PRECEDENTIAL
UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT ______________________
No. 23-2084 _______________________
UNITED STATES OF AMERICA
v.
ROBERT HAGGERTY, Appellant _______________________
On Appeal from the United States District Court for the Western District of Pennsylvania District Court No. 2-22-cr-00015-001 District Judge: Honorable J. Nicholas Ranjan __________________________
Argued April 16, 2024
Before: HARDIMAN, SMITH, and FISHER, Circuit Judges
(Filed: July 9, 2024)
Samantha Stern [ARGUED] Office of Federal Public Defender 1001 Liberty Avenue 1500 Liberty Center Pittsburgh, PA 152220 Counsel for Appellant
Adam N. Hallowell [ARGUED] Laura S. Irwin Office of United States Attorney 700 Grant Street Suite 4000 Pittsburgh, PA 15219 Counsel for Appellee
____________________________
OPINION OF THE COURT ____________________________
SMITH, Circuit Judge. I. INTRODUCTION In imposing a sentence on a defendant who has been found guilty of a child pornography-related offense, a district judge is required, under the United States Sentencing Guidelines, to enhance the applicable Guideline Sentencing range based on the number of “images” “involved” in the offense. Specifically, under U.S.S.G. § 2G2.2(b)(7)’s graduated sentencing enhancement scheme, that defendant’s
2 Guideline Sentencing range may be enhanced by up to five levels based on the number of images involved. The calculus is a simple one where the pornographic matter consists only of “still” images. But what about when a moving image—that is, a video—is involved in an offense? The Guideline itself does not answer that question. So may the judge look to the Commentary to the Guideline, which specifies that each video—no matter its length—constitutes 75 images for purposes of calculating the applicable sentencing enhancement? Whether we should defer to this commentary is the issue we now confront. We hold that “image,” in the moving picture or video context, unambiguously means “frame.” Deference to the Commentary’s 75-images rule is therefore unwarranted under Kisor v. Wilkie, 588 U.S. 558 (2019). Instead, the number of frames comprising a moving picture or video will determine the specific sentencing enhancement that a District Judge must apply. Because the case before us involved videos with over 14,000 total frames, Haggerty probably possessed the requisite number of images to warrant a five-level enhancement under the Guideline. But because the District Court did not use the frame-counting calculus we now hold is the correct one, we will vacate the District Court’s sentencing order and remand for resentencing in a manner consistent with our holding. II. BACKGROUND This appeal is brought by Robert Haggerty, a 62-year- old first-time offender. In February of 2022, a federal grand jury indicted Haggerty on three counts of receiving a visual depiction of a minor engaging in sexually explicit conduct, as 3 well as one count of possessing such depictions (including depictions of prepubescent minors and minors less than twelve years of age). See 18 U.S.C. §§ 2252(a)(2), (a)(4)(B), (b)(2). Haggerty admitted at his plea hearing that he had communicated with undercover detectives posing as underage girls in Montgomery County, Pennsylvania, using the online messaging platforms Skout and Kik Messenger. 1 Acting on information derived from the undercover operation, agents obtained and executed a federal warrant to search Haggerty’s house. There, they seized a Samsung tablet, which contained five still files and one video file depicting child sexual abuse. Agents arrested Haggerty and recovered a second Samsung tablet from his truck. Haggerty, on his own, informed the agents that the second tablet contained sexual abuse material depicting minors. Examination of the device revealed 92 still image files and 8 video files depicting child sexual abuse, including sadistic and masochistic content involving prepubescent children and even toddlers. Haggerty entered an open guilty plea to the indictment. The District Court applied multiple Guideline enhancements at sentencing, including a five-level enhancement under U.S.S.G. § 2G2.2(b)(7). 2
1 Skout is a mobile application that purports to facilitate online social interaction. About, SKOUT, https://perma.cc/9SN7- 2QYT. Kik Messenger is an instant messaging mobile application. Help, KIK, https://perma.cc/7AP8-5US3. 2 The U.S. Sentencing Commission has explained that “[e]ach type of crime is assigned a base offense level, 4 U.S.S.G. § 2G2.2(b)(7) provides for a graduated enhancement scheme, based on the number of “images” involved in a child-exploitation offense. Under that scheme, a defendant receives a two-level enhancement if the offense involved 10 to 149 images and up to a five-level enhancement for 600 or more images. U.S.S.G. §§ 2G2.2(b)(7)(A)-(D). However, the text of the Guideline does not explain how courts should determine the number of images contained in any given video. Instead, that direction appears in the Commentary to § 2G2.2(b)(7), which states that: “[e]ach video, video-clip, movie, or similar visual depiction shall be considered to have 75 images.” U.S.S.G. § 2G2.2 app. n.6(B)(ii). 3
which is the starting point for determining the seriousness of a particular offense.” U.S. SENT’G COMM’N, How the Sentencing Guidelines Work, AN OVERVIEW OF THE FEDERAL SENTENCING GUIDELINES 1, https://www.ussc.gov/sites/default/files/pdf/about/overvi ew/Overview_Federal_ Sentencing_Guidelines.pdf. Under the Sentencing Guidelines, there are “43 levels of offense seriousness — the more serious the crime, the higher the offense level.” Id. The Sentencing Guidelines provide enhancements, which raise the base offense level based on various criteria related to a given offense. 3 “If the length of the visual depiction is substantially more than 5 minutes, an upward departure may be warranted.” U.S.S.G. § 2G2.2 app. n.6(B)(ii).
5 Applying that commentary, the Presentence Investigation Report (“PSR”) calculated that the 8 videos on the tablet contained 600 images, bringing the total image count on the second tablet alone to 692 images. 4 Haggerty objected to the application of a five-level, number-of-images enhancement. He asserted that the Guideline is unambiguous and does not include videos. Thus, Haggerty argued that the sentencing court should not defer to the Commentary’s interpretation that a video contains 75 “images,” based on the standards set forth in Kisor v. Wilkie. Based on that argument, he contended that only the 92 still images found on his tablets should count toward the enhancement, not the 8 videos. He also acknowledged that “if ‘image’ is genuinely ambiguous . . . the only acceptable alternative interpretation would be to count a video as one image.” Appx. 73. So, he asserted, his offense involved no
4 The Government points out in its brief: Haggerty’s two tablets contained a total of 97 still images and 9 videos of child sexual abuse material. For unclear reasons, the [PSR’s] calculations included only the still images and videos on the tablet from Haggerty’s truck, and ignored the tablet from Haggerty’s residence. Compare PSR ¶ 30 with PSR ¶¶ 12, 14. Haggerty’s brief repeats this mistake. See Br. at 5, 16. Response Br. at 7 n.2. 6 more than 92 to 100 “images,” and he should receive no more than a two-level enhancement under § 2G2.2(b)(7)(A). In response, the Government asserted that the Guideline itself is ambiguous and that, under the standards set forth in Kisor, the Commentary was entitled to deference. Accordingly, the Government contended, the five-level enhancement should be applied in calculating Haggerty’s Guideline Sentencing range. The Government also argued, in the alternative, that the enhancement for 600 or more “images” would apply, even without deference to the Commentary. In other words, if “image” is synonymous with “frame,” the Government was prepared to present evidence that Haggerty’s videos had a total length of over 600 seconds. Using the standard frames-per-second rate of a video derived from the Motion Picture Association’s definition of video, the Government argued that, at a standard rate of 24 frames per second, the tablets contained over 14,000 frames, or “images,” of offending content. 5
5 The Government offered that
[it] . . . would present evidence that the eight videos found on the second Samsung tablet seized in the Defendant’s tractor trailer on October 7, 2021, contain[ed] 313 seconds of content depicting the sexual exploitation of eight separate minors, ranging in age from approximately three years old to approximately eight years old. In addition, 7 The District Court overruled Haggerty’s objection to the five-level, number-of-images enhancement. It determined, based on what it viewed as a persuasive majority opinion from a panel of the Sixth Circuit, that “construing ‘image’ to equal ‘visual depiction’ is not a reasonable interpretation of the Guideline.” Appx 1 (citing United States v. Phillips, 54 F.4th 374, 381 (6th Cir. 2022)). The District Court held that “the term ‘image’ is ambiguous,” and that the Commentary’s 75-image rule was entitled to deference. Appx 1. Alternatively, the District Court recognized that, even if it were to adopt a definition in which “image” in the video context meant each video frame, the enhancement for 600 or more images “would still apply based on the government’s representations of the duration of the videos that Mr. Haggerty possessed.” Appx 2 n.1. The District Court applied the five-level enhancement and
the Government would present evidence that the first Samsung tablet seized from the Defendant’s residence on that same date contains a video that is 311 seconds in length depicting the sexual exploitation of a minor female who is approximately 12 years old. As described above, because each second of video contains 24 frames, the 624 seconds of offending content in Defendant’s possession could count as 14,976 separate images, well over the 600 images needed to apply the five- level enhancement under § 2G2.2(b)(7).
Appx 87. 8 calculated a total offense level of 32. This offense level, together with Haggerty’s criminal history category of I, yielded an advisory Guideline range of 121 to 151 months in prison. Haggerty argued for a downward variance to a prison sentence of 60 months, while the Government requested a sentence within the Guideline range. Based on Haggerty’s age, health, lack of criminal history, and statement of remorse, the District Court varied downward from the advisory Guideline range to a prison sentence of 96 months, to be followed by five years of supervised release. The District Court also required Haggerty to pay $3,500 in restitution to one victim, ordering “that the first $1,000 of that restitution be paid within the next six months” following sentencing. Appx 171. Haggerty objected to the six-month partial payment deadline, but the District Court was unmoved. The Court noted that the PSR identified assets Haggerty and his wife owned jointly. The judge also pointed out that Haggerty had agreed to pay as much restitution as possible to the victim “up front,” and concluded that it was “appropriate for [the Court] to impose as much as [it] can that wouldn’t otherwise . . . create [an] undue . . . hardship.” Appx 173. This appeal followed. III. JURISDICTION The District Court had jurisdiction pursuant to 18 U.S.C. § 3231 (offenses against the laws of the United States). Haggerty timely filed a notice of appeal. We have jurisdiction pursuant to 28 U.S.C. § 1291 (final decisions of the District Court) and 18 U.S.C. § 3742(a) (appeal of a final sentence).
9 IV. STANDARD OF REVIEW We exercise plenary review of the District Court’s interpretation of the Sentencing Guidelines. United States v. Gray, 942 F.3d 627, 630-31 (3d Cir. 2019). And we review the District Court’s restitution payment order for abuse of discretion. United States v. Fallon, 61 F.4th 95, 125 (3d Cir. 2023). V. ANALYSIS A. SENTENCING ENHANCEMENT On appeal, Haggerty argues that ordinary usage and plain text require us to interpret “each image – whether still (a photograph) or moving (a video)” – as one image for purposes of § 2G2.2(b)(7). Opening Br. at 13. He further asserts that the Guideline’s context, history, and purpose support this reading. Id. Finally, he contends that we should give no deference to the Commentary. Id. Because we conclude that the term “image” unambiguously means “frame” when applied in the video context, we reject Haggerty’s arguments. Section 2G2.2(b)(7) of the Sentencing Guidelines applies to child-exploitation crimes and sets forth the following gradations for applying the enhancement: (7) If the offense involved – (A) at least 10 images, but fewer than 150, increase [the offense level] by 2 levels; (B) at least 150 images, but fewer than 300, increase by 3 levels;
10 (C) at least 300 images, but fewer than 600, increase by 4 levels; and (D) 600 or more images, increase by 5 levels. Ordinarily, the United States Sentencing Commission, “an independent commission in the judicial branch,” created by the Sentencing Reform Act provisions of the Comprehensive Crime Control Act of 1984, establishes the Sentencing Guidelines. 28 U.S.C. § 991(a); U.S.S.C., AN OVERVIEW OF THE UNITED STATES SENTENCING COMMISSION 1 (2011), https://perma.cc/Q949-QPB5; 28 U.S.C. § 994. To update the Guidelines, the Commission “promulgate[s] and submit[s] to Congress amendments.” Rule 4.1, U.S. SENTENCING COMMISSION RULES OF PRACTICE AND PROCEDURE (as amended Aug. 18, 2016). In doing so, it must comply with standard agency rule-making procedures “relating to publication in the Federal Register and public hearing procedure,” provided by 5 U.S.C. § 553, though it may promulgate commentary and policy statements, as well as related amendments, without complying with those procedures. 28 U.S.C. § 994(x); Rule 4.3, U.S. SENTENCING COMMISSION RULES OF PRACTICE AND PROCEDURE. Congress retains authority over the Guidelines, and it has, at times, specifically directed the Commission to establish particular Guidelines. U.S. SENTENCING GUIDELINES MANUAL ch. 1, Pt. A (quoting Kimbrough v. United States, 552 U.S. 85, 103 (2007) (“Congress has shown that it knows how to direct sentencing practices in express terms. For example, Congress has specifically required the Sentencing Commission to set
11 Guideline sentences for serious recidivist offenders ‘at or near’ the statutory maximum.”); 28 U.S.C. § 994(h). It was through Congress’s direct exercise of its legislative authority that § 2G2.2(b)(7), the provision at issue here, became part of the Sentencing Guidelines. Congress added this Guideline provision through the Prosecutorial Remedies and Tools Against the Exploitation of Children Today Act of 2003, Pub. L. No. 108-21, 117 Stat. 650 (2003) (“PROTECT Act”). The Act amended the Sentencing Guidelines to increase penalties “based on the amount of child pornography involved in the offense.” H.R. Rep. No. 108-66, at 59 (2003). 6 None of the courts that have considered the issue before us have concluded that a video is an image. 7 That said, jurists
6 While the Sentencing Commission has renumbered the number-of-images provision to its current position at § 2G2.2(b)(7), see U.S.S.G. App. C. amend. 664 (2004), there have been no amendments to the text of the provision since it took effect in 2003. See U.S. SENTENCING COMM’N, FEDERAL SENTENCING OF CHILD PORNOGRAPHY: NON-PRODUCTION OFFENSES 2-3 (June 2021), https://perma.cc/QQR3-RCPP (“NON-PRODUCTION OFFENSES”) (discussing that the Commission has put forward statutory or guideline recommendations for § 2G2.2, but Congress has not implemented them). 7 United States v. Phillips, 54 F.4th 374, 384 (6th Cir. 2022); United States v. Vandyke, Nos. 23-11268, 23- 12 have taken different approaches in determining whether the term “image” is ambiguous when applied to the video context for purposes of § 2G2.2(b)(7). Notably, when considering this issue, a majority of a Sixth Circuit panel determined that “image” is ambiguous in this context and that deference to the Commentary is warranted. Phillips, 54 F.4th at 384. A third member of that panel, Judge Larsen, concurred only in the judgment. She persuasively expressed the view that “image” unambiguously means “frame” in the video context. Id. at 391 (Larsen, J., concurring in the judgment). We agree, and hold that based on the term’s ordinary meaning, and supported by the Guideline’s structure, purpose, and history, “image” unambiguously means “frame” when applied to a video in the context of § 2G2.2(b)(7). i. Post-Kisor Approach The Supreme Court has declared that the United States Sentencing Commission’s Guidelines are legislative rules, while the Commentary interpreting the Guidelines consists of
11794, 2024 WL 505080, at *4 (11th Cir. Feb. 9, 2024) (per curiam) (unpublished); see also United States v. Carmody, Nos. 22-12539, 22-13542, 2023 WL 7014048, at *2-3 (11th Cir. Oct. 25, 2023) (per curiam) (unpublished); United States v. Pratt, No. 20-10328, 2021 WL 5918003, at *2 (9th Cir. Dec. 15, 2021) (unpublished) (deferring to the Commentary under Stinson, but alternatively concluding that § 2G2.2(b)(7) is genuinely ambiguous).
13 interpretive rules. 8 Stinson v. United States, 508 U.S. 36, 44- 45 (1993); United States v. Adair, 38 F.4th 341, 347-48 (3d Cir. 2022). 9 Within this paradigm, we must consider whether the Commentary to § 2G2.2(b)(7) warrants so-called Auer deference. The Supreme Court has recognized that there are times that agency regulations “may be genuinely ambiguous.” Kisor, 588 U.S. at 566. That is, “[t]hey may not directly or clearly address every issue,” and “when applied to some fact patterns, they may prove susceptible to more than one reasonable reading.” Id. And when a court must “apply the rule to some unanticipated or unresolved situation,” it “must make a judgment call.” Id. at 568. Kisor provides lower courts with
8 We recently noted the difference between these two types of rules: “The term ‘legislative rule’ generally refers to an agency rule promulgated through formal or informal (notice-and-comment) rulemaking, although certain subject-matter exceptions exist,” and “[t]o qualify as an interpretive rule, a rule must ‘derive a proposition from an existing document whose meaning compels or logically justifies the proposition.’” United States v. Adair, 38 F.4th 341, 347 nn.1-2 (3d Cir. 2022) (citation omitted). 9 As we noted in Adair, “[t]he paradigm applies only to the Commission’s interpretive commentary, not its commentary related to either background information or circumstances that may warrant a departure from a guideline.” Id. at 347-48. 14 guidance regarding whether to defer to an agency’s own interpretation of its regulations. Id. Federal court deference to an administrative agency’s interpretation of its own regulations is no recent jurisprudential invention. The current namesake case for this type of deference, Auer v. Robbins, 519 U.S. 452 (1997), was decided in 1997. Kisor, 588 U.S. at 569. Before Auer, courts referred to the doctrine as Seminole Rock deference, named after a 1945 case. Id. at 568 (citing Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414 (1945)). And the Supreme Court has recognized that “[d]eference to administrative agencies traces back to the late nineteenth century, and perhaps beyond.” Id. at 569 (citing United States v. Eaton, 169 U.S. 331, 343 (1898)). The Supreme Court roots this deference in “a presumption that Congress would generally want the agency to play the primary role in resolving regulatory ambiguities.” Id. That presumption rested upon assumptions that (1) “the agency that promulgated a rule is in the ‘better position [to] reconstruct’ its original meaning,” id. at 570 (internal citation omitted); (2) resolving such ambiguities “entails the exercise of judgment grounded in policy concerns,” id. (cleaned up); and (3) such a presumption promotes “consistency in federal regulatory law,” id. at 573. Kisor changed the jurisprudential landscape by providing instructions for when Auer deference should be applied. Before Kisor, Auer deference permitted courts to afford controlling weight to an agency’s interpretation of its own regulation (or Guideline) unless its interpretation was “plainly erroneous or inconsistent with the regulation” or Guideline. Stinson, 508 U.S. at 45 (quoting Seminole Rock, 325 U.S. at 414). Since Kisor, Auer deference applies only if a court 15 determines that a regulation is “genuinely ambiguous” after “exhaust[ing] all the ‘traditional tools’ of construction.” Kisor, 588 U.S. at 575 (quoting Chevron U.S.A. Inc. v. Nat. Res. Def. Council Inc., 467 U.S. 837, 843 n.9 (1984)). In determining whether a regulation or Guideline is “genuinely ambiguous,” we must “consider the text, structure, history, and purpose of a regulation, in all the ways [we] would if [we] had no agency to fall back on.” Id. at 575 (cleaned up). That means exhausting the contents of our “legal toolkit” before concluding that there is “no single right answer” to the interpretive question and that the question “is ‘more [one] of policy than of law.’” Id. (internal citation omitted). And if we determine that a regulation or Guideline is “genuinely ambiguous”—and before deferring to the agency’s interpretation—we then look to the “character and context” of the agency’s interpretation. Id. at 576. That inquiry requires us to determine whether the interpretation: (1) is “the agency’s ‘authoritative’ or ‘official position,’” (2) “implicate[s] its substantive expertise,” and (3) reflects its “fair and considered judgment.” Id. at 577-79 (internal citations omitted). Here, however, we need not undertake an “independent inquiry into whether the character and context” of the Commentary “entitles it to controlling weight.” Id. at 576. After fully engaging in the process mandated by Kisor, we conclude that “image” unambiguously means “frame” in the video context.
16 ii. The Text Conveys an Unambiguous Meaning When interpreting a Guideline issued by the U.S. Sentencing Commission, “[w]e start with the plain text and presume that words carry their ordinary meaning.” United States v. Caraballo, 88 F.4th 239, 246 (3d Cir. 2023) (internal citations omitted). Haggerty contends that a moving image should be quantitatively treated as if it were a still image. But such an interpretation is oxymoronic. If “image” needs the modifier “moving” to accurately describe what is depicted or displayed, then the term “image” can hardly be equated with the term “video.” Having rejected Haggerty’s plain text argument, we attempt “[t]o discern if [the] Guideline is ‘genuinely ambiguous.’” United States v. Mercado, 81 F.4th 352, 356 (3d Cir. 2023) (internal citations omitted). And in furtherance of that inquiry, we examine “contemporary ‘dictionary definitions while keeping in mind the whole statutory text, the purpose, and context of the statute, and relevant precedent.’” Id. (internal citations omitted). For example, the Oxford English Dictionary defines image as “[a] physical or digital representation of something, originally captured using a camera from visible light, and typically reproduced on paper, displayed on a screen, or stored as a computer file” or as “any picture or graphic . . . in printed form.” 10 And though other
10 Image, OXFORD ENGLISH DICTIONARY (2009), https://perma.cc/4V2P-GZP5 (last accessed May 2024).
17 definitions vary somewhat, they are to similar effect. 11 In short, none of the definitions we have found support interpreting “image” to naturally include videos. 12
11 The Cambridge Dictionary defines “image” as “any picture, especially one formed by a mirror or a lens [or] . . . through a camera . . . [or on a] computer or television screen.” Image, CAMBRIDGE DICTIONARY (2013), https://perma.cc/3JHT-54WR (last accessed May 2024). Merriam-Webster defines “image” as 1 a: a visual representation of something: such as (1): a likeness of an object produced on a photographic material[;] (2): a picture produced on an electronic display (such as a television or computer screen) b: the optical counterpart of an object produced by an optical device (such as a lens or mirror) or an electronic device Image, MERRIAM-WEBSTER, https://perma.cc/6TBN- HQWB (last accessed May 2024). Per the American Heritage Dictionary of the English Language, an “image” is “[a] representation of the form of a person or object, such as a painting or photograph,” and Webster’s New World College Dictionary provides that an “image” is “[a] representation or likeness of a person or thing, as in a drawing, painting, photograph, or sculpture.” Image, THE AMERICAN HERITAGE DICTIONARY OF THE ENGLISH 18 Further, as the Government points out, what we now hold to be the meaning of “image” in the video context accords with how dictionaries defined “image” at the time that § 2G2.2(b)(7) was drafted in 2003. 13 Notably, we have recently relied on the Oxford English Dictionary definition of “video,” i.e., “images for display on a television screen or other
LANGUAGE (2022), https://perma.cc/6QSA-8VUX (last accessed May 2024); Image Definition, WEBSTER’S NEW WORLD COLLEGE DICTIONARY (2014), https://perma.cc/E973-NKAV (last accessed May 2024). 12 The Merriam-Webster definition, set forth in n.11 supra, might possibly be stretched to include a video, but nowhere does it specifically make space for “moving” images within the definition. 13 See Image, MERRIAM-WEBSTER’S COLLEGIATE DICTIONARY (11th ed. 2003) (“a reproduction or imitation of the form of a person or thing”; “a visual representation of something” such as “a likeness of an object produced on a photographic material” or “a picture produced on an electronic display”); Image, AMERICAN HERITAGE DICTIONARY (4th ed. 2000) (“A reproduction of the form of a person or object”). Later definitions are also in accord. See, e.g., Image, NEW OXFORD AMERICAN DICTIONARY (3d ed. 2010) (“a visible impression obtained by a camera, telescope, microscope, or other device, or displayed on a computer or video screen”).
19 electronic device.” United States v. Heatherly, 985 F.3d 254, 272 (3d Cir. 2021) (quoting Video (def. 1a), OXFORD ENGLISH DICTIONARY (3d ed. 2016)). In doing so, we recognized that a video inherently contains multiple “images.” Ordinary usage makes plain that an “image” is a fixed visual representation. True, “the human eye cannot perceive any individual frame” when viewing a video or motion picture. Phillips, 54 F.4th at 382. Nonetheless, in a video or motion picture, a fixed visual representation of something is a frame, based on the term’s ordinary usage. See id. at 391 (Larsen, J., concurring in the judgment). In her concurring opinion, Judge Larsen hits the nail on the head when she explains: “‘Images’ means exactly what you’ll find in every dictionary—a ‘still representation’; and vis-à-vis a video, an ‘image’ is a ‘frame.’” Id. at 398 (Larsen, J., concurring in the judgment). Multiple dictionary definitions of “frame” support this interpretation. The Oxford English Dictionary defines “frame” as “one of the individual images on a strip of film; (later also) a single complete image in a series forming a television picture, film, or video sequence.” Id. at 391(Larsen, J., concurring in the judgment) (quoting Frame, Oxford English Dictionary (2022)). Other definitions are similarly worded. 14 Based on
14 Merriam-Webster defines “frame” as “one picture of the series on a length of film” or “a complete image for display.” Frame, MERRIAM-WEBSTER, https://perma.cc/37KU-CB89 (last accessed May 2024). The American Heritage Dictionary defines “frame” as “[o]ne of the set of still images that constitute a film or 20 ordinary usage and dictionary definitions of the terms, “image” is synonymous with “frame” in the video context. iii. Structure, History, and Purpose Support that “Image” Means “Frame” in the Video Context Consistent with the analysis called for by Kisor, we now consider the structure, history, and purpose of the Guideline. Bolstering our interpretation of the plain text, our examination of the structure, history, and purpose of § 2G2.2(b)(7) confirms that a video contains multiple images, thereby supporting our conclusion that an “image” constitutes a “frame” in the video context. We have described § 2G2.2(b)(7) as setting forth “gradations.” In other words, it establishes increased enhancement of a defendant’s sentence based on a numerical range of “images” involved in his offense. A defendant possessing at least 10 images but fewer than 150 is to receive a two-level enhancement; one possessing at least 150 but fewer than 300 would receive a three-level enhancement; and a defendant possessing at least 300 but fewer than 600 faces a
video.” Frame, THE AMERICAN HERITAGE DICTIONARY (2022), https://perma.cc/WUZ9-BVTU (last accessed May 2024). Webster’s New World provides the definition “the rectangular image on a film screen, or the particular objects or activity focused on by the camera.” Frame, WEBSTER’S NEW WORLD (2014), https://perma.cc/4M9B- 8E7A (last accessed May 2024).
21 four-level enhancement. 15 The maximum enhancement is reserved for possession of at least 600 images and calls for enhancing the sentence by five levels. Section 2G2.2(b)(7) does not define the term “image,” nor does it clarify how the term “image” applies to videos. Consequently, after Congress passed the PROTECT Act, the Sentencing Commission sought public comment concerning proposed amendments to the Guideline commentary, acknowledging the lack of definition for “what constitutes an ‘image’ for purposes of applying” the Guideline. Sentencing Guidelines for United States Courts, 68 Fed. Reg. 75340, 75353 (Dec. 30, 2003). The Commission explicitly asked for comments or instructions on “how to count images” and “[h]ow [] videos, films, or AVI files [should] be considered.” Id. It also inquired as to whether, in a video with multiple “scenes” depicting “the same minor engaging in sexually
15 U.S.S.G. § 2G2.2’s graduated enhancement scheme applies in the same manner based on number of images involved in possessing, trafficking, receiving, transporting, shipping, soliciting, or advertising material involving the sexual exploitation of a minor. The case before us concerns charges of possession and receipt of visual depictions involving the sexual exploitation of minors. As a form of shorthand, we use the term “possess” throughout this opinion when discussing the graduated enhancement scheme. . 22 explicit conduct with a different adult,[] each scene with a different adult” should “be considered a separate image.” Id. 16 The Commission received a plethora of comments. The Federal Sentencing Guidelines Committee of the Federal Public and Community Defenders proposed defining “image” as “a single item such as one photograph, or video” instead of “each frame in a video or each person depicted.” 17 In addressing what it described as the “complex” question of “how many ‘images’ are in a video/movie clip,” the Department of Justice (the “DOJ” or “the Department”) asserted that it “would be arbitrary” to treat “a video/movie and a still image as both being one image.” 18 It also explained that
16 Video possession in this context has become more common over time. The Sentencing Commission noted in 2021 that “the prevalence of videos in an offender’s collection is higher today than in the data provided in the 2012 Child Pornography Report.” NON-PRODUCTION OFFENSES at 30. 17 U.S. Sentencing Commission, The History of the Child Pornography Guidelines 43 (2009), available at https://perma.cc/4J8F-ALHP (“Sentencing Commission, History”).
Letter from Deborah J. Rhodes, Counselor to the 18
Assistant Attorney General, to the U.S. Sentencing 23 “[a] video/movie that contains even one second of sexually explicit conduct is a more serious item than a still image.” Rhodes Letter at 5. The Department of Justice further supported its position by citing the Motion Picture Association of America’s definition of “video as 24 frames per second,” where “[e]ach frame is equivalent to one still image.” Id. That translates into “a one minute video [being] equivalent to 1440 still images.” Id. The DOJ contended that “counting each minute of video as 1440 images would be inappropriate.” Id. Instead, the Department proposed that, “considering the increased harm caused by moving videos,” the Commission should apply “a two or three level enhancement for offenses that involve video clips (defined as any type of moving images).” Id. The Sentencing Commission has described this as “a more modest enhancement” than that which a frame-counting approach would prompt. Sentencing Commission, History at 43 n.201. After the public comment period, “[t]he Commission ultimately determined that because each video contained multiple images it should be counted as more than one image.” Id. at 43. The resulting commentary “instructs that each photograph, picture, computer or computer-generated image, or any similar visual depiction shall be considered one image” and “each video, video-clip, movie, or similar recording shall
Commission, at 5 (Mar 1, 2004), available at https://perma.cc/NCF7-ADKH (“Rhodes Letter”).
24 be considered to have 75 images for purposes of the specific offense characteristic.” U.S.S.G. App. C amend. 664 (2004). “Given that the image table enacted by Congress assigned a 2-level increase for between ten images and 150 images, and a 3-level increase for 150 to 300 images, the Commission adopted a definition of video that considered each video to contain 75 images, squarely in the middle of the 2- level increase range.” Sentencing Commission, History at 43- 44. The foregoing recital of the Commission’s decision- making process reflects its desire to achieve a compromise. In the legislative process, compromise is often essential to reaching a result. Yet no matter how salutary this compromise may have been in reaching a consensus, it was not an exercise in interpretation. The stage was set, then, for the controversy that confronts us. Before the PROTECT Act, the Sentencing Guidelines contained a provision that increased the defendant’s offense level by two points if the offense involved “ten or more . . . video tapes . . . containing a visual depiction involving the sexual exploitation of a minor.” U.S.S.G. § 2G2.4(b)(2) (2003). Through § 2G2.2(b)(7), which effectively superseded that provision, Congress made clear that the number of videos would no longer be the focus. Instead, to further the PROTECT Act’s goal of increasing punishment “based on the amount of child pornography involved in the offense,” the severity of an enhancement would be tied to the number of images involved. H.R. Rep. No. 108-66, at 59. And a video, of any length, almost always includes more content than a single still image. Thus, as the Government now argues, the purpose and history of the Guideline, as amended by the 25 PROTECT Act, “suggest[s] that videos should be treated as containing more than one ‘image’ to serve the Guideline’s purpose of punishing more culpable offenders more harshly.” Response Br. at 25. The Government’s primary position, however, is that the structure of § 2G2.2(b)(7) “weighs against treating every frame of a video as an ‘image.’” Response Br. at 24. Its argument that the Guideline’s structure indicates the term “image” is ambiguous stems from an apparent concern over the practical result of equating “frame” with “image.” Applying the frame interpretation to § 2G2.2(b)(7), it argues, will skew many sentences in the direction of the maximum number of images, 600 or more. Citing the Phillips majority, the Government argues that if “‘images’ meant ‘frames,’ then possessing any video would nearly automatically vault the offender to the top of the range, thereby obviating the purpose of prescribing different levels.” Response Br. at 23 (quoting Phillips, 54 F.4th at 383). But it is not our role as a court to subordinate an inquiry into the plain meaning of a congressionally imposed Guideline to subjective concerns about the severity of a sentence that results from application of that Guideline. See, e.g., Lamie v. U.S. Trustee, 540 U.S. 526, 536 (2004) (“We should prefer the plain meaning since that approach respects the words of Congress.”). Severe though resulting sentences may appear, § 2G2.2(b)(7) reflects a policy decision made by Congress—and the making of policy is solely within the province of the legislative branch. The judiciary’s role is to heed the text that Congress has provided and to interpret that text, faithfully, in a manner that is consistent with the words themselves and the purpose behind a particular congressional action. 26 Similarly, it is not our role to question Congress’s choice of language because we believe that such language would produce unreasonable sentence variations between defendants. The Government argues that “[t]he image table’s precision indicates that Congress sought to vary the penalty based on proportionate quantities of images, not quarter- second differences.” Response Br. at 24 (quoting Phillips, 54 F.4th at 383). Still, notwithstanding the image table’s precision, enhancements prompting longer prison sentences for more pornographic content—i.e., more frames of content— align with the Guideline’s history and purpose, discussed further infra. And where applying the Guideline would be overly harsh or unreasonable, sentencing judges maintain the discretion to vary downward from the Guideline. The Sentencing Guidelines are not mandatory. United States v. Booker, 543 U.S. 220, 245 (2005); Gall v. United States, 552 U.S. 38, 46 (2007). After a sentencing judge accurately calculates the applicable Guidelines’ range, and having given both parties the opportunity to argue for the sentence they believe to be appropriate, that judge “must make an individualized assessment based on the facts presented,” to determine whether justification exists for an outside-the- Guidelines sentence. Gall, 552 U.S. at 50. She “must adequately explain the chosen sentence,” which “allow[s] for meaningful appellate review and [] promote[s] the perception of fair sentencing.” Id. Thus, if a sentencing judge determines that applying the § 2G2.2(b)(7) enhancement would be overly harsh or unreasonable based on the facts of a given case, she has the discretion to vary downward from the Guideline range and impose what she considers to be a reasonable sentence.
27 Judges retain discretion to consider factors that can ameliorate a sentence that is overly harsh or unreasonable. And such variances have become common. The U.S. Sentencing Commission reported in 2021 that “[i]n fiscal year 2019, less than one-third (30.0%) of non-production child pornography offenders received a sentence within the guideline range,” and “[t]he majority (59.0%) of non- production child pornography offenders received a variance below the guideline range.” Child Pornography Report.” NON- PRODUCTION OFFENSES at 5. Finally, we reject the argument that reading “image” to mean “frame” in the video context is unreasonable because it would create a burdensome process for a court in calculating a Guideline Sentencing range. The Phillips majority expressed just such a concern. It claimed that “a frame-by-frame analysis of each video to ascertain the number of frames that include illegal images” would be “an onerous and unrealistic task given the multitude of frames in any one video and the many cases that involve multiple videos.” Phillips, 54 F.4th at 382. But the Government has conceded here that it is prepared to present evidence showing the number of frames containing explicit content in Haggerty’s videos, based on the standard frames- per-second rate. We fail to understand how parties could not do the same in other cases. If the Government is confident that it can meet such an evidentiary burden in this case, we are equally confident that the Article III judiciary—often tasked with making findings in far more complicated settings—is equally up to meeting its responsibility in sentencing child pornography offenders. The court would merely need to determine how many seconds within each video contain child 28 pornography—as defined under 18 U.S.C. § 2256(8)—and then apply the standard rate of 24 frames per second. The Government argues further that “[t]he purpose and history of § 2G2.2(b)(7) also do not resolve the ambiguity – although they do confirm that a video contains multiple images.” Id. It contends that “Congress’s goal of increasing punishments based on the amount of child pornography involved in the offense, would be served more precisely if each frame were not considered an ‘image.’” Id. (cleaned up). However, the Government fails to explain its rationale behind that assertion. The history surrounding § 2G2.2(b)(7), outlined above, shows that Congress’s purpose in passing the PROTECT Act was to ensure that defendants who possess greater amounts of child pornography would face sentences of greater length. That alone does not resolve the ambiguity surrounding what “image” means in the video context. But the plain meanings of “image” and “frame” resolve that ambiguity, and the history and purpose reinforce our conclusion that “image” and “frame” are synonymous for purposes of § 2G2.2(b)(7). After all, as the Government acknowledged, “it is reasonable to conclude ‘that videos and movies cause more harm and so should be weighed much more heavily than photos or pictures.’” Id. (quoting United States v. Lynde, 926 F.3d 275, 280 (6th Cir. 2019)). Possessing more frames of explicit content necessarily means possessing more pornographic material. Accordingly, concluding that an “image” is a “frame” in the video context aligns well with the PROTECT Act’s history and purpose.
29 iv. The Traditional Tools of Construction Yield a Definitive Meaning of the Term “Image” To summarize, we hold that, based on the ordinary meaning of both terms, an “image”—in the video context—is a “frame.” With that understanding, “[t]he [Guideline] then just means what it means—and the court must give it effect, as the court would any law.” Kisor, 588 U.S. at 575. As discussed above, the Guideline’s structure, history, and purpose further support that conclusion. We should not defer to the Commentary where there is “only one reasonable construction” of the Guideline. Id. And that is the case here. We will, therefore, vacate the sentencing order and remand for sentencing in accordance with this opinion. Haggerty further argues that the rule of lenity supports his interpretation of the Guideline. However, “[t]hat rule applies only when a [provision] contains a ‘grievous ambiguity or uncertainty,’ and ‘only if, after seizing everything from which aid can be derived,’ the Court ‘can make no more than a guess as to what Congress intended.’” Ocasio v. United States, 578 U.S. 282, 295 n.8 (2016) (quoting Muscarello v. United States, 524 U.S. 125, 138-39 (1998)). That is not our case. Because the term “image” is not ambiguous when applied to videos, resort to the rule of lenity is not appropriate here. While we concede that application of this Guideline by quantifying pornographic matter contained in a video or motion picture on the basis of the number of frames contained within it may yield sentences that seem unusually harsh, we cannot engage in an exercise of semantic selection with an
30 over-riding concern for results. We must give effect to the words Congress has chosen, and we must do so through the faithful use of the tools that Kisor reminds us are available. And where harshness suggests injustice, a sentencing judge should thoughtfully consider a downward variance. B. RESTITUTION Haggerty also challenges the District Court’s imposition of a requirement that he make a $1,000 partial payment within six months of the $3,500 restitution ordered. Haggerty argues that “the record was clear that no substantial payment toward restitution would be feasible until after [he] finished serving his term of incarceration.” Opening Br. at 14. Because the payment schedule was both procedurally and substantively proper, the District Court did not abuse its discretion in imposing the $1,000 payment requirement. “The issuance of a restitution order” for child- exploitation offenses like Haggerty’s “is mandatory.” 18 U.S.C. § 2259(b)(4)(A). The Probation Office is required by 18 U.S.C. § 3664 to include “to the extent practicable . . . information relating to the economic circumstances of each defendant” in the Presentence Report. 18 U.S.C. § 3664(a). The District Court must then resolve at sentencing “[a]ny dispute as to the proper amount or type of restitution” “by the preponderance of the evidence,” and “[t]he burden of demonstrating the financial resources of the defendant . . . shall be on the defendant.” Id. at § 3664(e). Further, the court must specify “the manner in which, and the schedule according to which, the restitution is to be paid.” Id. at § 3664(f)(2). The sentencing judge is to consider, among other things, “the
31 financial resources and other assets of the defendant, including whether any of these assets are jointly controlled.” Id. Payment schedules may include “partial payments at specified intervals.” Id. at § 3664(f)(3)(A). Because the District Court considered Haggerty’s jointly owned assets and his agreement to pay a victim as much restitution as possible, up front, it committed no abuse of discretion. See Fallon, 61 F.4th at 125-26. VI. CONCLUSION Because “image” unambiguously means frame in the context of § 2G2.2(b)(7), and Haggerty possessed well over 600 images, the District Court did not err in imposing a five- level sentencing enhancement based on § 2G2.2(b)(7). Nor did it err in imposing the restitution order as it did. We will vacate the District Court’s judgment and remand so that the District Court may resolve any factual disputes related to a frame-based sentencing analysis.