KEITH, Chief Justice.

OPINION

In this case we again consider the Minnesota Psychopathic Personality Commitment Act, Minn.Stat. §§ 526.09 and 526.10 (1992). Recently we focused on the constitutionality of the statute and found that it did not violate the substantive due process and equal protection guarantees of the federal and state constitutions. In re Blodgett, 510 N.W.2d 910 (Minn. 1994). We now address the reach and application of the statute.

Appellant Dennis Darol Linehan, age 53, served nearly 20 years in prison for a kidnapping conviction arising out of the abduction and murder of a 14 year old babysitter in June of 1965. Shortly before appellant's scheduled release date in 1992, Ramsey County petitioned for his commitment as a psychopathic personality and as a chemically dependent person. After the initial commitment hearing, the trial court ordered appellant provisionally committed to the Minnesota Security Hospital (MSH). Upon receipt of the 60-day report from MSH and a second commitment hearing, the trial court found appellant to be a psychopathic personality and ordered that he be committed to MSH indefinitely. The petition for appellant's commitment as a chemically dependent person was denied. The court of appeals upheld appellant's commitment as a psychopathic personality. Matter of Linehan, 503 N.W.2d 142 (Minn.App. 1993). We granted appellant's petition for further review.

In State ex rel. Pearson v. Probate Court of Ramsey County, 205 Minn. 545, 287 N.W. 297 (1939), aff'd 309 U.S. 270, 60 S.Ct. 523, 84 L.Ed. 744 (1940), this court narrowed the reach of the statutory definition of psychopathic personality to apply only to "those persons who, by a habitual course of misconduct in sexual matters, have evidenced an utter lack of power to control their sexual impulses and who, as a result, are likely to attack or otherwise inflict injury, loss, pain or other evil on the objects of their uncontrolled and uncontrollable desire." Id. at 555, 287 N.W. at 302. In Blodgett we determined that we should follow the Pearson definition and reverse cases that misapplied the statute as narrowed by that case. 510 N.W.2d at 915.

As we noted in Blodgett, the burden is on the state to prove by clear and convincing evidence, each of the three elements set out in Pearson. See 510 N.W.2d at 913-15. Appellant argues that the record fails to support, by clear and convincing evidence, the utter lack of control/uncontrollable element and the prediction of harm element of the Pearson test. We agree and reverse the lower courts.

Appellant has a history of involvement in the criminal justice system starting with bicycle theft at age 13. As a teenager appellant was placed in the State Training School at Red Wing and the National Training School in Washington, D.C. for various violations including taking indecent liberties with a 4 year old girl. On July 18, 1963 when appellant was 22 years old, he and two male acquaintances met 19 year old L.H. at a restaurant and invited her to a party. After L.H. got into the car with the three men, they drove to a field where appellant and one of the other men raped her. Criminal charges were brought and dismissed.

On the evening of June 10, 1965, appellant "window peeped" at the Shoreview, Minnesota residence where 14 year old B.I. was babysitting. Appellant loosened the exterior light to darken the entrance, knocked on the door, and told B.I. that someone in a nearby car wanted to talk to her. When B.I. opened the door, appellant took her from the house and placed her in his car with the intention of sexually molesting her. B.I. began screaming, hitting, and clawing at appellant. In an effort to quiet her, appellant hit B.I. several times, grabbed her by the throat, and choked her to death. Appellant hid B.I.'s body in a remote wooded area. He then moved the body to a dry well on the property of relatives. Later he hid the body in a shallow roadside grave.

In the latter part of July, the police investigation focused on appellant. Appellant pled guilty to kidnapping B.I. and the murder charges against him were dropped. He was sentenced to the maximum term of 40 years. Since B.I.'s murder, appellant has told numerous versions of the events surrounding her death. He, however, has never denied that he caused B.I.'s death or that he buried her body.

After B.I.'s death but before he was arrested for her murder, appellant committed two sexual assaults. While at a party in July of 1965, appellant grabbed W.L., age 22, forced her into a bedroom, and raped her. This incident was not reported to the police. A second assault occurred on July 15, 1965 when appellant went to a friend's house and got into bed with the friend's sisters, age 11 and 12. Appellant told the girls he had a knife and would stab them if they screamed. He licked the genital area of the older girl, M.J. When the younger sister ran out of the room, appellant went after her and then left the house. Appellant was in custody for this offense at the time he was implicated in B.I.'s murder.

Appellant escaped from prison on June 20, 1975. Eleven days later he was arrested in Niles, Michigan, for sexually assaulting 12 year old T.L. T.L. was hitchhiking with two friends when appellant picked them up and drove them to the beach. When T.L. left the beach alone, appellant pursued her. T.L. refused to get in appellant's car and ran away. Later, however, appellant found T.L., pushed her down an embankment, and jumped on top of her. Appellant told T.L. he wanted to have oral sex and intercourse with her and warned that if she screamed he would kill her. Appellant then put down his knife and tried to remove T.L.'s pants. The assault was interrupted when some people arrived on the scene. Appellant fled, but later was apprehended and convicted of attempted rape.

Appellant was imprisoned in Michigan from July 1, 1975 through September 26, 1980 at which time he was returned to the Minnesota State Prison in Stillwater. In October 1981, appellant was transferred to a prison in North Dakota for protective custody reasons. While there, he violated prison rules by buying merchandise, selling it to inmates, and charging interest. In October 1983 he was transferred back to Minnesota.

While in North Dakota, appellant started a sexual treatment program, but dropped out before completing it. Although he started the program a second time, he was transferred back to Minnesota before he could finish. In 1988, appellant completed the Atlantis Inpatient Program for Chemical Dependency. He also participated in and completed the inpatient Transitional Sex Offender Program (TSOP) at Lino Lakes in May 1993 and was scheduled to enter the outpatient phase upon his release from prison. Appellant is on parole until August 21, 1997. He was paroled to MSH until his commitment "is resolved."

At the first commitment hearing, L.H., W.L., M.J., T.L., B.I.'s mother, and appellant's ex-wife testified about the assaults that took place in 1963, 1965, and 1975. Most of the testimony, however, came from four experts — Dr. Richard Friberg, Dr. Hector Zeller, Dr. Nancy Steele, and Dr. John Austin.

Dr. Friberg, a licensed consulting psychologist, interviewed appellant extensively on two occasions and briefly on two or three others. He also reviewed appellant's psychological test results and all available records at Lino Lakes. At the outset of his testimony, Dr. Friberg explained that listening to the lay witnesses "tipped [him] in the direction of saying" appellant met the "statutory definition" of psychopathic personality.

Dr. Friberg diagnosed appellant as an anti-social personality, a disorder unlikely to change. Although Dr. Friberg testified that when appellant is under the influence of alcohol he is "extremely impulsive, self defeating and prone to making major errors in judgment," Dr. Friberg described appellant as a "stable, intact and fairly controlled personality." Dr. Friberg testified that appellant's behavior resulting in B.I.'s death "reflected a much more planful, controlled, goal directed kind of a situation than [appellant] would like us to believe." Dr. Friberg felt that appellant "was trying to give us the impression that things kind of went haywire * * * when, in fact, * * * it was a fairly planful situation."

Dr. Friberg was never asked for his opinion as to whether appellant met the Pearson criteria and there is no evidence that he was familiar with Pearson's narrowing construction of Minn.Stat. § 526.09. He acknowledged that the standard he used to identify appellant as a psychopathic personality was "whether or not Mr. Linehan is a psychopathic personality within the definition of the Minnesota statute." Moreover, many of the county attorney's questions and Dr. Friberg's answers mirrored the statutory language — "impulsiveness of behavior," "irresponsible," "dangerous," and "appreciates the consequences of his acts."

Dr. Zeller, the trial court's first appointed examiner, also based his opinion that appellant was a psychopathic personality on the statute, not Pearson. When the county attorney quoted the statutory definition and asked if appellant met the criteria, Dr. Zeller stated "I feel that almost each one of the conditions [set forth in the statutory language] applies to Mr. Linehan." Dr. Zeller admitted he had not read Pearson and was not aware that the Minnesota Supreme Court had narrowed the definition of psychopathic personality in that case. He also admitted that he did not understand a key distinction in the Pearson test — the difference between uncontrolled desire and uncontrollable desire.

Dr. Zeller agreed with Dr. Friberg that appellant planned much of his criminal behavior, including his escape in 1975. Dr. Zeller believed that appellant has the ability to control and take responsibility for his actions. He noted that while in prison, appellant has not had any out-of-control behaviors with respect to sexual matters.

Dr. Nancy Steele, a licensed consulting psychologist with expertise in treating sex offenders and the supervisor of TSOP, testified that appellant had successfully completed TSOP's inpatient program and was ready to move to the community under supervision. Based on her contact with appellant and her review of his records, Dr. Steele believed appellant did not meet the definition of psychopathic personality at the time of the hearing. Dr. Steele agreed with Dr. Friberg and Dr. Zeller that appellant's sexual behavior prior to B.I.'s death was "very much planned." She further testified that she had "no evidence that [appellant] is currently sexually impulsive, dangerous, or out of control."

Dr. John V. Austin, a licensed consulting psychologist and the trial court's second appointed examiner, interviewed the appellant, reviewed the MSH records, and wrote an evaluation for the review hearing. Dr. Austin had read and was conversant with Pearson as well as the Psychopathic Personality Statute. He concluded that appellant does not pose a danger to others in the foreseeable future and did not support the petition for commitment because appellant had not shown recent behavior of loss of impulse control, emotional stability, or lack of good judgment.

The trial court accepted the testimony of Dr. Friberg and Dr. Zeller as clear and convincing and used it to provisionally commit appellant as a psychopathic personality. The court rejected Dr. Steele's testimony on the grounds that she did not have an adequate opportunity to review appellant's case and that she was not the primary participant in appellant's treatment program. Dr. Austin's testimony also was rejected. In both the findings of fact and conclusions of law, the trial court quoted the statutory language but not the Pearson criteria.

At the second commitment hearing, the trial court heard from Dr. Zeller, Dr. Austin, and Dr. Darel John Hulsing, appellant's treating psychiatrist at MSH. Dr. Zeller reaffirmed his opinion that while appellant continues to exercise poor judgment, he has the power to control his impulses. Dr. Hulsing testified that appellant has been appropriate with him and agreed that appellant had progressed from the evaluation unit to the treatment unit in as short a period of time as any other resident. Dr. Hulsing believed that appellant had a "high" chance of repeating offenses "in the absence of treatment control, supervision."

The 60-day report from MSH states that appellant "has made no significant change or progress in his behavior since his admission to the Security Hospital and that, therefore, the conditions of his commitment continue." The report also states that the treatment team "feels that it is not in a position to predict any future dangerous acts that Mr. Linehan may or may not engage in due to limited experience with him."

In its conclusions of law accompanying the order committing appellant indefinitely, the trial court again set out the statutory definition of psychopathic personality. However, although none of the expert witnesses was asked for an opinion as to whether appellant met the Pearson criteria, the court concluded that appellant's conduct "shows an utter lack of power and ability to control his sexual impulses" and that appellant "continues to be dangerous to others by sexual attack."

The reach of the Psychopathic Personality Statute is limited by Pearson. Under Pearson, there must be (a) a "habitual course of misconduct in sexual matters" and (b) "an utter lack of power to control * * * sexual impulses" so that (c) it is likely the person will "attack or otherwise inflict injury, loss, pain or other evil on the objects of their uncontrolled and uncontrollable desire." Minnesota ex rel. Pearson v. Probate Court of Ramsey County, 309 U.S. 270, 274, 60 S.Ct. 523, 526, 84 L.Ed. 744 (1940), aff'g, 205 Minn. 545, 287 N.W. 297 (1939). Appellant's criminal history demonstrates that he meets the habitual course of misconduct element of the Pearson test. The question before us is whether the record supports, by clear and convincing evidence, the trial court's conclusion that appellant meets the second and third elements. This is a question of law which we review de novo. Frost-Benco Elec. Ass'n v. Minnesota Pub. Util. Comm'n, 358 N.W.2d 639, 642 (Minn. 1984).

As noted above, the trial court relied on the testimony of Dr. Friberg and Dr. Zeller in reaching the conclusion that appellant is a psychopathic personality. This testimony, however, fails to support the trial court's finding that appellant exhibits an utter lack of control over his sexual impulses. First, the testimony was given in the context of the statute, not Pearson. Dr. Friberg was never asked about Pearson, he used terms from the statute, and he referred to the statute throughout his testimony. Dr. Zeller was asked about Pearson but had not heard of the case. Moreover, when the utter lack of control/uncontrollable element was explained to him, Dr. Zeller did not understand the distinction. Second, both experts were inconsistent on the key issue of whether appellant could control his behavior. Although Dr. Friberg testified that appellant was "extremely impulsive" when under the influence of alcohol, he described appellant as a "stable, intact and fairly controlled personality" and described appellant's behavior as "planful" and "controlled." Dr. Zeller also testified that appellant has the ability to control and take responsibility for his actions.

It also appears that appellant may have his alcohol abuse, the precursor of his assaults, under control. Appellant successfully completed a chemical dependency treatment program four years ago and the trial court in this proceeding failed to find enough evidence to commit appellant as a chemically dependent person.

In its order committing appellant to the MSH for an indefinite period of time, the trial court used the Pearson language. It is not enough, however, for the trial court to use this language in a conclusory fashion when the expert testimony upon which it relies has been given in terms of the statutory definition. Neither the testimony of Dr. Friberg and Dr. Zeller nor appellant's behavior while incarcerated supports the finding of uncontrollability. There is, therefore, no clear and convincing evidence that appellant has an utter lack of power to control his sexual impulses.

Appellant also argues that the record fails to support, by clear and convincing evidence, the prediction of harm element of the Pearson test. He notes that even the county's experts, whose testimony provided the basis for the trial court's findings, testified that it was difficult to predict future dangerous behavior. In light of this difficulty and the loss of liberty a committed person will experience, appellant urges this court to adopt specific scientific standards for the prediction of dangerousness to ensure that individuals committed as psychopathic personalities pose a serious danger to the public. Specifically, appellant urges the use of base rate studies "identifying particular characteristics of research subjects who subsequently engaged in violent behavior." See John Monahan, Predicting Violent Behavior: An Assessment of Clinical Techniques (1981). Appellant also urges the use of risk assessment statements that address the likelihood of an individual or a particular class engaging in certain acts as opposed to "dichotomous" statements that simply opine as to whether particular behavior will or will not occur in the future.

Because we hold that the county did not prove the utter lack of control/uncontrollable element of the Pearson test, it is unnecessary to address whether there is clear and convincing evidence that appellant was likely to engage in future dangerous behavior. Dangerousness in the context of the Psychopathic Personality Statute is predicated on an utter lack of ability to control sexual impulses. Pearson, 205 Minn. at 555, 287 N.W. at 302.

Where utter uncontrollability of sexual impulses is found, however, the trial court, in predicting serious danger to the public, should consider the following factors if such evidence is presented: (a) the person's relevant demographic characteristics ( e.g., age, education, etc.); (b) the person's history of violent behavior (paying particular attention to recency, severity, and frequency of violent acts); (c) the base rate statistics for violent behavior among individuals of this person's background ( e.g., data showing the rate at which rapists recidivate, the correlation between age and criminal sexual activity, etc.); (d) the sources of stress in the environment (cognitive and affective factors which indicate that the person may be predisposed to cope with stress in a violent or nonviolent manner); (e) the similarity of the present or future context to those contexts in which the person has used violence in the past; and (f) the person's record with respect to sex therapy programs. In reviewing psychopathic personality commitments in the future, we will look to see whether these factors have been considered, particularly where, as here, there is a large gap of time between the petition for commitment and the appellant's last sexual misconduct.

The state has failed to prove by clear and convincing evidence that appellant meets the Pearson standard as applied to the Minnesota Psychopathic Personality Commitment Act; the commitment of appellant as a psychopathic personality is reversed and vacated.

Reversed; commitment vacated.