call it a dictum in that case, but we said it, we said it flatly. Yes. I think it was essential to the outcome in Heatherly. I mean, Heatherly was, Heatherly was talking about whether a defendant's, about videos necessarily meant a prior recording or a live. Exactly. Yes. And citing the Oxford English Dictionary, I might add. That's, that's evidence of the text. And it's also evidence of the point that Judge Fisher made that these videos on any sense are something that Congress and the commission would want to be punished more harshly than a single still image. They show motion, they show action, they can go sound. Some of them, some of them can be longer or shorter. These are clearly within Congress's intent of punishing the amount of child pornography, clearly something that should be punished more harshly than a single still image. And if, and if in fact we concluded on the textual analysis that an image that applied to video meant frames, that would mean that Congress made a choice that any video would get the fine five points because any video of any length is going to be over 600 frames. Any video of more than 24 seconds typically would be 600. Yes, would be 600. But why, why would we, why would we count every image in the video? Wouldn't we only count images of, of child sexual abuse material? You would, your honor. Yes. So if it's, if it's a, if it's a, I don't expect this is a realistic hypothetical, but if it were a 20 minute video and only four seconds of the 20 minute video involved child sexual abuse material, it would be four times 24. Correct, your honor. And we'd, we'd say that limitation comes from the guidelines saying the images have to be involved in the effect. All right. And in this case, if, if we were to, if we were to decide that image meant frame following judge Larson's concurrence in the Phillips case, remand would be required in this case, would it not? Because there has not been a finding of how many seconds or minutes of the video included child sexual abuse material. Oh, we disagree, your honor. First of all, that's not an argument that Hagerty ever made in the district court. Well, they don't, they don't have to make, they don't, they don't have to make that. Your honor, we made representations that there were over 14,000 explicit, there were over 14,000 frames of the videos. Right. But I have no idea how many of those 14,000 frames involved CSAM. Well, as your honor, as you're going to point it out in the typical case, you wouldn't expect you, you would expect a large number. You want us to write an opinion, affirming a sentence because we're going to assume that Mr. Hagerty is like typical consumer of CSAM. That's like quite remarkable. The evidence that suggests your videos that he had was that he had, he qualified for the 600. Well, that suggests that you probably went on remand, but I don't take that as any sort of persuasive argument that remand is required. Well, there was no, there was no response from Mr. Hagerty that, uh, that would require the district court to look at this evidence. It would require. Was the district court considering adopting judge Larson's view? It did, your honor. It said in its alternative holding. That wasn't the primary. That wasn't the primary holding, but it wasn't alternative holding. And that's enough for this court to affirm on that basis. If it reaches a judge Larson's concurrence, which is, which is again, a reasonable reading of the guideline. We're not saying that it's not, it's within the zone of ambiguity. Um, and again, that reflects Congress's purpose in punishing that. So you started out your argument saying that, uh, one, the, it was the, uh, it was, unarguably unambiguous. It was ambiguous that we have deference to the sentencing commission's commentary. Did you not? We do. Yes, your honor. And we, and just to, just to clarify, we argue it's unambiguous, unquestionably, uh, rejecting aggregates interpretation. Okay. So, and you have two alternative arguments that how the commission could have determined, uh, that the 600, uh, images number, uh, was what entitled somebody to get five points. Uh, do we do, do we do a Kaiser analysis, uh, or could we possibly do a Skidmore analysis? Well, your honor, we think we would prevail under either mode of analysis, but we think a Kaiser analysis is better because this is an instance where the commission is interpreting its own guideline. It's interpreting its own regulation. Um, the Kaiser rules apply, you know, if, as we argue, there is ambiguity between the every frame reading and some lower number of frames, then you would look at the, you would look at whether this is a reasonable reading from the guideline and whether it relies on the commission's data-driven expertise. And it does in this instance, in 2004, when it was considering these amendments to the commentary, the commission, uh, conducted studies of offenders. Was this data, was this data determination? I, I believe it was because it rests on the, it rests on the, the commission conducted studies between Congress's enactment of the Protect Act and the promulgation of this commentary. And that's described in the historical report from the Sentencing Commission that we cite and that Mr. Haggerty cites. And those reports, uh, describe the studies that were conducted. But these numbers just implement the Protect Act. This is, this is, this is Congress speaking. This is not the commission. The commission is speaking in the commentary. Correct. But the guidelines text, that's Congress. It is Congress, but it's within the commission's power to enact commentary as Congress anticipated. And it's within our power to, to say that the commentary doesn't fit with Congress's. Sure. Absolutely, Your Honor. In this instance, it does fit within that ambiguous reading. And I'd like to point out the tiers, the four-tiered structure of this guideline, because it's not simply a question of an on-off five-level enhancement for a very large number of images. There's this very precise four-tiered structure from Congress at 10, 150, 300, 600 are the cutoffs. If Congress had met for the number of frames in a video to count as the number of images in that video, any offender who possessed a video of any significant length, almost all offenders who possessed videos automatically end up at the five-level enhancement. It would have been very, it's a very unusual way for whoever drafted this guideline to say videos always get the five-level enhancement. Well, they didn't spend a lot of time on it. I'm sorry? Doesn't the legislative record indicate that they spent very little time on this? This was a piece of the Protect Act that didn't get a lot of debate or attention in Congress. Isn't that right? Perhaps, Your Honor, that's a suggestion, but certainly it would have been easy to say videos are a two-level enhancement or something like that, given that videos are already mentioned in one of the guidelines that were amended by the Protect Act. So this is really an elephant's and mousehole's argument, Your Honor, that because it would have been so unusual for the draft of this guideline to intend that it be applied to every frame of the video, that some instance that a lower number would have been required. And I would point out to Judge Smith's point, I mean, this is an area that Congress takes an interest in. Congress is certainly aware in the last 19 years that the commission's commentary has said an image equates to 75 or video equates to 75 images. I don't know why you think it would be so peculiar when you consider the context. When was the Protect Act passed? 2003. 2003, Your Honor. I mean, I don't remember 2003 having the same extraordinary proliferation of videos on I don't think it's fair to criticize them for not being clairvoyant about where the digital age might take us 20 years after the statute. Certainly, Your Honor. I mean, the internet and camcorders were around back then, but even if this was an oversight by Congress, Kaiser says this is exactly the kind of area that agencies are permitted to exercise their substantive policy expertise in applying a general text to a specific situation that is not directly, explicitly addressed. And I do want to talk about Judge Phillips's three properties in her concurrence in the judgment. We think that those properties are broadly consistent with Kaiser and Nasir, but to some extent they diverge and therefore should follow Kaiser and Nasir. As to her first property, Kaiser makes clear that agencies are supposed to be using their substantive policy expertise in reaching interpretations of the guidelines. It should be an interpretation. It has to be an interpretation. It must be an interpretation, but the interpretation. These plug numbers aren't interpretations. These plug numbers are estimates for how to, these are policy choices about how harshly to punish somebody who has video. The tears come from Congress, Your Honor, but given those tears, this 25 images per video interpretation is an interpretation of those numbers, an interpretation of that tiered structure. Wait, when you say the tiered structure, you're referring to A, B, C, and D? Correct. Yes. Okay. What those tears say is that the more images one has, the more level increase you're going to suffer, correct? Correct. Yes. Those don't speak at all to videos. They say nothing about videos. They don't explicitly. So this 75 number is just a, it's a policy-based plug number that the commission came up with. I'm very brief, Your Honor. We think that is an interpretation of images based on the ambiguity as to whether images refers to every frame or some smaller number, the camera angles. It seems like such an unlikely description of what happened here. Isn't the Occam's razor approach here that the commission knew that if they didn't come up with this plug number, that there would just be a remarkable gravitational pull for the five level increase because almost everyone using video would have at least 600 or more images. Isn't that the explanation as to what happened here? That may have been in the commission's mind. That's a policy choice. How did they pull 75 out? By reference to the number of tears, Your Honor, and in the knowledge that that is a reasonable number of- Is that reasonable? It is, Your Honor, because it's an estimate of the number of specific moments, motions, actions that might be perceived by the human eye in the typical video. And this is something that's just say on an overarching basis rather than leave that determination to individual district judges. But whether that exercise of the commission's interpret power is authorized under Kaiser, we say it is, or whether this unambiguously refers to every frame, this court should affirm the sentence here. Thank you, Mr. Halligan. Ms. Stern? The parties here agree that the frames approach is not consistent with the guidelines structure. And if you're in that world, then you must consider applying the rule of lenity. The rule of lenity requires- Did you argue the rule of lenity in the district court? Well, we didn't get there because it was an issue. The court was, excuse me, the district court was applying the commentary. And so the district court opted to apply the commentary and didn't engage with the text. So the answer is no? The answer is no. I believe the answer is no. But certainly this court, if it gets there, needs to consider the rule of lenity as it held in Fleming, the rule of lenity applies in the guideline context. If you're considering two different readings, the rule of lenity requires that you adopt the more lenient reading. Well, it has to be grievously ambiguous, right? Only Justice Gorsuch takes the lower approach, right? Majority of the Supreme Court made clear it has to be, quote, grievously ambiguous. Grievous ambiguity sounds a lot like genuine ambiguity from Kaiser. And so if you're there- Doesn't it sound a lot like Kaiser's requirement that you exhaust all of the tools in the interpretive toolkit before you give up and say this is ambiguous? And one of those tools is lenity. But I think you need to reach lenity, whether you go there at Kaiser step one or after considering and very possibly rejecting the commentary after Kaiser step two. If you're still weighing two possible readings, lenity absolutely has a role to play. And I think judicial restraint in this area is critically important. If this court adopts a frames approach, the commission cannot issue a fix. And that's because Congress here thrusted this image table on the commission. So- And if Congress were disinclined to fix it, why wouldn't we rely on the good offices of our federal district judges who have discretion in application of the 3553A factors to make sure that they impose a just and fair sentence regardless of what the guidelines range might suggest? This is a guideline already plagued by harshness. I think that would just compound the unfairness that district courts have already observed time and time again in this area. There's simply no basis to adopt that harsher reading that even the government doesn't support in its briefing. It was always considered as an alternative argument, but it's really out of step with the statutory context, out of step with the incremental structure, and as has noted multiple times, could result in a person who possesses one single video being vaulted to the full five-level enhancement. I think there's a- Does that depend on the amount of CSAM in the video? Can you address the remand issue? Mr. Hallowell says if we follow Judge Larson's interpretation that remand is not necessary here. What's your response? It's absolutely necessary. We have no fact-finding as to the number of frames involved here. If that is ultimately the approach needs to go back to the district court in the first instance to consider it. There's so many administrability problems with a frames approach. It might seem elegant, but I have a feeling if it goes back to the district court and that's the directive, it's going to be a mess. That's because- Well, it's an evidentiary mess though, right? That's what courts do. Courts hear evidence, they get an expert, you get an expert, and then the judge makes a decision. We're talking about a frame-by-frame analysis to determine whether each frame is- Well, first of all, is it a duplicate of something else? Is it actually unique? Is it sexually explicit? Is it not? Aren't you just by raising questions making Judge Hardiman's point? That is questions, issues, whether they involve credibility, whether they're simply factual questions or what district courts do all the time. Administrability questions can be difficult, but that doesn't mean they're incompetent. The issue is that this guideline gives no indication that it intended to take district courts down this path, down this deeply frame-specific parsing inquiry that's really not compelled by this text. It's clear to me that this guideline speaks in a much greater field of generality. I mean, maybe it's harder than I'm supposing you can help me with this, but if I were the district judge, I wouldn't necessarily look at every single frame. I would first ask the lawyers to see if they can agree on the speed of the film to figure out how many frames per second. Wouldn't that be a reasonable place to start? I imagine that is a place to start. And I don't think that's hard for people skilled in the industry to tell one fact finder how many frames per second this particular video has. And then once you establish the number of frames per second, then you do the math. You don't have to do the math. You just have to figure out how many frames per second this particular video has. But even with what Your Honor just posited, are we going by frame? Are we going by seconds? Those are two different... Well, I thought I had established once the expert tells the fact finder how many frames there are per second, then you do the math. You don't have to... Are you suggesting that these videos run 24 frames per second, one second, and in the next second, six frames per second? I mean, that just seems implausible to me, perhaps impossible. I don't know. No, but if we're doing a frames approach, I mean, I don't know if that video exists, but if we're doing a frames approach and we're looking, we're trying to assess the amount of child pornography, we're going to have to make a content-based determination of each frame. Does it qualify as sexually explicit? No, but if it's 24 frames per second and one second includes child pornography, then you know you got 24 frames. Perhaps. I don't know that it will actually be that easy. There's no question. I'm not suggesting it's as easy as a lot of other fact findings. I'm just suggesting that it doesn't seem to be as complicated as you're suggesting, but who knows? I don't know. I think it would be incredibly onerous. The sort of upper limit threshold is 600 images. We're asking that you reverse and remand with instructions to apply the two rather than five. Thank you very much, Ms. Stern. Thank you, Mr. Hallowell. Appreciate the briefing and argument. The court will take the matter under review.