dissenting.
[¶ 46] I respectfully dissent. On this record, civil commitment is nothing more than a mechanism for general deterrence.
I
[¶ 47] To involuntarily commit an individual under our statute, there must be a finding the individual has engaged in sexually predatory conduct and is diagnosable with a personality, sexual, or mental disorder that makes the individual “likely to engage in further acts of sexually predatory conduct which constitute a danger to the physical or mental health or safety of others.” N.D.C.C. § 25-03.3-01(8). This Court has said “likely to engage in further acts of sexually predatory conduct” means the individual’s “propensity towards sexual violence is of such a degree as to pose a threat to others.” Matter of G.R.H., 2006 ND 56, ¶ 16, 711 N.W.2d 587. Our statute must be read in light of the due process requirements enunciated in Kansas v. Crane, 534 U.S. 407, 122 S.Ct. 867, 151 L.Ed.2d 856 (2002). G.R.H., 2006 ND 56, ¶ 18, 711 N.W.2d 587. The individual must have a serious difficulty in controlling behavior. Crane, 534 U.S. at 413, 122 S.Ct. 867; see also G.R.H., at ¶ 18. To justify commitment, this difficulty must be severe enough to distinguish a sex offender from other dangerous, but typical, recidivists. Crane, 534 U.S. at 412-13, 122 S.Ct. 867; see also G.R.H., at ¶ 18. Otherwise, civil commitment quickly becomes a “ ‘mechanism for retribution or general deterrence’ — functions properly those of criminal law, not civil commitment.” Crane, 534 U.S. at 412, 122 S.Ct. 867 (citing Kansas v. Hendricks, 521 U.S. 346, 372-73,117 S.Ct. 2072, 138 L.Ed.2d 501 (1997) (Kennedy, J., concurring)). Therefore, before an individual can be civilly committed as a sexually dangerous individual, there must be a showing: (1) of prior sexually predatory conduct; (2) of some type of congenital or acquired condition manifested by a personality, sexual, or mental disorder; (3) that the condition makes the individual “likely to engage in further acts of sexually predatory conduct” meaning the individual’s propensity towards sexual violence is of such a degree as to pose a threat to others; and (4) that the individual must have a serious difficulty in controlling behavior.
[¶ 48] Other jurisdictions interpreting Crane have concluded merely being diagnosed with antisocial personality disorder is insufficient to civilly commit an individual. As I wrote in G.R.H., a diagnosis of antisocial personality disorder is not uncommon in individuals who have been in prison. 2006 ND 56, ¶ 40, 711 N.W.2d 587 (Kapsner, J., dissenting). “Approximately 40-60% of the male prison population are diagnosable with antisocial personality disorder.” Id. (citing Crane, 534 U.S. at 412, 122 S.Ct. 867; Moran, The Epidemiology of Antisocial Personality Disorder, 24 Social Psychiatry & Psychiatric Epidemiology 231, 234 (1999)). In G.R.H., Dr. Etherington testified that those percentages were actually much higher. G.R.H., at ¶ 40 (Kapsner, J., dissenting). In G.R.H., Dr. Etherington “testified that approximately 60%-75% of incarcerated individuals suffered from antisocial personality disorder.” Id. (Kapsner, J., dissenting). One jurisdiction held:
Under United States Supreme Court case law, a state cannot constitutionally confine a person solely based on antiso*584cial behavior. In order to civilly commit an individual, there must be at least clear and convincing evidence that the individual is “mentally ill” and “dangerous.”
In re Doe, 102 Hawai'i 528, 78 P.3d 341, 361-62 (Ct.App.2003); accord Foucha v. Louisiana, 504 U.S. 71, 82, 112 S.Ct. 1780, 118 L.Ed.2d 437 (1992) (disapproving of civil commitments based solely on antisocial personality diagnosis). It is imperative to note Anderson was not diagnosed with any sexual disorder. His only disorder for purposes of this commitment is having an antisocial personality.
[¶49] The majority, at ¶22, acknowledges the Crane requirements, but fails to adequately explain how the record establishes Anderson’s antisocial personality disorder distinguishes him from other dangerous, but typical, recidivists. Instead, the majority relies on a cursory conclusion derived from Dr. Etherington’s analysis of the PCL-R 2nd, which supports “a belief that Mr. Anderson has a more severe antisocial personality than the general prison population.” Majority, at ¶ 41 (emphasis omitted). The majority also summarily concludes: “In the expert opinion of three doctors, Anderson is not the typical recidivist in the ordinary criminal case.” Majority, at ¶42. The majority’s statement is unsupported by the record. Dr. Ethering-ton’s position, that Anderson’s disorder is more severe than the general prison population, cannot be contorted to mean Anderson’s behavioral issues are somehow more dangerous than other dangerous, but typical, recidivists. Dr. Belanger concludes Anderson has a severe antisocial personality disorder, and that his disorder renders him likely to engage in additional acts of sexually predatory conduct. Dr. Gulkin concludes Anderson’s “general sexual adjustment is problematic and his poor judgement/impulse control puts him at risk for further acts of sexually predatory behavior in unsupervised situations.”
[¶ 50] Dr. Etherington’s belief that Anderson’s disorder is more severe than the general prison population is not the standard pronounced in Crane, 534 U.S. at 413, 122 S.Ct. 867, or in G.R.H., 2006 ND 56, ¶ 18, 711 N.W.2d 587. There must be more than an acknowledgment that Anderson has a more severe antisocial personality than the general prison population. See Crane, at 413, 122 S.Ct. 867; G.R.H., at ¶ 18. Without this standard, nothing prevents the State from using civil commitment as a “ ‘mechanism for retribution or general deterrence,’ ” which is constitutionally impermissible. Crane, 534 U.S. at 412, 122 S.Ct. 867.
[¶ 51] Meeting the Crane standard is questionable given Anderson’s criminal history. Anderson’s first sexually-related criminal charge, occurring when he was seventeen, involved violence and force, although it was sentenced as a class B misdemeanor. His only other conviction for sexual conduct was a conviction of a class A misdemeanor, corruption of a minor, for engaging in sex with his sixteen-year-old girlfriend when he was twenty years old. It is acknowledged that this sex was consensual and criminal only because of the age of the female. The majority glosses over this fact and states that the later conviction “did not figure greatly into the psychological evaluations,” an astonishing statement in light of the Crane criteria. While clearly criminal conduct under our law, and one for which he has been appropriately criminally punished, this is not the type of crime that one ordinarily thinks of when contemplating the civil commitment of sexual predators. A twenty-year-old man having consensual sex with a sixteen-year-old girl may be criminal by virtue of her age, but in the absence of such a *585statute, could not even be considered an unusual occurrence.
[¶ 52] All three experts agreed Anderson has not been diagnosed with any sexual disorder. The trial court found: “Paraphilia is arousal through socially unacceptable sexual practices such as pedophilia or sadism and Anderson does not display abnormal sexual arousal patterns that rise to the level of a paraphilia.” However, Anderson has an antisocial personality disorder characterized by impul-sivity and irritability. Therefore, the logic goes, his next criminal act might include sexual misconduct. Anderson does have trouble with self-control as evidenced by his past criminal history and by the numerous altercations while incarcerated. This seems to be the bottom line of the evidence supporting his commitment, not that he is a sexually dangerous individual but that he is generally unable to control his behavior. As a result, he is being confined for an indefinite period of time. If his criminal history and his general behavior in the prison system are sufficient to demonstrate that it is likely he will fail again to conform to social norms, it is unnecessary to label him a “sexual predator.” Our criminal justice system is prepared to deal with the repeat offender by increasingly severe criminal sanctions. What our constitutional framework is not prepared to accept, however, is that we may civilly commit Anderson in anticipation that he will probably commit such acts and speciously call him a “sexual predator” in order to do so.
[¶ 53] I do not believe, and the record does not adequately reflect, that Anderson is more dangerous than the dangerous, but typical, recidivist. The commitment here amounts to nothing more than general deterrence.
II
[¶ 54] I reiterate my concern over the uncritical use of diagnostic tools and assessment tests on sex offenders. See Interest of P.F., 2006 ND 82, ¶¶ 26-30, 712 N.W.2d 610 (Kapsner, J., concurring). These tests and assessments are not substitutes for judicial decision-making or review. Id. at ¶ 29. Instead, their proper function in the civil commitment proceedings should be considered one factor to be weighed in determining whether an individual is likely to engage in further acts of sexually predatory conduct.
[¶ 55] In civil commitments, courts are asked to strike a delicate balance between protecting society from sexual violence and depriving an individual of his constitutionally guaranteed freedoms. This difficult position was noted by Eric Janus and Paul Meehl, who said:
If predictions about future violence are too optimistic, sexual violence may result. Unduly pessimistic predictions result in unnecessary, prolonged deprivations of liberty. In addition, sex offender commitments entail treatment that is expensive and intrusive, while sexual violence is exceedingly destructive. Thus, both types of prediction errors are costly in many ways.
Eric S. Janus & Paul E. Meehl, Assessing the Legal Standard for Predictions of Dangerousness in Sex Offender Commitment Proceedings, 3 Psychol. Pub. Pol’y & L. 33, 35 (1997) (footnotes omitted). Those concerns have been echoed through numerous articles. See, e.g., Eric S. Janus, Closing Pandora’s Box: Sexual Predators and the Politics of Sexual Violence, 34 Seton Hall L.Rev. 1233 (2004) (discussing constitutional implications of sexual predator commitment laws); John Q. La Fond, The Future of Involuntary Civil Commitment in the U.S.A. after Kansas v. Hendricks, 18 Behav. Sci. & L. 153,165 (2000) *586(noting sexual offender civil commitment statutes might bring about “the collapse of substantive due process” and lead state legislatures down a “slippery slope,” where the “next ‘crime of the month’ ” might justify civil commitment).
[¶ 56] To assist clinicians with making the “dangerousness” prediction, researchers have created actuarial risk assessments to help predict recidivism. These actuarial risk assessments are statistical models, which involve documenting “the existence of a statistical correlation between the presence of certain ‘risk factors’ and a particular outcome.” Fred S. Berlin et al., The Use of Actuarlais at Civil Commitment Hearings to Predict the Likelihood of Future Sexual Violence, 15 Sexual Abuse: J. Res. & Treatment 377, 377 (2003). In this case, as the majority discussed, psychometric “ ‘risk assessment inventories’ or RAIs were also used to determine the probability or actuarial risk that Anderson will engage in future sexually predatory conduct!.]” Majority, at ¶ 35. Specifically, three actuarial RAIs were performed: (1) the Rapid Risk Assessment for Sexual Offender Recidivism (“RRASOR”); (2) the Static-99; and (3) the Minnesota Sex Offender Screening Tool-Revised (“MnSOST-R”). A psychologist also performed the Psychotherapy Checklist-Revised (“PCL-R 2nd”) on Anderson, which attempts to measure “psychopathy,” which it defines as a combination of interpersonal patterns, emotional characteristics, and behavioral tendencies. See Majority, at ¶ 35. I am concerned the majority’s opinion might erroneously be read to imply these assessments can replace independent judicial decision-making and review. RAIs cannot and should not be used in that fashion.
[¶ 57] There is an ongoing debate in academic circles about the use of actuarial tests such as the RRASOR, Static-99, and MnSOST-R in civil commitment proceedings of sexually dangerous individuals. See, e.g., Eric S. Janus & Robert A. Prentky, Forensic Use of Actuarial Risk Assessment with Sex Offenders: Accuracy, Admissibility and Accountability, 40 Am.Crim. L.Rev. 1443 (2003); see also Niklas L&ngstrbm, Accuracy of Actuarial Procedures for Assessment of Sexual Offender Recidivism Risk May Vary Across Ethnicity, 16 Sexual Abuse: J. Res. & Treatment 107 (2004); Richard Wollert, Additional Flaws in the Minnesota Sex Offender Screening Tool — Revised: A Response to Doren and Dow (2002), Vol. 2, No. 4 J. Threat Assessment 65 (2003); Dennis M. Doren & Edward A. Dow, What “Shrinkage” of the MnSOST-R? A Response to Wollert (2002b), Vol. 2, No. 4 J. Threat Assessment 49 (2003); Richard Wollert, The Importance of Cross-Validation in Actuarial Test Construction: Shrinkage in the Risk Estimates for the Minnesota Sex Offender Screening Tool— Revised, Vol. 2, No. 1 J. Threat Assessment 87 (2002); Richard Wollert, An Analysis of the Argument That Clinicians Under-Predict Sexual Violence in Civil Commitment Cases, 19 Behav. Sci. & L. 171 (2001); Dennis M. Doren, Recidivism Base Rates, Predictions of Sex Offender Recidivism, and the “Sexual Predator” Commitment Laws, 16 Behav. Sci. & L. 97 (1998). Before discussing the proper use of actuarial assessments in civil commitments for sexually dangerous individuals, it is necessary to discuss what each RAI is and what each was designed to predict.
A
[¶ 58] The RRASOR was developed in 1997 by R. Karl Hanson, a researcher for the Department of the Solicitor General of Canada. The RRASOR was designed to “predict sex offense recidivism using a *587small number of easily scored variables.” R. Karl Hanson & David Thornton, Improving Risk Assessments for Sex Offenders: A Comparison of Three Actuarial Scales, 24 Law & Hum. Behav. 119, 120 (2000). The RRASOR measures four static variables that, according to Hanson and Thornton, correlated well to observed rates of recidivism. Id. The RRASOR factors include: (1) the subject’s prior sex offenses, (2) whether there were any unrelated victims, (3) whether any victims were male, and (4) whether the subject was less than 25 years old. See id. at 121.
[¶ 59] Due to the relative ease in scoring, the RRASOR has been widely used as a screening tool. Subsequent research by the RRASOR’s author, Hanson, and another researcher, Monique Bussiére, indicated the RRASOR can be accurate in predicting future sexual violence. See R. Karl Hanson & Monique T. Bussiére, Predicting Relapse: A Meta-Analysis of Sexual Offender Recidivism Studies, 66 J. Consulting & Clinical Psychol. 348 (1998). Further research by other psychologists has validated the RRASOR. See Berlin et al., supra, 15 Sexual Abuse: J. Res. & Treatment at 378 (discussing earlier research on the RRASOR). Other researchers have warned the RRASOR, like all actuarial assessments, is based on certain assumptions, and depending on the validity of those assumptions, the likelihood that a particular offender will commit future sexual violence may or may not be correct. Id.; see also Doren, supra, 16 Behav. Sci. & L. 97. Some have concluded the RRA-SOR has little practical import in the civil commitment context. Berlin et al., supra, 15 Sexual Abuse: J. Res. & Treatment at 378. According to Berlin, researchers developing the MnSOST-R found that “[n]inety-eight percent of what had determined whether or not persons were likely to recidivate [had] to do with factors other than those that had been assessed by the RRASOR_ Thus, although the RRA-SOR achieved statistical significance, it may not have had much practical significance.” Id. at 379.
B
[¶ 60] The Static-99, the most recent actuarial development, is a ten-item assessment developed by R. Karl Hanson and David Thornton. It rates an individual based on (1) the number of prior sex offenses; (2) the number of prior sentencing dates; (3) any convictions for non-contact sex offenses; (4) index case nonsexual violence; (5) prior nonsexual violence; (6) any unrelated victims; (7) any stranger victims; (8) any male victims; (9) current age of the offender; and (10) whether the individual is single. Berlin et al., supra, 15 Sexual Abuse: J. Res. & Treatment at 380. Since Hanson also developed the RRA-SOR, all four RRASOR factors are subsumed into the Static-99. See id. A “score of 6 or more is considered to be reflective of a ‘high risk of sexual recidivism.’ ” Id. The statistical tables reporting Hanson and Thornton’s follow-up data indicated the higher an individual’s Static-99 score, the higher percentage of individuals who eventually recidivated. See id. Therefore, the Static-99, like most actuarial risk assessments, can work effectively as an initial “screening tool that can identify a group of individuals ..., who when considered as a group, are at heightened risk [to recidivate].” Id. When offenders are considered individually, however, “the Static-99 cannot do much better than a coin flip.” Id. at 381.
C
[¶ 61] The MnSOST-R was developed for the Minnesota Department of Corrections to predict sexual recidivism in rapists and extrafamilial child molesters. Hanson & Thornton, supra, 24 Law & Hum. Be-*588hav. at 120. It measures sixteen variables, “addressing sexual and nonsexual criminal history, the victim’s age and relationship to the offender, substance abuse, unstable employment, age, and treatment history.” Id. The MnSOST-R considers both static and dynamic variables for recidivism. From the sixteen variables, the individual receives a “dangerousness” score from 1 to 10. “In practice, a MNSOST-R score of 8 or higher is often considered suggestive of a ‘high risk’ of recidivism.” Berlin et al., supra, 15 Sexual Abuse: J. Res. & Treatment at 380.
[¶ 62] There has been a significant academic discussion about the likelihood of recidivism based on the MnSOST-R. Specifically, “the MnSOST-R ... has not yet been fully cross-validated.” Terence W. Campbell, Assessing Sex Offenders: Problems and Pitfalls 98 (2004). Some researchers have noted the assessment’s results have been “greatly inflated,” see Berlin et al., supra, 15 Sexual Abuse: J. Res. & Treatment at 380 (citing Wollert, supra, Vol. 2, No. 1 J. Threat Assessment 87), while others support the predictive qualities of the MnSOST-R because it attempted to narrow its predictive accuracy. See Doren & Dow, supra, Vol. 2, No. 4 J. Threat Assessment 49. One researcher suggested “that use of the MnSOST-R be suspended at least until its performance estimates can be confirmed by proper cross-validation studies, in large samples, with naturally occurring and ‘middle-of-the-road’ recidivism rates.” Campbell, supra, at 104-05 (quotations omitted). Even MnSOST-R supporters noted:
Further research is needed before the extent to which the MnSOST-R generalizes to diverse samples is known. Both the MnSOST and MnSOST-R were constructed from preestablished groups of recidivists and nonrecidivists, which makes it difficult to directly translate the scores into recidivism rates.
Id. at 105 (quoting R. Karl Hanson, Who is Dangerous and When are They Safe: Risk Assessment with Sexual Offenders, printed in Protecting Society from Sexually Dangerous Offenders 69 (B.J. Winick & John Q. La Fond, eds., 2003)).
D
[¶ 63] The PCL-R 2nd is a supplementary assessment tool to measure an individual’s “psychopathy.” The original PCL-R was developed by Robert Hare in 1985. Campbell, supra, at 134. The PCL-R 2nd, developed in 2003, is “essentially the same instrument except for some minor changes.” Id. at 135. Conventional scoring of the PCL-R [and PCL-R 2nd] requires specialized training, file review, and a lengthy guided interview. Howard E. Barbaree et al., Evaluating the Predictive Accuracy of Six Risk Assessment Instruments for Adult Sex Offenders, 28 Crim. Just. & Behav. 490, 514 (2001) (discussing scoring of the PCL-R). Data used in this tool is obtained via an interview and review of the relevant file data. Campbell, supra, at 135. In developing the PCL-R 2nd, Hare commented “[a]mong offenders the PCL-R tends to be strongly associated with general and violent criminality, but only weakly or inconsistently with the number of convictions or charges for sexual offenses.” Robert D. Hare, The Hare Psychopathy Checklist-Revised-^nd Edition: Manual 138 (2003). Relying on Hare’s comments and independent research conducted on the PCL-R, some commentators have indicated the PCL-R or the PCL-R 2nd does “not successfully predict sexual recidivism.” Campbell, supra, at 137; Barbaree et al, supra, 28 Crim. Just. & Behav. at 514.
Ill
[¶ 64] As the researchers studying and developing these RAIs have noted, each of *589the tools has limitations. These actuarial assessments rely on aggregate or group data. See Janus & Prentky, supra, 40 Am.Crim. L.Rev. at 1476. Which means they “tell[] us the empirically measured rate of recidivism among a group of sex offenders who share a set of characteristics with the subject of the evaluation.” Id. The RAIs do not predict the likelihood of individual recidivism and thus tell us nothing about the likelihood that a particular person who shares the set of characteristics will reoffend. See id. As Justice Coyne of the Minnesota Supreme Court succinctly stated:
Not only are the statistics concerning the violent behavior of others irrelevant, but it seems to me wrong to confíne any person on the basis not of that person’s own prior conduct but on the basis of statistical evidence regarding the behavior of other people.
Matter of Linehan, 518 N.W.2d 609, 616 (Minn.1994) (Coyne, J., dissenting). Because the actuarial assessments rely solely on aggregate data, they have no probative value in determining whether a particular individual is likely to reoffend.
[¶ 65] Our civil commitment statute requires a determination that the individual’s propensity towards sexual violence is of such a degree as to pose a threat to others. This is, in part, why the RAIs have become so important as a starting point in the analysis of the commitment procedures for persons who have committed a sexual offense. They are a fairly mechanical tool that will identify a group of convicted persons who should be evaluated for consideration to be civilly committed. N.D.C.C. § 25-03.3-03.1. However, even when it is thoroughly understood that the RAIs do not assess the likelihood of the individual to reoffend, rather the likelihood of a group of people possessing similar characteristics to reoffend, the tools do not assess the imminence of the statistical reof-fense in any meaningful way.
[¶ 66] In Anderson’s case, the mental health professionals do not appear to be concerned with imminence whatsoever. Dr. Belanger testified:
My job, as I understand it, is to make an estimation of the risk of one or more acts of recidivist sexually predatory conduct over a lifetime. Not to make an estimation of apprehension or conviction but to make an estimation of the risk of one or more act of recidivist sexual predatory conduct over a lifetime and I’m to do this on the basis of my training and experience. This is good, strong, nomothetic evidence to the fact that when I say likely to engage that’s got numbers behind it.
(Emphasis added). It is a judicial function, not a psychological function, to decide whether measuring a statistical likelihood to sexually reoffend over a lifetime is evidence, clear and convincing, to justify the deprivation of an individual’s liberty.1 Such evidence does not exist in this record.
IV
[¶ 67] Each of the tests discussed in Part III indicated Anderson exhibited some likelihood to reoffend. Majority, at ¶ 37 (indicating that under the RRASOR, Anderson had a 14.2% likelihood of being *590reconvicted for a new sexual offense within 5 years and about a 21% probability that he would engage in recidivist acts in the next 10 years); Majority, at ¶ 39 (stating that under the MnSOST-R, Anderson’s score correlates to a 78% likelihood of committing another chargeable sexual offense in the next 6 years); Majority, at ¶ 40 (indicating that under the Static-99, Anderson had a 38-40% probability of recidivism over the next 15 years); and Majority, at ¶ 41 (noting Anderson’s PCL-R score was well above the threshold for finding a high degree of psychopathy, which supports a belief that Anderson’s antisocial personality is more severe than others in the general prison population). These actuarial tests were never designed to predict individual recidivism. Janus & Prentky, supra, 40 Am.Crim. L.Rev. at 1476-77. Although Anderson shares some characteristics of a group of people who have a likelihood to reoffend, importing the results to Anderson individually is seriously misleading.
[¶ 68] Notwithstanding Anderson’s scores on the various diagnostic tools and actuarial assessments, we have previously made clear that we will not engage in a “contest over percentage points” when it comes to determining whether an individual meets the requirements for civil commitment. Interest of 2002 ND 25, ¶ 18, 639 N.W.2d 473. Instead, we require a thorough examination by experts to make the initial recommendation of whether an individual poses a threat to society. Id. A certain test score on the RRASOR or Static-99 does not make an individual automatically committable. If we were to accept such logic, the judiciary would be without purpose. The court has the ultimate decision to determine whether the State has met its burden of producing clear and convincing evidence sufficient for commitment. A psychological test cannot act as a substitute for independent judicial review.
V
[¶ 69] Aside from depriving an individual of his life, taking his liberty is the most intrusive invasion of the government into a person’s life. Civil commitment is such an invasion. And it is an invasion for an indefinite period of time. Based upon this record, I cannot distinguish Anderson’s commitment from a mechanism for general deterrence which, Kansas v. Crane and Kansas v. Hendricks advise, is properly the function of the criminal, not the civil, law. Therefore, I dissent.
[¶ 70] Carol Ronning Kapsner
. One commentator has stated we should not be devoting extraordinary resources to the ordinarily dangerous, and we should not be civilly committing individuals based on this speculative assessment of non-imminent occurrences. However, that is a legislative policy matter, not a judicial matter. See Eric S. Janus, Minnesota's Sex Offender Commitment Program: Would an Empirically-Based Prevention Policy Be More Effective? 29 Wm. Mitchell L.Rev. 1083, 1101 (2003) (noting that commitments in Minnesota cost about $20 million per annum, and will increase more than three-fold by 2010).