(dissenting).
I respectfully dissent. We held in State ex rel. Morrow v. LaFleur, 590 N.W.2d 787, 789 (Minn.1999), that extending an inmate’s supervised release date due to his failure to participate in a sex offender treatment program is not “compulsion” in violation of the Fifth Amendment. “We are extremely reluctant to overrule our precedent” and require a “compelling reason” to do so. State v. Lee, 706 N.W.2d 491, 494 (Minn.2005) (internal quotations omitted). The majority nevertheless argues that McKune v. Lile, 536 U.S. 24, 122 S.Ct. 2017, 153 L.Ed.2d 47 (2002), and Carrillo v. Fabian, 701 N.W.2d 763 (Minn.2005), require us to overrule Morrow. I disagree with the majority’s reading of McKune and the application of Carrillo. Because we are reluctant to overrule precedent and no compelling reason to do so has been advanced, I would apply Morrow.
In this case, inmates Johnson and Henderson each claim that their Fifth Amendment privilege against self-incrimination was violated by the extension of periods of incarceration for refusal to admit crimes in a sex offender treatment program. In Morrow, we held that giving an inmate a choice between treatment and confinement for a larger portion of his sentence was not compulsion within the meaning of the Fifth Amendment. 590 N.W.2d at 792-93. We noted that “[cjompulsion is the touchstone of the Fifth Amendment,” and concluded that Morrow’s choice between treatment and confinement for a larger portion of his sentence did not “rise to the level of compulsion necessary in order to constitute a Fifth Amendment violation.” Id. at 792 (internal quotations omitted). We also said that “[a] prisoner’s supervised release date is conditional on compliance with prison rules, including participation in ‘treatment or other rehabilitative programs.’ ” Id. at 793 (quoting Minn.Stat. § 244.05, subd. lb(b) (1998)). Absent a compelling reason to overrule Morrow, Johnson’s and Henderson’s claims must fail.
In McKune v. Lile, the Supreme Court concluded that imposing less attractive prison conditions on an inmate as a result of the inmate’s refusal to admit past crimes as part of Kansas’s sex offender treatment program did not rise to the level of compelled self-incrimination. 536 U.S. at 29, 122 S.Ct. 2017. Because the inmate refused to participate in the program, he was transferred to a maximum-security unit and lost certain privileges, including visitation rights, work opportunities, and access to a personal television. Id. at 30-31,122 S.Ct. 2017.
McKune did not have a majority opinion. A four Justice plurality, borrowing the “atypical and significant hardship” test from Sandin v. Conner, 515 U.S. 472, 115 S.Ct. 2293, 132 L.Ed.2d 418 (1995), found no compulsion. McKune, 536 U.S. at 37, 41, 122 S.Ct. 2017. The plurality recognized that the Sandin test was not a “precise parallel” but provided “useful instruction” and a “reasonable means” of assessing compulsion. Id. In its strongest endorsement of the Sandin standard, the plurality stated that it “provides a reasonable means of assessing whether the re*313sponse of prison administrators to correctional and rehabilitative necessities are so out of the ordinary that one could sensibly say they rise to the level of unconstitutional compulsion.” Id. at 41, 122 S.Ct. 2017. Unlike the majority here, I do not read this somewhat equivocal statement as a requirement that future courts apply Sandin in Fifth Amendment cases.
The McKune plurality also did not say that any threatened consequences amounting to “atypical and significant hardships” violate the Fifth Amendment. The plurality’s language suggests that an atypical and significant hardship is a necessary, but not a sufficient, condition for compulsion: a prison rehabilitation program “does not violate the privilege against self-incrimination if the adverse consequences an inmate faces * * * do not constitute atypical and significant hardships in relation to the ordinary incidents of prison life.” 536 U.S. at 37-38, 122 S.Ct. 2017. It does not necessarily follow, as the majority concludes, that the threat of an atypical and significant hardship is ipso facto compulsion. The majority’s conclusion is certainly reasonable, but before overruling our own precedent based on a fractured Supreme Court decision with no clear holding, I would require an unequivocal statement from that decision of the standard we are to apply. I do not find such a standard in the McKune plurality.
Justice O’Connor concurred in the result, but rejected the plurality’s application of the “atypical and significant hardship” standard. Id. at 48-49, 122 S.Ct. 2017 (O’Connor, J., concurring). While she indicated that she favored a “broader” test, Justice O’Connor did not articulate what test she had in mind; rather, she described the particular consequences at issue in McKune and opined that they were not severe enough to be “compulsive on any reasonable test.” Id. at 48-49, 53-54, 122 S.Ct. 2017.
"When no single reasoning explains the result of a divided Supreme Court, we apply the “narrowest grounds doctrine” of Marks v. United States, 430 U.S. 188, 193, 97 S.Ct. 990, 51 L.Ed.2d 260 (1977): Because Justice O’Connor’s concurrence was confined to the facts of McKune, her conclusion that the program before the Court did not create compulsion is the holding of McKune. See, e.g., Ainsworth v. Stanley, 317 F.3d 1, 4 (1st Cir.2002) (describing Justice O’Connor as concurring on much narrower grounds); Searcy v. Simmons, 299 F.3d 1220, 1225 (10th Cir.2002) (“[W]e view [Justice O’Connor’s] concurrence as the holding of the Court in McKune.”). The majority argues that the plurality and Justice O’Connor agree that the Fifth Amendment test for compulsion is at least as broad as the Sandin test, so the teaching of McKune is that any consequence that meets the Sandin test is one that creates compulsion for Fifth Amendment purposes. Assuming that this is an accurate reading of the plurality, it places more weight on Justice O’Connor’s passing statement that she favors a “broader” test than it will bear. Justice O’Connor might favor a test qualitatively different than Sandin’s, that would find no compulsion where some “atypical and significant” consequences are threatened. Without any insight into Justice O’Connor’s thinking beyond the facts of McKune, I do not believe we are justified in finding an implied consensus among her concurrence and the plurality. Instead, as McKune offers “no clear guideposts” for compulsion claims, we should “resort to our own sound judgment, so long as it does not conflict with existing precedent.” Ainsworth, 317 F.3d at 4.
Whatever lessons can be drawn from McKune, they do not involve additional *314incarceration time, the issue confronting us here. Although both the plurality and Justice O’Connor hinted in McKune that additional incarceration time might rise to the level of compulsion for Fifth Amendment purposes, 536 U.S. at 38, 52, 122 S.Ct. 2017, these comments were dicta and the plurality was generally equivocal about how it might view an extended term of incarceration. The plurality, furthermore, cited with approval Minnesota v. Murphy, 465 U.S. 420, 104 S.Ct. 1136, 79 L.Ed.2d 409 (1984). McKune, 536 U.S. at 42-43, 122 S.Ct. 2017. The plurality read Murphy to find no Fifth Amendment violation where an inmate’s probation agreement required him to be truthful with his probation officer or face execution of his suspended 16-month prison sentence.7 536 U.S. at 43, 122 S.Ct. 2017. While Murphy involved the threat of additional prison time, the plurality mentioned it as a choice that did not amount to compulsion. 536 U.S. at 43, 122 S.Ct. 2017. In light of all this, I do not believe we have a compelling reason to extend the Sandin test into the realm of extension of supervised release dates and overrule Morrow.
Accepting for the sake of argument the majority’s divination of McKune, the majority still needs to go through Camilo v. Fabian to reach its conclusion that Morrow is no longer good law. In Carrillo, the issue was whether the Commissioner provided Carrillo with sufficient procedural due process prior to delaying his supervised release date for the purposes of a disciplinary offense. 701 N.W.2d at 766. In analyzing Carrillo’s procedural due process complaint, our initial inquiry was “whether Carrillo has a liberty interest in his supervised release date.” Id. at 768. Applying Sandin, we held that “any extension of an inmate’s period of imprisonment represents a significant departure from the basic conditions of the inmate’s sentence” and concluded that Carrillo had “a protected liberty interest in his supervised release date that triggers a right to procedural due process before that date can be extended.” 701 N.W.2d at 771, 773.
Carrillo did not involve a Fifth Amendment self-incrimination claim. Moreover, Carrillo did not explicitly overrule (or even mention) Morrow. While inmates undoubtedly enjoy a right to procedural due process before a supervised release date is extended, it does not follow that they enjoy greater Fifth Amendment self-incrimination protection if the release date is extended. Carrillo does not mean that prison authorities cannot extend a supervised release date; it only means that they cannot do so without affording a hearing to the inmate.
Carrillo’s due process analysis, moreover, does not help us determine whether an inmate has been compelled to testify against himself. Carrillo focused on Minnesota’s statutory penal scheme, which created a presumption that an inmate would be released from prison on a certain date, and concluded that an extension of incarceration represented “a significant departure from the basic conditions of the inmate’s sentence” requiring due process protection. 701 N.W.2d at 772-73. But the basic conditions of the inmate’s sentence should make no difference to a compulsion analysis. Where an inmate is asked to choose between self-incrimination and additional prison time, it seems to me that the level of compulsion is the same wheth*315er the prison time was a presumptive part of his sentence or not.
Numerous federal circuit courts have considered this issue and held that extension of a supervised release date for failure to participate in treatment is not compulsion. See Ainsworth, 317 F.3d at 2; Searcy, 299 F.3d at 1227; Asherman v. Meachum, 957 F.2d 978, 980 (2d Cir.1992).8 While the majority cites cases that have held differently, I find that the Supreme Court has not spoken clearly on this issue, nor is there a national consensus that would compel us to overturn Morrow. I would hold that extending an inmate’s supervised release date because of his failure to participate in a sex offender treatment program does not rise to the level of compulsion necessary to violate the inmate’s Fifth Amendment privilege against self-incrimination.
. As the majority correctly notes, the McKune plurality's recital of the facts in Murphy is not entirely accurate. The point remains, nevertheless, that the plurality was operating under the assumption from its reading of Murphy that the threat of additional prison time did not amount to compulsion.
. Although Asherman arose before McKune, it has never been overruled.