Skip navigation

Oklahoma Law Review Article on Sexually Violent Predator Committment, 2014

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
Dangerous Diagnoses, Risky Assumptions,
and the Failed Experiment of “Sexually Violent Predator” Commitment
Deirdre M. Smith †

246 Deering Avenue
Portland, Maine 04102
(207) 780-4376

Draft: Please do not quote from or cite to this article without the author’s express permission. This paper is a
draft of an article that is forthcoming in Oklahoma Law Review, 67 Okla. L. Rev. __. It is available for download
at no cost at the author’s SSRN page: .


Professor of Law and Director of the Cumberland Legal Aid Clinic, University of Maine School of Law. I am
grateful to the following people who read earlier drafts of this article and provided many helpful insights: David
Cluchey, Malick Ghachem, Barbara Herrnstein Smith, and Jenny Roberts. I also appreciate the comments and
reactions of the participants in the University of Maine School of Law Faculty Workshop, February 2014, and the
participants in the Association of American Law Schools Section on Clinical Legal Education Works in Progress
Session, May 2014. I am appreciative of Dean Peter Pitegoff for providing summer research support and of the staff
of the Donald L. Garbrecht Law Library for its research assistance.

Electronic copy available at:

In the 1997 opinion, Kansas v. Hendricks, the U.S. Supreme Court upheld a law that
presented a new model of civil commitment. The targets of these new commitment laws were
dubbed “Sexually Violent Predators,” and the Court upheld this form of indefinite detention on
the assumption that there is a psychiatrically distinct class of individuals who, unlike typical
recidivists, have a mental condition that impairs their ability to refrain from violent sexual
behavior. And, more specifically, the Court assumed that the justice system could reliably
identify the true “predators,” those for whom this unusual and extraordinary deprivation of
liberty is appropriate and legitimate, with the aid of testimony from mental health professionals.
This Article evaluates the extent to which those assumptions were correct and concludes that
they were seriously flawed and, therefore, the due process rationale used to uphold the SVP laws
is invalid. The category of the “Sexually Violent Predator” is a political and moral construct,
not a medical classification. The implementation of the laws has resulted in dangerous
distortions of both psychiatric expertise and important legal principles, and such distortions
reveal an urgent need to re-examine the Supreme Court’s core rationale in upholding the SVP
commitment experiment.


Electronic copy available at:



In 1990, the state legislature of Washington, in response to calls for action after a highlypublicized violent sexual crime committed against a young child by an offender with prior
convictions for violence against children, enacted a statute to enable the state to continue to
detain sex offenders after they had completed their criminal sentences. The targets of these new
laws were dubbed “Sexually Violent Predators,” a label intended to connote a sub-class of sex
offenders who run a high risk of recidivism after their release due to the presence of a mental
abnormality or personality disorder. Soon thereafter, a few other states, including Kansas,
enacted their own commitment laws modeled closely on Washington’s. The first person
committed under Kansas’s law, Leroy Hendricks, challenged the constitutionality of his
indefinite detention under due process, ex post facto, and double jeopardy principles in a case
that reached the United States Supreme Court. In the 1997 opinion in Kansas v. Hendricks, 1 the
Court upheld this new model of commitment. In the wake of that case, other states (for a total of
20 to date) and the federal government enacted “Sexually Violent Predator” (SVP) laws as well.
Since 1990, several thousand people have been committed under such laws, the vast majority of
whom remain in indefinite detention.
The core rationale of the Hendricks opinion, as well as that of the follow-up opinion in
Kansas v. Crane, 2 is that indefinite preventive detention is consistent with substantive due
process principles where a mental disorder limits the committed individual’s ability to control his
behavior. Although a finding of such mental disorder is, consequently, a constitutional
prerequisite for these indefinite commitments, the Court also conferred broad discretion on
legislatures regarding how states could satisfy this requirement. The Court based its opinions
regarding SVP laws on the assumption that there is a medically distinct class of individuals who
are not “typical recidivists” but who have a mental condition that impairs their ability to refrain
from violent sexual behavior and for whom this unusual and extraordinary deprivation of liberty
is appropriate and legitimate. And more specifically, the Court assumed that the justice system
could reliably distinguish between the two groups and, with the aid of mental health
professionals, could identify the true “predators.”
In this Article, I evaluate the extent to which those assumptions were correct, both at the
time of the SVP laws’ enactment and as they have been implemented. First, I consider
psychiatry’s own views of the relationship of mental pathology to sexual violence and of the
field’s ability to predict such violence. 3 Second, I review the key features of the psychiatric
expertise offered by prosecutors to support SVP commitment and analyze how courts have used

Kansas v. Hendricks, 521 U.S. 346 (1997).
Kansas v. Crane, 534 U.S. 407 (2002).
I will generally use the term "psychiatry" to refer to the professional field concerned with the identification of
mental illness in the Sexually Violent Predator (“SVP”) context because it is closely associated with the overall
development of mental pathology classification and nosology, such as through the Diagnostic and Statistical
Manuals of Mental Disorders published by the American Psychiatric Association. I refer to "psychology" in the
context of research regarding human behavior and the like. Parties in court proceedings quite often present expert
evidence through the testimony of forensic or clinical psychologists. See GARY B. MELTON ET AL., PSYCHOLOGICAL


such expertise to decide whom to commit under SVP laws. These examinations reveal that the
assumptions upon which the Court based the Hendricks-Crane rationale were erroneous.
The Court’s most consequential error was its failure to acknowledge that the category of
the “Sexually Violent Predator” is a political and moral construct, not a medical classification.
Mainstream psychiatry has never claimed that it can accurately predict who is at risk of
committing acts of sexual violence and has never conceptualized sexual aggression as the
product of volitional impairment. Indeed, the American Psychiatric Association (APA), the
leading professional organization in American psychiatry, and other voices from within the
mental health profession have vociferously opposed SVP laws since their enactment precisely
because of the role assigned to psychiatric expertise by such laws to identify those who should be
The controversies regarding admission of expert testimony in individual SVP cases
reveal the troubling consequences of the Supreme Court’s failure to heed the warnings of the
APA. Trial courts permit prosecution experts to offer diagnoses and predictions of risk in support
of these commitments notwithstanding the fact that such testimony often strays far from current
scientific understanding of the relationship of acts of sexual violence to psychopathology. In so
doing, such courts distort or disregard key values in our justice system, such as limiting the
admission of expert testimony to that based upon scientifically-sound methodology and reliable
facts and data. Rulings in such cases have become even more dubious in the years since the SVP
laws’ initial development, as the debate regarding the medical basis of SVP commitment has
only intensified, bringing to the surface the unsteady foundation upon which the medical and, by
extension, constitutional premise of SVP was based.
The SVP laws generally 4 and the Hendricks opinion specifically5 have been the target of
extensive criticism from scholars as well as from legal and mental health professionals. While
some have focused upon specific problems in the implementation of SVP laws, such as experts’
reliance upon controversial diagnoses or their use of actuarial instruments to assess risk, many in
both groups—scholars and mental health professionals—have pointed to the laws as inherently
flawed policy. 6 Although critical of the SVP laws, the scholars and medical professionals
offering these analyses generally assume that, in light of the Hendricks opinion, the question of
their constitutionality is now a settled matter. Taken together with a review of how the laws have
actually operated, however, these and related criticisms can be seen to make that assumption of
constitutionality itself questionable.
My approach in this Article is to analyze the SVP laws as a legislative experiment in
preventive detention endorsed in the Hendricks and Crane opinions by the Supreme Court
through a rationale based upon a set of dubious hypotheses and assumptions regarding psychiatry
and psychiatric testimony. The rationale first developed in Hendricks was strictly theoretical: the

Aman Ahluwalia, Civil Commitment of Sexually Violent Predators: The Search For A Limiting Principle, 4
PREDATOR LAWS AND THE RISE OF THE PREVENTIVE STATE 61–66 (2006); Stephen J. Morse, Uncontrollable Urges
and Irrational People, 88 VA. L. REV. 1025, 1076–77 (2003).
Michael L. Perlin, “There's No Success Like Failure/And Failure's No Success At All”: Exposing The Pretextuality
of Kansas v. Hendricks, 92 Nw. U.L. Rev. 1247, 1248 (1998).
See, e.g., Melissa Hamilton, Adjudicating Sex Crimes as Mental Disease, 33 PACE L. REV. 536 (2013).


Court was evaluating a new statutory model for indefinite preventive detention and Leroy
Hendricks was of the first people to challenge it. The Supreme Court expected mental health
professionals to help courts and fact finders discriminate between the typical recidivist and the
truly ill, thereby ensuring that the new laws did not reach too far. These expectations stemmed
largely from the longstanding use of psychiatric expertise by the courts to help answer difficult
questions about the mental status of persons appearing before them. However, the actual use of
such expertise in SVP proceedings reveals that such faith in psychiatry was in fact misplaced.
Other commentators have noted that the use of certain diagnoses in SVP proceedings
runs counter to the APA’s Diagnostic and Statistical Manual of Mental Disorders (“DSM”)
system of psychiatric classification. 7 What I demonstrate here, however, is that the problems
with the psychiatric evidence offered in these cases are far broader than occasional
misclassification and in fact stem from limitations inherent in the field of psychiatry itself.
Rationalization of SVP preventive detention based on the conviction that psychiatric testimony
will ensure that such detention adheres to due process principles reflects a fundamental
misunderstanding of psychiatric evaluation and diagnosis.
The problems seen in the use of expert evidence in these proceedings do not admit of
solution through specific fixes. Rather, they reveal that there are no means to implement SVP
laws consistent with notions of due process and individual liberty. A sexual predator is a legal
classification that depends upon medical delineation to be constitutionally sound. But because
there is no conceptualization in psychiatry resembling a “sexual predator,” the implications of
this incongruence go to the essential question of the constitutionality of the SVP laws. Written
opinions reveal that courts are basing SVP commitments largely upon the criminal records of the
respondents 8 because the expert opinions themselves are based upon little else. As a result, the
opinions offered by experts in SVP cases are not in fact “medical” but moral. And because such
conclusions are essentially normative ones, then we are improperly delegating such decisions to
psychiatry, an action that flies in the face of both legal principles and psychiatric practice. This is
not a mere problem of labels and professional realms; this experiment has resulted in the
indefinite detention of thousands of people at an enormous monetary cost to the jurisdictions and
an enormous personal cost to those committed as well as to their families.



In the early 1990s, intense public awareness and concerns about sexual abuse of children,
and widespread views of those who commit such as offenses as what has been called “the
ultimate other,” 9 gave rise to the notion of “the sexual predator.” 10 In the wake of media reports
of a spate of high-profile sexual crimes against children, some state legislatures passed measures

Robert A. Prentky, et al., Sexually Violent Predators in the Courtroom: Science on Trial, 12 PSYCHOL. PUB. POL'Y
& L. 357, 358 (2006); Hamilton, supra note 6, at 23–29.
Court opinions refer to those individuals who are the targets of SVP commitment petitions primarily as
“respondents” and occasionally as “defendants”; I will primarily use the former term.
Perlin, supra note 5, at 1248.
Amy Adler, To Catch a Predator, 21 COLUM. J. GENDER & L. 130, 130–31 (2012); ROGER LANCASTER, SEX


in an attempt to control these individuals. These new laws were based on the assumption that
these criminals had unusually high rates of recidivism and posed a special risk to the public.
They were sick, the laws’ supporters reasoned, with a condition that rendered them resistant to
typical forms of deterrence in the criminal justice system. 11 Policymakers concluded that these
unique attributes, combined with the particularized harm resulting from sexual abuse, warranted
unique measures. Legislatures enacted new or enhanced laws that addressed punishments for the
possession and viewing of child pornography, created registries and notifications requirements,
and at the extreme end, established programs for the indefinite detention via civil commitment of
those identified as “sexually violent predators” (“SVPs”). 12
The first such SVP commitment law was enacted by the State of Washington against the
backdrop of both the “sexual psychopath” laws of the mid-20th century and the heightened
attention to the problem of sexual violence committed by those previously in the criminal justice
system. Once this new form of commitment received the sanction of the U.S. Supreme Court in
Hendricks, the model spread, and there are now several well-established SVP commitment
programs across the country, through which thousands of people have been, and continue to be,
detained indefinitely.

The Origins of SVP Commitment


Rise and Fall of Sexual Psychopath Laws

The SVP laws conceived in the early 1990s were not the first that targeted sex offenders.
The first generation of laws permitting the detention of sex offenders were enacted between the
1930s and 1960s, although they had significant differences from contemporary SVP laws.
During the time that these earlier laws were in place, mainstream psychiatry explained that
“sexual psychopaths” were ill, which placed them in the realm of medicine in terms of both
identification and care. Since the laws assured the administration of treatment, rather than simple
detention, they were open-ended in terms of the length of hospitalization. 13 Men who were


Adler, supra note 10, at 131–32.
Both the laws and common parlance use a range of terms to describe those who commit, or are at risk of
committing, multiple crimes of sexual violence. I will use the abbreviation “SVP” throughout the article to describe
such category of classification as this was the one used by Washington in the first such law and it is the most
commonly used by other states. See ARIZ. REV. STAT. ANN. §§ 36-3701–3717 (“Sexually Violent Persons”); CAL.
WELF. & INST. CODE §§ 6600–6609.3 (“Sexually Violent Predators”); D.C. CODE §§ 22-3803–3811 (“sexual
psychopath[s]”); FLA. STAT. ANN. §§ 394.910–932 (“Sexually Violent Predators”); IOWA CODE ANN. §§ 229A.1–16
(“Sexually Violent Predators”); 725 ILL. COMP. STAT. ANN. 207/1–99; KAN. STAT. ANN. §§ 59-29a01–29a22
(“Sexually Violent Predators”); MASS. GEN. LAWS ANN. ch. 123A, §§ 1–16 (“Sexually Dangerous Persons”); MINN.
STAT. ANN. § 253B.185 (“sexually dangerous persons or persons with a sexual psychopathic personality”); MO.
ANN. STAT. §§ 632.480–513 (“Sexually Violent Predators”); NEB. REV. STAT. §§ 71-1201–1226 (“dangerous sex
offenders”); N.H. REV. STAT. ANN. §§ 135-E:1–24 (“Sexually Violent Predators”); N.J. STAT. ANN. §§ 30:4-27.24–
.38 (“sexually violent predator[s]”); N.Y. MENTAL HYG. LAW §§ 10.01–17 (“Sex Offenders Requiring Civil
Commitment or Supervision”); N.D. CENT. CODE §§ 25-03.3-01–24 (“Sexually dangerous individual[s]”); S.C.
CODE ANN. §§ 44-48-10–170 (“Sexually Violent Predator[s]”); VA. CODE ANN. §§ 37.2-900–920 (“Sexually
Violent Predators”); WASH. REV. CODE ANN. §§ 71.09.020 (“Sexually Violent Predator”).


charged with sex crimes could be sent for such treatment, rather than sentenced to prison, with
the prospect that the treatment would prevent recidivism. 14
The U.S. Supreme Court upheld this form of commitment in 1940,15 but the laws were
eventually subjected to widespread criticism. The detainees were held much longer than had
been expected, in part because of the reluctance of their treatment providers, who were not
confident that their patients would not recidivate, to approve their release. A growing number of
commentators within psychiatry attacked the legal classification of the “sexual psychopath,” as
there was no agreed-upon definition or basis to attach the label to any individual. It was also
clear that many of these hospitalized men were not mentally ill and that little if any treatment
was being provided in these hospitals. The laws were revealed as little more than extended
detention on a preventive basis. 16 Most such laws were either repealed or no longer used by the
early1980s. 17
The final nail in the coffin for the remaining laws came from a strong statement from
within the psychiatric establishment. 18 The Group for the Advancement of Psychiatry (GAP) 19
Committee on Forensic Psychiatry concluded in a 1977 report that there was little true prospect
for effective treatment of sexual offenders and that the “discrepancy between the promises in sex
statutes and performance have rarely been resolved.” 20 “In retrospect,” the GAP Committee
reported, “we view the sex psychopath statutes as social experiments that have failed and that
lack redeeming social value. These experiments have been carried out by the joint participation
of the psychiatric and legal professions with varying degrees of acquiescence by the general
public.” 21 The GAP Committee acknowledged that the “promises” made by psychiatry at the
time the laws were enacted went unfulfilled. The profession could not separate out the mentally
ill sex-offenders from the others, and there was little in the way of treatment that psychiatry
could provide the men once they were committed. The report goes on to state starkly and
The notion is naive and confusing that a hybrid amalgam of law and psychiatry can
validly label a person a “sex psychopath” or “sex offender” and then treat him in a
manner consistent with a guarantee of community safety. The mere assumption that such


State of Minn. ex rel. Pearson v. Probate Court of Ramsey Cnty., 309 U.S. 270, 274, 60 S. Ct. 523, 526 (1940).
The Minnesota statute upheld in that case required “proof of a ‘habitual course of misconduct in sexual matters' on
the part of the persons against whom a proceeding under the statute is directed, which has shown ‘an utter lack of
power to control their sexual impulses', and hence that they ‘are likely to attack or otherwise inflict injury, loss, pain
or other evil on the objects of their uncontrolled and uncontrollable desire’. “ The Supreme Court reasoned that
there was no violation of due process since such “underlying conditions, calling for evidence of past conduct
pointing to probable consequences, are as susceptible of proof as many of the criteria constantly applied in
prosecutions for crime.” Id. at 274, 526.
EWING, supra note 13, at 8.
APA, supra note 14, at 13-15.
EWING, supra note 13, at 9.
The GAP identifies itself as the “think tank” for American psychiatry. (
THE 80’S 935 (1977).
Id. at 840.


a heterogeneous legal classification could define treatability and make people amenable
to treatment is not only fallacious, it is startling. 22
Remarkably however, only a short time after the sexual psychopath laws were discarded, they
were resurrected in a new, more extreme form of experiment, one that was also “carried out by
the joint participation of the psychiatric and legal professions,” this time in complete disregard of
the psychiatric profession’s own conclusions.

Washington State Enacts First SVP Law and Creates a Model Statute

Under public pressure following a set of horrific and highly publicized cases of sexual
violence committed by offenders who had previously served time, state legislatures, led by
Washington in 1989, dusted off the basic concept of these earlier sexual psychopath laws but
transformed them in several importance respects. 23 Most notably, in their new incarnation, the
commitment of convicted offenders would occur not as an alternative sentence in lieu of a prison
sentence after conviction, as was the case for most of the earlier sexual psychopath laws, but as
an additional period of indefinite detention after a criminal sentence had been served. 24
Some commentators have noted that the current generation of SVP laws were enacted in
response to the rise of determinate sentencing, which gave states less control over release dates
for those convicted of crimes, including sex crimes, and the public perception that sentences for
sex crimes were too short. 25 Indeed, the enactment of the first SVP law in Washington State
involved precisely that scenario. Earl Shriner, a man with a history of repeated involvement in
the criminal justice system in connection with crimes against young people and officially
described as “mildly retarded,” was released from prison in 1998 after completing the term of his
sentence for kidnapping two girls. 26 Several months after his release, and while other charges
against him were pending, he was charged with raping and mutilating a young boy, apparently at
random, in Tacoma. 27
The public outrage in response to this crime was immediate, widespread, and intense. An
editorial in the Seattle Post-Intelligencer summed up the sense, shared by many, of how the
criminal justice system had failed Shriner’s latest young victim: “This case makes clear that a
class of criminal exists that is beyond reach of rehabilitation because of mental deficiencies. …
The legal system needs to be changed to make it possible to remove the criminally insane from
society, quickly and permanently. In such obvious cases as this, the law should err, if it errs at all,
on the side of protecting the innocent.” 28

Id. at 935.
EWING, supra note 13, at 9–10.
Ewing, supra note 13, at 10.
APA, supra note 14, at 34; John Q. La Fond, Sexually Violent Predator Laws and the Liberal State: An Ominous
Threat To Individual Liberty, 31 INTL. J. OF LAW & PSYCHIATRY 158, 160 (2008).
David Boerner, Confronting Violence: In the Act and in the Word, 15 U. PUGET SOUND L. REV. 525, 542 n.10
(1992). This article is an invaluable glimpse into the development of the Washington SVP law, which served as the
model for all current laws. It was written soon after the law’s enactment by David Boerner, a prosecutor and law
professor who was the lead drafter of the law (and who proposed the basic framework), and it provides a frank and
personal account of his thinking during the events leading to the enactment of the law.
Boerner, supra note 26, at 525-27.
Id. at 529 (quoting Editorial, SEATTLE POST-INTELLIGENCER (May 24, 1989)) (emphasis added).


Within days of Shriner’s arrest, Washington Governor Booth Gardner called for the
development of legislation to prevent people like Shriner from “fall[ing] through the cracks.”
Specifically, he stated: “[T]here should be a way to involuntarily commit people who have a
profile of an individual that is a known risk with a high degree of probability that they would
commit this type of crime.” 29 Soon thereafter and less than a week after the crime, Gardner
created a task force to study the Shriner case and draft legislation to address “gaps that exist
between civil and criminal commitments, particularly regarding predatory offenders,” gaps that
had presumably permitted Shriner the opportunity to commit his most recent crime. 30
The fact that the state had previously unsuccessfully attempted to commit Shriner
highlighted the limitations of the standard involuntary hospitalization statutes for, as it was said,
“quickly and permanently” removing the dangerous mentally ill from society. 31 In terms of their
purpose and outcomes, such laws were indeed a poor fit for the goal of detaining criminally
violent men like Shriner for an extended period of time or until they no longer posed a high risk
of committing sexually violent acts.
The central objective of contemporary involuntary hospitalization laws is to provide a
means to stabilize a person identified as having severe mental illness, such as schizophrenia or
bipolar disorder, and to administer treatment, usually in the form of psychotropic medications
such as anti-psychotics or mood stabilizers. 32 A series of U.S. Supreme Court and lower court
opinions in the 1960s and 1970s clarified the constitutional limitations on such deprivations of
liberty. According to these opinions, such involuntary hospitalization must be based upon a
showing that a person posed a danger to himself or others (as demonstrated through a recent
overt act) and that the hospitalization would end as soon as the acute danger had passed. 33
Involuntary hospitalization can occur only when there is a crisis as demonstrated by either threats
to others or, more commonly, an inability to care for one’s basic needs. If that threshold showing
is met, a court will order treatment in a secure community hospital or state hospital, with a
maximum length of hospitalization set by statute. As a result of reforms brought about by the
court opinions in concert with the “deinstitutionalization” movement, which ended the long-term
warehousing of the mentally ill, the average length of such hospitalizations is now measured in
days. 34

Boerner, supra note 26, at 530. The arrest of Shriner occurred six months after the murder in Seattle of Diane
Ballasiotes. A convicted sex offender participating in a work-release program was charged (and eventually
convicted) of her murder. Id. at 534.
Id. at 534-35. Other reasons given for the rise of the law include a rising perspective that government has a critical
role to prevent harm to its citizens, Eric S. Janus, Sexual Predator Commitment Laws: Lessons for Law and the
Behavioral Sciences, 18 BEHAV. SCI. LAW 5, 8 (2000), and the “victims’ rights” movement, Michael M. O’Hear,
Perpetual Panic, 21 FED. SENTENCING REP. 69, 74 (2008). Also, such laws were seen as an example of the growing
success of feminists to reform the legal responses to sexual violence. LANCASTER, supra note 10, at 14.
Boerner, supra note 26, at 533. Washington sexual psychopath law, which had been the subject of controversy
regarding its scope and implementation, had been repealed in 1984. Id. at 551-52.
La Fond, supra note 25, at 160–61.
See, e.g., O’Connor v. Donaldson, 422 U.S. 563 (1975); Lessard v. Schmidt, 349 F.Supp. 1078 (E.D. Wis. 1972).
Indeed, many states are moving in the direction of adopting involuntary outpatient treatment laws, where the
medication is administered without full-time hospitalization. Nisha C. Wagle et al., Outpatient Civil Commitment
Laws: An Overview, 26 MENTAL & PHYSICAL DISABILITY L. REP. 179 (2002). It should be added that recurring
hospitalizations are not uncommon.


The Washington Legislature noted in its findings that a “small but extremely dangerous
group of sexually violent predators exist who do not have a mental disease or defect that renders
them appropriate for” involuntary civil commitment under the “existing involuntary treatment”
law. 35 As the legislature saw it, the problem with existing involuntary commitment law was that
the state would not be able to meet the recent “overt act” requirement of such law when seeking
commitment of a person serving a sentence since such a person would not “have access to
potential victims.” 36 The legislature acknowledged that the target for the new SVP legislation
was not those who had “classic mental illness” as that was understood and used in traditional
commitment laws. 37 The Washington lawmakers were concerned about a different set of people:
those convicted of a sex crime who, because of some severe mental disorder, posed a high risk of
The social problem posed by the existence of such people could not be addressed by a
short-term hospitalization and the administration of medication, as such measures would
presumably do nothing to prevent their recurring criminal conduct. Only their long-term removal
from society and thus from potential victims would, it was thought, reduce the risk of future acts
of sexual violence. Those to be detained under the SVP laws were, after all, not pathetic people
sleeping under bridges or in their parents’ basements, as is often the case for involuntary
commitment, but people who were incarcerated, out on bail, or under supervision of some kind
because they had committed a violent sex crime. Rather than seeking, by such laws, to detain
someone who was already at large, the lawmakers wanted to prevent a return of such persons to
society. For these reasons, perhaps, the measures might have seemed less extreme than those that
entitled a police officer to pick someone off of the street and bring him to an emergency room
against his will.
Another notable distinguishing feature of the new SVPs laws is that the commitment is
indefinite, with the committed person having the burden of petitioning for review of his
commitment. 38 The Washington Legislature reasoned that no set time frame for detention could
be included in the statute because “the prognosis for curing sexually violent offenders is poor,
the treatment needs of this population are very long term, and the treatment modalities for this
population are very different" from those appropriate for individuals confined under the general
commitment laws. 39 With no clear treatment protocol for persons now referred to as
“predators,” 40 the treatment-oriented laws for standard commitment of the mentally ill were
clearly a poor fit.


WASH. REV. CODE § 71.09.010.
Id. See also Black v. Voss, 557 F.Supp.2d 1100, 1109–10 (D. Cal. 2008) (rejecting habeas corpus petition of
person committed under California SVP law and noting that the statute has no overt act requirement to establish
dangerousness under SVP commitment).
WASH. REV. CODE § 71.09.010. The reference to “classic mental illness” arose in the public testimony of
Professor Boehner, the lead drafter of the law. Young v. Weston, 898 F. Supp. 744, 750 n.3 (W.D. Wash. 1995).
One scholar has argued that this “new generation” of SVP laws is the product of a confluence of two criminal justice
trends: (1) a blurring of the civil-criminal distinction; and (2) increased use of “risk assessment,” particularly
through actuarial instruments and conclusions based upon what groups of individuals do (what he dubbed “actuarial
justice”). Ahluwalia, supra note 4, at 491.
La Fond, supra note 25, at 161, 164.
WASH. REV. CODE ANN. § 71.09.010.


As indicated by recorded legislative reasoning, SVP laws were based upon two critical
and commonly-held assumptions about those who commit sex crimes: first, that they are
criminals who “specialize” in a particular type of crime; and, second, that they have a
particularly high rate of recidivism because of a mental pathology—a compulsion of some sort-that leads to repeated acts of sexual violence. 41 Such specialization and compulsion rendered
these men “predators” and, the reasoning went, since their sexually violent conduct resulted from
a mental disorder, mental health professionals could identify those offenders likely to engage in
such conduct in the future.
SVP laws were thus also based on a third crucial assumption, this one about the role that
psychiatric diagnosis could play in ensuring that such laws would not have an overbroad reach.
The significance of this assumption is clear from this statement of the California Legislature,
made when it enacted its SVP law in 1995: “The Legislature finds and declares that a small but
extremely dangerous group of sexually violent predators that have diagnosable mental disorders
can be identified while they are incarcerated. These persons are not safe to be at large and if
released represent a danger to the health and safety of others in that they are likely to engage in
acts of sexual violence....” 42
None of the crucial assumptions about so-called sexually violent predators has a footing
in scientific or clinical findings, as discussed further in Part III.A below. 43 At the time the rise in
SVP laws occurred, data already indicated that the significant majority of sex crimes were in fact
committed not by stereotypical “predators” who stalked, lured, and pounced on random hapless
victims, but, rather, and particularly in the case of the sexual assault of children, 44 by men who
were family members and acquaintances of the victims. 45 Similarly, studies indicated that,
contrary to popular belief, 46 sexual offenders did not have unusually high levels of recidivism 47
or specialization with regard to victims. 48 Rare as they were, however, crimes such as Shriner’s
were so compelling that many members of the public were persuaded that children were at a high
risk of random victimization unless the state acted quickly to protect them.


The state of Washington conceded in one of the first legal challenges to these statutes, Young v. Weston, that the
treatment prospects for detainees was “poor” and therefore “prolonged incarceration is to be expected.” Young v.
Weston, 898 F. Supp. 744, 749 (D. Wash. 1995).
Lenore M. J. Simon, An Examination of the Assumptions of Specialization, Mental Disorder, and Dangerousness
in Sex Offenders, 18 BEHAV. SCI. LAW. 275, 277 (2000).
1995 Cal. Legis. Serv. 4611 (West) (emphasis added).
See infra notes 185–290 and accompanying text.
See OFFICE OF JUSTICE PROGRAMS BUREAU OF JUSTICE STATISTICS, “Criminal Victimization,” See also EWING, supra note 13, at xvi-xvii for general discussion on
these statistics
Paul Good and Jules Burstein, Modern Day Witch Hunt: The Troubling Role of Psychologists in Sexual Predator
Laws, 28 AM. J. FORENSIC. PSYCH. 23, 40 (2010) (noting significant number of erroneous statements about rates of
sex offender recidivism in the media, including statements to the effect that such rates more than 75% or near
Prison In 1994”
See infra notes 239–233 and accompanying text.


Washington’s “Community Protection Act of 1990” provided the model for the new
incarnation of sexual psychopath laws, not least in giving legal status to a new term, “sexually
violent predator,” which spread quickly through the common parlance. Governor Gardner’s use
of the phrase “predatory acts” in a press statement soon after Shriner’s arrest struck a chord with
prosecutor and law professor David Boerner, the lead drafter of the new law. Boerner saw it as a
way to specify the class of individuals to be reached by this unique form of indefinite
detention. 49 He defined the term “predatory acts” as those “directed towards strangers or
individuals with whom a relationship has been established or promoted for the primary purpose
of victimization” and he recommended that only those who engaged in such acts would be
eligible for commitment. 50 Since one who commits such “predatory acts” is a “predator,” that
category of persons, along with a putative medical diagnosis and rationale for detention, was
built directly into the statute. A “sexually violent predator” was defined by Washington’s new
law as: “any person who has been convicted of or charged with a crime of sexual violence and
who suffers from a mental abnormality or personality disorder which makes the person likely to
engage in predatory acts of sexual violence if not confined in a secure facility.” 51
Thus Washington’s SVP law set four prerequisites to civil commitment: (1) a history of
criminal sexual conduct, resulting in either a conviction or a charge; (2) the presence of a mental
disorder, personality disorder, or mental abnormality of some kind at the time the commitment
was under consideration; (3) a likelihood of engaging in sexual criminal behavior in the future;
and (4) a causal link between the disorder or abnormality and the risk. 52 These essential
requirements, although often phrased somewhat differently, can be found in all SVP laws.
The procedure established under the Washington SVP statute provides that proceedings
for indefinite detention can be initiated at the conclusion of a period of incarceration for a sex
crime committed as an adult or juvenile, after a person charged with such a crime has been found
not competent to stand trial or is acquitted on the basis of a finding of insanity, or after a person
who has been previously convicted of a sexual offense commits a “recent overt act.” 53 After a
probable cause hearing, the person is evaluated in custody. The trial on the commitment must
occur within 45 days of the filing of the petition. 54 The person is entitled to counsel and courtappointed experts to assist with his defense. 55 Either side may request a jury trial. 56 If the fact
finder concludes that the state has demonstrated beyond a reasonable doubt that the person is “a
sexually violent predator,” the person is committed to a “secure facility for care, control, and
treatment” until the mental abnormality or personality disorder “has so changed that the person is
safe to be at large.” 57

Legal Challenges to the New SVP Laws


Boerner, supra note 26, at 569.
Id. at 569.
WASH. REV. CODE ANN. § 71.09.020 (West) (emphasis added).
Id.; Janus, supra note 30, at 9.
WASH. REV. CODE ANN. § 71.09.030 (West)
WASH. REV. CODE ANN. § 71.09.050 (West).
WASH. REV. CODE ANN. § 71.09.040 (West)
Id. § 71.09.050.
Id. § 71.09.060(1). See also Young v. Weston, 898 F. Supp. 744, 747 (D. Wash. 1995) (summarizing key
requirements of SVP law).


Preventive detention is very limited in American law because it is seen as antithetical to
notions of liberty interest and the presumption of innocence. In each instance of preventive
detention, even where an individual is seen as posing a threat to public safety, there are generally
strict limitations on when it can be imposed and when it must be ended. For example, we permit
the pretrial detention of criminal defendants only where there is probable cause to believe that
they committed a crime and to the extent found necessary to secure their appearance at trial (thus
defendants are usually given the opportunity to post bail and be released). 58 The two exceptions
to our reluctance to impose long-term preventive detention target two of the groups probably
most feared and despised by the American public: one is enemy combatants seized on the
battlefield in foreign countries; the other is sex offenders. 59

Background of the Hendricks-Crane Litigation

The constitutionally of the Washington statute and of those SVP laws modeled after it
was immediately subjected to legal the target of legal challenges to its constitutionality on a
range of grounds, including the violations of constitutional guarantees of substantive due process
and the prohibitions of ex post facto laws and double jeopardy. Andre Young, one of the first
men committed under the Washington’s SVP law, challenged the constitutionality of the law in
both state 60 and federal 61 courts. The Washington Supreme Court upheld the law while the
federal district court held that it was unconstitutional. 62 These differing outcomes were among
the first of a series of sharply dividing judicial responses to the new law and to the similar SVP
laws enacted by the Kansas 63 and Wisconsin 64 legislatures soon thereafter.
The focus of the substantive due process challenges stemmed from the same theories that
had been used to limit the reach of other forms of involuntary commitment and preventive
detention: that using state power to deprive a person of liberty outside of the realm of criminal
punishment runs afoul of core values established through the due process clause. The Supreme
Court has acknowledged: “[T]he Due Process Clause contains a substantive component that bars
certain arbitrary, wrongful government actions ‘regardless of the fairness of the procedures used
to implement them.’” 65 Such guarantee against excessive government interference applies with
particular import in the context of involuntary detention, the Court has noted, because “freedom
from bodily restraint has always been at the core of the liberty protected by the Due Process
Clause from arbitrary governmental action.” 66 Accordingly, a court must subject such detention,


U.S. v. Salerno, 481 U.S. 739, 752, 107 S. Ct. 2095, 2104 (1987).
See Norman J. Finkel, Moral Monsters and Patriot Acts: Rights and Duties in the Worst of Times, 12 PSYCHOL.
PUB. POL'Y & L. 242, 243 (2006); Good & Burstein, supra note 46, at 42; Christopher Slobogin, Preventive
Detention in Europe and the United States ( )
In re Personal Restraint of Young, 122 Wash.2d 1, 857 P.2d 989 (1993).
Young v. Weston, 898 F. Supp. 744 (D. Wash. 1995).
Id. at 754; In re Personal Restraint of Young, 857 P.2d 989, 1018 (1993).
KAN. STAT. ANN. § 59-29a01 et seq. (1994).
WIS. STAT. ch. 980 (1993-94)
Zinermon v. Burch, 494 U.S. 113, 125, 110 S. Ct. 975, 983, 108 L.Ed.2d 100 (1990).
Foucha v. Louisiana, 504 U.S. 71, 80 (1992).


even if imposed pursuant to statute, to a rigorous review and invalidate it if it does not fall under
one of the few, narrow exceptions to the broad general prohibition of preventive detention. 67
When applying these principles to their review of the new SVP laws, the holdings of the
Washington and Wisconsin Supreme Courts included some fractured opinions and vehement
dissents. Most of the debates about whether the laws were consistent with the “substantive
component” of due process focused on the states’ open acknowledgment that the targets of the
new laws were people who did not have a mental illness that would subject them to commitment
under standard civil commitment laws and the fact that, in lieu of serious mental illness, the laws
used terminology such as “mental abnormality” and “personality disorder.” 68 Justice Shirley
Abrahamson of the Wisconsin Supreme Court was especially troubled by the nebulous language
of “mental abnormality” in the Wisconsin law. 69 That term, she observed, does not translate to
any well-settled or understood concept in psychiatry. 70
For the courts reviewing the constitutionality of the first SVP laws, a key source of
guidance was a then-recent opinion of the U.S. Supreme Court in Foucha v. Louisiana.71 The
Court held that a state could not continue to detain an insanity acquitee who no longer had a
mental illness on the basis of medical opinions that he had an “antisocial personality” and would
be a danger if released. 72 The Court rejected Louisiana’s argument that the state could continue
“to hold indefinitely any other insanity acquitee not mentally ill who could be shown to have a
personality disorder that may lead to criminal conduct.” 73 The Court ruled that, in the absence of
a mental illness, Louisiana’s detention of Foucha was contrary to fundamental notions of due
process. It noted:
The same would be true of any convicted criminal, even though he has completed his
prison term. It would also be only a step away from substituting confinements for
dangerousness for our present system which, with only narrow exceptions and aside from
permissible confinements for mental illness, incarcerates only those who are proved
beyond reasonable doubt to have violated a criminal law. 74


Id. at 81–86. See also O'Connor v. Donaldson, 422 U.S. 563, 575 (1975) (holding that involuntary commitment of
those who “are dangerous to no one and can live safely in freedom” is a violation of due process principles); cf.
United States v. Salerno, 481 U.S. 739, 749–50 (1987) (upholding pretrial detention under limited circumstances
where the government’s interest was compelling).
See, e.g., Weston, 898 F. Supp. At 749-50 (“The essential component missing from the Sexually Violent Predator
Statute is the requirement that the detainee be mentally ill.”).
State v. Post, State v. Oldakowski, 541 N.W.2d 115, 142–45 (Wisc.1995) (Abrahamson, J., dissenting).
La Fond, supra note 25, at 161.
Foucha v. Louisiana, 504 U.S. 71, 80 (1992).
Id. at 80.
Id. at 82. The Court’s holding here flowed explicitly from its earlier ruling in Addington v. Texas that: “to commit
an individual to a mental institution in a civil proceeding, the State is required by the Due Process Clause to prove
by clear and convincing evidence the two statutory preconditions to commitment: that the person sought to be
committed is mentally ill and that he requires hospitalization for his own welfare and protection of others.” Id. at 7576, (citing Addington v. Texas, 441 U.S. 418 (1979)).
Id. at 82–83 (emphasis added). The Court noted that other forms of preventive detention were narrowly tailored to
a specific legitimate need and a finite duration, such as pretrial detention in limited circumstances, which was upheld
in Salerno. Id. at 81, 83


Many concluded from this language that, in Foucha, the Court had made clear that
“dangerousness” alone was not a sufficient basis for preventive detention and that an
indispensable constitutional requirement for such detention was a finding of clear and convincing
evidence of “mental illness.”
In 1994, two years after the opinion in Foucha, Kansas enacted the “Sexually Violent
Predator Act.” Modeled closely on the Washington law, it required a finding of mental
abnormality or personality disorder as a prerequisite to commitment. 75 As defined under the
Kansas statute, a “mental abnormality” is “a congenital or acquired condition affecting the
emotional or volitional capacity which predisposes the person to commit sexually violent
offenses in a degree constituting such person a menace to the health and safety of others.” 76 The
law did not have a clear requirement for a finding of “mental illness.”
Leroy Hendricks, who was then presently serving a sentence for sexual victimization of
children, was the first person to be committed under Kansas’s new SVP law pursuant to a jury’s
determination. 77 If the State selected him for the first petition under the law on the assumption
that his case would be the first challenge to the new law, and therefore subject to particular
scrutiny, the State chose well; Hendricks he had a long history of multiple sexual offenses
against children and therefore exemplified the seemingly undeterrable “predator” the drafters and
public had in mind in enacting the law.
At trial, the State called as its expert witness Dr. Charles Befort, the chief psychologist at
Larned State Hospital. 78 Befort, who had performed an evaluation of Hendricks, testified that he
had concluded that it was “likely that Hendricks would engage in predatory acts of sexual
violence or sexual activity with children if permitted to do so.” 79 Befort based his opinion, as he
stated, on his view that “‘behavior is a good predictor of future behavior,” on his professional
knowledge that pedophiles tend to repeat their behavior, and on Hendricks’ poor understanding
of his behavior.” 80 Befort concluded that Hendricks was not mentally ill and did not have “a
personality disorder” but that “as he [Befort] interpreted the Act, pedophilia was a mental
abnormality.” 81 The psychiatrist who testified on behalf of Hendricks challenged Befort’s
testimony regarding the tendency of pedophiles to recidivism, observing that, “based on current
knowledge, ‘a psychiatrist or psychologist cannot predict whether an individual is more likely
than not to engage in a future act of sexual predation.’” 82 The jury found that Hendricks was a
“sexually violent predator” and, under the new Kansas statute, he was committed to Larned State
Hospital. 83


KAN. STAT. ANN. § 59-29a01– 59-29a24.
Id. § 59-29a02(b)
Kansas v. Hendricks, 521 U.S. 346, 350 (1997).
In The Matter of Hendricks, 912 P.2d 129, 131 (Kan. 1996). The State also called Hendricks himself as a witness
after the court ruled that, because the proceedings were civil rather than criminal, Hendricks had no right to invoke
the privilege against self-incrimination. See Allen v. Illinois, 478 U.S. 364, 373, 106 S. Ct. 2988, 2994 (1986).
Id. Befort conceded in his testimony that the statute’s definition of “mental abnormality” was “circular in that
certain behavior defines the condition which is used to predict the behavior.” Id. at 138.


In reviewing Hendricks’ appeal, the majority opinion of the Kansas Supreme Court noted
that the Kansas SVP law had been modeled on that enacted by Washington (including adopting
the same legislative “findings”) and that the latter was already the subject of constitutional
challenges. 84 Hendricks’s attorneys based their substantive due process argument on the key
holding in Foucha that mental illness was an indispensable requirement for indefinite detention
on the basis of dangerousness and that the Kansas law’s “mental abnormality or personality
disorder” standard fell short of that requirement. The Kansas Supreme Court agreed and held that
Kansas’s SVP law was invalid under Foucha (as well as an earlier civil commitment opinion,
Addington v. Texas 85) since it did not require a showing of an “illness.” 86 In so ruling, the
majority found the reasoning of the federal district court in Young v. Weston striking down the
Washington SVP law to be more persuasive than that of the Washington Supreme Court in its
opinion upholding the law. The term “mental abnormality,” it concluded, was not equivalent to
“mental illness.” The Kansas Supreme Court based this conclusion in part upon the testimony of
the State’s own expert witness, who had testified that the term was not a diagnosis but rather “a
phrase used by clinicians to discuss abnormality or deviance.” 87 The majority also contrasted that
description with the definition of “mental illness” found in the Kansas standard involuntary
commitment statute. 88

The Supreme Court Upholds the SVP Model of Commitment

Once these questions reached the United States Supreme Court, they received a quite
different reception by the five-justice majority that, in Kansas v. Hendricks, ultimately overruled
the Kansas Supreme Court and upheld the state’s SVP law. 89 On the question of whether
Kansas’s definition of SVP satisfied the “mental illness” element in Foucha, the parties took
significantly differing positions. The State noted in its brief that, in the line of cases requiring
“mental illness” as a matter of substantive due process, the Supreme Court had never defined the
term. 90 This was understandable, the State argued, since there is no universally accepted
definition of the term. What was more important for constitutional purposes, it claimed, was that
“mental health professionals [can] give the definition content by identifying specific mental
disorders that may or may not satisfy the definition.” 91 In Hendricks’ case, the State’s argument
continued, the respondent had a mental disorder of “pedophilia” as defined by the DSM, and
therefore the constitutional requirement had been satisfied. 92 Hendricks’ attorneys countered that
the “mental abnormality” language in the Kansas statute, when examined closely, was nothing
more than “pseudoclinical terminology” useful for “after-invented rationalizations.” 93 Indeed, the
Kansas legislature used the language specifically to empower the state to detain people who did


Id. at 132.
Addington v. Texas, 441 U.S. 418 (1979).
Id. at 138.
Id. (citing KAN. STAT. ANN. 59-2902(h) (since repealed)) (defining a person with “mental illness” as one who:
“(1) [i]s suffering from a severe mental disorder to the extent that such person is in need of treatment; (2) lacks
capacity to make an informed decision concerning treatment; and (3) is likely to cause harm to self or others.”)
Kansas v. Hendricks, 521 U.S. 346 (1997).
Kansas v. Hendricks, Brief of Petitioner at 39.
Id. at 40.
Id. at 41.
Kansas v. Hendricks, Brief of Respondent at 21-22.


not have a “mental illness” since those with such illnesses could be committed under the
standard commitment statute. 94
Justice Clarence Thomas, who had dissented in Foucha five years earlier, 95 wrote the
majority opinion reversing the Kansas Supreme Court and upholding the SVP law under all three
constitutional challenges raised by Hendricks’ attorneys: that the law violated his rights under
the due process clause and under prohibitions of ex post facto laws and double jeopardy. 96 With
respect to the substantive due process analysis, which is the focus of this Article, Justice Thomas
noted that the Court has long recognized the importance of the state’s authority to detain through
civil proceedings “those who are unable to control their behavior and who thereby pose a danger
to the public health and safety.” 97 The Court has upheld civil commitment of this sub-population,
he explained, so long as proper procedures and standards are followed. The prior cases clearly
established that dangerousness alone would not satisfy due process requirements; it was only
when commitment statutes coupled such a requirement with “proof of some additional factor,
such as a ‘mental illness’ or ‘mental abnormality’” that the laws would not impermissibly
infringe on a person’s liberty interests. 98 There must be a “link,” therefore, between an
individual’s potential to commit future violence and “the existence of a ‘mental abnormality’ or
‘personality disorder’ that makes it difficult, if not impossible, for the person to control his
dangerous behavior.” 99 Under this framework, Justice Thomas reasoned, the Kansas SVP law
met these essential features to be consistent with notions of due process. The law limited the
potential class subject to commitment to those with either a “mental abnormality” or “personality
disorder,” which, he wrote, sufficiently “narrows the class of persons eligible for confinement to
those who are unable to control their dangerousness.” 100
Thus, Justice Thomas dispensed with the interpretation of Foucha and Addington as
requiring a specific finding of mental illness (as opposed to a broader requirement of any form of
“mental abnormality”) as a prerequisite to involuntary civil commitment. The term “mental
illness,” he explained, has no “talismanic significance.” 101 Rather, the critical factor to satisfy
substantive due process is that such commitment laws “limit involuntary civil confinement to
those who suffer from a volitional impairment rendering them dangerous beyond their
control.” 102 He noted that the Court had never required states to adopt particular medical terms
for involuntary commitment statutes. Legislatures, he stated, are not required to adopt terms that
“mirror those advanced by the medical profession.” 103 Since, in Hendricks’ case, his
“pedophilia” diagnosis met the statute’s mental abnormality requirement and Hendricks had
conceded in his own testimony that he lacked control over his urges, Hendricks’ “condition”
easily met the constitutional requirements for commitment. 104 Justice Thomas acknowledged that
the record on appeal included evidence of extensive controversy within psychiatry regarding

Id. at 22.
Foucha, 504 U.S. at 102–124 (Thomas, J. dissenting).
Kansas v. Hendricks, 521 U.S. 346 (1997).
Id. at 357.
Id. at 358.
Id. at 358.
Hendricks, 521 U.S. at 358-59.
Id. at 358 (emphasis added).
Id. at 359.
Id. at 360.


whether pedophilia was a mental illness; but he indicated that such debates only support the
conclusion that legislatures should be provided the “widest latitude in drafting” SVP laws. 105
Justice Thomas then considered Hendricks’ remaining constitutional arguments that the law
violated both the ex post facto and double jeopardy prohibitions in the Constitution and, based on
the categorization of SVP commitment as a civil, not criminal, proceeding, rejected them. 106
Justice Kennedy joined the majority in Hendricks but wrote separately to underscore that
the Kansas SVP law could not be used for retribution, only for treatment. 107 He noted some
concern with the prospect that, given that “medical knowledge” at that time did not hold great
promise for treatment of pedophilia, there was real potential for Hendricks and others to be
detained for life. 108 He also acknowledged that the Court was permitting states to proceed into
uncharted waters with these laws and that, “if it were shown that mental abnormality is too
imprecise a category to offer a solid basis for concluding that civil detention is justified, our
precedents would not suffice to validate it.” 109 As Kennedy’s concurrence makes clear, the
Hendricks opinion endorsed pure preventive detention, with benefits flowing only to the
community presumably protected from the committed person, as consistent with substantive due
Justice Breyer wrote for the four-justice minority and dissented only with respect to the
majority’s analysis of the ex post facto clause argument. He largely agreed with the majority’s
substantive due process conclusion but adopted a slightly different analysis: Hendricks’
condition, he wrote (characterizing pedophilia as a “serious mental disorder”), 110 was essentially
akin to the well-established “irresistible impulse” concept in criminal and preventive detention
law. 111 The medical evidence at the hearing (as well as Hendricks’ own admission), he wrote,
clearly established Hendricks’ inability to control his conduct, which brought him squarely
within the scope of the statute’s limited reach. 112 The debate within psychiatry regarding the
limits of mental illness, he observed, can serve to inform a state legislature’s course of action,
but does not mean it that may not set out to act at all. 113
Five years later, in Kansas v. Crane, the Court revisited the Kansas statute in the appeal
of a different respondent and clarified its volition-oriented requirement. In an opinion authored
by Justice Breyer, 114 the Court held that the volitional requirement was a substantive and
meaningful limitation on a state’s power to commit under the law. 115 It also held that a finding
that a person may be detained under the SVP law does not require a finding that the person

Id. at 360 n.3.
Id. at 361–71. See generally Wayne A. Logan, The Ex Post Facto Clause and the Jurisprudence of Punishment,
35 AM. CRIM. L. REV. 1261 (1998).
Id. at 371–73 (Kennedy, J. concurring).
Id. at 372 (Kennedy, J., concurring).
Id. at 373 (Kennedy, J., concurring).
Id. at 375 (Breyer, J. concurring).
Id. at 375–76.
Id. at 376.
Id. at 375. Justice Breyer’s analysis of the substantive due process issue was joined by Justices Stevens and
Souter. Justice Ginsberg, who did not author an opinion, joined only those parts of Breyer’s dissent on the ex post
facto analysis, and not his due process analysis. Id. at 373.
Kansas v. Crane, 534 U.S. 407 (2002).
Id. at 412–13.


entirely lacks any control over his behavior, since it is unlikely that the state could ever meet
such standard. 116 A person’s “inability to control his behavior” is not, Justice Breyer wrote, a
standard subject to a requirement of “mathematical precision.” 117 Rather, a state must provide
“proof of [the respondent’s] serious difficulty in controlling [his] behavior.” 118 And “this, when
viewed in light of such features of the case as the nature of the psychiatric diagnosis, and the
severity of the mental abnormality itself, must be sufficient to distinguish the dangerous sexual
offender whose serious mental illness, abnormality, or disorder subjects him to civil commitment
from the dangerous but typical recidivist convicted in an ordinary criminal case.” 119
Significantly, the Crane majority went on to comment on the role of courts in setting
standards in these cases in which a deprivation of a liberty interest turns on a requirement of
some kind of mental condition or impairment. The Court acknowledged that its reading of
Hendricks’ requirements “provides a less precise constitutional standard than would those more
definite rules for which the parties have argued.” 120 “But,” the Court reasoned, “the
Constitution's safeguards of human liberty in the area of mental illness and the law are not
always best enforced through precise bright-line rules.” The Court explained this reasoning as
For one thing, the States retain considerable leeway in defining the mental abnormalities
and personality disorders that make an individual eligible for commitment. For another,
the science of psychiatry, which informs but does not control ultimate legal
determinations, is an ever-advancing science, whose distinctions do not seek precisely to
mirror those of the law. 121
Justice Scalia argued in dissent that the majority’s conceptualization of the “volitional
impairment” requirement had gutted the core holding of Hendricks and creating an unworkable
framework for implementation of SVP laws. 122 Although his critique was based upon a view that
states should have more leeway in such commitments, he accurately identified some of key
problems with the Court’s analysis that rendered it a poor foundation for ensuring that the sweep
of these laws would not be too broad

Core Assumptions Revealed by the Stated Rationales of Hendricks and Crane

As the Crane opinion makes clear, the Supreme Court upheld the SVP experiment under
a number of core assumptions about how courts would determine that individuals were
appropriately subject to indefinite detention. The Court saw an indispensable role for psychiatry
both in informing the determinations of courts and fact-finders applying these laws and also in
supplying the proof of volitional impairment. As one federal appeals court later put it: “Crane
held that the Constitution requires findings that separate inability to control from unwillingness
to control, that is, to separate the sick person from the vicious and amoral one. The Court

Id. at 411–12.
Id. at 413.
Id. at 413.
Id. at 413 (emphasis added).
Id. at 413.
Id. at 413.
Id. at 416–25 (Scalia, J., dissenting).


thought this rule necessary to prevent fear of recidivism from leading to indefinite preventive
detention.” 123
In Hendricks and Crane, the Court rationalizes this unusual form of preventive detention,
one that looked very different from standard involuntary commitment, by reframing SVP
commitment so that it seems not so very different from other commitment laws. The essential
component of all involuntary commitments is the presence of pathology of a kind that limits
one’s ability to regulate one’s behavior and choices. By using nebulous terms such as
“impairment,” “abnormality,” or “condition,” and by restricting detention only to those who
presumably already have impaired free will, 124 the majority opinions in these cases suggest that
we are not really depriving such persons of their “liberty.” Thus in Hendricks, the Court writes:
The precommitment requirement of a ‘mental abnormality’ or personality defect is
consistent with the requirements of … other statutes that we have upheld in that it
narrows the class of persons eligible for confinement to those who are unable to control
their dangerousness. 125
Without this requirement, SVP statutes would amount to no more than punishment, thus
implicating all of the constitutional protections afforded to those subjected to punishment,
including prohibitions on ex post facto laws and double jeopardy. 126 Without it, the indefinite
detention sanctioned by SVP laws is wholly noxious to our views of liberty. As one commentator
observed: “Hendricks teaches that the role of the mental disorder element is to limit civil
commitment and prevent it from swallowing the criminal law.” 127
The constitutionality of the SVP laws and their place within our legal system as a
procedure that is not inconsistent with core U.S. values hangs entirely, then, upon the finding of
the presence of a mental illness so severe that it deprives one of the ability to exercise choice and
volition. But how would this identification—of those who are unable to control their behavior
specifically due to mental impairment—be made? If such a distinction could not be made
accurately, trial courts would run the risk of detaining un-impaired citizens simply on the basis of
a perceived or feared risk. The Court was evidently confident that such delineation could take
place, specifically, that a trial court could turn to the expertise of psychiatrists and other mental
health professionals to identify when such pathology was present and, moreover, that the
identification could be made with sufficient precision to distinguish the volitionally impaired
from the “dangerous but typical recidivist.” 128 Depending on “the nature of the [respondent’s]
psychiatric diagnosis and the severity of [his] mental abnormality,” 129 such professionals, it was

Varner v. Monohan, 460 F.3d 861, 864 (7th Cir.2006) (emphasis added).
David L. Faigman, Making Moral Judgments Through Behavioural Science: The 'Substantial Lack of Volitional
Control' Requirement in Civil Commitments, 2 LAW PROB. & RISK 309, 314 (2003). Faigman criticizes the
“volitional impairment” requirement of Hendricks-Crane on the basis that “[T]here is no empirical/scientific basis
for determining when an act was (or, much less, will be) a product of 'free will'. Free will is a normative construct
that has no corresponding operational definition that can be tested.”). Id. at 319.
Kansas v. Hendricks, 521 U.S. at 358 (emphasis added).
Faigman, supra note 124, at 314.
Janus, supra note 30, at 13.
Faigman, supra note 124, at 314
Kansas v. Crane, 534 U.S. 407, 413 (2002).


assumed, would be able to identify the key features to be considered when assessing whether
someone was a true predator.
Confidence in the ability of psychiatry to make such identifications and distinctions grew
at the same time as psychiatry was being given an increasingly prominent role in legal
proceedings. An important factor here was the appearance of the third edition of DSM, published
in 1980 by the APA. This edition (DSM-III), which shed most Freudian concepts from its
nosology and instead focused on a biological basis for classifying mental disorders, quickly
became a fixture in courtrooms. 130 Its science-and-research orientation, in contrast to the
psychoanalysis-inspired prior editions, suggested a new and more reliable role for psychiatrists
to help courts make more scientifically informed findings and, as it were, unlock the minds of
Psychiatric evidence, including diagnostic assessment, became ubiquitous and routinized
in a vast array of legal proceedings. 131 Judges and others in the legal system were accustomed to
seeing mental health professionals offer opinions on a range of legal questions from parenting
and the extent and causes of psychological injuries to insanity, commitment, and sentencing.
These experts played a critical role in many cases, informing fact-finders on some of the most
difficult decisions , including whether a person was criminally responsible, and on decisions with
some of the most serious implications, such as which parent should raise a child.
On the urging of prosecutors, courts extended the role of psychiatric evidence from assessment
of past and present mental states to testimony bearing on the prediction of future conduct, and
courts became very protective of their continued ability to admit and consider such testimony.
In a crucial decision, Barefoot v. Estelle in 1983, the Supreme Court upheld the
admissibility of expert testimony on future dangerousness in the sentencing phase of a death
penalty case. 132 The State of Texas offered the testimony of two psychiatrists who stated, in
response to hypotheticals regarding the defendant, that the defendant “would probably commit
further acts of violence and represent a continuing threat to society.” 133 The Court majority
rejected the defendant’s arguments and Justice Blackmun’s dissenting opinion that, as
established by research, psychiatrists, with an accuracy rate of only about 1 in 3, were poor in
predicting future violent conduct. 134Significantly here, the APA, in its amicus filing in Barefoot,
sided with the defendant. Psychiatrists, it stated, have no expertise at predicting dangerousness
and are really no better than anyone else in doing so.
One of the Barefoot majority’s rationales in rejecting these arguments was that exclusion
of prediction testimony in this context would limit uses of such psychiatric testimony in other
contexts, including that of involuntary commitment: “Acceptance of petitioner's position that
expert testimony about future dangerousness is far too unreliable to be admissible would
immediately call into question those other contexts in which predictions of future behavior are



Barefoot v. Estelle, 463 U.S. 880 (1983).
Id. at 884.
Id. at 898-903.


constantly made.” 135 The majority contended that the tools of the adversarial process, such as
cross-examination and contrary expert opinion, would be sufficient check on the reliability of
such predictions. 136 Thus the Supreme Court paved the way for psychiatrists testifying in SVP
proceedings to have a central role in predicting dangerousness.
Perhaps in light of the outcome in Barefoot, and in contrast to the role played by
psychiatrists at the time the sexual psychopath laws were passed in the first half of the 20th
century, the psychiatric establishment was quick to distance itself from the SVP proposals right
from the inception of the trend. In 1995, the Washington State Psychiatric Association submitted
an amicus brief in the Young v. Weston litigation indicating that nothing in the state’s SVP
statute restricted its reach to those who were identified as mentally ill by psychiatry. Rather, in
limiting the application of the law to “sexually violent predators,” it established nothing more
than an “unacceptable tautology.” 137
The APA made similar arguments in the amicus brief it submitted in support of
Hendricks’ position before the U.S. Supreme Court. There, the APA argued that legislatures
should not be free to define “mental illness” freely; otherwise, it warned, “the limits on
deprivations of liberty to protect the public safety would quickly disappear.” 138 The APA also
argued that the definition of mental illness for involuntary commitment purposes should not be
tied to the diagnoses contained in the Diagnostic and Statistical Manual. As they explained, the
DSM’s “classification schemes are developed …to serve diagnostic and statistical functions,
forming a common (and always imperfect) language for gathering clinical data and for
communication among mental health professionals.” 139 The APA’s elaboration of this argument
is striking:
[The diagnoses set forth in the DSM are not] designed to identify those subject to various
legal standards, such as those for involuntary confinement. Thus, the authors of DSM-IV
caution that “[i]n most situations, the clinical diagnosis of a DSM-IV mental disorder is
not sufficient to establish the existence for legal purposes of a ‘mental disorder,’ ‘mental
disability,’ ‘mental disease,’ or ‘mental defect.”’ DSM-IV at xxiii. The authors further
caution that “a DSM-IV diagnosis does not carry any necessary implication regarding the
individual’s degree of control over the behaviors that may be associated with the
disorder.” Id. Not all individuals who come within a DSM-IV category suffer an
impairment that diminishes their autonomy, much less one justifying involuntary
confinement for the individual's own good. 140


Id. at 898.
Id. at 898-99. Ten years later, Justice Blackmun authored the majority opinion in Daubert v. Merrell Dow
Pharm., Inc., 509 U.S. 579 (1993), and noted that limitations of such tools on preventing unreliable expert testimony
from being given undue weight by a fact finder and therefor imposing on trial judges the responsibility of being a
“gatekeeper” to exclude such unreliable testimony from being admitted.
Young v. Weston, 898 F.Supp. 744, 750 (D. Wash. 1995) (quoting amicus brief of Washington Psychiatric
Brief for the American Psychiatric Association as Amicus Curiae in Support of Leroy Hendricks at 21, Kansas v.
Hendricks, 521 U.S. 346 (1997).
Id. at 22.
Id. at 23 (emphasis added).


In upholding the Kansas statute and the central model of SVP laws, the Supreme Court
majority rejected psychiatry’s strong words of caution about a law that drew a line ostensibly
based upon the identification of a mental disorder but couched in language completely alien to
the field that oversees such identifications and which, at the same time, conferred upon that field
a central role in ensuring the constitutionality of the application of such laws in the future.

The Spread of SVP Laws and Their Impact

The drafters of the original SVP law in the state of Washington apparently thought that
the imposition of such indefinite commitment would be limited to exceptional cases like those of
Earl Shriner or Leroy Hendricks, where the risk of recidivism was unquestionably high and
indications of committing future violence were obvious. 141 However, the numbers who have
been committed under such SVP laws, in Washington and elsewhere, suggest that they have been
subject to much broader use than in such cases. At the same time, there are important indications
that the laws have not operated as initially expected, with regard either to those committed under
them or to the protection of society.
With the question of the constitutionality of SVP laws resolved by the Court’s opinion in
Hendricks, several other states followed the lead of Washington and Kansas. Today, a total of
twenty states have adopted SVP laws. 142 The U.S. Congress adopted an SVP commitment
scheme as part of the “Adam Walsh Child Protection and Safety Act.” 143 The federal law applies
to those incarcerated by the U.S. Bureau of Prisons, so it involves a somewhat different set of
respondents, since most sexual abuse and assault cases are prosecuted in state courts. 144
However, one class that is prevalent in federal prisons are those charged with possession of child
pornography and, in some instances, a pornography charge serves as a predicate offense 145 or
even the sole predicate offense 146 for an SVP commitment under the Adam Walsh Act. 147 The
law was immediately challenged as being enacted outside of Congress’s authority, and that

Boerner, supra note 26, at 566.
ARIZ. REV. STAT. ANN. §§ 36-3701–3717 (2009); CAL. WELF. & INST. CODE §§ 6600–6609.3 (West 1998 & Supp.
2009); D.C. CODE §§ 22-3803–3811 (2001); FLA. STAT. ANN. §§ 394.910–.932 (West 2006 & Supp. 2009); 725 ILL.
COMP. STAT. ANN. 207/1–99 (West 2008); IOWA CODE ANN. §§ 229A.1–.16 (West 2006 & Supp. 2010); KAN. STAT.
ANN. §§ 59-29A01–29A22 (2005 & Supp. 2009); MASS. GEN. LAWS ANN. CH. 123A, §§ 1–16 (West 2003 & Supp.
2009); MINN. STAT. ANN. § 253B.185 (West 2007 & Supp. 2009); MO. ANN. STAT. §§ 632.480–.513 (West 2006);
NEB. REV. STAT. §§ 71-1201–1226 (2009); N.H. REV. STAT. ANN. §§ 135-E:1–24 (LexisNexis Supp. 2009); N.J.
STAT. ANN. §§ 30:4-27.24–.38 (West 2008); N.Y. MENTAL HYG. LAW §§ 10.01–.17 (McKinney Supp. 2010); N.D.
CENT. CODE §§ 25-03.3-01–24 (2002); S.C. CODE ANN. §§ 44-48-10–170 (2002 & Supp. 2008); VA. CODE ANN. §§
37.2-900–920 (2005 & SUPP. 2009); WASH. REV. CODE ANN. §§ 71.09.010–.903 (West 2008 & SUPP. 2010); WIS.
STAT. ANN. §§ 980.01–.14 (West 2007 & Supp. 2008).
Pub. L. No. 109–248, 120 Stat. 587 (2006), codified at 18 U.S.C. §§ 4247–48.
A notable exception is prosecution of crime committed in “Indian Country,” and one commentator has raised
concerns about the large number of Native Americans who have been subject to commitment under the federal law.
Karen Franklin, Appellate Court Rejects "Past As Prelude" Myth, IN THE NEWS (Feb 12, 2014)
See, e.g., U.S. v. Timms, 664 F.3d 436 (4th Cir. 2012); United States v. Wetmore, 766 F. Supp. 2d 319 (D. Mass.
2011) aff'd, 700 F.3d 570 (1st Cir. 2012) cert. denied, 133 S. Ct. 1652, 185 L. Ed. 2d 631 (U.S. 2013).
See, e.g., U.S. v. Volungus, 730 F.3d 40, 43–46 (1st Cir. 2013).
The Walsh Act provides that a federal prisoner can be “certified” as an SVP under the statute without a judicial
determination. The Court of Appeals for the Fourth Circuit has held that such determination must be subject to
review “within a reasonable period of time”147 and failure to provide access to such determination may constitute a
deprivation of due process. U.S. v. Broncheau, 645 F.3d 676, 687 n.10 (4th Cir. 2011).


litigation was finally resolved when the Supreme Court upheld the law in U.S. v. Comstock, in
2010. 148
As New York State was about to implement its own SVP law in 2007, the New York
Times published a three-part series examining the SVP commitment programs already in place
across the country. 149 The series’ authors made several findings that suggested that the operation
of SVP programs are falling far short of their promise. Notably, although nearly 3,000 people
had been committed under the 19 state SVP laws then in effect, (1) the programs are not
committing the most violent and dangerous offenders, since some rapists are being released
while exhibitionists are being committed; (2) the treatment programs are largely ineffective; (3)
few of those committed are ever released; and (4) few states have developed effective programs
for monitoring those men who are released. 150 In spite of these problematic findings,
commitment programs continue to expand. More recently, a 2013 survey of 18 state-based SVP
programs found that 4779 individuals are presently committed, with an additional 861 in
detention awaiting the outcome of SVP proceedings. 151
The expanding reach of SVP programs originates, in part, in the fact that states can and
do base SVP commitment petitions on a wide range of predicate offenses. In many states, SVP
laws permit indefinite commitment based on juvenile offenses, on offenses for which the person
was acquitted on the basis of insanity, or on uncharged conduct, all of which serve to further
widen the pool of those potentially subject to the laws. In Minnesota, for example, more than 7%
of those committed under that state’s SVP program were never convicted of an adult crime prior
to their commitment. 152 Moreover, courts have interpreted most of the statutory requirements for
prior convictions or criminal offenses quite broadly. Thus indefinite commitments as sexually
violent predators have been based on sexual offenses that do not involve any physical contact
with a victim, such as exhibitionism, indecent conduct, or possession of pornography. 153
The high number committed under SVP statutes also suggests that it may be difficult,
although not impossible, for a respondent to prevail in an SVP trial. 154 The state enjoys several
advantages in the conduct of such trials. The holding in Hendricks that the SVP schemes are civil
rather than criminal in nature, and therefore do not implicate either the ex post facto or double
jeopardy prohibitions, has had implications significant for shaping the implementation of the

United States v. Comstock, 560 U.S. 126, 130 S. Ct. 1949 (2010).
Monica Davey & Abby Goodnough, Doubts Rise as States Hold Sex Offenders After Prison, N.Y. TIMES (March
4, 2007).
Deirdre D’Orazio, SOCCPN [Sex Offender Civil Commitment Program Network] Annual Survey of Sex Offender
Civil Commitment Programs 2013
Chris Serres, Minnesota Sex Offenders: Are They Really the 'Worst of the Worst'? STAR TRIBUNE (Dec. 2, 2013)
( ) (profiling the case of a developmentally disabled man who was
committed at the age of 19 for acts of child molestation that he committed before the age of 14). Courts in several
other states, by contrast, have held that a sex offense committed as a juvenile cannot be a predicate crime for an SVP
commitment. See In re Detention of Geltz, 40 N.W.2d 273, 279–80(Iowa 2013) (reviewing case law on question).
See, e.g., Comm. v. Sauve, 53 N.E.2d 178 (Mass. 2011) (public masturbation in front of adult women).
Commitments have also been based on attempted sexual abuse or assault, where there was no actual physical contact
with a victim.
I have not located any empirical studies of rates of success of SVP commitment petitions.


law, and particularly the proceedings. Since, according to Hendricks, these are civil proceedings,
the protections afforded to criminal defendants under the Sixth Amendment and elsewhere do
not apply. These protections include rights with respect to burden of proof, 155 competency, 156
effective assistance of counsel, 157self-incrimination, 158 and confronting witnesses. 159
The promise of treatment under SVP statutes is tied to the mental abnormality rationale
of all forms of involuntary commitment. However, the treatment outcomes from SVP programs
have been uneven. Scores of those committed as SVPs receive little to no treatment whosoever
and some states have been involved in protracted litigation regarding access to treatment. One
such case is that brought by Andre Young, who initiated the initial challenge to Washington’s
law. (By the time his case reached the U.S. Supreme Court, the Court had already decided the
Hendricks case.) In dismissing Young’s challenge based upon an “as applied” theory, the Court
noted in dictum that, if a person is detained for purpose of “to incapacitate and treat,” then it
follows that “due process requires that the conditions and duration of confinement under the Act
bear some reasonable relation to the purpose for which persons are committed.” 160 However,
such language has provided no real guidance to lower courts evaluating right-to-treatment
claims. 161 Most state SVP laws do not offer immunity from prosecution for disclosure of
criminal conduct so the threat of self-incrimination is a real one. 162 Further, social scientists have
yet to reach anything approaching a consensus on the efficacy, in terms of preventing recidivism,
of the various kinds of inpatient treatment programs administrated to SVPs. 163
The burden on an SVP respondent, once committed, to obtain release from detention is
considerable. Proving a decrease in risk of re-offending is difficult, particularly when one has


See infra notes 292–298 and accompanying text. Several SVP laws, including Kansas and Washington’s, require
proof beyond a reasonable doubt.
See, e.g., In re Detention of Morgan, 253 P.3d 394, 403 (Wash. Ct. App. 2011); In re Commitment of Luttrell,
754 N.W.2d 249 (Wis. Ct. App. 2008).
See Michael Perlin & Heather Cucolo,“Far From the Turbulent Space”: Considering the Adequacy of Counsel in
the Representation of Individuals Accused of Being Sexually Violent Predators, 20–25
( ). However, all laws provide for some access to
Allen v. Illinois, 478 U.S. 364, 373, 106 S. Ct. 2988, 2994, 92 L. Ed. 2d 296 (1986).
See, e.g., United States v. Abregana, 574 F. Supp. 2d 1123, 1140 (D. Haw. 2008).
Seling v. Young, 531 U.S. 250, 265,121 S.Ct. 727 (2001).
Ever where some form of treatment is offered, many detainees refuse to participate in the treatment offered
because a condition of such treatment is full disclosure (checked by polygraph tests) of all sexual offenses, including
those which the detainee had previously denied under oath or for which the detainee was never charged or
convicted, thus exposing him to potential further criminal liability or extended commitment. Jeslyn Miller, Sex
Offender Civil Commitment: The Treatment Paradox, 98 CAL. L. REV. 2093, 2095 (2010); La Fond, supra note 25,
at 167–69
EWING, supra note 13, at 56. The Supreme Court has held that conditioning the constitutionally required
treatment on such disclosure (and removing privileges and increasing the level of detention as a penalty for refusing
treatment) does not run afoul of the Fifth Amendment’s guarantee against compelled self-incrimination. McKune v.
Lile, 536 U.S. 24 (2002). Justice Kennedy, writing on behalf of the same 5-4 majority that upheld the Kansas SVP
law in Hendricks (check this) that the treatment program did not truly compel self-incrimination because the
penalties imposed for refusing to participate in the treatment program were not severe and the state had a valid
objective in encouraging rehabilitation and deterring future sexual offenses by leaving the possibility of future
prosecution. Id. at 33–36.
EWING, supra note 13, at 52–55.


been denied opportunities either to demonstrate self-restraint or to receive effective treatment. 164
As a result, thousands of people who have been detained for lengthy periods of time have little
likelihood of ever being released. 165 Surveys of release rates suggest that most individuals are
committed for extended periods of time. 166 The New York Times’ 2007 study revealed that, of the
nearly 3,000 individuals who had been committed nationwide under SVP laws, only 50 had been
released on the assessment by a clinician and state-appointed evaluator that they were “ready”
for release. 167 The near impossibility of release means that there is a growing and aging group of
people who are living out their lives in detention. 168 The Times authors noted that Leroy
Hendricks, who was 72 years old in 2007, “spen[t] most days in a wheelchair or leaning on a
cane, because of diabetes, circulation ailments and the effects of a stroke” 169 and that those who
remained in detention included a 102-year-man with poor hearing.
Minnesota’s SVP program, established in 1993, provides perhaps the most extreme
example of the challenges of obtaining release. Between the program’s enactment date and 2012,
635 people (nearly all men) were committed under that state’s SVP law. 170 Not one was released
until 2012. 171 That state’s program has come under criticism for its failure to provide adequate
treatment for detained offenders as well as its stringent release requirements. 172 In 2012 the
British High Court refused to extradite a sex offender to Minnesota who faced possible SVP
commitment on the basis that such commitment would constitute a “flagrant denial” of his
human rights. 173 More recently, the U.S. District Court for District of Minnesota permitted a
class action lawsuit brought on behalf of those committed in the state’s program to go forward.
In its decision, the court noted: “Given the prison-like conditions described by Plaintiffs, and the
lack of treatment and essentially no-exit regime alleged in this case, it may well be that, with a
fully developed record, the Court will find the totality of the MSOP system to be unacceptably


Prentky et al., supra note 7, at 380–81 (noting that many programs are grossly inadequate, while at the same
time, a person’s lack of improvement in treatment is often used as a basis to extend their detention). See, e.g., In re
Commitment of West, 800 N.W.2d 929, 947–48 (Wisc. 2011) (holding that placing burden on committed person to
prove by clear and convincing evidence that he is no longer a “sexual violent person” in order to be released from
commitment does not violate due process).
La Fond, supra note 25, at 166–70; EWING, supra note 13, at 22.
Prentky et al., supra note 7, at 380. (“Those discharged or released range from 0 in North Dakota, New Jersey,
and Iowa to 1 in Minnesota, 4 in Massachusetts, 6 in Missouri, and fewer than 20 in Washington, Kansas, Illinois,
and Florida (Lieb &Gookin, 2005). The only states that have released a sufficient number of committed offenders to
permit a follow-up are Arizona (221), California (67), and Wisconsin (56).”); WASHINGTON STATE INSTITUTE FOR
PREDATORS: 2006 UPDATE, REVISED (2007) available at
Another 115 people had been released because of “legal technicalities, court rulings, terminal illness or old age.”
Davey & Goodnough, supra note 149.
Several studies have noted that the risk of recidivism for sexual violence decreases significantly for those over
the age of 60. U.S. v. Wilkinson, 646 F.Supp.2d 194, 208 (2009) (citing R. Karl Hanson, Recidivism and Age:
Follow- Up Data From 4,673 Sexual Offenders, 17 J. INTERPERSONAL VIOLENCE 1046, 1059 (2002)).
Davey & Goodnough, supra note 149.
Rupa Shenoy, Families of Sex Offenders Find Hope in Clarence Opheim's Release, MPRNews (Mar. 5, 2012)
( )
Mary Lynn Smith & Dave Hage, Minnesota Sex-Offender Lawsuit Takes Step Forward, STAR TRIBUNE (July 25,
John Aston, Court blocks Shawn Sullivan's U.S. Extradition, THE INDEPENDENT (June 28, 2012).


and unconstitutionally punitive.” 174 In short, as one commentator puts it in reference to the
realities of SVP laws: “Involuntary commitment is both incarceration and exile.” 175
Since so many that are committed under SVP laws remain in detention, these programs
are becoming increasingly financially burdensome on the states that adopted them. Estimates of
the cost to house each detainee range from $94,000 to $175,000 annually. 176 These figures do not
including capital expenditures to build new facilities for SVP program or the costs of litigation
for the initial petition or for requests for release. 177 One study suggests that the cost of detaining
a sex offender is four times more expensive than incarcerating a prisoner. 178
Notwithstanding the failure of SVP programs to achieve their ostensible purposes and the
extreme financial burden they impose on states prosecuting them, states continue to identify
individuals for SVP commitment at the conclusion of their prison sentences. Since the public has
become accustomed to SVP detention as the standard course for those convicted of sex crimes,
legislatures appear to have boxed themselves in. The likelihood of public outrage at the idea of
releasing “sexual predators” or not permitting their further detention makes such options
politically unfeasible. Indeed, a Florida newspaper criticized that state for not detaining enough
people under its SVP program, and the legislature responded by loosening the commitment
criteria even further. 179
It should be emphasized here that one consequence that SVP laws appear not to have had
is decreasing the overall incidence of sexual violence in those states that have enacted such laws.
It is difficult to assess empirically whether there are broad public safety benefits to SVP
programs (that is, beyond ensuring that specific individuals have no access to anyone outside of
the SVP detention facility), but some researchers have attempted to do so. In one recent study,

Karsjens v. Jesson, 2014 WL 667971, *25 (Civil No. 11–3659 (DWF/JJK) D. Minn. Feb. 20 2014)
Jenny Roberts, The Mythical Divide Between Collateral and Direct Consequences of Criminal Convictions:
Involuntary Commitment of "Sexually Violent Predators," 93 MINN. L. REV. 670, 708 (2008).
Davey & Goodnough, supra note 149. The Times study noted that wheelchairs, walkers, and high blood pressure
medication are among the growing costs for an increasing aging population of people in SVP detention. Florida’s
SVP detention center filled 229 prescriptions for arthritis medication one month, and 300 for blood pressure and
other heart problems. Id,
EWING, supra note 13, at 57–59. The latter include costs of court-appointed counsel and expert witnesses, which
are estimated to double those costs. Media reports have documented that the use of expert testimony for such
proceedings constitutes a significant portion of expense for such programs. For example, one 2010 report found that
the State of New York had spent $3 million paying for experts for both the State and respondents since that state’s
SVP program was launched in 2007. See also Gary Craig, Expert Opinion Among Civil Commitment's High Costs,
( ); Sally Kestin and
Dana Williams, Experts Cash in on Predator Law, SUN SENTINEL (Aug. 21, 2013) ( ); Christine Willimsen, State Wastes Millions Helping Sex Predators Avoid
Lockup, SEATTLE TIMES (Jan. 21, 2012).
Davey & Goodnough, supra note 149.
In 2013, the South Florida Sun Sentinel released a series of articles, collectively titled “Sex Predators
Unleashed,” that was highly critical of how many convicted sex offenders were not being committed under that
state’s SVP law and calling on state lawmakers to make it easier to detain such offenders. Sally Keston and Dana
Williams, Sex Predators Unleashed, SO. FLA. SUN SENTINEL (Aug 18, 2013). A follow up article in late December
described the Florida Legislature’s response to the paper’s investigation as crafting a “comprehensive overhaul” of
the state’s SVP law. Sally Keston and Dana Williams, Investigation Spurs Reform Of Sex Offender Laws, SO. FLA.
SUN SENTINEL (Dec. 29, 2013).


researchers concluded that “SVP laws have had no discernible impact on the incidence of sex
crimes.” 180 It may be added that, by enacting SVP laws and prosecuting these expensive
programs, policymakers are often shifting resources away from other arguably more relevant and
effective programs, including those aimed at providing probation officers, preventing domestic
violence and child abuse, and treating sex offenders. 181



As discussed in the prior section, the language in the Hendricks and Crane decisions
confirming the constitutionality of SVP laws confers progressively broad discretion on courts in
their application of statutory terms to meet the due process requirement of mental abnormality.
The Supreme Court reasoned in Crane that the science of psychiatry is “ever-advancing” and its
“distinctions do not seek precisely to mirror those of the law.” 182 It also made clear that it was
not going to establish specific principles to guide lower courts and legislatures, reasoning that
“the Constitution's safeguards of human liberty in the area of mental illness and the law are not
always best enforced through bright-line rules." 183 In effect, it invited policymakers and courts to
experiment with their approaches to establishing eligibility for SVP commitment.
The Hendricks-Crane rationale assumes that, however legislatures chose to precisely
define the contours of each state’s SVP commitment laws, mental health professionals would
reliably identify those whose medical conditions put them at higher risk of committing sexual
violence due to volitional impairment, thus ensuring that such laws would not sweep too broadly.
By framing the standards for commitment in terms of mental disorder and making findings of
volitional impairment from such disorders a constitutional requirement, the legislatures and court
have assigned psychiatry a central role in the implementation of SVP laws: providing expert
opinion on the likelihood future sexual violence stemming from mental conditions in specific
In effect, the constitutionality of the new scheme of SVP laws was saved by the promise
of psychiatry. The Court’s rationale is valid, however, only if it is based upon accurate
assumptions about the contributions psychiatry could make to ensure that the SVP laws did not
overreach. Justice Kennedy made that point explicitly in his Hendricks concurrence when he
noted that, if it turns out that “mental abnormality is too imprecise a category to offer a solid
basis for concluding that civil detention is justified.” then the constitutionality of the SVP
scheme would again be called into question. 184 It follows that, if the very concept of a mental
health predicate is highly imprecise, then the entire model of SVP law similarly falls short of
meeting due process requirements
Two major problems are involved here. One is that the place of sexual deviance is
unclear in the classification of psychopathology. The other is that psychiatry does not operate in

Tamara Rice Lave & Justin McCrary, Do Sexually Violent Predator Laws Violate Double Jeopardy and
Substantive Due Process: An Empirical Inquiry, 78 BROOKLYN L. REV. 1391, 1392 (2013) (emphasis added).
Id. at 1426–27; Good & Burstein, supra note 46, at 38.
Kansas v. Crane, 534 U.S. 407, 413 (2002).
Id. at 373 (Kennedy, J., concurring).


terms of predicting behavior. It is a profession whose orientation is to identify the disordered
primarily for the purposes of treating them—for relieving their suffering and improving their
functioning. As noted earlier, the psychiatric profession never claimed that it had the knowledge
or instruments to identify those at an especially high risk of committing acts of sexual violence.
and the past 24 years of SVP proceedings indicate that the Court’s evident assumption that it
could make that crucial identification was misplaced. The years since those opinions have, in
fact, borne out the warnings of the APA in its Hendricks amicus brief. What has become clear is
that the Supreme Court based its ruling regarding the class of “sexually violent predators” on a
legal, not psychiatric, construct, and its assignment of the role of determining such classification
to the field of psychiatry involved a distorted view of that field with dire consequences for those
targeted by the statutes.
This section will first review the historical and current approaches within psychiatry to
identifying disorders involving sexual arousal. Next it will examine how such approaches
became significantly distorted in SVP proceedings under the framework set forth in HendricksCrane. Particular attention will be given to the problem of reliance on psychiatry to predict
sexual violence. Finally, this section will review some the attempted fixes to this imperfect fit
between psychiatry and law in regard to prediction, primarily through proposed revisions to
psychiatric diagnoses and use of alternative methods of prediction.

Psychiatry’s View of Diagnosing and Predicting Sexual Violence

The holdings in Hendricks and Crane assigned psychiatric experts a central,
indispensable role in the prosecution of SVP commitments. The State cannot obtain an order for
detention without proving dangerousness, and such dangerousness must be couched in terms of
abnormality, or a “mental disorder that has some medical legitimacy.” 185 When experts speak of
mental pathology, particularly in courtrooms, they tend to do so in terms of diagnoses. However,
the diagnoses that, on their face, appear to identify those individuals who present the greatest
threat of sexual dangerousness are not in accord with the conceptualization of mental
abnormality or mental disorder evidently contemplated in the Court’s opinions and the SVP
statutes they upheld.

Role of Diagnosis and the DSM Generally in Psychiatric Assessment

As an initial matter, even the broad concept of “mental disorder” does not enjoy a
consensus definition within psychiatry. Beginning with the third edition the Diagnostic and
Statistical Manual of Mental Disorders (“DSM”), the APA’s standardized nosology, has offered
a definition for this term, although such definition has varied over the years. In one recent
edition, the editors acknowledged that, in making a diagnosis, the line between disordered and
non-disordered is elusive and variable: “The concept of mental disorder, like many other
concepts in medicine and science, lacks a consistent operational definition that covers all
situations.” 186 Recent editions of the DSM also feature cautionary language about using the

Eric S. Janus & Robert A. Prentky, Sexual Predator Laws: A Two-Decade Retrospective, 21 FED. SENTENCING
REP. 90, 93 (2008).
text Rev. 2000) (“DSM-IV-TR”) at xxx.


manual’s diagnostic classifications in legal situations, where such line-drawing has far greater
implications than in clinical settings. The “Cautionary Statement for Forensic Use” in the most
recent edition, published in 2013, advises: “When DSM-5 categories, criteria, and textual
description are employed for forensic purposes, there is a risk that diagnostic information will be
misused or misunderstood.” 187
However, the Supreme Court clearly anticipated that experts testifying in SVP cases
would frame their opinions, at least in part, in terms of diagnosis. Crane referenced diagnosis
specifically: “[W]hen viewed in light of such features of the case as the nature of the psychiatric
diagnosis, and the severity of the mental abnormality itself, [the proof] must be sufficient to
distinguish the dangerous sexual offender whose serious mental illness, abnormality, or disorder
subjects him to civil commitment from the dangerous but typical recidivist convicted in an
ordinary criminal case.” 188 And the Hendricks majority noted that the “mental abnormality”
requirement was met in Hendricks’ case because the respondent had a “disorder” listed in the
DSM. 189 However, in neither opinion did the Court indicate the full range of or kind of diagnoses
that would be sufficient for purposes of a constitutionally permissible preventive detention. In
the absence of such direction, uncertainties abound for those in both law and psychiatry. Indeed,
it appears that virtually any diagnosis attached by a single mental health professional could
suffice to justify indefinite commitment of someone as a sexually violent predator. 190
Psychiatric diagnoses have often been presented in insanity defense cases, but there are
important differences between these settings. In the context of determining criminal
responsibility, the diagnosis is one element of the reconstruction of a past frame of mind at a
given moment in time. The reconstruction is less dependent upon a specific label than on an
overall assessment of how the person’s mind was working (or not) at such moment. 191 More
importantly, in insanity defense cases, it is usually the defendant himself who puts a diagnosis in
evidence through his own expert testimony as part of a defense he raised. Absent a defendant’s
choice to assert such defense, there is no role for psychiatric testimony, including diagnoses, at
trial. By contrast, in SVP proceedings, because the diagnosis of mental abnormality is the basis
for adhering to due process protections, it is the linchpin for the deprivation of liberty. It is how
we rationalize preventive detention for a subset of the population. And if that is the case, then the
diagnosis itself must align with due process principles in court proceedings.
In Hendricks, the Court noted the lack of consensus among psychiatrists regarding where
the line between ill and not-ill, and also the lines among the specific illnesses, can be drawn. This
led the majority to conclude that legislatures, in framing the laws, and judges reviewing the
evidence in individual cases, should do the line-drawing. But the lack of consensus here should

ed. 2013) (“DSM-5”). The DSM-IV-TR’s “cautionary statement” was quoted in the APA’s amicus brief to the Court
in Hendricks, as noted supra note 140 and accompanying text.
Kansas v. Crane, 534 U.S. 407, 413 (2002) (emphasis added).
Kansas v. Hendricks, 521 U.S. 346, 360 (1997).
Allan Frances, et al., Defining Mental Disorder When It Really Counts: DSM-IV-TR and SVP/SDP Statutes, 36 J.
In fact, Stephen Morse has argued that insanity opinions could be based entirely on the defendant’s capacity at
the moment of the crime using descriptive rather than diagnostic terms. Stephen Morse, Crazy Behavior, Morals &
Science: An Analysis of Mental Health Law, 51 S. Cal. L. Rev. 527, 604–13 (1978).


have signaled to the courts that, for such a massive deprivation of liberty as an indefinite
preventive detention (for terms far longer than in the standard involuntary hospitalization
context), and especially where the burden is on the respondent to petition for release and show
that he has become well enough to be at large, the deciding factor should not be so variable and
subjective. Such reasoning also failed to account for the high degree of deference courts
generally grant to mental health experts and the limited ability of courts and juries to assess the
reliability of such expert’s opinions. The Court’s rulings, when implemented in the context of the
on-the-ground realities of trials, paved the way for scores of SVP commitments to be based upon
expert opinions with highly dubious scientific foundation.

Origins of Lack of Consensus Regarding Relation of Pathology to Sexual Deviance

In the case of SVP laws, mental health professionals are asked to make a very specific
finding of dangerousness: the person must be at risk for committing sexual violence, not any
kind of violence. Most civil commitment statutes have a blanket “harm to self or others”
requirement, which provides for a range of prognostication. The requirement of the specific risk
in SVP laws leads many to assume there must be the specific diagnosis tied to that specific risk.
Given this central role assigned to psychiatric diagnosis in SVP proceedings, we must consider
carefully what psychiatry has to say about the underlying pathology of those who engage in
sexual violence.
The history of pathologizing sexual attitudes and conduct is long, complicated and
inextricably caught up with cultural and ethical views—often tacit—that construct deviance and
perversion in contrast with a presumed normality. Other scholars have set out this history in
some detail, 192 and I will only summarize some key developments here, particularly as they
pertain to implications for the SVP statutory schemes. French philosopher Michel Foucault
compellingly argued that much that has been labeled pathology is in fact nothing more than
deviance from social norms predominant at a given time, including norms regarding sexuality
and proper gender behavior. 193 Contemporary historians of psychiatry generally regard supposed
pathological “conditions” as “constructions,” and often quite problematic ones. 194
Although Western societies, particularly through religious 195 and legal-political
institutions, had long identified and condemned a range of sexual behaviors as deviant, the
notion of such conduct as evidence of mental illness did not arise until the mid-19th century with
the increasing authority of psychiatry. 196 As new works about sexual deviance and perversion
appeared in the European medical literature, the criminalization of specific sexual acts also


See, e.g., Andreas De Block & Pieter R. Adriaens, Pathologizing Sexual Deviance: A History, 50 J. of SEX
RESEARCH, 276, 277 (2013). See generally JESSE BERING, PERV: THE SEXUAL DEVIANT IN ALL OF US (2013).
De Block & Adriaens, supra note 192, at 277.
Id. (“[P]sychiatrists’ and sexologists’ descriptions of new pathologies or types of persons should not be
considered as discoveries but rather as inventions or constructions.”).
The word “perversion” originates from a broader term “used to denote an aberration or a deviation from a divine
norm: any act that violated the laws of God was considered a perversion.” Id. at 278.


became more widespread. 197 In time, some psychiatrists were led to criticize the punishment of
such behaviors and to recommend that they be eliminated through treatment instead. 198
The publication in 1886 of Austrian psychiatrist Richard von Krafft-Ebing’s
Psychopathia Sexualis, which set forth a medically detailed account of specific pathologies, is
considered a watershed moment in the medicalization of sexual deviance. 199 The Psychopathia
Sexualis differed from prior accounts in its focus on the psychological origins of such conduct,
based in an individual’s personality, rather than on its origins in an individual’s anatomy. 200
Although the original work included extensive classification of pathological sexual feelings and
behavior, it was only in later versions that Kraft-Ebbing discussed pedophilia and other forms of
“paraphilia,” that is, sexual arousal not from heterosexual intercourse with adults but from nonstandard sources, such as objects, animals, contexts, and children. Krafft-Ebing, himself a
forensic psychiatrist, noted the implications of his research for criminal law, but he observed that
classifying conduct as normal, perverted, or criminal was not a simple matter. 201
Sigmund Freud, though clearly influenced by Krafft-Ebing’s approach, took a somewhat
different tack regarding sexual deviance versus normality. Most individuals, Freud maintained,
are “polymorphously perverse” during childhood, and a range of sexual interest remains quite
common among the population. 202 He wrote: “However infamous they may be, however sharply
they may be contrasted with normal sexual activity, quiet consideration will show that some
perverse trait or other is seldom absent from the sexual life of normal people.” 203 Accordingly,
these desires signal dysfunction only when they are the source of compulsion, fixation, and
exclusiveness such that they interfere with normative functioning. 204 While this psychoanalytic
approach further blurred the lines between normal and pathological sexuality, Freud, like KrafftEbing, assumed that such a distinction in fact existed 205 and, in his later work, maintained that
most perversions originated from an unresolved castration anxiety and early sexual trauma. 206
Many elements of these early debates have resurfaced in contemporary American psychiatry,
with significant implications for controversies regarding the extent to which psychopathology
can be linked to sexual violence.

The DSM and Paraphilias


Id. at 279.
Id. at 280; BERING, supra note 192, at 11.
De Block & Adriaens, supra note 192, at 280.
Id. at 281.
Id. at 282 (citing SIGMUND FREUD, THREE ESSAYS ON SEXUALITY 57 (1962)).
SIGMUND FREUD, THREE ESSAYS ON SEXUALITY 26–27 (1962) (noting that some “perversions” “are constituents
which are rarely absent from the sexual life of healthy people” and this presents “insoluable difficulties as soon as
we try to draw a sharp line to distinguish mere variations within the range of what is physiological from pathological
Jerome C. Wakefield, DSM-5 Proposed Diagnostic Criteria For Sexual Paraphilias: Tensions Between
Diagnostic Validity and Forensic Utility, 34 INTL. J. OF L. AND PSYCHIATRY 195, 199 (2011).
Id. See also David Bryden and Maran M. Grier, The Search For Rapists' “Real” Motives, 101 J. CRIM. L. &
CRIMINOLOGY 171, 174–76 (2011).


In the second half of the 20th century, the DSM became the leading source of psychiatric
classification. The first two editions, based primarily on psychoanalytic approaches, were
published in 1952 and 1968. They did refer to sexual disorders (the early editions lacked the
diagnostic criteria seen in more recent editions), but these were placed within the “personality
disorders” category, and the focus was on the relationship between the individual’s desires and
predominant social norms. 207 The texts did not place sexual perversions clearly within the realm
of mental illness but, rather, treated them as types of social deviance. 208
As remarked above, the DSM-III, published in 1980, was a significant departure from the
earlier editions. What was most notable in this edition was its presentation of specific diagnostic
criteria for each disorder. 209 The definitions and criteria it offered for disorders associated with
sexual deviance, particularly for “pedophilia,” became increasingly embroiled in controversy and
politics in subsequent editions. Starting with the DSM-III, the manual included a category called
“paraphilias” 210 (or, as they are referred to in the current edition, DSM-5, “paraphilic disorders”)
which are specific disorders associated with sexual attraction to people, things, or situations that
are considered deviant or non-normal. Under the category, the manual lists disorders such as
pedophilia, exhibitionism, and sadomasochism. Each edition presented a slightly different list of
disorders and a slightly different set of diagnostic criteria for each. The central debate or tension
pervading the development of these classifications was this: at what point does sexual attraction
or desire signal or implicate psychopathology?
Since the field of psychiatry is centrally concerned with identifying and treating those
whose mental disorders cause personal distress and impair functioning, many (including Freud,
as indicated above) have taken the position that it is only when persistent form of sexual
attraction leads to such distress or impairment is it appropriate to label it as a disorder. 211 Thus
the extent to which a subject’s sexual feelings deviated from social norms were less important
for making the diagnosis of the presence of a “disorder” than the existence of distress or
impairment of function for the subject himself or herself. This view stems in part from
psychiatry’s wariness of classifying certain types of sexual attraction as disordered in light of the
enormous controversy regarding the previous inclusion of homosexuality in the DSM’s list of
sexual disorders. The elimination of homosexuality from the list in 1973 led to a debate about
whether and which other forms of sexual deviation should be included in the manual, particularly
where such deviation did not cause any distress to the individual (the key rationale used for
removing homosexuality). The DSM-III included language in the forward noting a distinction


302 (2d. Ed. 1968); De Block & Adrieans, supra note 192, at 285-86.
De Block & Adriaens, supra note 192, at 286.
SHORTER, supra note 130, at 300-302.
For an excellent, detailed analysis and critique of the paraphilias and their use and implications in forensic
settings, including SVP proceedings, see generally Hamilton, supra note 6. The DSM-III-R list of paraphilias
included: Fetishism, Tranvestism, Zoophilia, Pedophilia, Exhibitionism, Voyeurism, Sexual Masochism, and Sexual
268-75 (3rd ed. 1980) (“DSM-III”).
DeBlock & Adriens, supra note 192, at 288-89; Wakefield, supra note 204, at 197, 200.


between deviance and disorder 212 and the lead editor of the manual, Robert Spitzer,
acknowledged that the term “disorder” “always involves a value judgment.” 213
This emphasis on personal distress and impaired functioning became more apparent with
the publication of the DSM-IV in 1994. Under the diagnostic criteria for the paraphilias, in the
absence of distress or limited functioning, conduct based upon such urges could be criminal but
not pathological. 214 With this revision, that edition further clarified that child sex offenders could
not be considered mentally ill unless their deviant behavior caused such distress or
impairment. 215 This modification, however, which moved the notion of paraphilia away from
the problematic normal-abnormal dichotomy, 216 elicited outrage among certain conservative
groups who claimed that this would de-pathologize non-distressed pedophiles 217 and give an
“ego-syntonic well-functioning paraphilic a free pass as far as disorder goes.” 218 Robert Spitzer
later referred to the blowback as a “public relations disaster,” and the amendment (referred to as
a “misinterpretation” by the editors) was reversed for those paraphilias “involving nonconsenting
victims” to allow a diagnosis of paraphilia based upon either the individual’s acting on paraphilic
urges with such victims or experiencing distress caused by such urges. 219 In the “text revision”
of DSM-IV six years later, the editors modified the criteria to make clear that acting on
paraphilic urges could itself satisfy the “harm” requirement for the diagnosis of pathology, even
if such activity was unaccompanied by “distress or interpersonal difficulty” for the person so
diagnosed. 220
Another significant change in the DSM-IV was to the “A Criterion” in the paraphilia
diagnoses, this one allowing clinicians to base a diagnosis on “recurrent, intense sexually
arousing fantasies, sexual urges, or behaviors.” 221 This revision was a technical adjustment
required by changes in wording made in other section of the diagnostic criteria for paraphilia. 222

DSM-III, supra note 210, at 6 (“When the disturbance is limited to a conflict between an individual and society,
this may represent social deviance, which may or may not be commendable, but is not by itself a mental disorder.”).
DeBlock & Adriaens, supra note 192, at 288.
Id. at 291. This change was part of a “system-wide effort” to incorporate “clinical significance criterion” to
diagnoses throughout the DSM-IV. Michael B. First, DSM-5 Proposals for Paraphilias: Suggestions for Reducing
False Positives Related to Use of Behavioral Manifestations, 39 ARCH. SEX. BEHAV. 1239, 1240 (2010).
ed. 1994) This modification was also consistent with revisions made throughout DSM-IV to ensure that only
conditions that caused harm, one of the essential components for a clinically-significant medical “disorder,” were
included. Wakefield, supra note 204, at 201–202.
Michael B. First & Robert L. Halon, Use of DSM Paraphilia Diagnoses in Sexually Violent Predator
Commitment Cases, 36 J. OF THE AM. ACAD. OF PSYCHIATRY AND L. 443, 445 (2008). The edition retained the list of
paraphilias, however, which now included: exhibitionism, fetishism, frotteurism, pedophilia, sexual sadism, sexual
masochism, transvestic fetishism, and voyeurism. DSM-IV, supra note 215, at 569–75. Each paraphilia had its own
diagnostic criteria.
Michael B. First & Allen Frances, Issues for DSM-V: Unintended Consequences of Small Changes: The Case of
Paraphilias, 165 AM. J. PSYCHIATRY 1240, 1240 (2008). The specific protest cited by the authors apparently came
from “Exodus International,” an anti-gay Christian organization. See
Wakefield, supra note 204, at 201.
DSM-IV-TR, supra note 186, at 566; First & Frances, supra note 217, at 1240.
DSM-IV-TR, supra note 186, at 566 (“The person has acted on these sexual urges, or the sexual urges or
fantasies cause marked distress or interpersonal difficulty.”).
DSM-IV, supra note 215, at 566 (emphasis added).
First, supra note 214, at 1240.


It was only in hindsight that the editors and other commentators noted that the use of “or
behaviors” as a disjunctive, in combination with the amendment regarding the “harm”
requirement, could serve prosecutors in SVP cases as a basis for assigning the diagnosis of
mental abnormality to sexual offenders “based only on their having committed sexual offenses
(e.g., rape).” 223 The DSM editors have asserted repeatedly that such a reading of the A Criterion
is inconsistent with the basic conceptualization of paraphilias in the DSM: criminal conduct
alone, even if it appears to be based on an underlying paraphilia, cannot establish a diagnosis for
such a paraphilia. 224 Given that the “core construct” of a paraphilia is the presence of “deviant
arousal,” a clinical diagnosis must be based upon information beyond an instance of criminal
conduct alone. 225 As one of the DSM-IV editors, Michael First, explained in a 2010 editorial:
“A paraphilia is … fundamentally a disturbed internal mental process (i.e., a deviant focus of
sexual arousal) which is conceptually distinguishable from its various clinical manifestations.” 226
Since the best indicators of a pattern of sexual arousal pattern are a patient’s “self-reports” of
fantasies, urges, and actions, obtained through a diagnostic interview, the criteria should not be
interpreted in a way that would permit a clinician to “skip this crucial step” in the diagnostic
process. 227 To base a diagnosis on a person’s acts alone, therefore, “conflate[s] the underlying
phenomenology of a paraphilia with its clinical manifestations.” 228
The paraphilias are not, strictly speaking, limited to the specific diagnostic labels, such as
“pedophilia” and “exhibitionism,” set forth in the DSM. Beginning with the DSM-III the
“paraphilias” category also included a catchall label: initially it was “Atypical Paraphilia” 229 and
then, beginning with the DSM-III-R, it was “Paraphilia Not Otherwise Specified.” 230 The
purpose of this label was to acknowledge that the disorders specified in the category “paraphilia”
did not represent the full range of nonconforming sexual interests, and it provided clinicians with
a term to use for someone whose particular disorder (such as a sexual interest in animals or in
rubbing against strangers) was not among the ones listed. Each edition of the DSM provided a
non-exhaustive list of examples of such other conditions. 231 Some were removed from the list in

First & Frances, supra note 217, at 1240; Frances et al., supra note 190, at 380; Wakefield, supra note 204, at
201. As the DSM-IV’s lead editor, Allen Frances, noted recently: “This one stupid slip contributed to the
unconstitutional preventive detention of thousands of sex offenders. I have no pity for criminals, but do have great
concern when their constitutional rights are violated just because I made a dumb wording mistake.” Allen Frances,
DSM-5 Writing Mistakes Will Cause Great Confusion, HUFFINGTON POST (June 11, 2013)
First & Halon, supra note 216, at 446–47 (“It had never been anticipated that any clinician would interpret the
addition of “or behaviors” in Criterion A as indicating that the deviant behavior, in the absence of evidence of the
presence of fantasies and urges causing the behavior, would justify a diagnosis of a paraphilia.”).
Id. The authors indicate that such other information can be gleaned from interviews, questionnaires, a detailed
history of the individual’s sexual behavior, use of pornography, and testing of physiological responses. Id. at 447–
48. See also Wakefield, supra note 204, at 198 (“[P]araphilias are disorders of sexual arousal and desire, not
matters of behavior and action undertaken for other reasons.”).
First, supra note 214, at 1240.
DSM-III, supra note 210, at 275.
Rev’d 1987) at 290.
The DSM-5 list under “Other Specified Paraphilic Disorders” includes the following examples: “telephone
scatologia (obsene phone calls), necrophilia (corpses), zoophilia (animals), coprophilia (feces), klismaphilia
(enemas), and urophilia (urine).” DSM-5, supra note 187, at 705.


successive editions and provided full criteria; and some were added to the list. 232 The historical
variability of the “NOS”—Not Otherwise Specified--category of paraphilias is evident, and its
diagnostic validity has never been subjected to research. 233
There are strong opinions throughout psychiatry about the validity of entire category of
paraphilias and the implications of their use as a basis for SVP commitment. 234 Because of the
obvious intersection with issues regarding the significance of social norms and of religious and
moral judgments about sexuality and sexual behavior, the paraphilias have always been among
the most controversial diagnoses in the DSM. 235 As noted above, the debate about the removal of
homosexuality from the list of paraphilias had a profound impact on all later discussions of the
inclusion, revision, or removal of diagnoses in that category. Several within the psychiatric
profession have continued to question whether there should be such a category at all. They have
asked what justification there could be for classifying particular forms of sexual desire as
disorders. 236 Scholars questioning the validity of the diagnosis of pedophilia as a pathology point
to the wide variation, both historically and among states and countries today, regarding the
minimum age of the sexual partner required to avoid prosecution for child sexual abuse. 237 The
fact that such widely variable considerations are often the basis for the indefinite detention of
individuals is of particular concern to these commentators.

Research Undermines Presumed Connections Between Mental Disorders and Sex

No less problematic than the lack of agreement around the proper basis upon which to
assign diagnoses of paraphilia in SVP cases is the lack of consensus about whether the presence
of a condition such as pedophilia could serve as a cause of past acts or the extent to which such a
condition could serve as a predictor of future criminal conduct including sexual abuse and rape.
Although it might appear that paraphilias are the type of mental disorder most obviously
associated with violent sexual behavior, they are far from an ideal fit. Several researchers have
found that sexually violent criminal conduct, and specifically child sexual abuse and rape, does
not in fact strongly correlate with the presence of a paraphilia. 238 While most SVP laws take a
“one size fits all” approach to offenders, research indicates that sex offenders are a “markedly


For example, Frotteurism (rubbing against strangers) was initially listed as an “atypical paraphilia” and zoophilia
(sexual interest in animals) was removed from the list of specific conditions into the “Not Otherwise Specified”
category. Compare DSM-III, supra note 210, at 270, 275 with DSM-IV-TR, supra note 186, at 570, 576.
Prentky et al., supra note 7, at 366.
See, e.g., Robert A. Prentky et al, Commentary: Muddy Diagnostic Waters in the SVP Courtroom, 36 J. AM.
ACAD. PSYCHIATRY AND THE LAW 455, 456–58 (2008).
Wakefield, supra note 204, at 195.
Id. at 195–96.
BERING, supra note 192, at 150–52.
First & Halon, supra note 216, at 446 (citing N. Dunsieth, et al: Psychiatric and Legal Features of 113 Men
Convicted of Sexual Offenses, 65 J CLIN. PSYCHIATRY 293–300 (2004)); Alan Felhouse & Lenore Simon,
Introduction to this Issue: Sex Offenders Part One, 18 BEH. SCI. & THE LAW 1, 2 (2000) (noting that the consensus
of clinicians who treat sex offenders is that “most sex offenders do not have a paraphilia”); Simon, supra note 41, at
294 (“deviant sexual fantasies do not exist in the majority of sex offenders.”) See also Prentky et al., supra note 7,
at 366 (noting that studies have shown that “a substantial proportion of rapists do not meet the criteria for any


heterogeneous group of criminals,” 239 As one scholar notes, this “primary pathology attributed to
sex offenders … is beginning to be discredited empirically.” 240
These empirical findings were the basis of Dr. First’s foremost concern about clinicians
basing dubious pedophilia diagnoses upon actions alone: that is, the risk of a significant number
of “false positive” diagnoses. 241 First noted that sexually violent behavior can have a great
number of underlying causes and that the paraphilias are limited to one specific kind of behavior:
persistent, deviant sexual arousal. Inappropriate sexual conduct, such as exhibitionism or sexual
contact with minors, could alternatively be caused by “a manifestation of disinhibition or poor
impulse control related to substance intoxication, a manic episode, or personality change due to a
dementing illness,” and also by “opportunism in a person with antisocial personality disorder.” 242
As one example of such findings, Dr. First noted a study of child sex offenders in which only
one-third had a pedophilic arousal response pattern. 243
The pathology of those who commit sexual offenses against other adults is even more
indeterminate with regard to specific diagnosis. A diagnosis of “sexual sadism” could apply to
all those who derive specific erotic pleasure from another person’s suffering, 244 but it certainly
does not apply to all rapists, even to those who commit multiple offenses. 245 At the time the
DSM-III-R was adopted, the editorial committee debated including a new diagnosis, “paraphilic
coercive disorder,” among the paraphilias. 246 This proposal immediately generated controversy.
Not only was there “little systematic research on the usefulness, reliability, validity, or definition
of the proposed disorder,” but many raised concerns about turning rape into a mental disorder. 247
The concern was not for the potential use of such a diagnostic category as a basis for preventive
detention but rather to excuse criminal conduct. 248 Ultimately, the absence of sufficient data to
support the existence of such a separate disorder led to the rejection of this proposal entirely. 249

Prentky et al, supra note 234, at 456.
Simon, supra note 41, at 284.
First, supra note 214, at 1240. Dr. First apparently gave a deposition in which he attempted to explain the DSM’s
paraphilias language was being interpreted and used in a way not intended by the editors, resulting in misdiagnose of
individuals with a paraphilia. The transcript of this deposition was offered as evidence in a petition to terminate an
SVP commitment based on a paraphilia diagnosis, but it was rejected by the trial court (which was upheld on
appeal) because Dr. First had not examined the petitioning individual. In re Detention of McGary, 231 P.3d 205,
208–210 (Wash. Ct. App. 2010).
First, supra note 214, at 1240. See also Fabian M. Saleh, et al., The Management of Sex Offenders: Perspectives
for Psychiatry, HARV. REV. PSYCH. 359, 361 (Nov/Dec 2010) (noting the a wide range of motivations and
“environmental precipitants” related to sexual violence).
First, supra note 214, at 1240 (citing M.C. Seto & M.L. Lalumiere, A Brief Screening Scale to Identify
Pedophilic Interests Among Child Molesters, 13 SEXUAL ABUSE: A JOURNAL OF RESEARCH AND TREATMENT, 15–25
DSM-IV-TR, supra note 186, at 573 (“the individual derives sexual excitement from the psychological or
physical suffering (including humiliation) of the victim”).
Simon, supra note 41, at 293.
Frances et al., supra note 190, at 380.
Id. Similarly, a diagnosis of pedophilia is specifically excluded from the Americans with Disabilities Act defining
of “disability” out of concern that individuals might seek some kind of “accommodation” for such disorder. 42
U.S.C. s 12211(b)(1); Adrienne L. Hiegel, Note: Sexual Exclusions: The Americans with Disabilities Act As A Moral
Code, 94 COLUM. L. REV. 1451, 1473–75 (1994). These are only a few examples of the inconsistent legal
implications of having a mental disorder.
It was not even retained as potential diagnosis for future study, as is done with some rejected diagnoses. Id.


The Absent Connection Between Psychiatric Assessment of Paraphilia and
Determination of “Volitional Impairment”
Of particular significance for SVP commitments is the fact a diagnosis of pedophilia or
other paraphilia, in addition to not being strongly correlated with acts of sexual violence, does
not necessarily involve a lack of “volition” or form of compulsion, as required under the
Hendricks-Crane analysis. As First and Halon write, a “diagnosis of a paraphilia does not imply
that the person has difficulty controlling his behavior.” 250 The defining feature of each of the
paraphilias is a particular source of sexual arousal (labeled as deviant), and, as noted above, a
great many people with such sexual interests, urges or fantasies never act on them. 251 As a
result, some researchers “liken it [a paraphilia] to an addiction, others to sexual orientation.” 252
Indeed, the DSM-IV-TR’s introductory language makes clear that none of the diagnoses
in the manual imply an assessment of volitional control:
[T]he fact that an individual’s presentation meets the criteria for a DSM-IV diagnosis
does not carry any necessary implication regarding the individual’s degree of control over
the behaviors that may be associated with the disorder. Even when diminished control
over one’s behavior is a feature of the disorder, having the diagnosis in itself does not
demonstrate that a particular individual is (or was) unable to control his or her behavior at
a particular time. 253
This language reflects psychiatry’s consistent attempts to stay clear of weighing in on questions
of “volition.” As one group of commentators noted: “Assessing volitionality is perhaps the most
hopeless of all diagnostic quagmires.” 254
The psychiatric field has long rejected the notion that it has a special ability to predict
future behavior and particularly dangerous conduct. 255 It has also been ambivalent about its
ability to understand and to identify volitional impairment, particularly in the criminal context. 256
Such concerns on the part of the psychiatric profession have led many states to eliminate
volitional impairment (frequently referred to as “irresistible impulse”) as a basis for the insanity
defense. As the APA famously noted in its caution about the limits of psychiatry in this regard:
“[T]he line between an irresistible impulse and an impulse not resisted is probably no sharper


First & Halon, supra note 216, at 450.
Casey Schwartz, What Science Reveals about Pedophilia, THE DAILY BEAST (Dec. 2, 2011).
DSM-IV-TR, supra note 186, at xxxiii. There is a category of disorders known as “Disruptive, Impulse-Control,
and Conduct Disorders” such as kleptomania and pyromania, DSM-5, supra note 187, at 476–79, but they are not
associated with acts of sexual violence and therefore would not be appropriate predicates for an SVP commitment
finding of mental abnormality that results in volitional impairment. Prentky et al., supra note 7, at 365.
Prentky et al., supra note 234, at 457.
First & Halon, supra note 216, at 451; John Monahan, Clinical and Actuarial Predictions of Violence, in


than that between twilight and dusk.” 257 With regard to SVP laws, the Association for the
Treatment of Sexual Abusers (a group of medical professionals), in its amicus brief to the
Supreme Court in Crane, stated that the concept of volitional impairment in SVP legal standards
is “meaningless and unworkable” 258 Like the “irresistible impulse” test for criminal responsibly,
the notion of “volitional impairment,” if it even exists, should similarly be rejected because of
the inability of experts to identify it.
The hesitation of psychiatrists to make predictions in the SVP context is based in part on
specific relevant research findings. Contrary to a common assumption, the recidivism rate among
sex offenders for committing a future sex offense is actually quite low as compared with many
other crimes. 259 Indeed, a failed prediction of future sex offense is more likely to yield a false
positive than a false negative. Research findings also call into doubt the assumption that the
source of the behavior of sex offenders is a specific abnormality or condition. 260 Sexual offenses
are often committed by those with criminal histories for other offenses, and convicted sex
offenders may recidivate through other forms of criminal or antisocial behavior. 261 As one
psychiatrist noted: “The possibility of forfeiture of liberty based not on current behavior, but
rather on prediction of potential for future offending, imposes a stark obligation on the evaluator
to ‘get it right.’” 262 However, the consensus of the field is that such predictions cannot be done
with “any precision.” 263
Just as statistical analysis reveals the absence of a strong correlation between a paraphilia
and sexual violence, 264 empirical studies also reveal that pedophilia—that is, the presence of
intense sexual attraction to children—does not in itself indicate that a person is likely to engage
in child sexual abuse. 265 Although commitments of several men under SVP laws (particularly in
the federal system) have been based solely upon a prior conviction for possession of child
pornography, it is far from clear that viewing child pornography is indicative of sexual
dangerousness. 266

American Psychiatric Association, American Psychiatric Association Statement on the Insanity Defense, 140 AM.
J. PSYCHIATRY 681–8 (1983).
Brief of Association for the Treatment of Sexual Abusers as Amicus Curiae at 3, Kansas v. Crane, 534 U.S. 407
Simon, supra note 41, at 302-306.
See, e.g. Cynthia Calkins Mercado, Sex Offender Management, Treatment, And Civil Commitment: An Evidence
Based Analysis Aimed At Reducing Sexual Violence, Report issued to Research Report Submitted to the National
Institute of Justice (Jan 2011) ( (noting that even among the
highest risk groups of sex offenders, recidivism rates were “quite low” and most sex crimes were not committed by
“known offenders”). See also Simon, supra note 41, at 284 (“Although some sex offenders are at high risk to
reoffend, there is no clear empirical basis for assessing which sex offenders present the most immediate risk for
reoffending. Also, there is no evidence that sex offenders are any more mentally disordered than general criminal
Simon, supra note 41, at 284.
Saleh et al, supra note 242, at 366.
See supra notes 238–233 and accompanying text. One researcher has argued that paraphilias are “taxonomically
useless” to identify those sex offenders who would qualify as SVPs.
Prentky et al., supra note 7, at 367.
See Emily Bazelon, Passive Pedophiles: Are Child Porn Viewers Less Dangerous Than We Thought? SLATE
(Apr. 25, 2013); BERING, supra note 192, at 174–76 (providing an overview of research findings regarding the lack
of strong correlation between viewing child pornography and engaging in child molestation). A 2013 study released
by the U.S. Sentencing Commission found that one in three people convicted of possessing [confirm] child



ASPD as Alternative Basis of Mental Disorder

Given that diagnoses of paraphilias do not appear, at least in the view of mainstream
psychiatry, to be useful tools for identifying a mental disorder or abnormality that could be a
predictor for a sex offender’s future acts of sexual violence, the question arises as to whether
some other diagnoses might fit that need. As Dr. First noted in the statement quoted above, many
other diagnoses are, in fact, more strongly associated with sexual violence than the presence of a
mental disorder. 267
The diagnosis that is most obviously applicable to those who commit acts of sexual
violence is Antisocial Personality Disorder (ASPD). 268 Indeed, ASPD is a diagnosis that, by
definition, could apply to most people incarcerated in the United States. 269 ASPD is often
characterized by a pattern of criminal behavior, including committing sex crimes against children
and nonconsenting adults. 270 In the case of sexual offenders, then, what is indicated by a
diagnosis of ASPD is that the acts of violence are indicative not so much of a paraphilia as of a
“pervasive pattern of disregard for and violation of the rights of others.” 271
There is dispute within psychiatry about whether personality disorders, particularly
ASPD, are true mental disorders or illnesses, and whether a diagnosis of such a disorder should
be offered in support of SVP commitments, either standing along or on conjunction with one or
more paraphilias. 272 Nothing in the Supreme Court’s precedent precludes basing an SVP on such
a diagnosis alone; there is no requirement that a person have a “sexual” disorder of some kind. 273
The diagnosis of ASPD could apply to a great many rapists and child molesters, some of whom
may also have paraphilias; but untangling such comorbidity is not straightforward. As a result, it
pornography had engaged in acts classified as “criminal sexually dangerous behavior,” a category that includes
“non-contact” crimes such as voyeurism and exhibitionism and that the post-sentence sexual recidivism rate of the
people so convicted was 7.4% (3.6% for “contact” offenses), which is lower than the rates for those specifically
TO CONGRESS) x, xv (Dec. 2012) (available at
First & Halon, supra note 216, at 448.
Studies have estimated that anywhere from 40% to 80% of the male prison population would meet the diagnostic
criteria.EWING, supra note 13, at 25; Thomas K. Zander, Civil Commitment Without Psychosis: The Law’s Reliance
on the Weakest Links in Psychodiagnosis, 1 J. OF SEXUAL OFFENDER CIVIL COMMITMENT: SCIENCE AND THE LAW,
17, 53 (2005); First & Halon, supra note 216, at 443-454
Simon, supra note 41, at 294 (noting that empirical findings indicate that “clinicians diagnose more convicted
child molesters with antisocial personality disorder than with pedophilia.”).
DSM-IV-TR, supra note 186, at 701. A diagnosis of ASPD was usually inadequate for commitment under the old
sexual psychopath laws, which focused on treatment of offenders, since those with ASPD are not generally regarded
as being amenable to treatment; rather, the ASPD is seen as a fixed personality feature. First & Halon, supra note
216, at 449.
See, e.g., Dean R. Cauley, The Diagnostic Issue of Antisocial Personality Disorder in Civil Commitment
Proceedings: A Response To Declue, 35 J. PSYCHIATRY & LAW 475 (2007); Gregory DeClue, Paraphilia NOS (nonconsenting) and Antisocial Personality Disorder, 34 J. OF PSYCHIATRY & LAW 495 (2006).
And indeed, this means that an SVP commitment could theoretically be based upon a diagnosis of substance
abuse, mood disorders, or schizophrenia if some causal link to sexually violent behavior could be made. Frances, et
al., supra note 190, at 382.


is exceedingly difficult for courts to identify whether the sexually offending behavior is merely
criminal 274 and therefore to draw that critical constitutional line required in Crane of separating
the typical recidivist sexual offender from the one who has “volitional impairment.”
The Supreme Court has never had to consider whether an ASPD diagnosis, standing
alone, would be constitutionally adequate for an SVP commitment, and courts are divided on this
question, since many SVP laws refer to “personality disorder” as well as mental abnormality. 275
The Court’s opinion in Foucha suggests that ASPD would not be enough for post-acquittal form
of commitment since, in that particular case, the acquitee had an “antisocial personality.” 276
ASPD, like other personality disorders, has never been regarded in the criminal law as a
volitional impairment sufficient to exempt an offender from criminal responsibility. Indeed, to
treat it as such would call into question the conviction and incarceration of most of this country’s
prison population. Such disorder likely contributes to much “typical” recidivism, and therefore a
commitment based upon ASPD alone would be constitutionally suspect since it would extend
this extraordinary deprivation of liberty to a far greater segment of the population than
substantive due process principles would permit. 277

Psychiatry’s Response to SVP Laws and Hendricks-Crane Rationale

The Court’s rationale in Hendricks-Crane assumes that there is a unique and distinctive
pathology among dangerous sex offenders. As argued above, the assumption has no footing in
current medical thinking either about the mental condition of such offenders or about the extent
to which a mental health professional can identify those at particularly high risk of reoffending.
In light of this unsettled connection between sexual violence and psychopathology and the
absence of a reliable method for clinicians to predict future violence, the American Psychiatric
Association has repeatedly attempted to draw attention to the divergence between the SVP laws
and scientific understanding.
The passage of the initial SVP laws in the early 1990s led the APA to appoint a Task
Force on Sexually Dangerous Offenders. The report it released in 1999 (two years after the
Hendricks opinion) was highly critical of such laws. 278 Members of the Task Force noted that
that the “question of whether all or some sexual offenders are mentally ill is complicated and
controversial” 279 and, similarly, that there was no consensus on the degree to which sex
offenders have control over their behavior. 280 Certainly some offenders have paraphilias, the

Frances, et al., supra note 190, at 381.
EWING, supra note 13, at 25.
Foucha v. Louisiana, 504 U.S. 71, 78(1992).
Despite many calls to revise the rather circular diagnostic criteria to address many of the resulting problems with
its use, it was left unchanged by the editors of the DSM-5. DSM-5, supra note 187, at 659. However, the field trials
leading to the release of DSM-5 revealed that the diagnosis has one of the lowest inter-rater reliability ratings (in the
“questionable agreement” range). Bret S. Stetka & Christoph U. Correll, A Guide to DSM-5: Field Trial Results,
MEDSCAPE PSYCHIATRY (May 21, 2013) ( Such results have led
some commentators to argue that such results should preclude any use of the disorder in forensic settings. Karen
Franklin, DSM-5: Forensic Applications (Part II of II), IN THE NEWS (May 30, 2013).
APA, supra note 14, at vii.
APA, supra note 14, at 5, 7.
Id. at 5.


report acknowledged, but it also noted that paraphilias occur fairly frequently in those who never
commit sex offenses. 281 Personality and substance abuse disorders, it continues, are far more
common in sex offenders than are paraphilias, and, significantly, the latter conditions do not
usually have “explanatory connection” to the offender’s behavior. 282 In short, the Task Force
Report stated, “psychiatric nosology does not contribute in a systematic way to clinical
understanding or treatment of sex offenders.” 283 The language of the Report’s conclusion was
[S]exual predator commitment laws represent a serious assault on the integrity of
psychiatry, particularly with regard to defining mental illness and the clinical conditions
for compulsory treatment. Moreover, by bending civil commitment to serve essentially
nonmedical purposes, sexual predator commitment statutes threaten to undermine the
legitimacy of the medical model of commitment.
[The SVP laws] establish a nonmedical definition of what purports to be a clinical
condition without regard to scientific and clinical knowledge. In so doing, legislators
have used psychiatric commitment to effect nonmedical societal ends that cannot be
openly avowed.… This represents a misuse of psychiatry. 284
The inability of psychiatrists to predict future violence was also a key point asserted by
the APA in its brief in Hendricks, but it was not the first time the organization made that point.
In the 1983 case Barefoot v. Estelle, in which the Supreme Court upheld the admissibility of
psychiatric evidence on the issue of future dangerousness in a death penalty case, 285 the APA had
stated in its amicus brief in the case that “[t]he unreliability of psychiatric predictions of longterm future dangerousness is by now an established fact within the profession.” 286 This was only
one of multiple occasions on which the Supreme Court has rejected the cautions of the mental
health profession and left in place laws and practices whose legitimacy hinges on the field’s use
of pathology to predict future conduct.
Although the APA is the world’s largest organization of professional psychiatrists and its
official statements are significant, 287 there are dissenting views in psychiatry with respect to the
role of psychopathology in sexual violence. Indeed, there are segments of the mental health
profession that support the SVP laws and provide the research and expert testimony to support
the commitment of individuals. I provide several examples of their views and opinions in the two
sections that follow.
Mental health professionals who support the SVP laws are primarily treatment providers
who specialize in sex offenders, including those in state SVP programs, but who do not work in

APA, supra note 14, at 44.
Id. at 9.
Id. at 9.
Id. at 173– 174 (emphasis added).
Barefoot v. Estelle, 463 U.S. 880, 896–99 (1983).
Barefoot, 463 U.S. at 920 (Blackmun, J., dissenting) (quoting Brief for American Psychiatric Association
as Amicus Curiae at 12).


the correctional or criminal setting. 288 As one researcher has noted, this context can distort
treatment-providers’ views of such offenders: it tends to support a view of specialization in
offenders’ behavior and also an assumption that the individuals treated experience “deviant
sexual arousal, which, if not treated, will result in future sex crimes.” 289 Because of a lack of
foundation in criminological research, mental-health policy decisions made by such treatment
providers regarding such offenders also lack such foundation and continue to be based upon
misplaced assumptions about those who commit sex crimes: in particular, the assumption that
such offenders are “mentally disordered, treatable, dangerous (if not treated), and at high risk to
reoffend with another sex crime.” 290
It is not surprising that mental-health professionals’ views on SVP law are influenced by
their professional perspectives, including their employment relationships. My concern here,
however, is with the existence of the debate itself: specifically, with the sharply divergent
positions among those professionals and with the vast discrepancy between, on the one hand, the
standard nosology of the psychiatric profession and steadfast position of its primary
organizations and, on the other hand, the role assumed for and assigned to psychiatry in the SVP
laws. The SVP laws set up a complex relationship between mental health professions and the
legal system. And, as we will see in the section that follows, while psychiatric expertise has been
increasingly brought into SVP proceedings to support individual commitments, much scientific
understanding of the causes and prediction of violent sexual behavior has become, in the process,
highly distorted.

Pathologizing Predators in the Courtroom

Both the state legislatures that developed the SVP laws and the Supreme Court in
upholding them always assumed that mental health professionals would play a central role in
SVP proceedings, offering opinions regarding the risk of recidivism posed by specific
individuals due to the presence of a mental abnormality or disorder of some kind that impaired
their ability to refrain from committing acts of sexual violence. Indeed, such professional
opinions have been seen as indispensible since laypersons lack the competence either to identify
mental conditions or to assess volitional impairments. As discussed in the preceding section,
however, there is scant scientific foundation for such assessments or predictions by mental health
professionals themselves, nor is there anything in psychiatric classification that corresponds to or
otherwise supports the crucial SVP concept of the “sexual predator.” These well-attested
difficulties have not prevented state prosecutors from offering mental-health expert testimony in
support of SVP petitions; and most courts readily admit such testimony, even over strenuous
objections from defense counsel, who often cite the controversies discussed above. Maintaining

Simon, supra note 41, at 277. While three of five of the amicus briefs submitted in Hendricks on behalf of mental
health associations supported striking down the law (the American Psychiatric Association, the Washington
Psychiatric Association, and the National Mental Health Association), the two who supported the law were directly
involved with the treatment of sex offenders, including the Menninger Foundation, which operated a psychiatric
hospital in Kansas at the time, and which was joined on the brief by a series of “victims’ rights” and law-and-order
organizations such as the New York Chapter of Parents of Murdered Children, Protecting Our Children, People
Against Violent Crime, and Victims Outreach, Inc. Felhouse & Simon, supra note 238, at 2. Apparently, significant
portions of the majority opinion in Hendricks was drawn from the Menninger Foundation’s amicus brief. Id.
Id. at 277, 279, 281.
Id. at 278.


the role of expert evidence to support commitments in SVP proceedings has required a distortion
of psychiatric understanding. It has also required a severe compromise of core values and
practices of our justice systems. 291
One of several areas of continued psychiatric uncertainty and legal strain in the
implementation of SVP laws is the degree of risk of future dangerousness that can serve as a
basis for indefinite detention. The Supreme Court in Addington v. Texas held that a state may
involuntarily commit a mentally ill individual using a “clear and convincing evidence”
standard. 292 This is a lower threshold of proof than the “beyond a reasonable doubt” standard
usually reserved for the criminal context. One of the rationales of the lower threshold,
notwithstanding the liberty interest at stake, is the relative imprecision of psychiatric evidence,
which generally serves as the primary proof offered in support of such commitments. The
Court’s opinion states:
Whether the individual is mentally ill and dangerous to either himself or others and is in
need of confined therapy turns on the meaning of the facts which must be interpreted by
expert psychiatrists and psychologists. Given the lack of certainty and the fallibility of
psychiatric diagnosis, there is a serious question as to whether a state could ever prove
beyond a reasonable doubt that an individual is both mentally ill and likely to be
dangerous…. The subtleties and nuances of psychiatric diagnosis render certainties
virtually beyond reach in most situations…. Psychiatric diagnosis ... is to a large extent
based on medical ‘impressions' drawn from subjective analysis and filtered through the
experience of the diagnostician. 293
Understandably, some commentators have argued that the very fact that psychiatric diagnoses
are imprecise and ambiguous suggests that only the “beyond a reasonable doubt” standard will
adequately ensure fairness and due process in commitment proceedings. 294 However, the Court
also reasoned in this case that permitting states to use a lower standard of proof is
constitutionally acceptable because of the defined limitations and objectives of involuntary
hospitalization: such commitment, it maintained, is limited to people with severe mental illness
who pose a danger to themselves or others, and employing higher standard of proof could “erect
an unreasonable barrier to needed medical treatment.” 295 Such reasoning, of course, has only
limited application in the SVP context, where public safety, rather than the treatment, is the
foremost objective. Nonetheless, the Supreme Court recently reiterated that the intermediate
standard of proof in civil involuntary commitment proceedings meets due process requirements,
even for indefinite commitment of SVPs.296

Finkel, supra note 59, at 243 (explaining how “the worst of times,” including the occurrence of horrible crimes,
operates like a hydraulic pressure which can “distort clear concepts and bend established principles, as well as
foreshorten perspective such that history's lessons no longer help frame current issues”); David L. Faigman et al., 2
MOD. SCI. EVIDENCE § 11:23 (2011-2012 Edition).
Addington v. Texas, 441 U.S. 418, 427–33 (1979). See generally Alexander Tsesis, Due Process in Civil
Commitments, 68 WASH. & LEE L. REV. 253 (2011).
Addington, 441 U.S. at 429 (emphasis in original).
Tsesis, supra note 292, at 282–300.
Id. at 432.
U.S. v. Comstock, 130 S. Ct. 1949, 1954 (2010), in which the Court upheld the SVP provisions of the Adam
Walsh Act law against a range of constitutional challenges.


To date, no decision has clarified precisely how dangerous to himself or others a person
must be to satisfy that standard for involuntary commitment. The concept of dangerousness is
itself quite vague and subject to a range of conceptualizations and analyses. 297 For example, if a
fact finder is asked to conclude “beyond a reasonable doubt” that an individual is “likely” to
commit future acts of sexual violence (the typical standard in most SVP laws), it is not clear
whether the fact finder must have no reasonable doubt that there is at least a 35%, a 50%, or a
75% chance the defendant will reoffend. 298 The dangers of securing involuntary commitments on
such uncertain grounds serve only to compound the significant problems presented by the
evidence admitted to support the central determination in SVP proceedings, that is, that the
offender is “a sexually violent predator.”

One example of the distortions: McGee v. Bartow

The language in the Supreme Court’s opinions in Hendricks and Crane confers
progressively broad discretion on lawmakers to devise the specific terms used to meet the due
process requirement of a mental condition for involuntary civil commitment. 299 The language
also encouraged experimentation and diverse approaches by legislatures and courts in regard to
the implementation of the SVP laws. A 2010 opinion of the Court of Appeals for the Seventh
Circuit, McGee v. Bartow, demonstrates the implications of the Supreme Court’s approaches and
deference. 300
Michael McGee was committed in Wisconsin courts under that state’s SVP statute, which
was adopted in 1994 and modeled closely on Washington’s. Having exhausted his direct appeals
for release through state courts, McGee then filed a petition for habeas corpus in the federal
district court. 301 Because this was a habeas case, it had to meet a particularly high standard,
namely, that his continued detention was in violation of federal law, including the U.S.
Constitution, rather than simply in violation of the applicable state law. 302
McGee’s only criminal conviction and sentencing had been in 1987, when he was
convicted of burglary and the sexual assault of a woman during the course of the burglary. He
served 5 years in prison and was released on parole. In 1992, while on parole, he was accused of
two more sexual assaults, had his parole revoked, and served out the remaining three years of his
sentence. 303 Neither of the two subsequent allegations of sexual assault, one by a woman and
another involving an adolescent male, led to a conviction. 304 The state then filed a petition to
commit McGee under the Wisconsin SVP law. He was committed in 1995 based on a jury
verdict but released in 1999 when the commitment was reversed on a finding of ineffective

See generally Frederick E. Vars, Delineating Sexual Dangerousness, 50 HOUSTON L. REV. 855, 860–73 (2013);
Shoba Sreenivasan et al., Expert Testimony in Sexually Violent Predator Commitments: Conceptualizing Legal
Standards of “Mental Disorder” and “Likely to Reoffend,” 31 J AM ACAD PSYCHIATRY LAW 471, 477–78, (2003);
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 (1997).
See Finkel, supra note 59, at 259.
See supra notes 89–122 and accompanying text.
McGee v Bartow, 593 F3d 556 (7th Cir 2010).
Id. at 558.
Id. at 571–72.
Id. at 558–59.


assistance of counsel: his attorney had failed to discover important evidence that could have
undermined the credibility of the two accusers from the 1992 allegations. 305 A year later, in
2000, he was rearrested for failing a drug test and having contact with one of the 1992 alleged
victims. The state sought to commit him again. 306
At the bench trial during this second commitment hearing, the state’s case was based
largely upon the testimony of two forensic psychologists. One was Department of Corrections
psychologist, Dr. Caton Roberts, who offered his opinion that McGee had a “personality disorder
NOS [Not Otherwise Specified] with antisocial features” and a “substantial probability to
reoffend sexually if not detained and treated.” Roberts’ opinion was based not on a clinical
examination of McGee but on fifteen hours of review of McGee’s “record.” 307The second expert
to testify was Dr. Cynthia Marsh, who diagnosed McGee with “Paraphilia NOS- nonconsent”
and Personality Disorder NOS with antisocial features. 308 Her diagnosis was also based only
upon a review of records. Specifically, Marsh testified that she based her diagnosis primarily on
Mr. McGee’s “history,” including the (contested) 1992 allegations, and that she employed three
actuarial risk-assessment tools. From these, she concluded that McGee was “much more likely
than not to reoffend in a sexually violent manner.” 309
McGee’s attorneys claimed on appeal that the diagnoses that served as the bases for
satisfying the “mental illness” requirement were insufficient as a matter of due process.
Specifically, they alleged that the diagnoses used were not generally accepted as being either
valid or reliable within psychiatry (as noted earlier, the paraphilia category “nonconsent”
invoked by Marsh had in fact been explicitly rejected by the APA) and that the labels did not
have any standardized diagnostic criteria. 310
There was little case law upon which the Court of Appeals could evaluate such
arguments. Accordingly, the Seventh Circuit devised a specific standard for evaluating the
constitutional adequacy of a diagnosis used to commit an individual. To prove that use of a
diagnosis violated due process principles, the panel held, a petitioner must demonstrate that the
diagnosis was “devoid of content, or … near-universal in its rejection by mental health
professionals.” 311 The panel later re-stated the standard as being a determination of whether the
diagnosis was “empty of scientific pedigree.” 312
In explaining the standard, the panel devoted a considerable amount of the opinion to
reviewing the text of the DSM and noted the editors’ cautions about use of the manual in the
forensic context, particularly by “untrained professionals” (i.e. lawyers and judges), to answer
ultimate questions. 313 The panel also noted that, while nothing in Supreme Court precedent
expressly requires a valid DSM diagnosis as a prerequisite to a SVP commitment, such

Id. at 559.
Id. at 559–60.
Id. at 560, 572.
Id. at 560.
Id. at 574.
Id. at 577.
Id. at 581.
Id. at 578.


diagnostic labels could be useful tools when applied with “prudence and caution.” 314 However,
the court did not explain either what such prudence and caution involved or how its own
application of the text demonstrated such qualities. Indeed, what the panel concluded, based
upon language in Hendricks regarding the broad discretion conferred to states to develop their
own conceptualizations of mental abnormality without being tied to medical terminology, was
that neither the absence of a diagnosis from the DSM nor the existence of robust controversy
about the category among mental health professionals was a basis to disregard such a diagnostic
label entirely. Rather, it stated, such facts bear only on the weight to be assigned to the label as
part of the overall fact finding, not on its admissibility as evidence. 315 In short, a heated debate
within the field regarding a diagnostic label’s validity and reliability is not enough to exclude it
from serving as a basis for indefinite detention.
The McGee opinion illustrates many of the key problems with the role of psychiatric
evidence in SVP proceedings and, thereby, demonstrates the fundamental flaw in the Supreme
Court’s assumption that such testimony would ensure that SVP laws would not sweep too
broadly. McGee’s primary challenge was to the state’s experts’ reliance on a set of diagnoses
that were scientifically controversial and did not reflect any settled scientific understanding.
Other aspects of the experts’ opinions in McGee reveal a range of additional concerns seen more
generally in reported SVP cases, including basing opinions on inadmissible facts and data, such
as uncharged alleged criminal conduct, rather than on clinical examinations, and using actuarial
risk assessment tools.
Examination of the foundations of prosecution experts’ opinion of the likelihood of future
acts of sexual violence in SVP proceedings reveals that they are based largely upon the
respondent’s past behavior (alleged as well as proven) rather than, as required by the HendricksCrane rationale, an individualized medical assessment. This is because mental health
professionals, in attempting to assess whether a person is likely to commit acts of sexual violence
due to a volitional impairment stemming from a mental disorder, have little else but past
behavior to go on in the absence of scientific guidance for making such an assessment. But this
means that they predict future behavior based upon past behavior the same way we all do, and
not upon any particular form of expertise. The perpetuation of these unreliable and misleading
practices is facilitated by courts’ reluctance to assert their roles as “gatekeepers” with regard to
just such expert testimony.

Misuse of Diagnostic Labels

A core role of the diagnoses in SVP proceedings is to explain the basis for an expert’s
overall assessment that the respondent is likely to commit future acts of sexual violence. This
stems from the statutory requirement, given central importance in the Supreme Court’s due
process analysis in Hendricks and Crane, that the defendant have an identifiable “mental
disorder” or “mental abnormality.” 316 Although paraphilia diagnoses have a limited role in the
clinical setting and, as stressed and discussed above, are highly controversial within the field of
psychiatry, they enjoy broad acceptance by courts in SVP proceedings. As other commentators

Id. at 579
Id. at 580–81.
Id. at 581.


have noted, while there is an established history of presenting psychiatric evidence of specific
forms of psychopathology in support of involuntary commitment–-for example, schizophrenia
and other disorders characterized by psychosis–-SVP commitment, by contrast, is generally
based upon diagnoses, such as pedophilia and ASPD, that are “among the most controversial,
and that have the most questionable validity, of all the mental disorders in the DSM.” 317 As we
saw from the earlier discussion, the DSM's language regarding paraphilias is itself the product of
negotiation and public relations management, and is subject to a range of interpretations.
If used in a manner consistent with the DSM editors’ intentions, the diagnosis of a
“paraphilia” addresses only the (abnormal) circumstances that occasion sexual arousal; it does
not indicate an impaired ability to refrain from acting on the desires involved. Because existing
DSM diagnoses have limited use for identifying the reference of the forensic term “sexual
predator,” some experts testifying on behalf of states in SVP proceedings offer alternative
presumptively “diagnostic” labels that either strain the DSM criteria’s language beyond their
intended clinical application or fall outside of the diagnostic scheme entirely. 318 In so doing, as
in McGee, the experts essentially pathologize past criminal conduct.
The questionable nature of the invocation of such strained diagnoses in prosecuting SVP
cases is compounded when the catchall “NOS” (not otherwise specified) categories are invoked
or when forensic experts dispense altogether with the DSM’s criteria. 319 With regard to NOS
diagnoses in SVP proceedings, one commenter has observed: “Paraphilia NOS is a ‘proxy’ for
the rejected diagnosis of paraphilic coercive disorder, and has offered legislators and mental
health professionals carte blanche to invent criteria by which to deprive sex offenders of their
freedom after they have completed their sentences.” 320
The psychiatric validity of SVP diagnoses is put in further doubt by the asymmetrical
pattern of their invocation in courts. A survey of the reports of psychiatric experts in 28 SVP
cases conducted by Dr. Allan Frances, one of the editors of DSM-IV, found that, while
government experts usually gave an initial diagnosis of Paraphilia-NOS, defense experts usually
did not. 321 Frances concludes that the diagnosis was, in his word, “justified” in only 2 of those
cases whereas, in the other 26 cases, the respondents’ “sexual offenses had been opportunistic
crimes forming part of a pattern of generalized criminal behavior, very often facilitated by
substance intoxication.” 322 Government evaluators, Frances observes, seemed to base the
Paraphilia-NOS diagnosis not on an overall pattern of behavior suggestive of fundamental
pathology but only or primarily on the fact of prior conviction for sexual crimes. Several other
studies of psychiatric reports have also noted strong geographic variation in the rates at which

Zander, supra note 269, at 72.
EWING, supra note 13, at 24.
Id. The initial idea behind the “Paraphilia NOS” label or diagnosis—which is used almost exclusively in SVP
proceedings—has been credited to Dennis Doren, the lead forensic evaluator in Wisconsin’s SVP program. Good &
Burstein, supra note 46, at 27-28 (referring to DENNIS M. DOREN, EVALUATING SEX OFFENDERS: A MANUAL FOR
Allan Frances, DSM-5 Rejects Coercive Paraphilia: Once Again Confirming That Rape Is Not A Mental
Disorder, PSYCHIATRIC TIMES (May 12, 2011) ( ).
, Allan Frances, My Review of 28 Sexually Violent Predator Cases, (March 29, 2012)


various diagnoses (for example, paraphilia-NOS as compared to pedophilia) are used to support
SVP petitions.323 Variability of these kinds casts further doubt on the independent reliability, or
scientifically objective validity, of such diagnoses and have further fueled the significant ethical
concern within psychiatry about the forensic use of Paraphilia-NOS diagnoses. 324
As seen in McGee, even where a court is made aware that an examiner’s use of a
psychiatric diagnosis is patently inconsistent with the DSM’s language and with commentary
within the psychiatric field, the court is unlikely to reject the use of the diagnosis as a basis for
satisfying the mental disorder or abnormality requirement for SVP commitment. 325 The McGee
court squarely acknowledged that there was “heated professional debate” about use of the
diagnostic label Paraphilia NOS Nonconsent 326 and that McGee’s position that “the consensus
professional view that [such] diagnosis is invalid” is “not without support in the professional
literature.” 327 It even noted that the lack of diagnostic standards for the label “results in poor
diagnostic reliability.” 328 Nevertheless, the court denied McGee’s claim that his commitment,
based upon such contested diagnoses, amounted to a violation of his due process rights. In
denying his claim, the court concluded that the fact that the use of the label did find some
support in the medical literature took it outside of the realm of a diagnosis “empty of scientific
pedigree” or “near-universal” in rejection. 329


See, e.g., Shan Jumper et al., Diagnostic Profiles of Civilly Committed Sexual Offenders in Illinois and Other
Reporting Jurisdictions: What We Know So Far, 56 INTL. J. OF OFFENDER THERAPY & COMPARATIVE CRIMINOLOGY
838 (2012) (finding that revealed that pedophilia was diagnosed in persons targeted for commitment under Illinois’s
law at a “significantly higher rate” (59%) than those in proceedings in Minnesota, Texas, Wisconsin, and Florida,
and Paraphilia-NOS was diagnosed in Illinois more frequently (51%) than in Wisconsin (37.5%), and fifty-six
percent of sex offenders in SVP proceedings in Arizona have been diagnosed with Paraphilia NOS and nearly twothirds with Pedophilia). The jurisdictions included in the study were: Illinois, Texas, Florida, Wisconsin,
Washington, California, Arizona, and Minnesota. The study included all persons targeted for commitment since the
“vast majority” of those detained for commitment under the statute are eventually committed. Id. at 841. See also
Richard W. Elwood, et al., Diagnostic and Risk Profiles of Men Detained Under Wisconsin's Sexually Violent
Person Law, 54 INTL. J. OFFENDER THER. COMP CRIMINOL. 187, 193 (2010) (concluding that pedophilia was
diagnosed in Wisconsin at a higher rate than in Florida but at a lower rate than in Washington or Arizona); Julia E.
McLawsen, et al., Civilly Committed Sex Offenders: A Description and Interstate Comparison of Populations, 18
PSYCH, PUB. POL. & L. 453, 461 (2012) (analyzing diagnostic trends under Nebraska’s SVP law).
See, e.g., Allen Frances, Rape, Psychiatry, and Constitutional Rights—Hard Cases Make For Very Bad Law,
PSYCHIATRIC TIMES (Sept. 1, 2010) (arguing that “The most disturbing turbulence at the boundary between
psychiatry and the law is the misuse of a makeshift psychiatric diagnosis (“Paraphilia Not Otherwise Specified,
nonconsent”) to justify the involuntary, indefinite psychiatric commitment of rapists. This is a disguised form of
preventive detention (often for life), a violation of due process, and an abuse of psychiatry.”). See also Good &
Burstein, supra note 46, at 24–28 (criticizing the use of “fictitious mental disorders” by forensic evaluators
testifying in SVP proceedings); Prentky et al., supra note 7, at 369 (“Force-fitting a diagnosis or creating a new
DSM diagnosis to justify commitment is clearly unethical for psychologists.”); First & Halon, supra note 216, at 444
(“We contend that, during the process of adjudication of SVP commitment trials, profound and avoidable errors are
made by some mental health professionals who invalidly diagnose paraphilia, assert that there is volitional
impairment based solely on the fact that the offender has a paraphilia diagnosis, and thus wrongly claim that the
statutorily defined SVP commitment criteria are adequately addressed by the clinical diagnoses.”).
See generally Hamilton, supra note 7, at 40–51.
McGee v Bartow, 593 F3d 556, 591 (7th Cir 2010).
Id. at 580.
Id. at 581.


Several courts have been faced with similar questions about the admissibility of opinions
that include diagnostic labels attached to the catchall “Paraphilia NOS.” In addition to the
Paraphilia NOS – nonconsent label seen in McGee and other cases, 330 another such label created
and used almost exclusively by prosecution experts in SVP proceedings is “Paraphilia NOS –
Hebephilia,” a term used to indicate sexual interest in adolescents. 331 “Pedophilia,” under the
DSM’s criteria, can only be applied to those who have persistent sexual interest in children under
the age of 14. Like “nonconsent,” the term “hebephilia” appears nowhere in the DSM, and there
is no disorder recognized in the manual for sexual interest in teens. 332 In U.S. v. Carta, 333 the
Court of Appeals for the First Circuit reversed a district court’s denial of a commitment petition
brought under the Adam Walsh Act. The district court had ruled that “Paraphilia NOS –
Hebephilia,” which was one of the labels for the respondent’s mental abnormality offered in
support of the government’s petition, was not generally recognized as a serious mental illness
that could support an involuntary commitment. 334 The disorder was characterized by the
government’s testifying expert as a “‘sexual preference for “young teens ... ‘till about age
seventeen.”’ 335 In reversing such ruling, the appeals court acknowledged that the DSM contains
no reference to hebephilia or a sexual interest in teens, but reasoned that the specific diagnosis
offered in support of the commitment in that case was simply “Paraphilia NOS,” which does
appear in the DSM, and that the government’s expert had used the term “hebephilia” as a way to
describe the object of the respondent’s fixation, namely adolescents. 336 It also held that, in any
event, the “serious mental illness” requirement of the SVP statute “is not limited to either the
consensus of the medical community or to maladies identified in the DSM. 337


See e.g., Brown v. Watters, 599 F.3d 602, 607, 612 (7th Cir. 2010) (reaching same conclusion in appeal raising
same due process claim as in McGee where state’s testifying expert admitted that the “clinical indicators” he used to
arrive at the paraphilia NOS- Nonconsent diagnosis did not appear in the DSM and were not accepted by any
professional organization).
For a critical and detailed examination of the development and use of this label in SVP proceedings see Karen
Franklin, Hebephilia: Quintessence of Diagnostic Pretextuality, 28 BEHAV. SCI. & L. 751 (2010).
Id. at 751, 760–61.
U.S. v. Carta, 592 F.3d 34 (1st Cir. 2010).
U.S. v. Carta, 620 F.Supp.2d 210, 217 (D. Mass. 2009).
Id. Another judge in the District of Massachusetts also excluded expert testimony based upon a “hebephilia”
diagnosis. U.S. v. Shields, No. 07-12056, 2008 WL 544940, at *2 (D. Mass. Feb.26, 2008) (ruling that “hebephilia”
could not in itself serve as a serious mental disorder for purpose of commitment under the Adam Walsh Act and that
there was insufficient evidence of the applicability of Paraphilia- NOS in that case). However, that same judge later
admitted evidence of a hebephilia diagnosis, based upon the appeals court opinion in Carta. United States v.
Wetmore, 766 F. Supp. 2d 319, 332 (D. Mass. 2011) (basing commitment, in part, on expert testimony of
“paraphilia not otherwise specified, characterized by hebephilia”) aff'd, 700 F.3d 570 (1st Cir. 2012) cert. denied,
133 S. Ct. 1652, 185 L. Ed. 2d 631 (U.S. 2013).
U.S. v. Carta, 592 F.3d 34, 41 (1st Cir. 2010). On remand, Carta was committed after a seven-day trial, No. 07–
12064–PBS, 2011 WL 2680734 (D. Mass. July 7, 2011), and that ruling was affirmed on appeal. U.S. v. Carta, 690
F.3d 1 (1st Cir. 2012).
Carta, 592 F.3d at 39–40. By contrast, while the court in U.S. v. Neuhauser, 2012 WL 174363 *2 (E.D. N.C.
2012), admitted testimony that the respondent should be committed based upon a diagnosis of hebephilia, it later
concluded that, in light of the fact that “a large number of clinical psychologists believe is not a diagnosis at all, at
least for forensic purposes,” it was “inappropriate” to base a commitment upon such diagnosis. The court also
observed in its opinion: “It is important to note that Mr. Neuhauser's sexual orientation toward pubescent boys,
which he openly admitted in his testimony is, standing alone, insufficient to justify his civil commitment under the
Adam Walsh Act.” Id. at *4


Most courts, when presented with testimony from a government witness applying a label
that purports to be an expansion on the catchall Paraphilia NOS as central evidence of the
respondent’s “mental illness or abnormality,” have admitted and based commitments on such
evidence. They have done so even where the respondent’s expert directly challenged the
scientific basis for use of such label and testified about the considerable controversy about it
within psychiatry. 338 One New Jersey Superior Court opinion noted that the state’s expert had
acknowledged that the Paraphilia NOS diagnosis is used by examiners “in order to code for rape
or coercive or non-consent sex”; the commitment was nonetheless affirmed on appeal. 339 Most
such courts adopt the reasoning of that in Carta: the fact that “Paraphilia NOS” itself is in the
DSM (albeit without criteria established or confirmed by research or field trials) is apparently
sufficient to permit a prosecution expert to claim any form of persistent sexual interest not
described in the DSM as appropriately falling under that catchall label. 340
As noted earlier, some prosecutors have attempted to meet the “mental disorder or
abnormality” requirement of an SVP statute with a diagnosis of Antisocial Personality Disorder
(ASPD), 341 and respondents frequently challenge such use under the holding and analysis in
Foucha. 342 For example, in Brown v. Watters, a federal court habeas case brought by a man
committed under Wisconsin’s SVP law, the respondent presented expert testimony to challenge
the ASPD diagnosis used by the state’s expert witness. 343 Specifically, his forensic psychiatrist
testified that ASPD is a “‘circular diagnosis’ that is ‘descriptive of many criminals, but doesn't
really tell [an evaluator] much’” and that “the psychiatric profession does not generally view
individuals with ASPD ‘as people who have serious difficulty in controlling their behavior.’” 344
The district and appeals courts concluded that, as with the controversies regarding paraphilias, a
fact finder may consider such differing views when determining the weight to be assigned to the
diagnosis, but the existence of debate within the psychiatric community does not itself provide a
basis to exclude a diagnosis. 345


See, e.g., New York v Shannon S., 980 N.E.2d 510, [need pin cite for reporter] (N.Y. 2012); In re Det. of
Hutchcroft, 824 N.W.2d 562 (table), slip op. at ** 2–3 (Iowa Ct. App. 2012) ; In re Detention of Lieberman, 955
N.E.2d 118, 134–35 (Ill. Ct. App. 2011); In re A.M., 787 N.W.2d 752 (2010). See, e.g., Lieberman v. Kirby, 2011
WL 6131176 (N.D. Ill. 2011). cf. U.S. v. Abregana, 574 F.Supp.2d 1145, 1159 (D. Hawaii 2008) (concluding that
paraphilia NOS – hebephilia is a mental disorder, but was not a “serious mental disorder” in the respondent’s case
for purposes of commitment under the Adam Walsh Act).
In re Civil Commitment of D.X.B., 2006 WL 488641 (N.J. Ct. App. 2006) slip op. at 4. See also U.S. v. Graham,
683 F. Supp. 2d 129, 141–46 (D. Mass. 2010) (noting that, notwithstanding the significant controversy regarding the
validity of the diagnosis “Paraphilia NOS- nonconsent,” it could be an appropriate diagnosis in some cases).
See, e.g. Shannon S., 980 N.E.2d 510, [need pin cite for reporter] (N.Y. 2012); In re Det. of Hutchcroft,, 824
N.W.2d 562 (table), slip op. at *3.
See supra notes 268–277 and accompanying text.
See, e.g., Adams v. Bartow, 330 F.3d 957, 959 (7th Cir. 2003); Linehan v. Milczark, 315 F.3d 920, 928
(8th Cir. 2003).
Brown v. Watters, 599 F.3d 602, 612 (7th Cir. 2010).
Id. at 607.
Id. at 613–14. The 7th Circuit also concluded that the respondent had misread the holding on Foucha and that in
any event Crane provided the key authority on the question of the adequacy of a diagnosis in an SVP commitment
proceeding. Id. at 613. Mr. Brown was also unsuccessful on his claim that the state should be judicially estopped
from using ASPD as a basis for commitment where state law precludes a criminal defendant from using the
diagnosis as a basis for an insanity defense. Id. at 615–16.


Courts do differ, however, in their treatment of ASPD diagnoses as bases for SVP
commitment. For example, a federal district court judge in Massachusetts rejected the use of
ASPD as the predicate mental disorder in an SVP case brought under the Adam Walsh Act. In In
re Wilkinson, the court denied the Government’s petition (the respondent was nearing the end of
a 16 year sentence for being a felon in possession of a firearm, and two of the sex crimes had
occurred 25 years prior or longer) and concluded: “The government has not proven that
Antisocial Personality Disorder alone ever causes a person to have serious difficulty in
controlling his conduct. In essence, the evidence indicates that individuals with severe forms of
that disorder may often make unlawful choices, but they are able to control their conduct. 346
Significantly here, the court had conducted a careful review of the literature regarding ASPD and
SVP proceedings and concluded that there was little support for an SVP commitment on that
diagnosis alone, without some additional finding of a sexual disorder indicating limited
volitional control. 347 Indeed, given that studies estimate that a large majority of the prison
population at any given time could be diagnosed with ASPD, using it as the sole predicate
diagnosis would violate the limitations required in Crane, that is, that the individual subject to
the SVP commitment not be a mere recidivist but someone with an identifiable pathology
affecting volitional control of sexual violence. 348
Where a government expert in an SVP proceeding bases an opinion upon a DSM
paraphilia diagnosis such as pedophilia, it is often the case that, notwithstanding the DSM
editors’ clarifying statements to the contrary, the diagnoses are based largely upon a respondent’s
past criminal behavior or other conduct rather than (or even in the absence of) evidence of
persistent, intense urges or fantasies. 349 In these situations, “legal criteria for a crime and the
psychiatric criteria for mental disorder tend to converge,” which runs counter to the DSM
editors’ caution that social deviance in itself should not be thought to constitute a mental
disorder. 350 The editors of DSM-IV attempted to limit the forensic implications of the paraphilias
by stating in an editorial that assigning a diagnosis based solely on a person’s criminal history
was incorrect: “Defining paraphilia based on acts alone blurs the distinction between mental
disorder and ordinary criminality. Decisions regarding possible lifelong psychiatric commitment
should not be made based on a misreading of a poorly worded DSM-IV criterion item.” 351 As
discussed below, the editors’ recommendation that this confusion be alleviated through text
revisions in the DSM-5 went unheeded. 352


U.S. v. Wilkinson, 646 F.Supp.2d 194, 196 (D. Mass 2009).
Id. at 202–207.
Prentky et al., supra note 7, at 368.
Wakefield, supra note 204, at 202. The practice of basing diagnoses of paraphilia solely on past criminal
behavior has met with mixed responses from courts, generally depending upon the extent to which the defense
expert convincingly explains the error in interpretation and application of the DSM criteria or upon whether or not
the court, for whatever reasons, exercises discretion in following the DSM.See, e.g., U.S. v. Springer, 715 F.3d 535,
546–47 (4th Cir. 2013) (affirming dismissal of SVP petition despite testimony of government experts that respondent
had pedophilia based upon his prior sexual acts with children).
Id. See also Allan Frances, DSM-5 Writing Mistakes Will Cause Great Confusion, HUFFINGTON POST (June 11,
2013) (noting that the use of “or” in the DSM-IV-TR B Criterion is his “greatest regret” about that edition because
“[t]his one stupid slip contributed to the unconstitutional preventive detention of thousands of sex offenders.”).
See infra notes 462–468 and accompanying text.


Aside from the DSM editors’ cautionary statements, there is a significant additional
reason to question testifying experts’ diagnostic impressions using labels such as ASPD or
paraphilia-NOS in SVP proceedings. The results of studies of “inter-rater reliability” (the extent
to which two experts will arrive at the same diagnosis when evaluating the same offender) in the
SVP context are unsettling. A study of evaluators applying DSM criteria to those identified for
commitment under Florida’s SVP law revealed a reliability level in the “poor” range; this result
was consistent with earlier studies of SVP evaluators. 353 The author of the Florida study
attributes the findings both to “evaluator bias” and, more significantly, to the fact that
“practitioners are faced with diagnostic criteria that contradict both empirical research and
clinical conceptualization.” 354 Similarly, the authors of a 2013 study of 350 SVP evaluations
conducted in New Jersey found low reliability, that is, only “poor to fair agreement” among
clinicians as to the presence of the paraphilias and other disorders on which the commitments
were based. 355 The authors remarked that such high levels of inconsistency are a “widespread
issue” across states and diagnostic categories. 356 What one commentator calls the DSM’s
“idiosyncrasies and shortcomings” have a significant impact on the reliability of expert opinion
offered in SVP proceedings and, thereby, on the justification of the indefinite commitment of
respondents. 357

Basing Opinions on Records and Inadmissible Evidence

It is clear from the court’s description in McGee that the opinions of the prosecution
experts were not derived from methods and sources of information generally associated with
sound and reliable medical assessments. The experts were permitted to testify as to their
diagnostic opinions of Mr. McGee and their assessments of his volitional impairment solely on
the basis of information compiled and furnished to them by government attorneys and without
ever having examined the respondent. Such practices are common in SVP proceedings, often
because the respondent refuses to be examined. Government experts, in such cases, typically
review criminal investigation reports and alleged victims’ statements 358 (including information
that would be inadmissible in a criminal proceeding 359) and utilize these accounts of conduct to
identify “symptoms.” 360 Such practice by forensic psychiatrists has ben condemned by other
mental health professionals as a specific violation of professional ethics. 361

Jill S. Levenson, Reliability of Sexually Violent Predator Civil Commitment Criteria in Florida, 28 LAW AND
HUMAN BEHAVIOR, 357, 363–64 (2004).
Id. at 366. Other studies have generally documented the extent to which diagnostic assessment by mental health
professionals exhibits unconscious biases and the operation of other cognitive mechanisms that can lead to distorted
Anthony D. Perillo et al., Examining The Scope Of Questionable Diagnostic Reliability In Sexually Violent
Predator(SVP) Evaluations, 37 INTL. J. OF LAW AND PSYCHIATRY 190, 193–94 (2014).
Id. at 196.
Levenson, supra note 353, at 366.
Rebecca L. Jackson, et al., The Adequacy and Accuracy of Sexually Violent Predator Evaluations:
Contextualized Risk Assessment in Clinical Practice. 3 Intl J. of Forensic Mental Health 115, 125 (2004).
See, e.g., State v. Mark S., 87 A.D.3d 73, 78–79 (N.Y. Ct. App. 2011). See also Colleen D. Duffy, The
Admissibility of Expert Opinion and the Bases of Expert Opinion in Sex Offender Civil Management Trials in New
York, 75 Alb. L. Rev. 763, 763 (2012).
See Hamilton, supra note 7, at 50.
Prentky et al., supra note 7, at 370.


The McGee panel placed great stock in the DSM’s recognition of the role of “clinical
judgment” in cases of mental disorder where precise DSM criteria are not met, such as when
clinicians apply an “NOS” (i.e. not otherwise specified) label. “Clinical judgment” is defined in
one medical dictionary as “the application of information based on actual observation of a
patient combined with subjective and objective data that lead to a conclusion.” 362 What the panel
in McGee failed to note was that the two testifying forensic experts had in fact never had the
opportunity to use their “clinical judgment” when arriving at their conclusions about McGee’s
condition, including what they testified as to his diagnosis and volitional impairment, since they
had never observed the “patient.” Rather, they had simply reviewed evidence acquired by
others, namely, law enforcement officials, and had drawn their conclusions therefrom. Here
again, the testimony of experts in McGee was hardly unique for SVP proceedings. A survey of
evaluation methods by forensic experts in such proceedings found that “documentation” of that
kind, that is, police reports, treatment records, and institutional records, were the most important
sources they considered in assessing respondents for SVP commitment. 363
Because of evidence rules, such as Federal Rule of Evidence 703, that permit an expert to
base an opinion on inadmissible facts and data where others in the field reasonably rely on such
sources, the use of inadmissible evidence to arrive at an opinion does not in itself generally lead
to the exclusion of such opinion at trial. 364 Such evidence rules can also, in some instances,
permit such otherwise inadmissible facts and data themselves to be admitted to explain or
support an opinion. 365 However, one appellate court, applying principles of due process because
the proceeding “may result in a serious deprivation of the defendant’s interest in liberty,” has
specifically held that an expert witness for the state in an SVP cannot based his or her opinion
upon inadmissible hearsay even if it would otherwise be admissible under rules similar to FRE
703. 366 Rather, the court stated, “because hearsay can permeate the evidence used to commit a
sex offender, a victim's hearsay statements in police reports or presentence reports must have
special indicia of reliability to satisfy due process” before they can serve as the basis for the
expert’s opinion. 367
In some SVP proceedings, the information about the respondent’s past criminal activity
provided to the expert witnesses, and even to the fact finder, had never been tested through the
adversarial process in a criminal trial. For example, in McGee, the predicate conviction on which
the SVP petition against the respondent was based dated from 1987, more than twelve years
before the trial on petition. However, at the trial, the state also offered evidence of alleged
conduct that was the basis of his probation violations, even though McGee had never been
charged or convicted for such conduct. Other courts as well have permitted evidence of

MOSBY'S MEDICAL DICTIONARY 380 (9th ed. 2013) (emphasis added).
Rebecca L. Jackson & Derek T. Hess, Evaluation for Civil Commitment of Sex Offenders: A Survey of Experts,
19 SEX ABUSE 425, 431 (2007).
FED. R. EVID. 703.
Id. (“[I]f the facts or data would otherwise be inadmissible, the proponent of the opinion may disclose them to the
jury only if their probative value in helping the jury evaluate the opinion substantially outweighs their prejudicial
effect.”). See generally Duffy, supra note (reviewing the application of the “professional reliability” exception to
the hearsay rule in SVP proceedings in several states).
In re Interest of A.M., Jr., 797 N.W.2d 233, 261 (Neb. 2011) cert. denied, 132 S. Ct. 341, 181 L. Ed. 2d 214
(U.S. 2011).
Id. See also Jenkins v. State, 803 So. 2d 783, 786–87 (Fla. Dist. Ct. App. 2001) (holding that SVP commitment
cannot be based upon hearsay evidence).


uncharged but alleged criminal conduct to be admitted and considered as part of SVP
proceedings. 368 Thus a Washington appeals court affirmed the commitment of a man who had
been convicted of three rapes where the trial court in his commitment hearing had admitted the
testimony of a “criminal justice professor” who had concluded, based upon an analysis of
uncharged crimes bearing the respondent’s modus operandi in a database, that the man could
have committed an additional 17 unsolved sexual assaults. 369
Ironically, although courts permit experts to base opinions regarding dangerousness on
criminal conduct alone, at least one court made a point of noting that a lack of criminal conduct
(specifically, violence against persons) is insufficient to demonstrate that a person does not pose
a high risk of committing acts of violence in the future. In one recent SVP case, U.S. v.
Volungus, 370 the primary predicate offense was possession of child pornography; there was no
evidence that the defendant had actually molested any children. 371 The respondent acknowledged
at his SVP trial that he was attracted to children and the evidence showed that he was obsessed
with child pornography. At trial and on appeal, he challenged the Government’s expert’s
conclusion that his diagnosis of pedophilia supported a finding that he posed a high risk for
engaging in molestation. Specifically, he argued (and offered expert testimony in support) that,
despite his strong sexual attraction to children, he had in fact exercised control over acting on his
urges by not committing acts of molestation. 372 The trial and appeals courts rejected such
arguments and concluded that his pedophilia and pornography use were evidence of a
“trajectory” that “would cause him serious difficulty in refraining from child molestation in the
future.” 373 Such inferences run counter to the research findings discussed earlier regarding the
lack of any clear causal links between attraction to children and engaging in acts of sexual
molestation against them.
The disturbing trends seen in the methods used by expert testifying on behalf of the
government in SVP cases reflect the fact that mental health professionals have no scientific
foundation on which to assess “volitional impairment,” and therefore necessarily base their
conclusions largely on the respondents’ history of criminal behavior. Indeed, courts apply little
scrutiny to an expert’s assessment of the respondent’s volitional impairment as such. 374 Where

See, e.g., In re Detention of Coe, 250 P.3d 1056, 1067–68 (Wash. Ct. App. 2011); In re Williams, 253 P.3d 327,
337 (Kan. 2011); Boyce v. Com., 691 S.E.2d 782, 785–86 (Va. 2010); In re Care & Treatment of Miller, 210 P.3d
625, 633 (Kan. 2009).
In re Detention of Coe, 250 P.3d 1056, 1060–65 (Wash. Ct. App. 2011).
U.S. v. Volungus, 730 F.3d 40 (1st. Cir. 2013).
Id. at 42–45. The respondent had been convicted 10 years earlier of “attempted molestation” for having online
contact with someone he thought was a 14-year old girl, but was in fact the fictional creation of a undercover FBI
agent. Id. at 42–43.
Id. at 48–49.
Id. at 48. The appeals court conflated an “inability to control attraction,” which is not sufficient to support an
SVP commitment under Hendricks-Crane, and an inability to control one’s behavior. Those on a gluten-free diet
may have an uncontrolled attraction to chocolate cake, yet manage to avoid eating it out of concerns of adverse
consequences of doing so.
For example, the New York Appellate Division upheld an SVP commitment against a challenge based on
insufficient evidence where the state’s expert opined that the respondent had difficulty controlling his behavior
because he was aware that he “had a problem” with exposing himself to people yet continued to do so. State v.
Richard VV., 903 N.Y.S.2d 184, 186 (2010). Curiously, the forensic expert also considered the fact that the
respondent met most of the diagnostic criteria for ASPD to be further indication that he was unable to control his
behavior, id., although there is nothing in that diagnosis that is associated with volitional impairment. See also Eric


experts rely primarily upon law enforcement or prosecution files, including information such as
witness statements and criminal history, to render an opinion about volitional impairment, they
are engaging in the essentially same process and using the same information to reach a
conclusion as ordinary lay fact finders do when they receive and evaluate evidence offered by
the state at a trial. This raises the question of what “helpful” opinion testimony such experts are
really bringing to the courtroom and, conversely, whether they are simply doing the fact finder’s
job (albeit from an arguably biased perspective) under the guise of offering their “expertise.” 375
Given the variability and unreliability of expert testimony stressed here, it is not
surprising that, overall, mental health professionals’ predictions of recidivism by SVPs appear to
be no more accurate than those made by laypersons on the basis of general knowledge. Empirical
studies confirm what psychiatrists themselves have long stated to be the case: their predictions of
recidivism by SVPs are little better than chance. 376 A 2004 study concluded that experts were
accurate in predicting future sexual violence about one-half of the time. 377 This study also
confirmed many other concerns about the reliability of expert opinion in SVP cases, such as the
emotional impact of reviewing victims’ statements and other information in criminal records on
the development of an evaluator’s opinion and the existence of an overall bias towards “locking
up” prior offenders regardless of what is otherwise reasonably determined as actual risk. 378
These findings are consistent with prior studies of clinical judgment that have long
established that, due to the operation of a range of cognitive biases, such judgment, even by
intelligent, ethical and well-trained professionals, is significantly inaccurate. 379 For example,
where a professional fails to grasp the complexity of the circumstances that can lead to various
outcomes, the degree of confidence she feels in her conclusion, rather than being a measure of its
accuracy, may indicate just the opposite. 380 Also it appears that the very act of predicting the
likelihood of a rare event, because it involves visualizing the possibility of that event, leads to
overestimating the risk of its occurrence. 381 As psychologist Daniel Kahneman has observed:
“Errors of prediction are inevitable because the world is unpredictable” and yet “we resist our
limited ability to predict the future.” 382 We are easily misled by both hindsight bias (i.e., we
overestimate the extent to which we can identify causal relationships but base decisions on the
assumption that we have identified them correctly) and by a “readiness to ascribe propensity to
behavior” (i.e., we see behaviors that may be strongly affected by context as reflections of

S. Janus, Sex Offender Commitments: Debunking the Official Narrative and Revealing The Rules-In-Use, 8 STAN. L.
& POL'Y REV. 71, 83–84 (1997).
See FED. R. EVID. 702 (“A witness who is qualified as an expert by knowledge, skill, experience, training, or
education may testify in the form of an opinion or otherwise if … the expert’s scientific, technical, or other
specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue;...”).
Jackson, et al., supra note 358, at 124; Erica Beecher-Monas & Edgar Garcia-Rill, Danger At The Edge Of
Chaos: Predicting Violent Behavior In A Post-Daubert World, 24 CARDOZO L. REV. 1845, 1869–70 (2003).
Jackson, et al., supra note 358, at 124, 127.
Id. at 125. Another factor in the poor results was the fact that most of the terms in the applicable legal standards
were not sufficiently “operationalized,” meaning that the specific terms are poorly defined (if they are defined at
all). Id.
DANIEL KAHNEMAN, THINKING, FAST AND SLOW, 238–42 (2011); TAVRIS, ET AL., supra note 354, at 97-126.
KAHNEMAN, supra note 379, at 212.
Id. at 333.
Id. at 217–220.


underlying inclinations). 383 Both of these general cognitive tendencies can influence the thinking
of testifying experts, and both can influence the way fact finders weigh expert testimony in
making SVP commitment determinations.

Using Actuarial Tools

Expert opinion evidence offered by prosecutors in SVP cases is not always based upon
diagnostic assessment alone. The appeals court opinion in McGee notes that both of the State’s
experts also used actuarial risk assessment (“ARA”) instruments to arrive at their conclusions
about the respondent’s specific degree of risk of recidivism. Because McGee did not challenge
such use on appeal, the description of their testimony on the role of such tools is very limited. 384
Dr. Marsh testified regarding the scores she assigned to McGee under the three tools she used to
arrive at her conclusion, and she indicated that “subjects with scores similar to Mr. McGee’s in
each of these instruments reoffended at rates of between forty-eight and fifty-four percent over a
six- to fifteen-year period following release.” 385
The McGee opinion does not specify which ARA tools were used or described in
testimony by the testifying experts, but they were likely among those commonly used by forensic
examiners offering evidence in support of SVP commitment. The appropriateness of the use of
tools such as the “Static-99,” Rapid Risk Assessment for Sexual Offense Recidivism (RRASOR)
or Sex Offender Risk Appraisal Guide (SORAG) as a basis for expert opinions in support of SVP
commitment is an unsettled question in the courts. Some forensic examiners have advocated
greater use of ARA tools, which they characterize as especially objective, to address the
problems of bias and low inter-rater reliability accompanying clinical judgment and diagnostic
assessment described above. 386 A growing number of experts use risk-prediction actuarial tools
to inform their opinions and to support their testimony about the risk of recidivism, the “final and
most nebulous” part of the SVP analysis, 387 posed by a respondent. One study of evaluation
methods found that the vast majority of forensic evaluators used one or more tools as part of the
assessment process. 388 The guidelines issues by the Association for the Treatment of Sex
Offenders require use of such tools, although no single tool has emerged as the preferred. 389
These instruments are generally developed from studies of sex offenders that isolate a
number of specific “factors” including the number of sex offense convictions and characteristics
of the individual’s victims (age, gender, and relationship to the individual), associated with those
who recidivate. 390 Those factors are assembled into what are essentially checklists. Many of the
instruments can be completed without evaluating the individual but simply from reviewing
records, including court records. The results indicate what percentage of those individuals in the
study who share the offender’s factors went on to commit new crimes (sometimes identified by

Id. at 199–201.
McGee v Bartow, 593 F3d 556, 559-60 (7th Cir 2010).
Id. at 560.
Prentky et al, supra note 7, at 372–72.
Jackson & Hess, supra note 363, at 428 (noting that 95.1% of respondents used such instruments and 75% listed
them as “essential” to the evaluation process).
Id. at 434.
Id. at 426.
EWING, supra note 13, at 36–37.


arrests rather than convictions). After the factors are entered, the tool yields a score that places
the individual in a risk range, such as “high risk,” and perhaps offers a percentage of likelihood
of reoffending. 391 Thus, the tools are not psychological tests, 392 nor are they predictors of an
individual’s specific likelihood to re-offend. 393 The expert witness testifies that the actuarial
analysis of objective factors places the respondent at a specific level of risk of reoffending, 394
although such conclusion is not keyed to any legal criteria. 395 The tools also shed no light on the
questions of abnormality or volitional impairment.
Some commentators have advocated for the complete replacement of clinical judgment
with the use of actuarial instruments, given results of studies suggesting such practice would
yield improved accuracy. 396 Noted behavioral psychologist Paul Meehl argued decades ago that
clinical judgment is inferior to actuarial analysis, 397 and his findings have been replicated and
reinforced many times since his initial studies. 398 Empirical studies have shown that ARAs are
specifically better predictors of recidivism than “clinical judgment” alone, 399 a standard that does
not seem to be all that difficult given the exceptionally poor ability of forensic examiners to
predict recidivism. 400
However, as other commentators have stressed, there are reasons to approach the use of
ARAs in SVP proceedings with considerable caution. The use of ARAs is highly controversial
among legal and mental health professionals and several limitations of the effectiveness have
been noted. 401 One of the biggest shortcomings of the Static-99 and similar instruments is that
they assess risk based upon a series of “static” factors that do not change (such as the age of first
offense, characteristics of the victims etc.) over an offender’s lifetime. They therefore may fail to
take into account for dynamic factors such as life circumstances and participation in treatment.
because they are based upon the assumption that one’s risk never changes, even if one makes
choices to address the underlying propensity. 402 As a result, other than perhaps a decrease due to
aging, a person’s score will not change significantly a person’s score could be the same the day
of release from incarceration and 10 year later, even after leading an entirely law-abiding life


Jackson & Hess, supra note 363, at 439.
Indeed, one study of evaluation procedures noted how less frequently psychological testing is used in the SVP
context as compared with other forensic evaluations, such as for insanity and competency. Jackson & Hess, supra
note 363, at 437.
U.S. v. Hall, 664 F.3d 456, 464 (2012).
See, e.g., Lieberman v. Kirby, 2011 WL 6131176 (N.D. Ill. 2011).
Jackson & Hess, supra note 363, at 439.
Beecher-Monas & Garcia-Rill, supra note 376, at 1871; see generally 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).
EVIDENCE 94-95 (1954).
Janus & Prentky, supra note 396, at 1455.
Daniel A. Krauss, et al., Dangerously Misunderstood: Representative Jurors' Reactions to Expert Testimony on
Future Dangerousness in a Sexually Violent Predator Trial. 18 PSYCHOLOGY, PUBLIC POLICY & LAW 18, 20 (2012);
Prentky et al, supra note 7, at 372; Janus & Prentky, supra note 396, at 1455–58.
Such findings are consistent with studies of accuracy of many different kinds of predication across disciplines.
See generally KAHNEMAN, supra note 379, at 222.
Saleh, et al., supra note 242, at 366; Krauss et al., supra note 399, at 20.
Krauss et al., supra note 399, at 20.


during the interim. 403 Such an approach to risk assessment fails to take into account not only the
passage of time, but also the events that occurred (or did not occur) during such time, thus
rendering any such assessment severely liable to inaccuracy. 404 Some instruments do not even
consider the mitigating effect of age on risk of recidivism.405 A few scholars have advocated for
a uniform use of “dynamic risk factors” before a final risk assessment is made using ARAs, 406
although research has not yet suggested how best to integrate such factors. 407
The SVP laws and the call for risk assessment as the core question in the proceedings
have spawned a cottage industry of developing new instruments, each of which promises to be
more precise that those developed (and in use) before it. 408 However, no consensus in the field
has emerged regarding which test is most applicable and appropriate in the SVP commitment
setting, 409 or for predicting dangerousness generally, 410 and there are some sharp differences in
opinion and approach among psychologists who have developed and used various instruments. 411
Many commonly used ARAs have been criticized for being unreliable. For example, the SVR-20
(at least as of 2000) used only broad categories of risk (high, medium, and low), and there were
no inter-rater reliability rates for specific factors. 412 There is also no consensus what level of
predictive validity is sufficient for the instruments to be considered a useful tool for predicting
recidivism. 413
ARAs, even at their best, can still be used well or poorly. 414 Although the instruments are
said to be objective, the evaluators who administer them are not immune from common failings
of human judgment and bias, and the concept of “risk” is itself a construct subject to variable
understandings. 415 A simple difference in how the outcome of a risk is presented, in terms of a
probability versus a frequency, can affect how high a professional assesses the risk. 416 Also, the
objective factuality of some of the individual factors considered in the instruments may not be as
clear as initially assumed. For example, a factor such as participation in or compliance with

For an example of how the use of an ARA can have an impact on risk assessment of a person who commits a
crime at a young age, see Nora Hertel, Sex Offender Awaits Second Chance, WISCONSIN WATCH (Feb. 4, 2014)
(available at )
Prentky et al., supra note 7, at 378.
Id. at 375.
Id. at 383–85.
Good & Bursetin, supra n note 46, at 30.
Prentky et al, supra note 7, at 371–72; Faigman, supra note 291, at § 11:28.
Black v. Voss, 557 F.Supp.2d 1100, 1105–1107 (D. Cal. 2008) (noting that respondent’s expert criticized the
government’s experts for using the STATIC-99 test to assess risk for reoffending because that test addressed
criminal activity, not sexual deviancy, and advocated use of the Rapid Risk Assessment of Sexual Offense
Recidivism (“RRASOR”) test instead).
M. Neil Brown & Ronda R. Harrison-Spoerl, Putting Expert Testimony In Its Epistemological Place: What
Predictions of Dangerousness in Court Can Teach Us, 91 MARQ. L. REV. 1119, 1198–1204 (2008).
Prentky et al., supra note 7, at 373–80.
Terence W. Campbell, Sexual Predator Evaluations and Phrenology: Considering Issues of Evidentiary
Reliability, 18 Behav. Sci. Law 111, 120–21 (2000).
Good & Burstein, supra note 46, at 34.
Janus & Prentky, supra note 396, at 1493–97.
Beecher-Monas & Garcia-Rill, supra note 376, at 1871.
Risks phrased in the form of the probable occurrence of specific event are evidently less “vivid” than one phrased
in the form of a frequency. KAHNEMAN, supra note 379, at 330 (“Experienced forensic psychologists and
psychiatrists are not immune to the effects of the format in which risks are expressed.”).


treatment can be a complex question where there is limited access to treatment, 417 where the
treatment is cursory, or where the treatment requires disclosure or other actions by the committed
person that could lead to lengthier commitment in the absence of Fifth Amendment protections.
The use of instruments or set “factors” can also lead to “cherry picking” the factors to be
considered in the analysis, which can also lead to skewed results. 418 Some scholars suggest that
experts’ practice of making individualized “adjustments” to scores may little more than “dressing
up clinical judgment with actuarial science.” 419 Given such problems, several scholars have
suggested that the use of ARAs by examiners in SVP proceedings is unethical. 420
Testimony based upon ARA tools has received a mixed reaction in the courts. Some
courts resist admitting opinions based on such tools more than they resist admitting those based
solely upon diagnostic impressions. 421 In at least one case, a court rejected the forensic expert
testimony because the ARA employed failed to take into account events in the respondent’s life
that had transpired since the “factors” used in the assessment. 422 Some courts are uncertain about
how much weight is appropriate to give to the specific scores from such tests. For example, in In
re Williams, a Kansas appeals court reversed an SVP commitment because the government’s
expert had testified that the respondent’s score, which was lower than a 50% chance of
reoffending, was too low to sustain such a commitment. 423 The Kansas Supreme Court reversed
that ruling, however, arguing that there was other evidence to support a finding that the
respondent was likely to engage in acts of sexual violence. 424 Finally, some courts have excluded
testimony based on ARA results altogether because of concerns about unfair prejudice. 425
Despite the shortcomings of ARAs, many courts have embraced the tools, seeing them as
akin to psychological tests or as amounting to an objective predictor of a particular offender’s
individual likelihood of re-offending. 426 In U.S. v. Shields, for example, the Court of Appeals for
the First Circuit upheld a commitment order based upon expert testimony employing ARA tools
even though the Government’s experts conceded such tools were only “moderate” predictors of
recidivism and that there were significant reliability problems with the results of the tools used in
that particular case (which, among other things, were based on data obtained entirely outside of
the U.S.). 427 The appeals court concluded that the weight to be given to such evidence should be
left to the fact finder. 428


Prentky et al., supra note 7, at 379.
Id. at 378–79; Good & Burstein, supra note 46, at 30-31 (arguing that ARAs for SVPs may be “systematically
Prentky et al., supra note 7, at 380.
Campbell, supra note 412, at 128.
Krauss, et al., supra note 399, at 37.
Com. v. Squire, 685 S.E.2d 631, 62–33 (Va. 2009) (affirming dismissal of SVP petition despite expert testimony
that actuarial tests placed the respondent in the highest risk category).
In re Care & Treatment of Williams, No. 99, 235, 2009 WL 2762455, at *3 (Kan.App.2009).
In re Willliams, 253 P.3d 327 (Kan. 2011).
Janus & Prentky, supra note 396, at 1487–92.
EWING, supra note 13, at 40–44.
U.S. v. Shields, 649 F.3d 78, (1st Cir 2011)
Id. at 89–90. In that case, the trial court used an advisory jury, which concluded that there was insufficient
evidence of likelihood of the respondent reoffending. However, the court ultimately concluded that the Government
had met its burden. Id. at 85.



Sparse Use of Daubert-Frye Analysis

Allan Frances has implored: “SVP courts must insist on good science.” 429 The key
holding in 1923 U.S. District Court case of Frye v. U.S., which was widely adopted by state
courts, required judges to consider a theory’s “general acceptance” in the relevant scientific
community before allowing its admission. 430 The U.S. Supreme Court’s landmark opinion in
Daubert v. Merrill Dow Pharmaceuticals 431 requires the court to act as a “gatekeeper” with
regard to the scientific evidence presented; that is, to make its own determination of reliability of
such evidence, based in part upon such general acceptance as well as on the presence of other
indicators of “good science.” 432 The controversial nature of psychiatric diagnoses discussed
above, combined with the significant liberty interest at stake in SVP proceedings, suggest that
trial courts in such proceedings should exercise particular vigilance in the “gatekeeping” role
assigned to them by these opinions and by the court rules and opinions following their lead.
However, the case law suggests a significant abdication of this responsibility by the courts.
Legal scholars vary widely in their opinions of the type of gatekeeping scrutiny that
courts should afford to expert testimony by mental health professionals generally, and this range
of legal opinion has implications for SVP cases. At one extreme, some commentators argue that
psychiatry has little to offer courts in such cases. For example, Samantha Godwin has labeled
psychiatry a “pseudoscience” that lacks sufficient reliability to be considered at all in involuntary
commitment hearings. 433 Other scholars have suggested that, while there may be some utility for
mental health testimony in a range of legal contexts, diagnoses themselves should not generally
be admitted. 434 Still other scholars suggest that standards for admissibility of expert evidence
should be relaxed for mental health testimony, that courts should use an “informed speculation”
approach, particularly for evidence offered by a criminal defendant to excuse criminal
conduct. 435
Courts as well are divided on the role of the underlying principles and rationales of cases
such as Daubert and Frye in settling these issues of admissibility of expert psychiatric opinion as

Prentky et al., supra note 7, at 386.
Frye v. United States, 293 F. 1013, 1014 (D.D.C. 1923).
Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993).
One instance in which a court noted that expert testimony fell short of the Frye test and therefore could not serve
as a basis for an SVP commitment is one of the very few reported opinions involving a female respondent. In re
Coffel, 117 S.W.3d 116, 129 (Mo. Ct. App. 2003)
Samantha Godwin, Bad Science Makes Bad Law: How The Deference Afforded To Psychiatry Undermines Civil
Liberties, 10 SEATTLE J. FOR SOC. JUST. 647, 647 (2012). The most significant deficiency Godwin identifies is the
lack of validity of the “somatic reality” of psychiatric diagnoses, since they are based entirely on symptomatology,
not scientific testing. Id. at 662.
Deirdre M. Smith, Diagnosing Liability: The Legal History of Posttraumatic Stress Disorder, 84 Temple L. Rev.
1, 69 (2011); Daniel W. Shuman, Persistent Reexperiences in Psychiatry and Law: Current and Future Trends for
ASSESSMENT 1, 7 (Robert I. Simon ed., 2d ed. 2003) (“Both Daubert and the DSM make clear that it is not
appropriate to assume that a psychiatric diagnosis is relevant to, let alone dispositive of, an issue in a case.”); Daniel
W. Shuman, Softened Science in the Courtroom: Forensic Implications of a Value- Laden Classification, in
ed., 2002); Morse, Crazy Behavior, supra note 191 at 601-04.
Richard Bonnie & Christopher Slobogin, The Role of Mental Health Professionals in the Criminal Process: The
Case for Informed Speculation, 66 VA. L. REV. 427 (1980).


evidence in SVP proceedings. Indeed, the Daubert opinion was not cited at all by the McGee
court, despite McGee’s direct attack on the scientific basis of the state’s experts. The Washington
Supreme Court addressed the question of the applicability of Frye to the admissibility of expert
testimony shortly after enactment of its SVP law. In In re Young, the court rejected the
respondent’s argument that the state’s expert should not have been permitted to based an opinion
upon a diagnostic label that did not appear in the DSM. 436 Quoting a law review article by
Alexander Brooks, the court reasoned:
“The fact that pathologically driven rape, for example, is not yet listed in the DSM–III–R
does not invalidate such a diagnosis. … What is critical for our purposes is that
psychiatric and psychological clinicians who testify in good faith as to mental
abnormality are able to identify sexual pathologies that are as real and meaningful as
other pathologies already listed in the DSM.” 437
Such “good faith” approaches to the admissibility of psychiatric evidence, however, should raise
significant concerns in both the law and medical fields. One group of commentators noted that
courts should be wary of the use of new or “stretched” diagnoses with “no empirical track record
providing evidence for such a linkage.” 438 “Perhaps worse,” they caution, “we are conferring on
unvalidated diagnoses the presumptive medical authority of the DSM.” 439
On the other hand, and in better accord with such recommended caution, some courts
have urged trial courts to apply additional scrutiny to expert opinion evidence offered in support
of SVP commitments. For example, an Illinois appeals court held that a novel diagnosis such as
Paraphilia NOS – Hebephilia must be subject to a Frye hearing before offered to a fact finder. 440
The analysis in In re Detention of New began with finding that expert testimony based upon a
diagnosis “presupposes a mental condition exists as a matter of scientific evidence.” 441 The
court noted the considerable controversy over the “hebephilia” label and concluded that “[a]
Frye hearing is appropriate to determine whether an emerging diagnosis is an actual illness or
disorder.” The court observed, strikingly, that “[j]ustice does not put the fact finder in the
position of culling good science from bad.” 442 The court correctly noted that, above all, the
reasoning of Justice Kennedy’s concurrence in Hendricks mandated a scrutiny of the science
offered in support of an SVP commitment. Since SVP laws are ostensibly based upon a need for
treatment, not retribution, the court reasoned, “if a respondent in an SVP proceeding does not
suffer from an actual mental disorder, then there is nothing to cure, and commitment is
pointless.” 443

In re Pers. Restraint of Young, 857 P.2d 989 (Wash. 1993)
Id. at 1001 (quoting Alexander D. Brooks, The Constitutionality and Morality of Civilly Committing Sexually
Violent Predators, 15 U. PUGET SOUND L. REV. 709, 733 (1991–92)). More recently another Washington appeals
court, in In re Detention of Berry, noted that many courts have held that the Frye rule has no application to the
question of whether a diagnosis of Paraphilia-NOS may be admitted in an SVP proceeding. In re Detention of Berry,
248 P.3d 592, 595–96 (Wash. Ct. App. 2011).
Prentky, et al., supra note 7, at 370.
In re Detention of New, 992 N.E.2d 519, 5287–31(Ill Ct. App. 2013), appeal allowed by 2 N.E.3d 1045 (Ill.
Id. at 528.
Id. at 529.
Id. at 530.


On balance, however, there is little question that, even in the era of Daubert and similar
rules designed to ensure that only reliable expert testimony is admitted, clinical psychiatric
testimony is rarely excluded under such approaches. By the time of the Daubert opinion, the role
of psychiatric testimony was so embedded in legal decision-making that it was inconceivable to
courts that they should scrutinize or might reverse such practice. 444 Indeed, as the Supreme Court
noted in Barefoot v. Estelle: “The suggestion that no psychiatrist's testimony may be presented
with respect to a defendant's future dangerousness is somewhat like asking us to disinvent the
wheel.” 445
The analysis in McGee is remarkable for how far it strays from the core principles set
forth in the Daubert opinion. Presumably he panel did not apply that standard because of the
specific posture of the case. McGee was not a direct appeal challenging the lower court’s
evidentiary rulings on such testimony and there is no indication that McGee raised Daubert-like
challenges in his original proceeding. Rather, because his attorneys brought a habeas petition, the
court considered only whether there was a constitutional violation. The evidence rules, and cases
interpreting them such as Daubert, impose a more specific and therefore a higher standard for
admissibility than does the Constitution.446 But courts routinely follow the lower standard when,
as in the SVP context, they analyze admissibility to determine the constitutionality of an ongoing
deprivation of someone’s liberty. Barefoot in particular, which upheld the use of psychiatric
evidence about future dangerousness in the face of research suggesting the low reliability of such
predictions, suggests a very low standard for admissibility of expert evidence. 447 Such lack of
scrutiny of expert evidence is highly questionable where the state is offering such evidence to
rationalize indefinite detention.
That most courts distinguish between the admissibility standards regarding expert
testimony in the evidence rules and due process jurisprudence raise the question of whether the
admission of expert testimony in a manner apparently inconsistent with the approach required
under Daubert can itself implicate due process. No court has addressed that question squarely,
and the Seventh Circuit was not asked to do so in McGee. However, in cases where a person’s
constitutional rights to liberty are so directly implicated, there clearly are due process
implications for a court’s role as gatekeeper regarding expert opinion. 448 Courts should take into
account in their due process analyses that these invented or extended diagnoses or ARAs,
employed almost exclusively in the SVP commitment (rather than clinical) context, would not
pass either a Daubert or a Frye gatekeeping standard. Indeed, these made-for-trial expert

Barefoot v. Estelle, 463 U.S. 880, 896 (1983).
Paul Giannelli, The Supreme Court’s Criminal “Daubert” Cases, 33 SETON HALL L. REV. 1071, 1072–76
Id. at 1091–92. Giannelli also rejects the reasoning that the standard could be lower because it was an analysis
under the constitution, not the rules of evidence; the “death is different” principle necessarily means that evidence
offered in support of the death penalty should have to meet higher, not lower, standards of reliability. Id. at 1092.
The Supreme Court has not considered this issue, or the continuing validity of Barefoot v. Estelle, in light of
Daubert. See Brown v. Watters, 599 F.3d 602, 616 (7th Cir. 2010) (rejecting argument of SVP respondent based on
Daubert-Frye in an appeal of SVP commitment because “neither purports to set a constitutional floor for the
admissibility of scientific evidence.”); Giannelli, supra note 446, at 1091–92; Beecher-Monas & Garcia-Rill, supra
note 376, at 1859.


opinions appear to be precisely the kind of testimony that the Ninth Circuit excluded in
Daubert. 449
As discussed below, the call to include some of these extended diagnoses in the DSM-5 is
inextricably intertwined with arguments about the usefulness of such diagnoses in SVP
proceedings. This fact should signal to courts that expert opinions in such proceedings do little
more than use medicalized terminology to tell courts and juries what to conclude. Also, given the
consistent rejection from mainstream psychiatry and lack of peer-reviewed research supporting a
methodology of prediction, there is serious question as to whether any expert prediction of future
dangerousness could pass a strict Daubert test. 450
While courts admit expert testimony regarding future dangerousness (whether based upon
clinical judgment, ARAs, or both), they leave the determination of the weight to be assigned to
such testimony to the fact finder, which is often a jury or, in some states, an elected judge. 451
There are two fundamental problems with the reasoning behind such practice. First, it ignores
the limited ability of laypersons to assess critically the opinions of expert witnesses, one of the
core rationales for the Daubert “gatekeeping” requirement. 452The ability to uncover and assess
problems in reliability can be especially challenging for laypersons with respect to the often
“ipse dixit” opinions 453 offered by mental health professionals.
The second problem concerns the nature of SVP proceedings and the specific task
assigned to fact finders: determining whether a convicted sex offender should be permitted to be
at large in society. It seems unlikely that a fact finder could render a decision on such question
without a fear of repercussions if a conclusion that a respondent had low risk of committing
future acts of sexual violence was wrong. 454 SVP commitment is a decision that puts the fact
finder between an offender and a potential “next victim.” There has been little research on the
question of the extent to which jurors are influenced in their decision-making by expert
testimony on future risk. 455 Nonetheless, it is difficult to imagine how a jury of laypersons, upon

Daubert v. Merrell Dow Pharm., Inc., 43 F.3d 1311, 1317 (9th Cir. 1995) (affirming exclusion of expert
testimony that was based solely upon research conducted for purposes of litigation).
Beecher-Monas & Garcia-Rill, supra note 376, at 1857.
Wisconsin, Minnesota, and Washington, for example, elect trial court judges. Even courts that reject the
government expert’s opinion in an SVP proceeding generally do so under weight or “credibility” principles (after
admitting the testimony) rather than excluding the opinion under either a Daubert (or rule 702) or due process
analysis. See, e.g., U.S. v. Wilkinson, 646 F.Supp.2d 194, 201 (D. Mass. 2009).
Daubert, 509 U.S. at 595 ““Expert evidence can be both powerful and quite misleading because of the difficulty
in evaluating it. Because of this risk, the judge in weighing possible prejudice against probative force under Rule
403 of the present rules exercises more control over experts than over lay witnesses.”) (quoting Hon. Jack B.
Weinstein, Rule 702 of the Fed. Rules of Evidence Is Sound; It Should Not be Amended, 138 F.R.D. 631, 632
See General Electric Co. v. Joiner, 522 U.S. 136, 146 (1997) (noting that an expert’s opinion is not sufficiently
reliable to be admitted when it is “connected to the existing data only by the ipse dixit of the expert”).
Cf People v. Shazier, 212 Cal. App. 4th 520, 531, 151 Cal. Rptr. 3d 215, 224 (2012) (vacating SVP commitment
due to prosecutorial misconduct because, in part, prosecutor’s closing argument included references to the proximity
of schools to where respondent would be living and asking jurors to consider what their friends’ and family
members’ reactions would be if they denied the commitment), review granted and opinion superseded, 298 P.3d 178
(Cal. 2013).
Krauss et al., supra note 399, at 21. Florida courts specifically permit use of the term “sexually violent predator”
in SVP commitment proceedings, notwithstanding concerns raised by the defense bar that the term is “extremely


hearing an expert opine on the basis of an ARA instrument that a child rapist has a 33% chance
of reoffending (i.e. raping another child) would not commit that person. Many people would
likely follow thinking along the lines of former Vice President Dick Cheney’s “one percent
doctrine” and conclude that, in such cases, any amount of risk is too much to accept. 456 Judges
are not immune from similar concerns about the implications of their rulings. One Circuit Judge
on the Court of Appeals for the Fourth Circuit, dissenting from an opinion affirming a district
court’s denial of an SVP petition, wrote: “Though we may never learn the consequences of a
poor predictive judgment on our part, I fear that some young child somewhere will experience
them,” and noted that there are “sad and scarring consequences of a guess gone awry. 457 This
judge likely articulated the mental calculations made by many juries and jurists involved with
these cases. 458
This review of law and practice in SVP proceedings has demonstrated that the prevalent
use of psychiatric evidence in such proceedings is a distortion of both medical views of
pathology of sexual violence—including appropriate diagnostic methods and prediction of future
conduct—and also of legal principles regarding the admissibility of expert opinion, including
cases where such opinion is based upon unreliable methodology or upon data that runs counter to
predominant views of the field and poses risk of misuse by, or misleading of, fact finder. 459
These fundamental and extensive distortions of both sound science and just law are the inevitable
and unavoidable result of the courts’ experiment with SVP laws. These distortions also
demonstrate the accuracy of warnings issued at the outset of the SVP experiment by many in the
field of psychiatry of the dangers presented by the laws themselves.

Fixing the Science to Fit the Courtroom

The opinions in Hendricks and Crane assumed that there was a “bright line separating an
SVP/SDP mental disorder from ordinary criminal behavior.” Such line-drawing, however, “tests
a no-man’s land between psychiatry and the law.” 460 Many scholars and commentators in the
fields of both law and psychiatry believe that the forensic use of psychiatric evidence, and
particularly diagnoses, is unscientific and grossly misleading. Accordingly, there have been
many calls to fix the problem, sometimes by fixing the science.

inflammatory, prejudicial, and misleading” and would deprive respondents of due process. Standard Jury
Instructions-Criminal Cases (99-2), 777 So. 2d 366, 367–68 (Fla. 2000). The committee developing the jury
instructions agreed, however, that the term should not be overused to the extent that it becomes a “feature” of the
trial. Id.
Cheney stated: “If there's a 1% chance that Pakistani scientists are helping al-Qaeda build or develop a nuclear
weapon, we have to treat it as a certainty in terms of our response.” Ron Suskind, One Percent Doctrine: The Untold
Story of al-Qaeda's Plot to Attack the Subway, TIME [pin cite needed] (June 16, 2006).
U.S. v. Springer, 715 F.3d 535, 551(4th Cir. 2013) (Wilkinson, J. dissenting) (emphasis added).
There have not been empirical studies of the rates of commitment in bench versus jury trials, but there are
anecdotal press reports of jurors rejecting SVP commitment petitions. See, e.g., Karen Franklin, Another One Bites
The Dust: Hollow SVP Prosecution No Match For Jurors’ Common Sense, IN THE NEWS (Oct 27, 2012) (the blog
author was one of the defense experts in that case)
Hamilton, supra note 6, at 50; see also Prentky et al., supra note 234, at 456.
Frances, et al., supra note 190, at 383.


Commentators who maintain that science does have something to offer in SVP
proceedings tend to speak of the “disturbing frequency” of the “bad science” that appears in such
proceedings. 461 This conception of the problem of the use of psychiatric testimony in SVP cases
suggests that there may be a role for “good” (or at least “better”) science and, indeed, there have
been many suggestions and proposals for ways to improve the nature of the forensic science
evidence admitted in such proceedings. The proposed fixes include changing the way clinical
diagnoses are approached, changing the diagnoses themselves, and either supplementing or
replacing the diagnostic assessments with the use of actuarial tools. However, none of these
modifications would erase the core problem set up by the Hendricks-Crane rationale: that, in a
highly adversarial context, with very high stakes for the individual and society, courts are asked
to look to the conclusions of psychiatric examiners to answer a moral, normative question.

Addressing Problems with Diagnoses

The specific nature of “problem” of diagnostic labels in SVP proceedings differs
depending upon one’s perspective. Some psychiatric commentators, such as Allan Frances,
complain that experts testifying for the states misuse existing diagnoses such as Pedophilia or
ASPD, or invent diagnoses such as Paraphilia NOS – Nonconsent, which have not been set forth
in the DSM or otherwise been sanctioned by psychiatry. Due to the “uniquely negative outcome,
namely inappropriate and potentially indefinite civil commitment to a secure forensic psychiatric
facility,” these commentators are concerned about the potential for large numbers of “false
positive” diagnoses. 462 Accordingly, there were calls to revise DSM language to eliminate any
potential for such behavior-based approach to diagnosis. 463
Mental health professionals offering testimony for the states in SVP proceedings, by
contrast, see the problem in terms of a failure of the DSM or the field of psychiatry to provide
forensically usable categories. Some of these experts regard the science as failing to reflect the
reality of mental conditions underlying acts of sexual violence. They are concerned about
ambiguities that lead to court challenges to their testimony and/or present potential barriers to
fact finders receiving their opinions. This group, therefore, advocated for revisions to the
paraphilias in the DSM-5 so that there would be a clearer basis in the psychiatric nosology for
identifying the mental disorders most commonly seen in SVPs. Preserving the potential for an
approach to assigning a paraphilia diagnosis based on prior behavior was essential to those
offering testimony in support of commitments. Respondents are often uncooperative with
evaluators 464 or, for any number of other reasons, clinical evaluation may be impossible or not
involved in the diagnostic process. Most DSM diagnoses, however, are based upon the
assumption that they will be used in therapeutic, not forensic, settings and that many of the
criteria will be determined through a clinical interview. 465


Prentky et al., supra note 7, at 361.
First, supra note 214, at1239 (internal citations omitted).
Id. at 1242.
First, supra note 214, at 1240–41.
therapeutic context, the most important tool for diagnosis and assessment is “the clinical interview––a dialogue with
the patient exploring present mental state, past experiences, and desires for the future.”).


The array of views regarding the use and validity of DSM labels reflects the adversarial
setting of SVP proceedings, and it should come as no surprise that the outcome of the debate
over the proposed changes resolved nothing and left the paraphilias essentially unchanged. The
proposals for change did, however, garner fierce debate and a flood of papers and editorials
appeared while they were under consideration. 466 The varied commentaries brought to the
surface many of controversies about psychiatry’s role in these commitments discussed above.
The outcome of the debate was a compromise that resulted in maintaining essentially the
same approach of the DSM-IV-TR. 467 The publication simply maintained the tension between
deviance and disorder with which psychiatry has been increasingly aware. By making minimal
changes to the paraphilias listed in the DSM-IV, the APA rejected many revisions proposed by
those who support the state in SVP commitment proceedings, such as adding the categories
hebephilia or paraphilic coercive disorder. Although many of the proposals to include new
diagnoses were rejected, Allan Frances remains concerned that the revised paraphilias section are
“an ambiguous hodgepodge [which] will surely be misused in sexually violent predator hearings
where every word is given legal spin.” 468
The DSM-5’s editors evidently shared Frances’s concern to some extent (he was an
editor of an earlier edition himself), but they also did not want to see the influence of the manual
wane, including in legal settings. The new DSM’s “Cautionary Statement for Forensic Use” is
longer than the previous one, more explicit in its explanation of the limited purpose for which the
manual was devised (i.e., assisting clinicians with assessment and treatment in clinical settings),
and now has a clearer title. 469 But the statement begins with a sales pitch for its use in forensic
contexts. For example, it states that, “[w]hen used appropriately,” the “diagnoses and diagnostic
information” in the manual can “assist legal decision makers” in involuntary commitment cases
where the “presence of a mental disorder is a predicate.” 470 The manual may also, it states,
“facilitate legal decision makers’ understanding of the relevant characteristics of mental
disorders.” Especially significantly here, it also suggests that “diagnostic information about


See, e.g., Howard Zonada, Sexual Disorders: New and Expanded Proposals for the DSM-5—Do We
Need Them? 39 J AM. ACAD. PSYCHIATRY L. 245 (2011); Fred S. Berlin, Commentary on Pedophilia Diagnostic
Criteria in DSM-5, 39 J AM. ACAD. PSYCHIATRY L 242 (2011); John Matthew Fabian, Diagnosing and Litigating
Hebephilia in Sexually Violent Predator Civil Commitment Proceedings, 39 J AM. ACAD. PSYCHIATRY L 496
(2011); Robert Prentky & Howard Barbaree, Commentary: Hebephilia—A Would-be Paraphilia Caught in the
Twilight Zone Between Prepubescence and Adulthood, 39 J AM. ACAD. PSYCHIATRY L 242 (2011); First, supra note
214, at 1239; First & Halon, supra note 216, at 451-52; Franklin, supra note 331; Wakefield, supra note 204, at
DSM-5, supra note 187, at 697. Pedophilia is now “Pedophilic Disorder” but the diagnostic criteria themselves
are unchanged. The category of “Paraphilia Not Otherwise Specified” has been replaced with “Other Specified
Paraphilic Disorder” and has more extensive explanatory text than that in DSM-IV. Id. at 705. There is also a new
category for “Unspecified Paraphilic Disorder,” which is used in similar contexts as the “Other Specified” disorders
but the “clinician chooses not to specify the reason that the criteria are not met for a specific paraphilic disorder,”
such as where there is insufficient information for a more specific diagnosis. Id.
Allan Frances, DSM-5 Badly Flunks the Writing Test, PSYCHIATRIC TIMES (June 13, 2013)
( )
DSM-5, supra note 187, at 25. Previously the language was simply titled “Cautionary Statement.” DSM-IV,
supra note 215, at xxxvii.
DSM-5, supra note 187, at 25.


longitudinal course may improve decision making when the legal issue concerns an individual’s
mental functioning at a past or future point in time.” 471
The new DSM statement does subsequently lay out cautions about taking such forensic
use too far and, in places, the language appears to be addressed specifically to the experts and
judges involved in SVP proceedings. Thus the statement cautions that foremost is the risk of
misunderstanding “because of the imperfect fit between the questions of ultimate concern to the
law and the information contained in clinical diagnosis.” 472 It also emphasizes that “in most
situations” more information about the individual is “usually required beyond that contained” in
the diagnosis. Use of the manual for assessment by “insufficiently trained individuals is not
advised” it states, and it notes that “a diagnosis does not carry any necessary implications
regarding … the individual’s degree of control over behaviors that may be associated with the
disorder.” 473 Given, however, that similar cautionary language has been disregarded with some
regularity in SVP proceedings (as discussed above), such cautions are likely to have little effect
on the widespread use of psychiatric diagnoses in court settings, even to resolve factual questions
regarding volitional impairment associated with mental abnormality.

Using Actuarial Tools as a Check on or to Replace Clinical Judgment

Some legal scholars and some in the mental health profession have advocated use of
ARA instruments either in addition to 474 or in place of diagnostic assessment and clinical
judgment, 475 as noted above. 476 The appeal of such tools is obvious: they would permit testifying
experts to offer more accurate predictions while avoiding the unsettled realm of psychiatric
diagnoses. However, in addition to problems with reliability of such tools (discussed in the
immediately prior section), there are fundamental conceptual and moral problems as well. The
most significant problem with the use of ARAs in SVP proceedings is that these tools are
designed only to assess the risk of recidivism, not, as required by the Hendricks-Crane standard,
the existence of volitional impairment. Nor are ARAs designed to assess the presence of “mental
disorder,” another core requirement of the SVP statutes and a component of their constitutional
floor. 477 Moreover, because these instruments largely use information that can be gleaned simply
from a review of a respondent’s records alone, without an interview, the forensic examiners
employing them, like those who misuse paraphilia diagnoses as discussed above, are
constructing a condition of underlying volitional impairment based solely upon a selective record
of past actions.


Id. (emphasis added).
Some researchers have proposed used of “Guided Clinical Risk Assessments,” which use a number of factors that
associated with recidivism but are not necessarily static, such as low self-esteem and “general psychological distress.
However, studies have not demonstrated these to be sufficiently reliable for forensic use. Campbell, supra note 412,
at 120.
See, e.g., Robin Wilson, et al., Pedophilia: An Evaluation of Diagnostic and Risk Prediction Methods, 23
SEXUAL ABUSE: A JOURNAL OF RESEARCH AND TREATMENT, 260, 271 (advocating exclusive use of ARAs in SVP
See supra notes 386– 400 and accompanying text.
First & Halon, supra note 216, at 450–51.


There are studies showing that prediction rates based on actuarial tools are better than
those based on clinical judgment, and such studies are invoked by various social scientists and
others who advocate replacing clinical judgment with such tools,in order to ensure more accurate
assessments. 478 There is also, however, a general wariness about using statistics to predict
individual human behavior and, as noted by many social scientists, a resistance to doing so. Thus
Daniel Kahneman observes: “The debate about the virtues of clinical and statistical prediction
has always had a moral dimension… The aversion to algorithms making decisions that affect
humans is rooted in the strong preference that many people have for the natural over the
synthetic or artificial.” 479 Significantly, this aversion appears to be even stronger when the
“decisions are consequential.” 480
Although these emotional responses to the general use of actuarial tools to make
predictions about human outcomes strike many researchers as irrational, the “moral dimension”
of such reactions bears special consideration in the context of a legal proceeding such as SVP
commitment. In Phillip K. Dick’s short story, The Minority Report, the specter of using
“science” to determine what we will do in the future and then detaining individual people as a
result of such “precrime predictions” was evoked to paint a frightening dystopian picture. 481
Using statistically gathered numbers to assess the likelihood of individual human behavior,
especially as the sole basis for an indefinite commitment, is patently inconsistent with a justice
system that emphasizes individualized treatment rather than determinations based on groupbased behavior, such as “guilt by association.” 482 Indeed, such “moral dimensions” have a
central place in our legal system, and the fact that there is such discomfort at using actuarial
methods to determine whether to remove someone from society indefinitely is indicative that
such methods are out of place in SVP proceedings.
The sharpness of the debates regarding the use of psychiatric diagnostic assessments
and/or ARA instruments in SVP proceedings, with strong but conflicting evidence on both sides,
encourages a significant third perspective; that is, the view that the entire SVP commitment
model, with the essential role it assigns to forensic assessment of the likelihood of recidivism, is
inherently unworkable. Since findings of mental abnormality and dangerousness are
constitutionally required in such proceedings, the question of whether we can reliably assess the
relevant pathology and risk has direct implications for the committed persons’ liberty
interests. 483 What these debates reveal is that neither approach—clinical judgment or actuarial

See, e.g., MEEHL, supra note 397.
KAHNEMAN, supra note 379, at 228.
Janus & Meehl, supra note 297, at 60–61. Cf. Reno v. Flores, 507 U.S. 292, 345 (1993) (noting that “the Due
Process Clause establishes a powerful presumption against unnecessary official detention that is not based on an
individualized evaluation of its justification”) (Stevens, J., dissenting). David Faigman recently examined the
difficulty of offering expert opinion regarding an individual based upon research findings about a group: “In terms
of scientific inference, reasoning from the group to an individual case presents considerable challenges and, simply
put, is not a regular part of the basic scientific enterprise. In the courtroom, it is the enterprise.” David L. Faigman,
et al., Group to Individual (G2i) Inference in Scientific Expert Testimony (Oct 2013) forthcoming 81 U. CHI. L. REV. ___ (2014) (emphasis added) [need final pin
Prentky et al., supra note 7, at 371; Janus & Prentky, supra note 396, at 1458. This is not to suggest that clinical
judgment and ARAs are the only methods proposed for predicting risk of sexual violence. For example legal


instruments—is sufficiently reliable to ensure that the SVP laws are not sweeping too broadly.
The making of predictions itself, not the methodology used to make them, is the problem.
Given that all the proposed fixes to the invocation of psychiatric science in SVP
proceedings fall short of addressing the fundamental problems seen in the case law, the question
for legal scholars and analysts becomes whether the courtroom can be fixed to fit the existing
science instead. 484 While some degree of judicial leniency regarding the admissibility of expert
testimony by mental health professionals is arguably appropriate for many kinds of cases,
especially when a personal injury plaintiff or a criminal defendant raises the issue of mental
injury or disorder, there are compelling reasons to apply far more scrutiny to such evidence in
SVP cases. One is certainly the high-stakes outcomes of such cases. No less significant is the
central role assigned by the laws themselves to mental health professionals in preserving due
process under the Hendricks-Crane rationale.
A few rulings by courts suggest that a more assertive role by trial judges as gatekeepers
could prevent due process violations in individual cases, and several legal scholars have made
recommendations along these lines. 485 It remains true, however, that courts overwhelmingly
admit the suspect science and leave it to the fact finder to decide how much weight to give such
expert opinion. Most courts, like the McGee trial court, leave the issues of the validity of the
methods used to arrive at expert opinions—including use of the diagnostic labels and ARAs—
entirely to the assessment of the fact finder. Lower courts’ implementation of Hendricks-Crane
has made clear that they are uninterested playing a more active role in screening out such expert
testimony. If the current legal framework is retained, we should expect that the same tendencies
will continue to prevail.
Indeed, much would be at stake were trials courts to assume the role of aggressive
gatekeeper in SVP proceedings. The irreconcilable conflict between the known limits of the
science of psychiatry and the statutory requirements of the SVP laws could result in the
exclusion of a significant amount of evidence offered in support of commitment and thereby
reveal the inherent unworkability of the SVP commitment model. In other words, serious judicial
gatekeeping in the SVP context would effectively nullify the laws. Trial courts are generally
reluctant to undermine the objectives of elected legislators, especially when such policies have
broad public support and, as here, have been upheld by the Supreme Court. Accordingly, it is

scholar Adam Lamparello has advocated use of neuroscience to predict violent behavior. Adam Lampaello, Using
Cognitive Neuroscience To Predict Future Dangerousness, 42 COLUM. HUM. RTS. L. REV. 481 (2011). However, at
this time, there have been no studies of the use evaluating brain activity through functional MRI imaging to predict
such violence. Moreover, it is by no means clear that such technology will correct any of problems inherent in the
SVP commitment model discussed herein. Steven Erickson, The Limits of Neurolaw, 11 HOUS. J. HEALTH L. &
POL'Y 303 (2012); Daniel S. Goldberg, Against Reductionism in Law and Neuroscience, 11 HOUS. J. HEALTH L. &
POL'Y 321 (2012).
Samuel Brakel and Allan Frances debated this point in Psychiatrists and Law, Psychiatric Times (Nov. 19, 2010)
( ).
Hamilton, supra note 6, at 52; Prentky et al., supra note 234, at 458. See also Vars, supra note 297, at 895–97
(arguing that due process requires that courts commit individuals only upon a finding that there is at least a 75% risk
that the person will commit an act of sexual violence within the next five years).


unlikely that trial courts could be convinced to widely and consistently reject psychiatric
evidence in SVP commitment proceedings. 486



The SVP commitment laws have no shortage of critics from within law, psychiatry, and
other fields. 487 Many criticize the ways the laws are implemented; others argue that they reflect
failed, flawed, and misplaced policies that score political points or that they are based on myths
about sex offenders and unfounded assumptions about the potential for their treatment and
rehabilitation. 488 Most of these criticisms, however, though well taken in themselves, do not
directly address the constitutionality of the laws. Rather, in light of the Hendricks-Crane rulings,
critics commonly assume that the question of their constitutionality has been settled.
In this Article, my focus has been the validity of the rationale of the opinions that are
thought to have settled that question. As discussed in Part II, that rationale, as delivered in the
Hendricks-Crane holding, is based upon the integrity of using a mental-illness model for the
deprivation of liberty permitted in SVP laws. By extension, the medical, and therefore legal,
legitimacy of the prosecution of these laws depends upon the testimony of mental health
professionals weighing in on the question of respondents’ pathology and volitional control. That
testimony, however, is inherently problematic: unreliable at best and, at worst, hollow.
Since the crucial medical opinions offered in SVP proceedings regarding who is a
“predator” with a “volitional impairment”—as distinct from a “typical recidivist”—are routinely
based on conclusions drawn from reviewing the record of a respondent’s prior acts of sexual
violence, those opinions are, in effect, tautologies. 489 The term “sexual predator” has no
psychiatric meaning; it is used simply to name a group of sexual offenders from whom we want
to protect the public. It is like the term “weed,” which has no botanical meaning but which we
use simply to refer to plants of which we want to rid our gardens. In the absence of a scientific

And of course, absent further action from the Supreme Court, Barefoot v. Estelle remains good law, at least in
theory. The Court was recently presented with a petition for certiorari that could have provided an opportunity to
revisit Barefoot v. Estelle and the standard for admissibility of expert psychiatric evidence on future dangerousness,
but it declined to hear the case. Coble v. State of Texas, 330 S.W.3d 253 (Texas Ct. App. 2010), cert denied. ___
U.S. ___,131 S. Ct. 3030 (2011). Accordingly, the Court appears uninterested in providing courts any further
guidance on the admissibility of such evidence anytime soon.
See, e.g., Cucolo & Perlin, supra note 157, at 5–17; JANUS, supra note 4, at 87–92 (arguing that the laws are
antifeminist because they perpetuate a number of harmful myths about rape and child abuse, such as that such acts
are largely committed by “predators” rather than relatives and acquaintances of the victims); LANCASTER, supra note
10, at 233–34 (tracing the “sex panic” underlying many modern sex offender laws to less overt expressions of
homophobia and racism).
See Simon, supra note 41, at 281. Simon summed up her assessment of SVP laws as follows: “[T]hese legal
policies and mental health practices targeting offenders who commit sex crimes thrive despite the absence of
empirical evidence that sex offenders are distinguishable from other offenders; that sex offenders are any more
mentally disordered (and treatable) and dangerous than other offenders; and that mental health professionals are
competent to make predictions of dangerousness.” Id.
See also La Fond, supra note 25, at 162 (“The primary evidence for all of these elements – mental disorder,
volitional impairment, and dangerousness – is the same; an offender's past history of committing sex crime(s).
Simply put, a sex offender who has committed a qualifying sex crime thereby provides evidence that is legally
sufficient to be committed as a SVP.”).


basis for determining whether or not a person is a “sexual predator,” the task that has been
assigned to prosecution forensic experts in SVP proceedings is to make a normative
determination, and, thereby, results in a delegation to psychiatry inconsistent with core notions of
due process. Accordingly, the question of the constitutionality of such laws is in fact far from
The dangers and implications of attempting to align psychiatry with the problematic
concept of a “sexually violent predator” has been recognized by some judges. In a 2010
concurring opinion in an SVP appeal, Justice Richard Sanders wrote:
[I]f the scientific community does not recognize such a condition [as Paraphilia NOS –
Nonconsent], much less possess any methodology to identify individuals with such a
condition, the statutory test [for SVP commitment] cannot be met. Without a
scientifically recognized condition that compels a person to commit sex offenses, civil
confinement also runs afoul of the constitution … Where a person is deprived of his or
her freedom based upon opinion testimony lacking scientific credibility, reliability, and
accepted methodology, courts must step forward and announce with the courage of a
small child that the Emperor wears no clothes. 490
This is a remarkable acknowledgement—and call to action—regarding the fundamental problem
with these laws. However, the entire opinion, including this concurrence, was later withdrawn
upon a motion for consideration by the State. 491
Courts appear to be stuck in a box of their own creation. As captured in Minority Report,
the ability to predict future crime or violence holds tantalizing appeal for a society. Even if we
lack the technology available in the story, we are inclined to think that many instances of
horrifying criminal violence could have been prevented if someone, especially some scientist,
psychiatrist or other expert, had recognized its likelihood and taken steps to prevent it. As
scientists themselves have repeatedly told us, however, and as courts cannot fail to
acknowledge, 492 our general presumption regarding the ability of scientists, and specifically of
those in the psychiatric profession, to predict future violence far exceeds their actual ability.
Despite these acknowledged limits, however, and the constitutional values at stake when they are
disregarded, courts continue to uphold statutes based upon just such mistaken assumptions. The
SVP schemes are not the only examples of this problem but perhaps the most stark and farreaching ones. The Supreme Court has never identified a constitutionally acceptable error rate
for predictions of future violence, although its pre-Daubert opinion in Barefoot suggested that a
quite high error rate would be acceptable. 493 Such a low standard for acceptability gives courts
and legislators broad freedom to take significant legal actions based upon an assessment of risk
and to use psychiatry as a means to identify such risks. Courts have permitted legislators to

The Washington Supreme Court ordered a hearing on a committed person’s petition for release. State v.
McCuistion, 238 P.3d 1147, 1155 (Wash. 2010) (Sanders, J., concurring), opinion withdrawn upon reconsideration
May 20, 2011, as noted in State v. McCuistion, 275 P.3d 1092, 1097 (Wash. 2012).
State v. McCuistion, 275 P.3d 1092, 1097 (Wash. 2012).
See, e.g., Barefoot v. Estelle, 463 U.S. 880, 899 (1983); United States v. Umana, 707 F. Supp. 2d 621, 634
(W.D.N.C. 2010); United States v. Taveras, 424 F. Supp. 2d 446 (E.D. Tex. 2006). Adam Lamparello, Using
Cognitive Neuroscience to Predict Future Dangerousness, 42 COLUM. HUM. RTS. L. REV. 481, 488–92 (2011).
Jackson, et al., supra note 358, at 126.


effectively delegate a crucial normative question to the field of psychiatry and, in so doing, have
disregarded the field’s own disavowal of its ability to fulfill that role competently and ethically.
These objectionable and harmful patterns of delegation must be changed from within the
law. Nearly forty years ago, the noted circuit court Judge David Bazelon cautioned courts about
delegating “delicate questions of state intervention” to mental health professionals. In comments
that bear particularly on the questions examined in this Article, he explained:
[S]tate intervention involves a serious compromise of individual rights and hence a
difficult balancing of power between the state and the individual, where the stakes are
highest for human and personal rights. Courts have traditionally been the protector of
individual rights against state power, and there is no reason why the particularly difficult
problems in the area of state intervention are any different. We cannot delegate this
responsibility to the medical professions. Those disciplines are, naturally enough,
oriented toward helping people by treating them. Their value system assumes that
disturbed or disturbing individuals need treatment, that medical disciplines can provide it,
and that attempts to resist it are misguided or delusionary. The medical disciplines can no
more judge the legitimacy of state intervention into the lives of disturbed or disturbing
individuals than a prosecutor can judge the guilt of a person he has accused. 494
The Supreme Court, in deciding Kansas v Hendricks, did not heed Judge Bazelon’s caution or
give full consideration to the implications of drawing the line at mental abnormality. In light of
what we have learned from the enforcement of these laws, it is clear that courts must revisit their
The social implications of SVP laws bear some emphasis. By pathologizing and not
merely condemning the rapist and molester, and by relying upon a psychiatric and not merely
moral construction of sexual violence, these laws and their prosecution fuel a stigmatizing view
of mental illness more generally: the view, that is, that being labeled with a psychiatric diagnoses
signals that one may be dangerously “out of control,” and therefore a threat to society. Indeed,
language in Hendricks directly supports this view:
A finding of dangerousness, standing alone, is ordinarily not a sufficient ground upon
which to justify indefinite involuntary commitment. We have sustained civil commitment
statutes when they have coupled proof of dangerousness with the proof of some
additional factor, such as a ‘mental illness’ or ‘mental abnormality’. These added
statutory requirements serve to limit involuntary civil confinement to those who suffer
from a volitional impairment rendering them dangerous beyond their control. 495
Such reasoning links acts of violence and mental illness in a misleading and damaging way.
Most sexual offenders do not have serious mental disorders, as discussed above. But the Court’s
longstanding pronouncement that illness can serve as a basis for detention encouraged
lawmakers and courts to pathologize sex offenders to permit their removal from society in a

David L. Bazelon, Institutionalization, Deinstitutionalization and the Adversary Process, 75 COLUM. L. REV.
897, 910 (1975).
Kansas v. Hendricks, 521 U.S. 346, 346–47 (1997) (emphasis added).


manner inconsistent with notions of due process. 496 In this respect, SVP laws reflect the dual
problematic trends of criminalizing the mentally ill and pathologizing criminals.
The use of paraphilias, that is, deviant sexual arousal, as the basis for most SVP
commitments is particularly troubling given the controversy regarding whether such conditions
should even be listed as disorders for clinical purposes. Some observers suggest that
commitments made on such basis carry broad legal implications. Jerome Wakefield, for
example, has flagged what he regards as “a dangerous slippery slope implicit in these legal
developments.” 497 He reasoned:
A pluralistic society is based on respect for human difference and acceptance of the
enormous range of normal variation in tastes and desires. If sexual peculiarities that are
labeled disorders and are offensive to others can be the grounds for civil commitment on
the basis of the harm they do to the public, then it is not clear why other peculiarities that
may be labeled disorders and may be out of control of the afflicted individual – such as,
say, depression or anxiety that detracts from the efficiency of others and thus harms them
– need remain constitutionally immune to such provisions in the future. 498
SVP commitment laws carry implications for the field of psychiatry as well. Many within
the psychiatric field, conscious of their limited knowledge of the nature of sexual offenses and
offenders, are exceedingly uncomfortable with the role assigned to them by the laws. The task
given to forensic experts in SVP proceedings can be even more challenging than the typical
dangerousness predication. Not only is the expert being asked to make an assessment of a
person’s long-term risk for sexual violence, but such determination must be made of someone
who has been incarcerated, sometimes for a lengthy period of time, making prediction of his
future behavior in public especially difficult. 499 Psychiatrists also note that danger-prediction as a
predicate to detention strays far from the central role of psychiatry, which is to alleviate mental
suffering and distress. 500 Employing a host-parasite metaphor, psychiatrist James L. Knoll warns
that SVP laws put psychiatry at risk of becoming “co-opted by a political agenda.” The
prosecution of such an agenda through these laws, Knoll observes, would jeopardize the
“autonomous functioning, and thus the reliability, of the science,” and transform psychiatry into
“a new organism entirely—one that serves the ends of the criminal justice system.” 501
The constitutional infirmities of the SVP laws revealed in this Article serve as compelling
reasons for their legislative repeal. Moreover, as noted earlier and certainly of significance to
legislators, the laws are expensive and of questionable safety benefit to the public. States heeded
the advice of the GAP report in the 1970s and repealed the “sexual psychopath” laws. They
should once again take seriously psychiatry’s disavowal of its ability to identify predators. At


Janus, supra note 30, at 15.
Wakefield, supra note 204, at 197.
Prentky, et al., supra note 7, at 358.
Jerome C. Wakefield, False Positives in Psychiatric Diagnosis: Implications for Human Freedom, 31 THEOR.
MED. BIOETH. 5, 9 (2010) (“Treatment of disorder is the essential defining mission of psychiatry.”).
James L. Knoll IV, The Political Diagnosis: Psychiatry in the Service of the Law, PSYCHIATRIC TIMES (May 13,
2010) ( )


this time, however, there is no indication of any jurisdictions moving to repeal or significantly
reform its SVP commitment laws. 502
If state policymakers hesitate to change the SVP laws out of fear of political backlash, a
somewhat “quieter” option for states is to slow the rate of commitment under such laws and
increase the rate of release of those committed previously. This is the route presently being
followed by the State of Wisconsin. The state has committed nearly 500 individuals since the
state enacted its SVP law in 1994. 503 It released only 31 between 1994 and 2009, but released
114 in the 4 years between 2009 and 2013. It took such steps in light of recent research
suggested that recidivism risks for “certain types of individuals” were lower “than previously
thought. 504 Those who were released received treatment and monitoring in the community, and
the legislature enacted new laws to expand the community-monitoring program. 505
States could also consider programs that frequently obviate the need for commitment
altogether, such as sentencing options for sexually violent crimes that leave questions of mental
illness out of the equation. 506 For example, states could adopt supervised release laws, such as
Maine’s, which provides for an extended period of community supervision in lieu of probation as
part of a sentence for a sex offense. 507 Although Maine’s law was aimed at preventing recidivism
among sex offenders specifically, its use depends not upon a determination of a mental disorder
but upon whether the defendant is a “repeat sex offender” as defined under the law 508 and the
application of a series of other factors. 509 Currently, few courts evaluating SVP petitions
consider whether there are existing alternatives that may minimize a risk of recidivism. 510 If
more such programs were in place, it could provide an argument against commitment in
individual cases.
Regardless, however, of whether the states decide to follow such alternatives to SVP
commitment proceedings, there is a central role and responsibility for the Supreme Court with
respect to these laws. Given the demonstrably dubious basis of the Hendricks-Crane rationale in
light how that reasoning has played out in actual SVP commitments and the multiple,
exceedingly serious implications of leaving the holding in place, the Court must revisit the
constitutionality of the SVP laws.


Cucolo & Perlin, supra note 157, at 9–10.
Nora Hertel, Wisconsin Freeing More Sex Offenders From Mental Lockup, WISCONSIN WATCH (Feb. 2, 2014),
See, e.g., John Fabian, Kansas v. Hendricks, Crane And Beyond: “Mental Abnormality,” and “Sexual
Dangerousness”: Volitional vs. Emotional Abnormality and The Debate Between Community Safety And Civil
Liberties, 29 WM. MITCHELL L. REV. 1367, 1418–20 (2003).
17-A M.R.S. §§ 1231-1233. State v. Cook, 2011 ME 94 ¶ 24, 26 A.3d 834, 843–44. The sentencing court may
impose any of a number of conditions of such release, including limiting contact with the victim and other children,
undergoing evaluation and treatment, and community monitoring.
17-A M.R.S. §§ 1231(2)(A), 1252(4-B)(A).
State v. Cook, 2011 ME 94 at ¶¶ 27-29, 26 A.3d at 844–45.
One of the few courts to engage in this analysis is the district court of Massachusetts in U.S. v. Wilkinson, 646
F.Supp.2d 194 (2009), which considered the fact that the respondent was facing charges for a probation violation in
state court as well as supervised release through the federal probation office. Id. at 208.


While the Court is appropriately loathe to overrule itself, it can follow the example it set
when it overruled Bowers v Hardwick 511 with Lawrence v Texas. 512 The justices noted in
Lawrence that striking down the Texas sodomy law at issue in that case would place it squarely
in conflict with the precedent it had set seventeen years earlier in Bowers, when it upheld
Georgia’s law and that “[t]he doctrine of stare decisis is essential to the respect accorded to the
judgments of the Court and to the stability of the law.” 513 As they concluded, however, such
doctrine “is not an inexorable command; rather, it ‘is a principle of policy and not a mechanical
formula of adherence to the latest decision.’” 514
Significantly here, in applying these judicial principles to the constitutionality of sodomy
laws, the Court noted that, in the time since the Bowers opinion had been issued, there had been
several scholarly “criticisms of the historical premises relied upon by the majority and
concurring opinions in Bowers.” 515 Upon reexamination of those premises, the Court found that
the earlier opinion had been based upon erroneous or at least overstated historical grounds 516 and
that “the rationale of Bowers does not withstand careful analysis.” 517 Here, a comparable
analysis mandates that the Court acknowledge that its earlier opinions on SVP were based upon
erroneous medical grounds and that its core rationale “does not withstand careful analysis.”


The responsibility to make rationally informed policy rests, of course, with lawmakers. In
many ways, it is hard to fault the drafters and supporters of the first SVP laws, particularly those
acting in the immediate wake of almost inconceivably horrifying crimes such as Earl Shriner’s.
But once a policy is enacted, even if it was based largely upon immediate public outrage, fear,
and avoidance of risk, it is nearly impossible to undo. The fear and sense of high risk, even if
later understood by lawmakers themselves to be exaggerated, may still be potent among many
segments of the public—often, as in the case of the “sexually violent predator,” stoked by myths
and exploitative media representations, and reinforced by the existence of the laws themselves.
In light of this political reality, the courts have a significant role to play in the evaluation of the
basis for laws enacted in response to specific outrage-evoking events.
The Earl Shriner case had particular characteristics that shaped the SVP laws. Given
Shriner’s prior involvement with the criminal justice system and the unsuccessful attempt to use
the standard involuntary commitment procedures to keep him away from potential victims, the
public and the policymakers who served them were persuaded that the state’s laws contained a
gaping omission. Reports of his crimes fed the widespread public perception that child sexual
abuse is rampant and that our criminal justice system is powerless to control it. There was and

478 U. S. 186 (1986).
539 U.S. 558 (2003).
Id. at 577.
Id. at 567–58. The Court also noted that the Bowers opinion had not induced any “individual or societal reliance
on Bowers of the sort that could counsel against overturning its holding once there are compelling reasons to do so.”
Id. at 577.
Id. at 571 (“In summary, the historical grounds relied upon in Bowers are more complex than the majority
opinion and the concurring opinion by Chief Justice Burger indicate. Their historical premises are not without doubt
and, at the very least, are overstated.”).
Id. at 571.


remains a general belief that sex offenders have high rates of recidivism, are mentally ill, cannot
control their impulses, and cannot be successfully treated or supervised in the community. With a
previously convicted offender like Shriner, there seemed to be clear warning signs right there.
Viewed retrospectively after his subsequent acts of violence, Shriner appeared to many observers
to be clearly a sexual criminal who was all but certain to re-offend after his release. It also
seemed that the state should be given a mechanism to act on such signs to prevent the
reoccurrence of such crimes by other convicted offenders, specifically a law that would “lock
them away” if such signs were identified by experts as indicating that offenders posed a distinct
risk of victimizing children and others.
Clarity of hindsight, however, is often taken for intrinsic predictability, and our general
intuitions about risk—even the instructed intuitions of experts—are often grossly inaccurate.
Here, in the public and legislative reactions to the Earl Shriner case, the mistakes were many and
mutually reinforcing. The first mistake was to generalize improperly from the particular
circumstances of Shriner’s acts. While Shriner’s crime against a random victim led to the
construction in the anxious public mind of the sex offender as a kind of “bogeyman,” always
lying in wait, always ready to strike whatever innocent children were near, research has shown
that sexual violence is generally highly circumstantial and contingent, that it occurs under a
range of contextual and individual conditions, and that it most often involves victims who have
prior family, social, or institutional relationships to the perpetrator.
The second key mistake was the assumption by the public and legislators that mental
health experts could identify such people and prevent sexual violence through a process of legal
commitment. As demonstrated in this Article, however, psychiatry lacks the knowledge and the
instruments either to identify who is most likely to commit future acts of sexual violence or to
predict the likelihood of violence by a specific individual. The implementation of the SVP laws
has been likened by two forensic psychiatrists to the Salem Witch trials of the 17th century. 518In
an essay making the comparison, they go on to argue that the suggestion that clinicians can
identify the true predators among us creates a dangerous and false sense of security for the
public. Commitment of large numbers of sexual offenders under SVP laws does not enhance
public safety. The laws reflect the public’s fears and groundless beliefs, 519 not the realities of
either sexual violence or the capacities of mental health experts. SVP laws are dangerous,
damaging, and unconstitutional, and the experiment must be shut down.


Good & Burstein, supra note 46, at 24.
LANCASTER, supra note 10, at 78. Similarly, the common perception of a sex offender or predator is one who
lurks around schools, playgrounds, and candy stores waiting to lure trusting children into their cars or residences.
Such stereotypes lead to community notification laws, sex offender registries, and restrictions on offenders’
residence. In fact, the overwhelming number of cases of sexual abuse are committed by family members or “trusted”
adults such as teachers, clergy, and coaches. See, e.g., Cucolo & Perlin, supra note 157, at 5–17; LANCASTER, supra
note 10, at 78.