Opinion
23-2359
08-01-2024
Darrin S. Rick Petitioner - Appellee v. Jodi Harpstead, Commissioner, Minnesota Department of Human Services Respondent - Appellant Eric Steven Janus; L. Maaike Helmus Amici on Behalf of Appellee
Submitted: February 13, 2024
Appeal from United States District Court for the District of Minnesota
Before SMITH, Chief Judge, BENTON and STRAS, Circuit Judges.
Judge Smith completed his term as chief judge of the circuit on March 10, 2024. See 28 U.S.C. § 45(a)(3)(A).
STRAS, Circuit Judge.
Does a claim of "actual innocence" relieve someone who is civilly committed from filing a federal habeas petition within one year? See 28 U.S.C. § 2244(d)(1) (one-year statute of limitations). The answer is no, so we reverse the district court's grant of habeas relief.
I.
In 1993, Darrin Rick pleaded guilty to criminal sexual conduct after abusing four developmentally disabled girls and one seven-year-old boy. See Minn. Stat. § 609.342, subd. 1a. He started sex-offender treatment and faith-based therapy programs while in prison but dropped out each time. At the end of his sentence, Hennepin County petitioned to civilly commit him. See id. § 253B.185, subd. 1 (2004) (describing the process for indefinitely committing a "sexually dangerous person[] . . . to a secure treatment facility").
Three psychologists examined him, two appointed by a Minnesota district court and one retained by Hennepin County. All three agreed that he satisfied the statutory criteria for commitment as a "sexually dangerous person." Id. § 253B.02, subd. 18c (2004). One key area of consensus was that he was "likely" to commit additional "acts of harmful sexual conduct." Id., subd. 18c(3); see In re Linehan, 557 N.W.2d 171, 180 (Minn. 1996) (construing "likely" to mean "highly likely"), vacated sub nom. Linehan v. Minnesota, 522 U.S. 1011 (1997); see also In re Linehan, 594 N.W.2d 867, 878 (Minn. 1999) (reaffirming the highly likely requirement on remand).
The court committed Rick to the Minnesota Sex Offender Program. See generally Hince v. O'Keefe, 632 N.W.2d 577, 579-80 (Minn. 2001) (explaining that sexually dangerous people "are committed for an indeterminate amount of time" in "Sex Offender Program facilities"). It found that he "ha[d] engaged in harmful sexual contact" and, as the experts had concluded, "was at a moderate risk of reoffending." Combined with his failure to "complet[e] sex[-]offender treatment," these findings led to the decision to civilly commit him. See In re Civil Commitment of Ince, 847 N.W.2d 13, 20 (Minn. 2014) (requiring clear and convincing evidence). In 2007, after a winding appeals process, the Minnesota Supreme Court declined further review.
Not much happened for the next dozen or so years until Rick asked a different forensic psychologist to review his case. In her lengthy report, she relied on recent studies to conclude that the actuarial tools used to justify his commitment overestimated the risk of recidivism. It turned out that, due to improvements in "external controls" and "support systems," sex offenders who were released from prison around the same time as Rick ended up reoffending far less often than predicted. Instead of the 25% risk that the tools estimated for him, the actual risk was somewhere around 6%. Not to mention that the experts had overemphasized his failure to complete treatment, which was "already accounted for in [the] score." (Emphasis omitted.)
Rick sent the report to two of the psychologists who had examined him years earlier. Both changed their minds. One explained that the "sexual recidivism risk data available in 2004 was considerably less sophisticated or discriminating than what is available today." Had the modern data been available at the time, he "would have opined that Mr. Rick did not meet the statutory criteria necessary for commitment." The second had a similar view: Rick's "commitment . . . was inappropriate in 2004" because, "based on current actuarial scoring, [he] ha[d] a low likelihood of sexual recidivism."
Armed with these new expert reports, Rick filed a federal habeas petition. See 28 U.S.C. § 2254; see also Duncan v. Walker, 533 U.S. 167, 176 (2001) ("[F]ederal habeas corpus review may be available to challenge the legality of a state court order of civil commitment . . . ."). Minnesota argued, however, that Rick's petition came nearly a decade late, well after the one-year statute of limitations had expired. See 28 U.S.C. § 2244(d)(1)(D).
The district court entertained the petition anyway under the actual-innocence exception, which provides a gateway for claims if "a constitutional violation has probably resulted in the conviction of [some]one who is actually innocent." Murray v. Carrier, 477 U.S. 478, 496 (1986). The underlying constitutional violation, in the court's view, was the reliance on the withdrawn expert reports and the now-discredited actuarial data, which together had rendered Rick's civil-commitment proceeding so unfair that it violated his due-process rights. The due-process violation did double duty: it both allowed him to avoid the statute of limitations and made him eligible for habeas relief. The court granted the relief he sought, and now Minnesota challenges each step of the ruling.
II.
Our analysis begins and ends with the statute of limitations. Multiple procedural doctrines and filing rules "promote federal-state comity" by heavily "circumscribing]" the availability of federal habeas review. Shinn v. Ramirez, 596 U.S. 366, 378 (2022). A gateway through some of those procedural barriers is the actual-innocence exception, which "allow[s] a prisoner to pursue his constitutional claims . . . on the merits." McQuiggin v. Perkins, 569 U.S. 383, 392 (2013). It applies only when new evidence makes it "more likely than not that no reasonable juror would have convicted [the petitioner]." Id. at 395 (alteration in original) (citation omitted). In those circumstances, "the principles of comity and finality . . . must yield" to remedy a "fundamentally unjust incarceration." Murray, 477 U.S. at 495 (citation omitted). Otherwise, according to the Supreme Court, there will have been a miscarriage of justice. See Schlup v. Delo, 513 U.S. 298, 315 (1995).
Since Murray, the Supreme Court has revisited the actual-innocence exception several times. Some general principles emerge. The first is that it provides a pathway to relief in two situations: when a prisoner confronts a procedural bar, like failing to "present[] [a] claim to [a] state court," Edwards v. Carpenter, 529 U.S. 446, 453 (2000) (emphasis removed); see House v. Bell, 547 U.S. 518, 521-22 (2006), or has filed a federal habeas petition beyond the one-year statute of limitations, see McQuiggin, 569 U.S. at 398; see also 28 U.S.C. § 2244(d)(1). The other one is that it is "narrow," "rare," and "extraordinary," Schlup, 513 U.S. at 315, 321 (citations omitted), in that it "applies" only when the innocent are" convicted," a "severely confined category," McQuiggin, 569 U.S. at 394-95 (emphasis added) (citation omitted).
Criminal guilt or innocence has always been the common denominator, regardless of the circumstances. See id. at 398 n.3 ("[W]hat is at stake is a State's incarceration of an individual for a crime . . . ." (emphasis added)); Sawyer v. Whitley, 505 U.S. 333, 345 (1992) (describing withheld evidence as going to a capital defendant's "guilt or innocence of the crime of first-degree murder and the aggravating circumstance[s]" (emphasis added)). No Supreme Court case suggests, much less holds, that it extends to other situations. And although some courts have contemplated whether it might, none has held that it does, at least in the civil-commitment context. See Schmidt v. McCulloch, 823 F.3d 1135, 1137 (7th Cir. 2016) (observing that the petitioner's claim was "close enough" to actual innocence that the court could "assume that the excuse applies in the civil[-]commitment context" (emphasis added)); Levine v. Torvik, 986 F.2d 1506, 1517 n.9 (6th Cir. 1993) (suggesting that the "rare" actual-innocence exception could be applied when a constitutional violation "result[s] in the confinement of one who is actually not mentally ill"), overruled in part on other grounds by Thompson v. Keohane, 516 U.S. 99, 111 (1995).
Rick recognizes that he cannot prevail unless we "extend" the actual-innocence exception to this new situation. It is not a task to undertake lightly. When it comes to judge-made rules, after all, we are supposed to "exercise restraint" and "add[] to or expand[] them only when necessary." Dretke v. Haley, 541 U.S. 386, 394 (2004) (discussing the actual-innocence exception). For several reasons, extending the actual-innocence exception to civil commitments is not "necessary," in part because of "the many threshold legal questions" it raises. Id. at 395.
One of those is what to do with the fact that Rick has already admitted his criminal guilt. He did so by pleading guilty in 1993. The only question now is whether, as he describes it, he is "innocent" of the risk of committing future sex crimes. See Sawyer, 505 U.S. at 339 (explaining that the actual-innocence exception "is concerned with actual as compared to legal innocence").
But the question is itself a hypothetical: had Rick not spent the last two decades in the Minnesota Sex Offender Program, would he likely have committed other sex crimes? Unlike other applications of the actual-innocence exception, it is a prediction-would he likely have or not-rather than a determination of an historic fact-did he or didn't he. And the evidence used to figure it out would be a collection of actuarial tools that, as Rick himself argues, have changed substantially over just the past 15 years. See Ince, 847 N.W.2d at 23 (recognizing that "dangerousness prediction methodology is complex and contested" and requires "a careful balancing of all the relevant facts" (citations omitted)). In short, if we accept Rick's argument, the exception would no longer be just about guilt or innocence. See Bousley v. United States, 523 U.S. 614, 623 (1998) ("'[A]ctual innocence' means factual innocence, not mere legal insufficiency." (emphasis added)).
Another problem is that the actuarial tools are only one factor among many. See Ince, 847 N.W.2d at 23-24. Predicting dangerousness requires consideration of more than just "base[-]rate statistics." In re Linehan, 518 N.W.2d 609, 614 (Minn. 1994). Other factors include "relevant demographic characteristics," a "person's history of violent behavior," "sources of stress in the environment," "the similarity of the present or future" circumstances to those surrounding past "violence," and compliance with "sex[-]therapy programs." Id. "[I]solat[ing] the most important factors" in this list and deciding "the weight to be attributed to each," Ince, 847 N.W.2d at 23-24 (citation omitted), is, as the magistrate judge put it, "fundamentally different and far more complex" than determining guilt or innocence in a run-of-the-mill criminal case.
Not to mention that hindsight bias becomes a risk. A faithful application of the exception requires a court to transport itself back in time and close its eyes to post-commitment conduct. See House, 547 U.S. at 537-38. Yet "the distorting effects of hindsight" are difficult to "eliminate" in the civil-commitment context, Strickland v. Washington, 466 U.S. 668, 689 (1984), because future conduct is the entire focus of the prediction.
Consider an example. Suppose that new evidence demonstrates an individual never should have been committed in the first place, perhaps because, like here, the initial prediction of future dangerousness was too high. According to Rick, the individual is "actually innocent." But suppose further that, while confined, the individual committed inappropriate sexual acts demonstrating dangerousness. Unlike the typical question in actual-innocence cases, which is did he do it or not, it is not clear how to handle this situation. After all, under Minnesota's definition of "sexually dangerous person," the subsequent conduct is relevant to whether the confined individual "is likely to engage in acts of harmful sexual conduct." Minn. Stat. § 253B.02, subd. 18c(3) (2004); see Presidio Enters., Inc. v. Warner Bros. Distrib. Corp., 784 F.2d 674, 680 (5th Cir. 1986) ("A statement about the future can be verified only in the future; but then, of course, it is no longer a statement about the future . . . ."). This simple example shows that there is no "analogous framework" in the civil-commitment context that is both "workable" and "objective." Sawyer, 505 U.S. at 341.
Presumably, that is why Minnesota's periodic-review process focuses on whether the confined individual is still dangerous. Questions like whether an individual "is capable of making an acceptable adjustment to open society, is no longer dangerous to the public, and is no longer in need of treatment and supervision" become the focus. Minn. Stat. § 253D.31 (emphases added); Karsjens v. Piper, 845 F.3d 394, 409-10 (8th Cir. 2017) (describing the "extensive process and the protections to persons committed" in Minnesota). Periodic review is surely no substitute for habeas, in part because it has a different objective. See Duncan, 533 U.S. at 176-77. But the "availability of [an]other remed[y]"-one that is available no matter how long an individual has been committed-is reason enough to "exercise restraint" in this area. Dretke, 541 U.S. at 394-95; see Minn. Stat. § 253D.27, subd. 2 (authorizing the filing of a discharge petition six months after a previous denial).
Finally, importing the actual-innocence exception into civil-commitment cases would invite an endless stream of challenges. See Shinn, 596 U.S. at 391 (discussing the problems posed by "[s]erial relitigation"). Each time science improves, as it did here, it would potentially open the door to a new habeas petition claiming actual innocence. An "extraordinary remedy" would turn into an all-too-ordinary one. Id. at 377 (citation omitted); see Embrey v. Hershberger, 131 F.3d 739, 741 (8th Cir. 1997) (en banc) (noting that "[i]t would be ironic indeed" if actual innocence were used to "expand[]" the habeas remedy rather than "constrict[] it"). And the casualties would be "finality, comity, and the orderly administration of justice." Shinn, 596 U.S. at 379 (citation omitted).
III.
We accordingly reverse the judgment of the district court and remand for the denial of Rick's petition.