Amicus Curiae Brief — Smith v. Doe

Ci JMMIT 4

ia Te ANECU he

August, 2002.

OF THE UNITED STATES

ere

——

JF TRE \LASSA

TEE FOR PUBL IC COUNSEL SERVIC

“yore Th TOFR aN

nr i ae \ ‘ —_, “wr L

— f -—- }

- a/s mo oa SU:iE 4

od eC

Petit ners

re ‘

+

J swe

Respondents

- > - \ rT.

=—— ~ ~~

“ =

~ 7 7 rage

- “ ee

-_~

—

nh ae nrens; QT

Dr}? cSrP ~ ~

a ae « ‘s & Ae NAM ees TA

,

.

rapnyr . ry er me 4

Xe ~- — _- - ‘ _- ’ ihe

_ ‘

| bbe >) . Ds xr

= deta. i.e ~s

~ r :

Q\ap - : ~ _ se a°%-e

we le ee ee ~ — ha Diz es

‘ =

COMMITTEE FOR PUBLIC

-COUNSEL SERVICES

44 Bromfield Street

~. Boston, Massachusetts 02108

(617) 482-6212

= 7 on -

QUESTION PRESENTED

Does the Alaska Sex Offender Registration Act violate

the prohibition against ex post facto laws set forth in Article I,

$10 of the United States Constitution”

1 ill

TABLE OF CONTENTS C. Information on longer term recidivism

rates is extremely limited ................. 12

TABLE OF AUTHIORITEES ...cscsceesus seen iv

EE 13

INTEREST OF THE MASSACHUSETTS

COMMITTEE FOR PUBLIC COUNSEL SERVICES ... 1 | i) Colieenia ...................0eee. 14

ARGUMENT (iii) Hanson and Bussiere Meta-Analysis ... 15

l. EMPIRICAL EVIDENCE DOES NOT

SUPPORT THE CLAIM OF ALASKA AND

ITS AMICI CURIAE THAT SEX OFFENDERS

Il. EMPIRICAL EVIDENCE SUGGESTS THAT

DISSEMINATION OF SEX OFFENDER

REGISTRY INFORMATION MAY UNDER-

RELEASED FROM INCARCERATION ARE CS EEE co ccccccccccccoccce 18

AT UNUSUALLY HIGH RISK OF COM-

MITTING A NEW SEX OFFENSE ............ 2 en ood cee ee ececes 19

A. A imat S% (ii) Washington Study ................. 20

rearrested for a new rape during the

three years following their release Tee 21

(iv) U.S. Department of Justice Study

of Notification in Wisconsin ......... 21

Report cocesseces Sccecedosseeeeseer 4

EEE EE 24

Renee .nccsncssessusneveeneeee 6

Pe en 8

~

S =

tN _

= S

N ~

w E

ae

°o ° .

= = ;

re wae

o ao °

BoB:

a. a:

oO re) °

7 wv :

Ww

rr mmm

iV

TABLE OF AUTHORITIES

Cases

Doe v. Attorney General,

SR, DGG: cucectdiceeesckbaveane bonus 17

Doe v. Otte,

259 F.3d 979 (9th Cir. 2001),

cert. granted, 122 S. Ct. 1062 (2002) ........... 2,3, 18

Kennedy v. Mendoza-Martinez,

te: Pn ci neuenaneé oeeteeueeeea 3, 18

McKune v. Lile.

Ch Pee ..cancncetecnceuess 14

Constitutional Authority

United States Constitution,

REINS, GED cc nccccccccccccsccccsseccccensees: i

Statutes

Alaska Admin. Code tit.13, §09.050(2000) ............. 2

ee eB cc occuccccoucsecencetesse 2

1994 Alaska Sess. Laws 41 §1 .............50005- 17, 18

1994 Alaska Sess. Laws 41, §12(a) ..... 2... 6... e eee 2

Alaska Stat. §12.63.010(5)(1) ... 2.2... eee eee ee ee eee 2

Vv

Alaska Stat. §§12.63.020, 12.63.100 .................. 2

Mass. G.L. c.6, §178C-P

er eee l

TD. vcecesdatdcekscanuckoucet l

Other Authorities

Bureau of Justice Statistics,

Bureau of Justice Statistics,

National Conf. On Sex Offender

Bureau of Justice Statistics,

Recidivi Pri Rel

a 6, 7, 8,9

Bureau of Justice Statistics,

Sex Offenses and Offenders (February 1997) .... 4,5, 8,

Bureau of Justice Statistics,

Sexual Assault of Young Children as

R 1 to Law Enfi - Victi

i istics (July 2002) .... 10

v1

B. Schwartz and H. Cellini, Sex Offender

Recidivism and Risk Factors in the Involuntary

Commitment Process, in Schwartz, ed., The Sex

Offender: Theoretical Advances, Treating Special

Populations and Legal Developments III (1999) ... 16, 17

Center for Sex Offender Management,

Recidivism of Sex Offenders 7 (May 2001) ......... 16

C. Wells and E. Motley, Reinforcing the

Myth of the Crazed Rapist: A Feminist

Critique of Recent Rape Legislation,

Bp es Es SE OE awh eeseccéivescoeeve 23

D. Schram & C. Milloy, Community Notification:

s Study of Offender Cl ———

Recidivism (October 1995) ..........0-02.00000:. 20

G. Adkins, D. Huff, & P. Stageberg,

J. Comparet-Cassani, A Primer on the Civil

Trial of a Sexually Violent Predator,

37 San Diego Law Review 1057 (2000) ............. 8

N.J. Dept. of Correction, Report - 1990

ee

Vii

R.K. Hanson and M. Bussiere, Predicting Relapse:

A Meta-Analysis of Sexual Offender Recidivism

Studies, 66 Journal of Consulting and Clinical

Papemeeesy SOB (ISGS) 2... ccc ce cccccccvccces

R. K. Hanson et at., Long-Term Recidivism

of Child Molesters, 61 J. of Consulting and

clinical Psychology 646 (1993) ................

l

INTEREST OF THE MASSACHUSETTS COMMITTEE FOR

PUBLIC COUNSEL SERVICES.

The Massachusetts Committee for Public Counsel

Services (CPCS) is the Massachusetts agency charged with

representing indigenf parties in a variety of criminal and civil

matters. CPCS is statutorily mandated to provide counsel to

indigent former sex offenders who are subject to the

registration and notification provisions of the Massachusetts

Sex Offender Registration and Notification Act. Mass.G.L. c.6,

§$178C-P (2002 edition). The instant case, which raises a

question concerning the constitutionality of the Alaska sex

offender registration and notification act, is likely to have

important ramifications for the rights and interests of

Massachusetts former sex offenders subject to the

Massachusetts act.

CPCS also represents indigent defendants charged with

sex offenses. Sex offender registration and notification has

dramatically changed the consequences of a conviction for a

sex offense. This Court's opinion in the instant case is likely to

have considerable impact upon the future handling and

disposition of sex offense charges.

CPCS submits this amicus brief to challenge the validity

of certain empirical claims by petitioners and its amici United

States and Attorneys General of Forty-One States, including

Pursuant to Supreme Court Rule 37.6, counsel for the

Massachusetts Committee for Public Counsel Services (CPCS) states that

no part of this brief was written by counsel for any party. No person,

agency, or organization other than CPCS has made a monetary contribution

to the preparation and filing of this brief. All parties have consented to the

appearance of CPCS as amicus curiae in this matter. Letters of consent

have been filed with the Clerk.

2

claims regarding the rate of recidivism of sex offenders and the

usefulness of public notification in preventing sex offender

recidivism.

ARGUMENT

I. EMPIRICAL EVIDENCE DOES NOT

SUPPORT THE CLAIM OF ALASKA AND

ITS AMICI CURIAE THAT SEX OFFENDERS

RELEASED FROM INCARCERATION ARE

AT UNUSUALLY HIGH RISK OF

COMMITTING A NEW SEX OFFENSE.

The Alaska sex offender registration act, 1994 Alaska

Sess. Laws 41 (hereinafter "the act"), provides that every sex

offender, regardless of whether he or she has been rehabilitated,

must register with law enforcement authorities either for a

period of fifteen years from the date of unconditional discharge

or, in the case of two or more sex offenses, for his or her

lifetime. Alaska Stat. §§12.63.020, 12.63.100. Doe v. Otte, 259

F.3d 979, 982-983 (9th Cir. 2001), cert. granted, 122 S. Ct.

1062 (2002). The act applies retroactively to persons, such as

John Doe I and John Doe II, convicted before its enactment.

1994 Alaska Sess. Laws 41, §12(a).

Regulations implementing the act require that

registration information, which includes, inter alia, offenders’

names, home and work addresses, photographs, and description

of convictions, be made available on the internet at a state-

maintained web site. Alaska Stat. §12.63.010(b)(1); Alaska

Admin. Code tit.13, §09.050 (2000). As a consequence, not

only persons in the Does' home towns may obtain access to

their registry information, but also persons elsewhere in Alaska,

3

in New York, Paris, Tokyo or in any location where there is

internet access.

The Court of Appeals held that the Alaska act, as

applied to the Does, violates the ex post facto clause of the

United States Constitution. Doe v. Otte, supra, 259 F.3d at 995.

Reviewing the act through the lens of the seven factors set forth

in Kennedy v. Mendoza-Martinez, 372 U.S. 144 (1963), the

Court found that the act was excessive in relation to its purpose

of protecting public safety. Doe v. Otte, supra, 259 F.3d at 991-

993. Broad, world-wide dissemination of registry information

for all offenders, including those who have been rehabilitated,

the Court concluded, is "unquestionably punitive." Id. at 993.

Alaska and its amici Attorneys General? and Solicitor

General respond that world-wide dissemination of registry

information, even for rehabilitated offenders, is an appropriate

response to what they characterize as the dramatically high rate

at which convicted sex offenders commit new sex offenses. A

closer inspection of the sources they cite for this proposition

regarding sex offender recidivism, however, reveals that the

"once a sex offender always a sex offender" mantra is in fact a

myth.

A. Approximately 2.5% of rapists are

rearrested for a new rape during the

tendon Genie on

from prison.

The Bureau of Justice Statistics of the United States

Department of Justice (hereinafter "BJS") has issued two

. Amici curiae Attorneys General of Forty-One States are referred

to throughout this brief as "Attorneys General." Amicus curiae United

States is referred to "Solicitor General."

4

reports within the last decade addressing recidivism by rapists.

The most recent of these reports, issued in 2002, indicates that

the rate of rearrest of rapists for a new rape in the three years

following release from prison is 2.5%.

(i) 1997 Bureau of Justice Statistics Report

In 1997, the BJS issued a report on rape and sexual

assault, which included a review of recidivism statistics for

persons convicted of rape.’ BJS, Sex Offenses and Offenders

(February 1997) (hereinafter "Sex Offenses” or "1997 BJS

study").* The study followed 2,214 convicted rapists from 11

States for a period of 3 years after their release from prison in

1983, during which time they were on probation. Sex Offenses

25-26. The BJS recorded the following results with regard to

rearrest (not reconviction) of these offenders:

1. 51.5% were rearrested and charged

with a new crime

2. 27.5% were rearrested and charged

with a new violent crime

3. 7.7% were rearrested and charged

with rape

. "Rape" and "sexual assault" are broadly defined, and include most

offenses requiring registration under the Alaska statute. See BJS, Sex

Offenses and Offenders (Sex Offenses) 1 n.1 (1997). The report included

statistics for recidivism by rapists, but no statistics for recidivism by

persons convicted of sexual assault.

. All BJS studies may be found at http://www.ojp.usdoj.gov/bjs.

5

4. 10.7% were rearrested and charged

with nonsexual assault

~n

8.5% were rearrested and charged with

robbery

6. 2.8% were rearrested and charged with

murder

Sex Offenses 26, figure 27.

Fewer than | of 10 rapists was rearrested and charged with a

new rape.

The rate of rearrest of rapists for a new felony is low,

relative to the rearrest rate of other violent offenders. The 1997

BJS study reported that:

...rapists had a lower rate of re-arrest for a new

felony and a lower rate of re-arrest for a violent

felony than most categories of probationers with

convictions for violence. While about 41% of

violent probationers were re-arrested within 3

years of placement on probation for a new

felony offense, an estimated 19.5% of rapists

were re-arrested for a new felony within 3 years.

Sex Offenses 25-26.

Rapists were rearrested and charged with a new felony

less than half as frequently as other violent offenders.

6

(ii) 2002 Bureau of Justice Statistics Report

A larger, and much more current,’ study was conducted

by the BJS in 2002. BJS, Recidivism of Prisoners Released in

1994 (June 2002) (hereinafter "Recidivism" or "2002 BJS

study"). This study tracked recidivism rates for 272,111 former

inmates for a period of 3 years following their release from

prison in 1994. The persons included in the pool represented

two-thirds of all persons released from prison in 15 States, and

included 3,138 rapists. Recidivism 1, 8-9.

The 2002 BJS study found that rapists were rearrested

for a new rape at an even lower rate than was reported in the

1997 BJS study. During the 3 years following release from

prison in 1994:

1. 46% of rapists were rearrested and

charged with a new offense

2. 18.6% were rearrested and charged

with a new violent offense

3. 2.5% were rearrested and charged

with rape

4. 11.2% were rearrested and charged

with a drug offense

5. 8.7% were rearrested and charged with

a nonsexual assault

’ The 2002 study collected statistics on inmates released 11 years

after the inmates for whom statistics were collected in the 1997 study.

7

6. 14.8% were rearrested and charged

with a property offense

Recidivism 9, table 10.

See also ibid. ("of the 3,138 released rapists, 78 were arrested

for rape and the remaining 3,060 were either rearrested for

something else or not rearrested").°

The Solicitor General and the Attorneys General are

incorrect in asserting in their amicus briefs that sex offenders

"are the most likely to reoffend,"’ "have a dramatically higher

recidivism rate for their crimes than any other type of violent

felon," and "have an abnormally high rate of recidivism." Brief

of the Solicitor General (hereinafter Br.SG) 2. The claim that

° The 2002 BJS study presents data on the rates of rearrest of violent

felons, in addition to rapists. Recidivism 9, table 10. Unlike the 1997 BJS

study, however, it does not include a direct comparison of the rearrest rate

for violent felonies of rapists with the rearrest rate for violent felonies of

other violent felons. The rate of rearrest of all violent felons (including

rapists) for a new violent felony is reported as 27.5%, as compared to a rate

of 18.6% for rapists. Recidivism 9, table 10.

. In its amicus brief, the Solicitor General claims that sex offenders

are "'the least likely to be curec’’; '[]the most likely to reoffend’; and ‘[]prey

on the most innocent members of our society." Br.SG 2. These sweeping

conclusions are quotes from a political speech delivered by Texas State

Senator Florence Shapiro at a conference on sex offender registries. See

BJS, National Conf, On Sex Offender Registries 93 (April 1998). State

Senator Shapiro cites no sources in support of these claims.

8

"{rjapists have a recidivism rate of 49.4% to 63.8%,"* Brief of

Attorneys General (hereinafter Br.AG) 10, is strikingly at odds

with the statistics set forth in the 1997 and 2002 BJS reports.

(iii) The "Odds" that a Rapist Will Reoffend

Although amici Solicitor General and Attorneys General

both cite to the 1997 and 2002 BJS reports on recidivism by sex

offenders and by criminal offenders, neither acknowledges the

very low rate of recidivism for rapists reported in those studies.

Instead, they make much ado over the difference between the

odds that a former sex offender will commit a new rape and the

odds that a former felon, not convicted of rape, will be arrested

for rape. Br.AG 10.

An individual rapist released from prison is more likely

than an individual nonrapist released from prison to commit a

rape during the three years following his release. According to

the 2002 BJS report, a rapist's odds are 4.2 times a nonrapist's

odds of being rearrested for rape. Recidivism 10. The 1997

BJS study reports that the odds differ by a factor of 10.5. Sex

Offenses 27.

. The amicus brief of the Attorneys General cites as authority for

this statistic J. Comparet-Cassani, A Primer on the Civil Trial of a Sexually

Violent Predator, 37 San Diego Law Review 1057, 1072 n.80 (2000). The

Comparet-Cassani article reports only general recidivism rates of sex

offenders, i.e., rearrest rates of sex offenders for any new crime, and is not

limited to rearrest rates for new sex offenses. The California study upon

which Comparet-Cassani relies for her statistical claims in fact reports that

over a 15 year follow-up period, "19.7% of [sex offenders] went on to

oe with a worn gd sex ape offense.” R oe

Offenders 7 (1988). “This California study i is dovened infra at 14-15.

9

This, however, does not mean that former rapists are

highly likely to commit new rapes. Fewer than 3 of 100 former

rapists were rearrested for rape in the three years following

release from prison. Recidivism 9, table 10. The 2002 BJS

study indicates that a smaller percentage of nonrapists than of

rapists were rearrested and charged with rape during the three

year period. It also indicates, however, that the absolute

number of nonrapists arresved for rape was substantially larger

than the absolute number of rapists rearrested for rape.

Recidivism 9, table 10.

1. rapists arrested and charged with rape:

2.5% (number - 78)

2. drug offenders arrested and charged

with rape: .3% (number - 266)

3. burglars arrested for rape: .8%

(number - 330)

4. offenders against public order arrested

for rape: .4% (number - 104).

Ibid.”

Of the 778 felons arrested for rape after release from

incarceration, only 78 had a prior rape conviction.

’ These numbers were derived by multiplying the rate of arrests for

rapes by the number of inmates released for each category of prior offense.

For example, 41,257 burglars (Recidivism 9, table 10) were released from

prison at the outset of the study period. The percentage arrested for rape

during the following three years was .8%. ibid. When multiplied, these two

numbers indicate that 330 former burglars were arrested for rape during the

three year period.

10

B. Most offenders against children are

family members rather than prior

frend ae .

Meany victims of rape and sexual assault are persons

under 18 years of age. Sex Offenses iii ("44% of rape victims

were under the age of 18"). Their stories of abuse are often

horrific. Worldwide dissemination of the registry information

of former sex offenders cannot, however, reasonably be judged

an effective deterrent to future sex crimes against children.

In 2000, the BJS collected and analyzed data concerning

the sexual assault of persons under 18 years of age (hereinafter

“children"). — Sexnel_Anmmlt_of- Young {Sean 20

Characteristics "Guly, 2000) (hereinafter Sexual Assault of

or "2000 BJS study"). The study reviewed the

case files of 118,753 children who were victims of rape or

sexual assault between 1991 and 1996." Id. 1. Data was

gathered from 12 states. Ibid. The following statistics were

reported:

1. for victims under the age of 6, 49% of

the perpetrators were family members

2. for victims between the ages of 6 and

11, 42% of the perpetrators were

family members

3. for victims under the age of 6, 3% of

the perpetrators were strangers

= “Rape” and "sexual assault" are broadly defined, but exclude

statutory rape. Sexual Assault of Young Children 13.

11

4. for victims between the ages of 6 and

11, 5% of the perpetrators were

strangers

Sexual Assault of Young Children 10.

The 1997 BJS study likewise reported that in 90% of the

rapes of children younger than 12, the child knew the offender.

Sex Offenses iii. In 43% of these cases, the offender was a

family member. Id. 11.

No recidivism statistics were available for the BJS

studies, and none were included in any of the three reports. The

absence of statistics on rearrest of persons convicted of sex

offenses against children is a major impediment to determining

their rate of recidivism.

The Solicitor General claims that those who commit sex

offenses against children present an even higher risk of

recidivism than sex offenders generally. Br.SG 4. The

authorities cited do not support this proposition. BJS, Child

Victimizers: Violent Offenders and Their Victims 9-10 (1996)

(hereinafter Child Victimizers), for example, does not report

recidivism statistics for child molesters, but rather addresses

characteristics of violent offenders against children generally.

It notes that "[o}ffenders with multiple child-victims were most

likely to be serving time for robbery, aggravated assault,

negligent manslaughter, or murder." Id. 9.

R. Karl Hanson et al., Long-Term Recidivism of Child

Molesters, 61 J. of Consulting and Clinical Psychology 646

(1993), cited at Br.SG 4, n.3, studied a group of 197 Canadian

child molesters released between 1958 and 1974. Id. 648.

Hanson did not report statistics on the recidivism rate of child

12

molesters for new instances of child molestation, but only for

violent offenses. Id. at 648. Even so, Hanson found that:

1. the rate of reconviction of child

molesters for a new sexual or violent

offense during the first five years of

the study was 5.2%

2. the rate of reconviction of child

molesters for a new sexual or violent

offense during each following year

was approximatly 1.8% per year

Id. at 648.

This means that ten years after release, approximately 12.9%

of child molesters were convicted for either a new sexual or

violent offense.

Statistics aside, it is highly unlikely that internet

disclosure of registry information will be of particular use in

protecting many of these children, given the large percentage of

offenders who are family members. Consulting a registry of

convicted offenders will at best provide a false sense of security

based on an uninformed assumption that the most likely child

molesters have been included on the list.

C. ; sales

Information on longer term reeidis sm ty limited

Amicus Solicitor General urges that long term

recidivism rates must be considered when assessing recidivism

by sex offenders because sex offenders continue to reoffend

13

many years after release from prison. Br. SG4 n.3. This is

easier said than done.

The BJS appears to have published no long-term study

of recidivism by sex offenders. Two states have completed

studies that track sex offense recidivism rates in their

jurisdictions for more than three years after release from prison.

These studies, although limited by the small number of subjects

studied, are nevertheless of some interest. In addition, two

Canadian researchers have published a meta-analysis'' which,

although not a long term study, is worthy of notice because of

the very large number of offenders included.

(i) New Jersey

In 2002, the New Jersey Department of Corrections

issued a study, which tracked 79 offenders released from the

Adult Diagnostic and Treatment Center in 1990. N.J. Dept. of

Correction, Report - 1990 Recidivism Study” (January 21,

2002)(hereinafter "New Jersey Study"). All had been found to

be repetitive and compulsive sex offenders. The offenders were

tracked for a period of ten years. The following results were

reported.

1. 26.5% were reconvicted for nonsexual

offenses

. A meta-analysis is a study which collects and analyzes statistics

reported in other studies.

= This study in included in materials lodged by the New Jersey

Public Defender et al. along with its amicus brief in support of

respondents .

14

2. 3.8% were reconvicted for sexual

offenses

New Jersey Study 3.

One of the three reconvictions for a sexual offense occurred

during the first five years after release, and two more during the

second five year period. Id. 5.

(ii) California

In 1988, the California Department of Justice issued a

report on the effectiveness of its sex offender registration

statute. The report inc'uded recidivism statistics for a group of

1,362 sex offenders released in 1973 and tracked for a period of

15 years. Lewis, R., Effectiveness of Statutory Requirements

for the Registration of Sex Offenders (1988) (hereinafter

"California Study")” 1, 7. The following results were reported.

1. 19.7% of sex offenders were

arrested for a new sex offense“

" This study may be found on the internet at

http://caag.state.ca.us/cjsc.

- The higher rates found in California as compared to New Jersey

may reflect not only the five-year difference in the follow-up period, but

may also reflect the fact that the New Jersey offenders received sex

offender treatment while incarcerated. See McKune v. Lile, 2002 U.S.

LEXIS 4206, *19 (2002) ("[t}herapists and correctional officers widely

agree that clinical rehabilitative programs can enable sex offenders to

manage their impulses and in this way reduce recidivism"). It is also

significant that the California study reports rates of rearrest as opposed to

reconviction, and that the California study was conducted twelve years

before the New Jersey study.

EONS TO IT CN DOL LEE LEI LIO LNA. 5 MO A EN A Se RE

OT OE EE

15

2. although the percent of nonsexual

offenders arrested for a sex offense

was lower than 19.7%, the actual

number of violent offenders, of

property offenders and of each other

category of nonsexual offenders

(excluding drug offenders) arrested for

a sex offense was significantly higher

than the actual number of sex

offenders arrested for a sex offense’

California Study 8, table 2.

(iii) Hanson and Bussiere Meta-Analysis

In 1998, two Canadian researchers published a meta-

analysis of numerous previously published studies, addressing

the recidivism rates of sex offenders. R.K. Hanson and M.

Bussiere, Predicting Relapse: A Meta-Analysis of Sexual

Offender Recidivism Studies, 66 Journal of Consulting and

Clinical Psychology 348 (1998) [hereinafter "the meta-

analysis"]. Hanson and Bussiere gathered information from 61

studies of sex offender recidivism, which cumulatively reported

recidivism statistics for 28,972 sex offenders. Id. at 349-351.

a The actual number of offenders is derived by multiplying the rate

of arrests for sex offenses by the number of persons released for each

category of offense. For example, former property offenders were arrested

for a sex offense at the rate of arrest 2.7%. California Study 8, table 2. The

number of property offenders released in 1973 was 22,417. Ibid. When

multiplied, these two numbers indicate that 605 property offenders were

arrested for sex crimes during the 15 year study.

16

The studies surveyed were conducted in six North

American and European countries, as well as Australia. Thirty

studies were of American sex offenders. Ibid. Adult and

adolescent offenders, and offenders against children and adults,

were included. Ibid. Recidivism was measured by

reconvictions, arrests, self-reports and parole violations for a

new sex Offense. Ibid.

The meta-analysis found that the recidivism rate of sex

offenders for new sex offenses in the 4 to 5 years following

release from incarceration was 13.4%. Although the follow-up

period was not lengthy, the meta-analysis is of particular

interest because of the large number of sex offenders it

encompassed.

EEE

Many researchers on sex offender recidivism have noted

substantial inconsistencies among studies, both as to rates of

recidivism and as to factors that predict recidivism. Center for

Sex Offender Management,'* Recidivism of Sex Offenders 7

(May 2001). B. Schwartz and H. Cellini, Sex Offender

Recidivism and Risk Factors in the Involuntary Commitment

Process, in Schwartz, ed., The Sex Offender: Theoretical

Advances, Treating Special Populations and Legal

Developments III 8-4 (1999). Some of the variation has been

attributed to differences in the quality of the studies. Schwartz

and Cellini 8-4 (recidivism studies differ widely in the quality

of the research design, including criteria for success or failure,

presence of control group, length of follow-up time, and

correctness of statistics used in analyzing time at risk).

= Publications of the Center for Sex Offender Management may be

found at www.csom.org.

17

A second important factor, however, is the pool of

offenders studied.

One possible reason for the amount of

disagreement in these studies is that they may

use a heterogeneous group of sex offenders in

their analysis. Mixing an antisocial rapist with

a socially skilled fixated pedophile with a

developmentally disabled exhibitionist may

indeed produce a hodgepodge of results.

Schwartz and Cellini 8-6.

Alaska has made no attempt to distinguish subgroups of

sex offenders or individual sex offenders who may be at much

lower risk of reoffending from other subgroups or individuals

who pose a higher risk. It has instead opted for the

"hodgepodge" approach, with the result that even

incontrovertibly rehabilitated former offenders are subject to its

harsh provisions.

eEEEE

The Alaska sex offender registration act is grounded on

the legislative finding that "sex offenders pose a high risk of

reoffending after release from custody." 1994 Alaska Sess.

Laws c.41, §1. This finding is not supported by empirical

evidence. The fact that the legislature has declared it to be so

does not transform its "finding" into fact. "Where government

must show an interest of more than the usual degree of urgency

to justify imposition on the individuai, the simple invocation of

its legislative judgment, buttressed perhaps by a showing of a

rational basis for that judgment, will not do." Doe v. Attorney

General, 426 Mass. 136, 150 (1997) (Fried, J., concurring)

(holding that Massachusetts sex offenders have state

18

constitutionally protected liberty and privacy rights in avoiding

registration and dissemination and are entitled to due process

prior to encumbering those rights). Because of the importance

of the interests of sex offenders in avoiding the highly

damaging consequences of public dissemination of registry

information, actual empirical evidence that sex offenders pose

a high” risk of reoffending by committing a new sex offense is

required.

Il. EMPIRICAL EVIDENCE SUGGESTS THAT

DISSEMINATION OF SEX OFFENDER

REGISTRY INFORMATION MAY

UNDERMINE PUBLIC SAFETY.

The Alaska sex offender registration and notification act

was also premised on a finding that "release of certain

information about sex offenders to public agencies and the

general public will assist in protecting the public safety." 1994

Alaska Sess. Laws 41, §1. It is far from obvious that this

premise is valid. In fact, available evidence, while limited,

suggests that dissemination has the opposite effect.

Without evidence that the act serves a legitimate public

safety purpose, and without evidence that it is not excessive in

relation to the need to protect public safety, the legislature’ s

declaration that the act is regulatory and not punitive is suspect.

Doe v. Otte, supra, 259 F.3d at 991, citing Mendoza-Martinez,

supra, 372 U.S. at 169.

a The Alaska act does not define what is meant by a "high" risk of

reoffending.

19

(i) lowa Study

In December, 2000, lowa published a report comparing

the recidivism rates of sex offenders subject to the notification

law to the recidivism rates of sex offenders convicted of

comparable sex crimes, but not subject to the notification law

because their release date was prior to the effective date of the

statute. G. Adkins, D. Huff, & P. Stageberg, The lowa Sex

Offender Registry and Recidivism (December 2000)

(hereinafter lowa Study)."*

The notification group consisted of 233 sex offenders

released from prison and placed on the registry during 1995-96.

Iowa Study 2. The group not subject to notification consisted

of 201 sex offenders released from prison or discharged from

probation during 1995-96, who would have been required to

register if the registry law had been in effect at the time of their

offenses. The study followed these offenders for 4.3 years.

During the 4.3 year follow-up period:

1. 3% of the notification offenders were

reconvicted of sex offenses

2. 3.5% of the group not subject to notification

were reconvicted of sex offenses

Iowa Study 10.

This difference in reconviction rates is not statistically

significant. lowa Study 19. Notification appears to have had

no effect on recidivism by sex offenders.

- This study is available on the internet at

http://www state.ia.us/government/dhr/cjjp/pdfs/SexOffenderReeport.pdf.

20

(ii) Washington Study

In 1995, a study was published assessing the effect of

public notification of sex offender registration on sex offender

pence in the state of Wertuagee. D. Schram & C. Milloy,

tificati tudy of Offender teristi

and Recidivism (October 1995) (hereinafter Washington

Study).’* The recidivism rate of 90 adult offenders subject to

the highest level of notification was compared with the

recidivism rate of 90 offenders released prior to the effective

date of the law.

After 4.5 years,

1. 19% of the group subject to

notification were rearrested for a sex

offense

2. 22% of the group not subject to

notification were rearrested for a sex

offense

Washington Study 3. This difference is not statistically

significant. Ibid. "[T]he findings suggest that community

notification had little effect on recidivism as measured by new

arrests for sex offenses or other types of criminal behavior."

Ibid.

” This study may be found on the internet at

http://www. wsipp.wa.gov/crime/pdf/chrrec.pdf.

21

(iii) California Study

In 1988, California published a study which, in addition

to providing recidivism statistics for sex offenders, surveyed

criminal justice agencies, including law enforcement, probation,

parole and the district attorneys offices, concerning their views

of the effectiveness of offender registration in preventing sex

offender recidivism. R. Lewis, Effectiveness of Statutory

Requirements for the Registration of Sex Offenders 9-10 (1988)

(hereinafter California Study).” Researchers found that

"approximately half of the respondents do not believe that the

sex registration system is effective in preventing sex offenders

from committing sex offenses or other crimes." (emphasis in

original) California Study 10.

(iv) U.S. Department of Justice Study of

Notification in Wisconsin

In December, 2000, the National Institute of Justice, an

agency of the U.S. Department of Justice, issued a study of the

impact of the sex offender notification law in Wisconsin. See

R. Zevitz and M. Farkas, Sex Offender Community

Notification: Assessing the Impact in Wisconsin (December,

2000) (hereinafter Wisconsin Study).”" The study solicited

information from many sources concerning the effectiveness of

the act, including the impressions of sex offenders to whom it

is applicable. Personal interviews were conducted with thirty

"high risk" sex offenders.

= The internet site at which this study may be found as set forth in

footnote 8.

- This study may be found on the internet at http://www.ncjrs.org.

22

All but one interviewee stated that the

community notification process adversely

affected their transition from prison to the

outside world. Loss of employment, exclusion

from residence, and the breakup of personal

relationships were frequently cited

consequences of expanded notification actions

and ensuing detrimental publicity []. Seventy-

seven percent told of being humiliated in their

daily lives, ostracized by neighbors and lifetime

acquaintances, and harassed or threatened by

nearby residents or strangers. Although only

one interviewee was on the receiving end of

what might be described as vigilante action, all

expressed various degrees of concern for their

own safety.

Wisconsin Study 9.

The study notes that most of the thirty offenders

believed that the law would not prevent sex offense recidivism,

but "would have the opposite effect." Wisconsin Study at 10.

"Many drew from their own embittered experience with

community notification to suggest that the tremendous pressure

placed on sex offenders by the public and the media would

drive many of them back to prison." Ibid. See also id. 11 ("The

pressure placed on many of these individuals by community

notification needs to be further examined as a factor in their

success or failure under community supervision").

The negative repercussions for public safety of the daily

humiliation, ostracizing by friends and neighbors, and difficulty

in finding housing and employment predicted by sex offenders

in the Wisconsin study should not be ignored. A 1998 study of

dynamic factors affecting sex offender recidivism found that

23

sex offenders are at the highest risk of reoffending when they

show sharp increases in dysphoric moods, particularly anger."

R.K. Hanson and A. Harris,

Recidivism 1 (1998). Inflicting a high level of stress on a daily

basis through the operation of a public notification law is not a

sensible approach to reducing recidivism by sex offenders.

The declaration by the Alaska Legislature that its sex

offender notification act protects the public safety does not

make it so. A successful approach to sex assault prevention

must be guided by fact and not by fantasy.” Public notification

is not part of the solution to sex offender recidivism.

22

Even the most avid proponents of measures to reduce the incidence

of sexual assault have called for the repeal of sex offender notification laws,

on the ground that they ultimately undermine rather than serve the public

safety concerns that motivated them. C. Wells and E. Motley, Reinforcing

the Myth of the Crazed Rapist: A Feminist Critique of Recent Rape

Legislation, 81 B.U. Law Rev. 127, 130-135, 192-194 (2001).

24

CONCLUSION

For the reasons stated above, and for the reasons set

forth in the brief of petitioners John Doe I, Jane Doe and John

Doe II, and in the briefs of their amici curiae, this Court should

affirm the ruling of the Court of Appeals for the Ninth Circuit.

Respectfully submitted,

CAROL A. DONOVAN

Counsel of Record

Special Litigation Director

COMMITTEE FOR PUBLIC

COUNSEL SERVICES

44 Bromfield Street

Boston, Massachusetts 02108

(617) 482-6212

Dated: August, 2002.