concurring in part and dissenting in part.
The district court improperly added four points to Dennis Lay’s offense level at sentencing — two for unduly influencing a minor and two for using a computer. The application of these enhancements contravenes the clear, plain language of the Guidelines and should be reversed. These two enhancements added approximately three years to Lay’s sentence. Additionally, the twenty-year term of supervised release, which will not expire until Lay is 82 (if he lives that long), is substantively unreasonable and should also be reversed. I, therefore, respectfully dissent.
Section 2G1.3(b)(2)(B) of the Sentencing Guidelines provides for a two-point sentence enhancement if the defendant “unduly influenced a minor to engage in prohibited sexual conduct.” The application notes state that “[i]n determining whether subsection (b)(2)(B) applies, the court should closely consider the facts of the case to determine whether a participant’s influence over the minor compromised the voluntariness of the minor’s behavior.” There is no indication here that the voluntariness of the minor’s behavior was compromised, and much uncontroverted evidence that the victim was a voluntary participant in, and partial instigator of, the planned sexual conduct. This voluntariness in no way vitiates Lay’s conviction or his culpability for his actions. But it does preclude the application of this sentencing enhancement.
The district court also added a two-point enhancement under § 2G1.3(b)(3)(B). The majority concedes that this enhancement was clearly inapplicable, but dismisses this as a mere citation error, concluding that the district court really meant to apply § 2G1.3(b)(3)(A). Even if we were permitted to swap in the enhancement that the district court “should” have used, it would not change the result. Section 2G1.3(b)(3)(A) provides for a two-point sentence enhancement “[i]f the offense involved the use of a computer ... [to] persuade, induce, entice, coerce, or facilitate the travel of, the minor to engage in prohibited sexual conduct.” The majority admits that there is no evidence that Lay ever discussed sexual activity over the computer with the victim. Indeed, all evidence indicates that their relationship did not become romantic or sexual in nature until after they had stopped communicating over the Internet. Since the enhancement requires that the computer be used to “persuade, induce, entice, coerce, or facilitate the travel of, the minor to engage in prohibited sexual conduct,” that should be the end of the inquiry.
Finally, after serving seven years in prison, Lay will have to spend the next twenty years of his life asking the probation office for permission any time he wants to use a computer, apply for a job, or move to a new residence, among many other severe restrictions. The combination of the onerousness and length of Lay’s *452term of supervised release renders it substantively unreasonable.
I. Unduly Influencing a Minor
I agree with the majority that the application of the two-point enhancement for unduly influencing a minor did not constitute impermissible double counting. But I cannot agree that the facts of this case permit its application here. The authoritative commentary makes explicit what is clearly implied by the use of the term “unduly influenced” in the text: “the court should closely consider the facts of the case to determine whether a participant’s influence over the minor compromised the voluntariness of the minor’s behavior.” It is true that a rebuttable presumption of undue influence applies here, since Lay was 38 years older than his victim. But when uncontroverted evidence indicates that the voluntariness of the minor’s behavior was never compromised, that presumption has been rebutted.
The district court erred by focusing exclusively on Lay’s conduct and completely overlooking the dispositive question: was the voluntariness of the victim’s behavior compromised by the defendant’s actions? The district court justified the enhancement by stating, “I would indicate that the court believes that clearly the defendant through the use of his packages, his money, his cell phone as well as his conversations, admitted conversations with this young, troubled 15-year-old, clearly contrary to what has been argued here, the manipulator was the defendant.” These observations bear no logical relationship to the question of the victim’s voluntariness. It may be possible in some cases to use the defendant’s behavior to draw inferences about the voluntariness of the victim’s behavior. But there has been no attempt to do so here. The district court simply listed Lay’s actions, ignored the victim’s behavior entirely, and concluded without analysis that the enhancement must apply.
The majority’s approach is slightly different, but no less erroneous. Unlike the district court, it acknowledges that “evidence suggested that [the victim] proposed a meeting with Lay, that [he] lied to Lay about being abused, and that [he] initiated the original communication with Lay.” But the majority concludes that “these facts are consistent with a victim who has been influenced by a sexual predator.” This sets up an unappealing Catch-22: If the facts show that the victim’s participation was involuntary, the enhancement for undue influence will apply. And if the facts show that the victim’s participation was voluntary, well, that just proves that he must have been “influenced by a sexual predator.”
The majority states that the district court’s factual finding that Lay sent the victim gifts is not clearly erroneous. That is true, but beside the point — the question is whether it was legal error to conclude from these facts that the victim was unduly influenced, and this question requires us to examine the voluntariness of the victim’s behavior, which the majority fails to do. Instead, it characterizes this as a purely factual question that we may reverse only if we find clear error. But the applicability of the enhancement is a mixed question of fact and law. I do not dispute the district court’s findings about either Lay’s or the victim’s behavior; but I disagree that those facts can serve as the basis for the legal conclusion that Lay unduly influenced the victim within the meaning of § 2G1.3(b)(2)(B).
Besides contravening the text and authoritative application notes, the majority opinion stands in stark contrast to all reported decisions from the Courts of Appeals interpreting the undue-influence enhancements, including binding precedent *453from this Court. In United States v. Chriswell, 401 F.3d 459 (6th Cir.2005), we interpreted U.S.S.G. § 2A3.2(b)(2)(B), which uses the same language and has the same application notes as § 2G1.3(b)(2)(B). We considered the Oxford English Dictionary definitions of the words “unduly” (“without due cause or justification; without proper regard to right and wrong; unrightfully, improperly”) and “influence” (“to exert influence upon, to affect by influence; to affect the mind or action of’), in conjunction with the application notes, and concluded that “the court must engage in a victim-focused inquiry when applying this subsection in every case.” Chriswell, 401 F.3d at 469. We also noted that this approach was consistent with the background commentary to § 2A3.2, which “was added to the Guidelines in 2000 to capture those cases where ‘coercion, enticement, or other forms of undue influence by the defendant ... compromised the voluntariness of the victim’s behavior and, accordingly, increased the defendant’s culpability for the crime.’ ” Id. at 463-64 (quoting U.S. Sentencing Guidelines Manual § 2A3.2 cmt. background (2000)). Although the Chris-well court differed in certain respects from previous decisions by the Seventh and Eleventh Circuits, the three courts were unanimous on the only point that is relevant here: when there is a real child victim (as opposed to a case involving a police officer posing as a child), the applicability of the undue-influence enhancement must be determined by focusing on the victim’s behavior and asking whether its voluntariness was compromised by the defendant’s actions. See United States v. Mitchell, 353 F.3d 552, 558 (7th Cir.2003) (surveying various legal definitions of undue influence and concluding that each “requires an actual target of influence and contemplates a situation where the ‘influencer’ has succeeded in altering the behavior of the target”); United States v. Root, 296 F.3d 1222, 1234 (11th Cir.2002) (noting that “the voluntariness of a real child victim’s actions would be dispositive if an undercover agent were not involved”); id. at 1236 (Kennedy, J., concurring in part and dissenting in part) (noting that the enhancement “requires examination of the effect on the victim”).
More recently, in United States v. Myers, 481 F.3d 1107 (8th Cir.2007), the Eighth Circuit affirmed the district court’s conclusion that the § 2G1.3(b)(2)(B) enhancement should not apply, despite the presumption of undue influence that existed because of the discrepancy in age between defendant and victim. The court noted that Myers had put on evidence showing that the victim had “possessed some inclination to leave home before she even encountered Myers and had contemplated running away with” another man. Id. at 1112. Furthermore, the victim “characterized the plan for her to run away and marry Myers as ‘both of [their] ideas,’ agreed that Myers ‘didn’t have to do anything to convince [her] to go to Kentucky,’ and stated that she anticipated that she would at some point have sex with Myers.” Id. The court affirmed the district court’s conclusion that “despite the difference in their ages, Myers did nothing that compromised [the victim’s] volition, however misguided it may have been.” Id.
The Myers approach focuses on the victim’s volition, and whether it was overborne, and accepts the possibility that a victim could have been a willing participant, even if that willingness seems misguided or grotesque to the court. This is the proper approach to take in these cases, and the one that the majority should have followed here. Indeed, it is not clear how a presumption of undue influence could be rebutted, except by showing that the victim’s actions were consistently voluntary. The argument in favor of focusing on the *454defendant’s behavior, presumably, is that when the defendant takes certain actions— for example, plying the victim with gifts or flattery, emphasizing his greater knowledge and life experience — we should infer that the victim’s actions in response were not truly voluntary, even if they appear to be to outside observers, or even to the victim himself. But, at a minimum, a sentencing judge should be explicit that he is considering the defendant’s actions for the purpose of drawing inferences about the voluntariness of the victim’s behavior. Simply citing the defendant’s actions disapprovingly does not meet the requirements of § 2G1.3(b)(2)(B) and its application notes.
Had we followed the proper approach, the outcome here would be clear. All evidence suggests that the victim initiated the relationship with Lay, called Lay frequently, deceived Lay about his home life and living situation in order to win his sympathy, was the first to suggest that Lay visit him, and planned an alibi that would let him get away to spend the weekend with Lay (with an expectation that they would have sexual relations then). The victim professed to be in love with Lay and wanted to go to California to live with him. When informed of the police investigation, the victim stated a desire not to get Lay in trouble, thought the police should apologize to Lay, and mentioned the possibility of suicide, apparently at the prospect of Lay’s arrest. This evidence indicates that the victim’s volition was not compromised, and suffices to rebut the presumption that Lay unduly influenced the victim.
Traveling interstate for the purpose of engaging in sexual activity with a minor is a crime, whether the minor’s involvement is consensual or not. Properly applied, though, the enhancement for undue influence sensibly distinguishes cases involving consensual conduct from those more culpable cases in which the defendant’s actions compromise the voluntariness of the minor’s behavior. The majority eviscerates this distinction by allowing district courts to apply the enhancement to cases like this one where all the evidence suggests that the victim’s participation in the relationship was entirely voluntary.
II. Use of a Computer
The district court increased Lay’s sentence under U.S.S.G. § 2G1.3(b)(3)(B), which provides a two-point enhancement “[i]f the offense involved the use of a computer or an interactive computer service to ... entice, encourage, offer, or solicit a person to engage in prohibited sexual conduct with the minor.” As the majority concedes, this enhancement is clearly inapplicable, since its plain language requires the defendant to have used a computer to interact with “a person” other than the minor. The majority waives this away, however, as “an inconsequential citation error.” But this is not a simple matter of the parties quoting the language of one provision while inadvertently citing to another. The Presentence Report, the district court, and the prosecution (at sentencing and on appeal) refer repeatedly to the language of § 2G1.3(b)(3)(B), including the direct references to “a person” other than the minor. It was a mistake, but not a typographical or inconsequential one; it was a substantive error that stems from the underlying problem in this case — the district court’s disregard for the text of the enhancements it was applying.
Assuming arguendo, though, that the majority is correct — ’that the parties’ consistent and repeated quotation of the language of § 2G1.3(b)(3)(B) was in fact just an uncanny series of slips of the pen, rendering us free to substitute § 2G1.3(b)(3)(A) in its place — the enhancement is still inapplicable. Section *4552G1.3(b)(3)(A) provides for a two-point enhancement “[i]f the offense involved the use of a computer or an interactive computer service to ... persuade, induce, entice, coerce, or facilitate the travel of, the minor to engage in prohibited sexual conduct.” Ignoring the final clause of this provision, the district court held that “the mere fact that the defendant used a computer to contact this 15-year-old and to communicate with [the victim] and to entice [the victim] to begin this relationship was sufficient for the offense characteristics to be applicable.” The Sentencing Commission is certainly free to create an enhancement that applies whenever a defendant uses a computer to start a relationship that subsequently becomes sexual. But it has not done so. The enhancement only applies when the defendant uses a computer to entice (etc.) a minor to engage in prohibited sexual conduct. The record clearly indicates that this did not happen here.
Lay testified that his first contact with the victim occurred some time in July 2006, when he received an unsolicited invitation from the victim to chat online. The victim’s online profile said that the victim was 25 years old. There is nothing in the record about the content of this chat session. Soon thereafter, the victim sent Lay a series of emails, discussing quotidian events like “a picnic at a park, [and] things that ... the victim was doing over the weekend.” Lay would usually send short responses, “two or three lines” long, saying, “ ‘Wow, that was nice. It must have been exciting,’ ” or something similar. After communicating “off an on” like this until “mid to late August or early September,” 1 their correspondence shifted entirely to the telephone, at the victim’s request. At the outset, their phone conversations were of the same nature as their email correspondence — that is, they talked “primarily” about the victim’s interest in soccer, plus “every day things” like the weather and the minor tribulations of Lay’s job as a truck driver. Lay testified that over “a period of several months,” their relationship gradually became “more of an intimate friendship,” and at some point during this time Lay realized that the victim was younger than he had first believed. When asked directly whether any of the conversations that dealt with his visit to Ohio was “ever carried out by way of a computer,” Lay answered “No, sir, never.” None of these statements was called into question on cross-examination. Nor does the victim’s statement suggest anything different, except to the extent that he said that their correspondence was conducted entirely over the phone beginning in late July, rather than August or September, as Lay recalled. Nowhere in the record is there any evidence that their online correspondence was at all sexual in nature. In other words, nothing in the record supports the conclusion that Lay used a computer to “persuade, induce, entice, coerce, or facilitate the travel of, the minor to engage in prohibited sexual conduct.”
The majority offers three arguments to avoid what should be the obvious conclusion. First, it conjectures that Lay conducted the online correspondence with “the apparent intention of having prohibited sexual relations” with the minor. *456Where this factual allegation comes from is entirely unclear. But even if it is true, Lay’s subjective intentions while corresponding with the victim are irrelevant to the question of whether Lay enticed him to engage in sexual conduct. Second, the majority says that disallowing the enhancement here would contradict the purpose of the enhancement. If we imagine that the purpose of the enhancement is to lengthen the sentence of any offender who meets his victim online, then that might be true. But the enhancement’s purpose, which is best articulated by the text and application notes, is to lengthen the sentence of an offender who uses a computer to entice a minor to engage in prohibited sexual conduct, since a computer facilitates anonymous, one-to-many communication in a way that, say, a phone conversation does not. Finally, the majority distinguishes this from “a case in which a defendant and a victim, after an innocent encounter via computer, developed a relationship offline that evolved into an inappropriate sexual relationship.” But the only distinction between that hypothetical case (in which the majority would apparently find the enhancement inapplicable) and this one is the number of innocent online correspondences that occurred before the relationship developed offline into an inappropriate sexual one. Again, this distinction bears no relation to the language of the enhancement, and its rationale is completely unexplained. The applicability of the enhancement should turn on how the computer was used, not on the number of correspondences.
The first paragraph of the majority opinion states that “none of Lay’s discussions with the minor about sexual activity occurred via computer.” This fact should have been dispositive, and the enhancement for using a computer to entice a minor to engage in prohibited sexual conduct should not apply.
III. Substantive Reasonableness
Lay’s twenty-year term of supervised release will start to run when his seven-year prison sentence ends, at which time he will be 62 years old. In addition to the standard conditions of supervised release, the district court imposed a litany of special conditions, including several related to Lay’s new status as a sex offender. Among other things, Lay must not: have any contact with any person under the age of 18, unless his or her parent or legal guardian is present or Lay notifies the parent or guardian about his conviction; associate or correspond with anyone who “has a sexual interest [in] or attraction to minors”; reside within direct view of schools, parks, playgrounds, youth centers, or other facilities used primarily by minors; or use any online computer service — including at work or in an educational capacity — without prior written approval. Lay also must: get prior approval for all residential and employment decisions; install monitoring software on his computers; provide the probation office with his passwords and Internet history; submit to periodic unannounced searches of his computer; and submit to warrantless searches of his person, residence, vehicle, and place of business.
These conditions will undoubtedly make it very difficult for Lay to find a job or reintegrate himself into society. Perhaps it could be said that the conditions themselves, though extremely onerous, are “reasonably related to the dual goals of probation, the rehabilitation of the defendant and the protection of the public.” United States v. Ritter, 118 F.3d 502, 504 (6th Cir.1997). But for twenty years? A condition of supervised release must “involve[] no greater deprivation of liberty than is reasonably necessary for the purposes” of § 3553. 18 U.S.C. § 3553(d). I cannot see how it is reasonably necessary to require a man in his sixties, seventies, *457and eighties, with no prior criminal history, to sacrifice so many of his liberties in the name of protecting the public. I would conclude that the twenty-year sentence of supervised release, with its long list of liberty-infringing conditions, is substantively unreasonable.
IV. Conclusion
Lay’s actions were clearly criminal, as he acknowledged. But in our rush to condemn the wrongness of his actions, we should still strive to apply the law disinterestedly, looking closely at the facts to determine whether they fit within the language of the Sentencing Guidelines. Because all relevant evidence points to the conclusion that the voluntariness of the victim’s actions in this case was never compromised, and that Lay never used a computer to entice the victim to engage in prohibited sexual conduct, the two sentencing enhancements at issue should be reversed. And because Lay’s twenty-year term of supervised release is greater than necessary to comply with the goals of sentencing, his sentence should be vacated as substantively unreasonable. For these reasons, I respectfully dissent.
. The date range that Lay gave — beginning some time in July and ending between mid August and early September — permits but in no way requires the majority’s conclusion that Lay and the victim corresponded online for "two months.” As I explain below, the length of the correspondence is irrelevant to the question at hand. But the majority seems to believe otherwise, so I note in passing that its conclusions about the duration are only ambiguously supported by the record.