Jaycee Dugard v. United States

FOR PUBLICATION

UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT

JAYCEE DUGARD, individually, No. 13-17596
and as Guardian Ad Litem, for
her Minor Child, D.C. No.
Plaintiff-Appellant, 3:11-cv-04718-CTB

v.
OPINION
UNITED STATES OF AMERICA,
Defendant-Appellee.

Appeal from the United States District Court
for the Northern District of California
Carlos T. Bea, Circuit Judge, Presiding

Argued and Submitted January 12, 2016
Pasadena, California

Filed August 26, 2016

Before: Richard R. Clifton and John B. Owens, Circuit
Judges, and William E. Smith,* Chief District Judge.

Opinion by Judge Owens;
Dissent by Judge Smith

*
The Honorable William E. Smith, United States Chief District Judge
for the District of Rhode Island, sitting by designation.
2 DUGARD V. UNITED STATES

SUMMARY**

Federal Tort Claims Act / California Law

The panel affirmed the district court’s judgment in favor
of the United States, and agreed with the district court that the
Federal Tort Claims Act (“FTCA”) and its interaction with
California law precluded Jaycee Dugard’s recovery for the
incompetence of the parole office that was supposedly
supervising federal parolee Phillip Garrido, who committed
unspeakable crimes against Dugard for 18 years.

The FTCA, a limited waiver of the United States’
sovereign immunity, provides that the United States shall be
liable “in the same manner and to the same extent as a private
individual under like circumstances” under applicable state
law. 28 U.S.C. § 2674.

The panel held that in locating an analogous private party
under the FTCA, it was appropriate to look to cases involving
public entities or public immunities, so long as the policies
underlying them were applicable to private parties in the state
as well. The panel further held that because the extent of the
federal government’s liability under the FTCA was described
with reference to state law, when the state has described a
public policy for limiting liability for private parties, the
government’s liability under the FTCA would shrink as well.
The panel concluded that limiting liability here, for officials
involved in the release and rehabilitation of criminal
offenders, was consistent with California’s policies

**
This summary constitutes no part of the opinion of the court. It has
been prepared by court staff for the convenience of the reader.
 DUGARD V. UNITED STATES 3

encouraging criminal rehabilitation for public and private
parties alike. The panel concluded that because a private
individual in like circumstances would not be liable under
California law, the United States cannot be held liable under
the FTCA for the conduct of the parole officer here.

Chief District Judge Smith dissented, and would reverse
the summary judgment in favor of the United States. Judge
Smith wrote that the majority’s decision turned on its
understanding that the most analogous California cases for
the purposes of determining FTCA liability were those
involving private rehabilitation centers, and in his view, this
was an incorrect application of the FTCA and misinterpreted
the body of California tort law.

COUNSEL

Jonathan P. Steinsapir (argued), Dale F. Kinsella, Amber
Holley Melius, and David W. Swift, Kinsella Weitzman Iser
Kump & Aldisert, LLP, Santa Monica, California, for
Plaintiff-Appellant.

Patrick G. Nemeroff (argued) and Mark B. Stern, Attorneys,
Appellate Staff, Civil Division, Department of Justice,
Washington, D.C., for Defendant-Appellee.
4 DUGARD V. UNITED STATES

OPINION

OWENS, Circuit Judge:

Phillip Garrido, a parolee with a terrible history of drug-
fueled sexual violence, committed unspeakable crimes
against Jaycee Dugard for 18 years. State and federal
authorities missed many opportunities to stop these tragic
events. Ms. Dugard received a large cash settlement from the
State of California for its incompetence, and seeks similar
compensation from the Federal Government.

While our hearts are with Ms. Dugard, the law is not. We
agree with the district court that the Federal Tort Claims Act
and its interaction with California law precludes her recovery
for the incompetence of the parole office that was supposedly
supervising Garrido. We have jurisdiction under 28 U.S.C.
§ 1291, and we affirm.

I. FACTS AND PROCEDURAL HISTORY

A. Garrido’s Criminal History, Convictions, and
Kidnapping of Dugard

The sickening circumstances of this case have been
detailed many times in various forms, so we limit our
discussion to the essential facts.

Garrido’s predatory pattern of sexual abuse, often in
conjunction with drug use, began in the 1970s. In 1972, he
was arrested, though not convicted, for drugging and raping
a juvenile. He was using LSD and marijuana during this
time. In 1976, Garrido was charged with kidnapping and
raping a woman in South Lake Tahoe, California.
 DUGARD V. UNITED STATES 5

That same year, Garrido kidnapped a second woman in
South Lake Tahoe, California, and drove her into Nevada,
where he hid her in a shed and raped her. He was charged in
federal court with kidnapping, and was convicted and
sentenced to 50 years imprisonment in 1977. During his trial,
Garrido testified and explained that, while under the influence
of drugs, he had uncontrollable deviant and violent sexual
compulsions. In 1988, he was released on federal parole. His
federal parole was set to expire in March 1999, after which
the State of California would take over supervision.1

Due to his history of sexual violence while intoxicated,
Garrido’s federal parole terms required him to undergo
regular drug testing and counseling. They also banned drug
use and excessive alcohol consumption. Parole officers
recognized that “the potential for causing great physical harm
is present if [Garrido] becomes unstable as a result of drug
use.” Medical professionals described him as “a time bomb”
and “like a pot boiling with no outlet valve.” Despite these
prophetic warnings and mandatory reporting obligations, in
the 30 months after his release on parole, officers did not
report approximately 70 documented drug-related parole
violations, including methamphetamine abuse, to the Parole
Commission. These violations included Garrido drinking

1
The Sentencing Reform Act of 1984 abolished parole for all crimes
committed after November 1, 1987. See Marsh v. Taylor, 925 F.2d 1131,
1132 (9th Cir. 1991); see also Fassler v. U.S. Parole Comm’n, 964 F.2d
877, 878 (9th Cir. 1991); Farese v. Story, 823 F.2d 975, 976–77 (6th Cir.
1987) (per curiam). Prior to that, the Parole Commission determined if
and when a prisoner could be released on parole, and then ultimately
decided if a parolees’ performance outside warranted a return to prison.
Post-incarceration supervision of federal offenders now takes the form of
“supervised release,” with district judges ultimately monitoring the
performance of offenders with the assistance of federal probation officers.
6 DUGARD V. UNITED STATES

excessive amounts of water to dilute urine samples used in
drug tests.

On June 10, 1991, while Garrido was on federal parole,
Garrido and his wife kidnapped Dugard near her home in
South Lake Tahoe.2 She was only 11. For the next 18 years,
Garrido held Dugard captive, sometimes in chains, in a shed
that he built in his backyard. Often on methamphetamine
binges, he repeatedly raped and drugged her. Dugard gave
birth to two of his children without any medical treatment or
prenatal care. Dugard and her children remained captive until
their discovery on August 26, 2009.

B. Proceedings Before the District Court3

In September 2011, Dugard filed a complaint against the
United States under the Federal Tort Claims Act (FTCA), on
her own behalf and as guardian for her two minor children
(A. Dugard and G. Dugard).4 In December 2012, Dugard
filed a Second Amended Complaint. In it, she alleged that
the government negligently performed numerous mandatory
duties when supervising Garrido, including duties to report
parole violations in the years prior to her kidnapping. But for
the government’s negligence, she alleged, Garrido’s parole

2
Garrido had also been prohibited from travel to Lake Tahoe as a
condition of his parole.
3
The Honorable Carlos T. Bea, United States Circuit Judge for the
Ninth Circuit, sat by designation as the district court judge, pursuant to
28 U.S.C. § 291(b).
4
Only Jaycee and A. Dugard remain as plaintiffs (we will refer to them
collectively as “Dugard”). G. Dugard voluntarily dismissed her claims.
 DUGARD V. UNITED STATES 7

would have been revoked and he would not have been able to
kidnap her in 1991.

Following discovery, the government filed a motion to
dismiss and for summary judgment. The government argued
that the FTCA bars Dugard’s claims because there is no
liability for private individuals in like circumstances under
California law, as required to sustain an FTCA claim under
28 U.S.C. § 2674.5

The district court granted the motion, finding that the
proper state-law analogy to the government officials in this
case was to private parties providing criminal rehabilitative
services. California law holds that such private parties “do
not owe a duty of reasonable care to control others to the
entire world or to the general public.” Instead, “such duty of
reasonable care to control is owed by such rehabilitative
service providers only to a very small group of specifically
identifiable and foreseeable victims.” Since Dugard did not
allege that she was a specifically identifiable victim, the
“United States owed no duty of reasonable care to Plaintiffs
in this case.”

5
The government also argued that there were two additional bases on
which it would prevail: (1) that the discretionary function exception bars
Dugard’s claims; and (2) that the parole officer’s failure to follow
mandatory duties did not proximately cause plaintiffs’ injuries. Because
we agree with the district court that the government faces no liability
under the FTCA, we need not address the discretionary function exception
or proximate cause issues here.
8 DUGARD V. UNITED STATES

II. DISCUSSION

A. Standard of Review

We review de novo a district court’s dismissal of an
FTCA claim because a private person under like
circumstances would not be liable under state law. Xue Lu v.
Powell, 621 F.3d 944, 947 (9th Cir. 2010).

B. The FTCA Forecloses Federal Liability In This
Case

The FTCA, a limited waiver of the United States’
sovereign immunity, provides that the United States shall be
liable “in the same manner and to the same extent as a private
individual under like circumstances” under applicable state
law. 28 U.S.C. § 2674; see also id. § 1346(b)(1). Although
the federal government “could never be exactly like a private
actor, a court’s job in applying the standard is to find the most
reasonable analogy.” LaBarge v. Mariposa Cty., 798 F.2d
364, 367 (9th Cir. 1986).

The most analogous cases to this situation involve the
liability of private criminal rehabilitation facilities. Under
California law, private companies that operate rehabilitation
programs do not owe a duty of care to the public at large for
the conduct of inmates or parolees under their supervision.
See Cardenas v. Eggleston Youth Ctr., 238 Cal. Rptr. 251,
252–53 (Ct. App. 1987) (holding that a private rehabilitation
facility owes no duty of care to “members of the community
in which it is located for the criminal conduct of its
residents”); Beauchene v. Synanon Found., Inc., 151 Cal.
Rptr. 796, 798–99 (Ct. App. 1979) (holding that a private
rehabilitation center owed no duty to the plaintiff to control
 DUGARD V. UNITED STATES 9

the behavior of a convict who escaped the facility and shot
the plaintiff). Instead, the private facilities owe a duty only
to individuals that are foreseeable and specifically identifiable
victims of their wards’ conduct. See Vu v. Singer Co.,
706 F.2d 1027, 1029 (9th Cir. 1983) (discussing the duty to
warn under California law and concluding that it “clearly”
requires a “foreseeable and specifically identifiable” victim);
Rice v. Ctr. Point, Inc., 65 Cal. Rptr. 3d 312, 316 (Ct. App.
2007) (explaining that a duty exists only where the “injury is
foreseeable and the intended victim is identifiable”). This
rule is a result of California’s strong public policy to
encourage “innovative criminal offender release and
rehabilitation programs.” Beauchene, 151 Cal. Rptr. at 799.

As Dugard has not argued, and submits no facts to
suggest, that she was a specifically identifiable victim, she
would not have a viable claim against an analogous private
person under California law.

Dugard contends that Beauchene, Cardenas, Rice, and Vu
are inappropriate private person analogues because the duties
described therein originally arose out of a statutory immunity
granted only to public entities—Cal. Gov. Code Section
845.8.6 As she puts it, using these cases to locate the most

6
The relevant statutory provision states:

Neither a public entity nor a public employee is liable
for:

(a) Any injury resulting from determining whether to
parole or release a prisoner or from determining the
terms and conditions of his parole or release or from
determining whether to revoke his parole or release.
10 DUGARD V. UNITED STATES

analogous private parties “backdoor[s] inapplicable state
government immunities into the FTCA analysis.” As support,
she points to the Supreme Court’s instruction that the United
States’ liability under the FTCA may not be defined by
reference to the liability of state or local entities. See United
States v. Olson, 546 U.S. 43, 45 (2005) (rejecting argument
that the federal government’s liability should be compared to
the liability of state and local mine inspectors); Rayonier Inc.
v. United States, 352 U.S. 315, 318–19 (1957) (rejecting
argument that the federal government’s liability should be
determined by the liability of state and local firefighters);
Indian Towing Co. v. United States, 350 U.S. 61, 65 (1955)
(rejecting argument that the federal government’s liability
should be compared to the liability of a municipal
corporation).

In locating an analogous private party under the FTCA,
however, it is appropriate to look to cases involving public
entities or public immunities, so long as the policies
underlying them are applicable to private parties in the state
as well. See Xue Lu, 621 F.3d at 947 (“A public-entity case
. . . can offer some guidance, to the extent it illuminates
general principles of respondeat superior liability that apply
in California to public and private employers alike.”). The
public policy to limit the liability of institutions involved in

(b) Any injury caused by:

(1) An escaping or escaped prisoner;

(2) An escaping or escaped arrested person; or

(3) A person resisting arrest.

Cal. Gov. Code. § 845.8.
 DUGARD V. UNITED STATES 11

criminal rehabilitative endeavors is one that California courts
have expressly chosen to expand beyond public officials and
apply to private parties involved in similar efforts. See
Beauchene, 151 Cal. Rptr. at 799 (explaining that, although
the defendant “concededly is not a ‘public entity or public
employee’ . . . the same public policy that moved the
Legislature to immunize Public release and rehabilitation
programs from liability to encourage such innovations in the
interests of criminal justice compels the conclusion that
[defendant’s] Private release and rehabilitation program owed
no legal duty to” the plaintiff (emphasis added)); see also Vu,
706 F.2d at 1029–30; Rice, 65 Cal. Rptr. 3d at 313; Cardenas,
238 Cal. Rptr. at 252–53. That application to private parties
takes this case out of the ambit of Olson, Rayonier, and
Indian Towing. Because the extent of the government’s
liability under the FTCA is described with reference to state
law, when the state has described a public policy for limiting
liability for private parties, we expect the government’s
liability under the FTCA will shrink as well. Cf. Anderson v.
United States, 55 F.3d 1379, 1381 (9th Cir. 1995) (refusing
to limit the liability of the government for the negligence of
the U.S. Forest Service because California courts have “not
described any such public policy in the area of firesetting and
fire control by private persons” that would limit liability).
Limiting liability here, for officials involved in the release
and rehabilitation of criminal offenders, is consistent with
California’s policies encouraging criminal rehabilitation for
public and private parties alike.

The dissent agrees that due to the “imperative policy
objective of encouraging innovative release and rehabilitation
programs for criminal offenders,” California courts have
limited the liability of private criminal rehabilitation centers.
Dissent at 28 (quoting Cardenas, 238 Cal. Rptr. at 253); see
12 DUGARD V. UNITED STATES

also Dissent at 19–20 (quoting Beauchene, 151 Cal. Rptr. at
799). We disagree only on whether that policy consideration
is applicable to the government actors here. We hold that it
is. The United States parole system dealt with “a wide range
of individuals with differing problems.” Probation Division,
Administrative Office of the United States Courts, The
Supervision Process, Publication 106, at 21 (1983).
Accordingly, like the criminal rehabilitative centers in
Beauchene, Cardenas, Rice, and Vu, it furthered many
different rehabilitative goals, including those “rang[ing] from
protecting the public from further crime to providing the
offender with needed education, vocational training, medical
care, and other correctional treatment.” Publication 106, at 1;
see Vu, 706 F.2d at 1028 (“vocational training, work
experience and educational programs”); Rice, 65 Cal. Rptr.
3d at 313 (drug rehabilitation program); Cardenas, 238 Cal.
Rptr. at 251 (youth group home facility); Beauchene, 151 Cal.
Rptr. at 797 (drug and alcohol rehabilitation). Imposition of
broader liability logically forces this range of goals to shift to
the former (protecting the public), and away from the latter
(education, vocational training, medical, and correctional
treatment). This force would disrupt the ability of probation
officers to develop a “knowledgeable, sensitive, and flexible
approach” suited to each offender, Publication 106, at 21,
thereby discouraging “innovative” rehabilitative efforts and
“encouraging the detention of prisoners in disregard of their
rights and society’s needs.” Beauchene, 151 Cal. Rptr. at
799. The public policy concerns that militate against
expansive liability for private criminal rehabilitation
programs are therefore directly applicable here.

The dissent instead would have us rely on a line of
California cases, particularly involving medical professionals,
which has held that a doctor or therapist has a duty to act
 DUGARD V. UNITED STATES 13

reasonably to avoid injury to a third party where a patient
expresses an intent to harm or exhibits actions likely to cause
harm to a third party (which need not be a foreseeable and
specifically identifiable victim). See Bragg v. Valdez, 3 Cal.
Rptr. 3d 804, 810–11 (Ct. App. 2003) (holding that a
psychotherapist had a duty to the plaintiff not to discharge a
mentally ill patient because of his lack of insurance, where
patient killed plaintiff after discharge); Reisner v. Regents of
Univ. of Cal., 37 Cal. Rptr. 2d 518, 520 (Ct. App. 1995)
(holding that the doctor had a duty to the future sexual partner
of his teenage patient to warn the patient or “others likely to
apprise the [plaintiff] of the danger” that his patient had likely
been infected with HIV through a contaminated blood
transfusion); Myers v. Quesenberry, 193 Cal. Rptr. 733,
734–35 (Ct. App. 1983) (holding that a doctor had a duty to
a driver struck by his patient’s car to warn his patient that she
should not drive in an uncontrolled diabetic condition
complicated by a recently lost pregnancy). These cases rely
heavily on Tarasoff v. Regents of University of California,
551 P.2d 334 (Cal. 1976), a landmark case in which the
California Supreme Court held that, “[w]hen a therapist
determines . . . that his patient presents a serious danger of
violence to another, he incurs an obligation to use reasonable
care to protect the intended victim against such danger.” Id.
at 340.

The balance of policy considerations struck in those
cases—between the danger of assault or injury to the public
on the one hand, and treatment on the other hand—is struck
differently from the criminal rehabilitation context. Compare
Tarasoff, 551 P.2d at 346–47 (tipping in favor of public
safety over concerns related to confidentiality in treatment),
with Beauchene, 151 Cal. Rptr. at 798–99 (tipping in favor of
treatment and rehabilitation over public safety). “‘Duty’ is
14 DUGARD V. UNITED STATES

‘an expression of the sum total of those considerations of
policy which lead the law to say that the particular plaintiff is
entitled to protection’ against the defendant’s conduct.”
Beauchene, 151 Cal. Rptr. at 798 (quoting William L.
Prosser, Torts (4th ed. 1971)); see also Tarasoff, 551 P.2d at
342. The “sum total” of the considerations in the criminal
rehabilitation context is simply different from the calculus in
the cases on which the dissent would have us rely.

Because a private individual in like circumstances would
not be liable under California law, the United States cannot
be held liable under the FTCA for the conduct of the parole
officer here.

III. CONCLUSION

“Each member of the general public who chances to come
into contact with a parolee or probationer must risk that the
rehabilitative effort will fail.” Beauchene, 151 Cal. Rptr. at
799. Despite this increased danger, rehabilitation is an
endeavor the State of California values. Unless we adopt a
“throw away the key” approach to convicts, tragic crimes by
parolees and probationers inevitably will occur. No judge
wants to deny Dugard relief, but the FTCA, through the lens
of California law and its focus on rehabilitative efforts, does
not permit relief under the circumstances here.

AFFIRMED.
 DUGARD V. UNITED STATES 15

SMITH, Chief District Judge, dissenting:

As the majority states, our hearts are with Ms. Dugard.
But for the incompetence of both California and federal
officials, the unspeakable crimes committed by Garrido
would never have occurred. Ms. Dugard has been
compensated by California; and in my view she should have
her day in court against the federal government as well. I
differ with the majority’s conclusion that the law does not
allow her this, and therefore I respectfully dissent.

The majority’s decision turns on its understanding that the
most analogous California cases for the purposes of
determining liability under the Federal Tort Claims Act
(“FTCA”) are those involving private rehabilitation centers.
In my view, this is an incorrect application of the FTCA and
misinterprets the body of California tort law of which the so-
called private rehabilitation center cases are an inextricable
part.

As the majority notes, the FTCA is a limited waiver of
sovereign immunity and provides that the United States is
liable for the negligent conduct of its employees “in the same
manner and to the same extent as a private individual under
like circumstances.” 28 U.S.C. § 2674; see also 28 U.S.C.
§ 1346(b). When determining whether a state actor is
analogous to a private entity, courts must “interpret these
words to mean what they say, namely, that the United States
waives sovereign immunity ‘under circumstances’ where
local law would make a ‘private person’ liable in tort.”
United States v. Olson, 546 U.S. 43, 44 (2005) (emphasis in
original) (quoting 28 U.S.C. § 1346(b)(1)). And, when
conducting this analysis, common sense dictates that “where
the government and its private party counterpart diverge in a
16 DUGARD V. UNITED STATES

significant respect, that circumstance must be taken into
account in determining what is the ‘most reasonable
analogy.’” Bush v. Eagle-Picher Indus., Inc., 927 F.2d 445,
452 (9th Cir. 1991) (quoting LaBarge v. Mariposa County,
798 F.2d 364, 367 (9th Cir. 1986)), abrogated on other
grounds by Scheuring v. Traylor Bros., 476 F.3d 781, 783
(9th Cir. 2007).

Here, on the surface, the “private rehabilitation center
cases” cited by the government and relied on by the majority1
provide a tempting analogue to the United States Parole
Commission because both deal broadly with individuals who
have broken the law and are in some stage of incarceration,
transition on parole/probation, or supervision; however, upon
closer examination, it is clear that these cases represent an
exception to the general rule under California tort law that,
where there is a special relationship, there is a duty to warn
or control that extends to foreseeable, but not readily
identifiable victims, provided that the action required would
be reasonable and not futile. As I explain in detail below, the
courts in the private rehabilitation center cases effectively
granted those facilities a form of immunity—essentially a
carbon copy of the immunity enjoyed by their public
counterparts—because they found that the public policy
considerations were so strong as to preclude a finding of a
duty. Irrespective of the majority’s suggestion to the
contrary, no such considerations are present in this case, and
there is virtually no policy reason to extend this narrow
judicially created immunity to U.S. Probation. Thus, it is the

1
See generally Vu v. Singer Co., 706 F.2d 1027 (9th Cir. 1983); Rice v.
Ctr. Point, Inc., 154 Cal. App. 4th 949 (Ct. App. 2007); Cardenas v.
Eggleston Youth Ctr., 193 Cal. App. 3d 331 (Ct. App. 1987); Beauchene
v. Synanon Found., Inc., 88 Cal. App. 3d 342 (Ct. App. 1979).
 DUGARD V. UNITED STATES 17

general body of “special relationship” cases, rather than the
private rehabilitation center exception, that provides a closer
private party analogue, and summary judgment in favor of the
government is not warranted.

I. The Duty to Control or Warn under California
Law

In order to set the stage for the private person analogy
analysis on which this case turns, and to understand where I
part company with the majority, a chronological review of the
development of California tort law concerning the duty to
control and/or warn is necessary.

In 1966, in Poncher v. Brackett, 246 Cal. App. 2d 769
(Ct. App. 1966), a California Court of Appeals held that
grandparents had a duty to an unidentifiable victim of their
violent grandson. The court explained that:

The governing principle is expressed in
section 319 of the Restatement Second of
Torts as follows: “One who takes charge of a
third person whom he knows or should know
to be likely to cause bodily harm to others if
not controlled is under a duty to exercise
reasonable care to control the third person to
prevent him from doing such harm.”

Id. at 772–73. The court clarified that “[t]he ability to control
the child, rather than the relationship as such, is the basis for
a finding of liability on the part of a parent.” Id. at 772. That
is, it was not because it was a grandparent-child relationship
in particular that the duty attached, but rather because it was
18 DUGARD V. UNITED STATES

some type of “special relationship” in which there was an
ability to control.

In Johnson v. State, 69 Cal. 2d 782 (1968), the Supreme
Court of California found that state employees had a duty to
warn foster parents of the violent criminal history of a
teenage parolee placed in their care. In so holding, the Court
found that the state employees did not have immunity under
Government Code section 845.8, which provides that:
“Neither a public entity nor a public employee is liable for:
(a) Any injury resulting from determining whether to parole
or release a prisoner or from determining the terms and
conditions of his parole or release or from determining
whether to revoke his parole or release.” Id. at 795 n.9
(quoting Cal. Gov’t Code § 845.8); see id. at 795–96. In
giving this narrow application to section 845.8, the court
reasoned that “[o]nce an official reaches the decision to
parole to a given family, however, the determination as to
whether to warn the foster parents of latent dangers facing
them presents no such reasons for immunity.” Id. at 795.
The court added that “[t]his analysis, allowing immunity for
basic policy decisions . . . but rejecting it for the ministerial
implementation of that basic policy, receives support from a
long line of cases in California and in federal courts,
interpreting similar ‘discretionary’ language in section 2680,
subdivision (a), of the Federal Tort Claims Act.” Id. at 796.

Eight years later, in the landmark case Tarasoff v. Regents
of University of California, 17 Cal. 3d 425 (1976), the
Supreme Court of California considered whether a therapist
had a duty to warn a person threatened by a patient,
notwithstanding the general confidentiality obligations of that
profession. There, the therapist had notified the police of a
threat made by the patient, but did not notify the specific
 DUGARD V. UNITED STATES 19

person about whom the threat had been made—Tatiana
Tarasoff—or her parents. Id. at 430–31. The court held “that
defendant therapists cannot escape liability merely because
Tatiana herself was not their patient” because the “special
relationship” between therapist and patient created a duty “to
use reasonable care to protect the intended victim against
such danger.” Id. at 431. What constitutes “reasonable care,”
the court said, depends on the circumstances: “it may call for
him to warn the intended victim or others likely to apprise the
victim of the danger, to notify the police, or to take whatever
other steps are reasonably necessary under the
circumstances.” Id. at 431 (emphasis added).

In 1979, Beauchene v. Synanon Foundation, Inc., 88 Cal.
App. 3d 342 (Ct. App. 1979), considered whether a private
rehabilitation facility could be held liable for the criminal
conduct of a convict who, after being released into the
defendant’s care, escaped and injured the plaintiff. The
plaintiff argued that the facility had been negligent in
screening the convict and that he should never have been let
into the program. Id. at 345. The issue in Beauchene was
whether the facility owed the plaintiff a duty based on the
“special relationship . . . between the defendant and the active
wrongdoer.” Id. at 347 (citing Tarasoff, 17 Cal. 3d at 435).
The court found no liability on public policy grounds. Id. at
348. The court first explained that under California law:

Principal policy considerations in deciding
whether a duty exists include “the
foreseeability of harm to the plaintiff, the
degree of certainty that the plaintiff suffered
injury, the closeness of the connection
between the defendant’s conduct and the
injury suffered, the moral blame attached to
20 DUGARD V. UNITED STATES

the defendant’s conduct, the policy of
preventing future harm, the extent of the
burden to the defendant and consequences to
the community of imposing a duty to exercise
care with resulting liability for breach, and the
availability, cost, and prevalence of insurance
for the risk involved.”

Id. at 347 (quoting Rowland v. Christian, 69 Cal. 2d 108, 113
(1968)). With that framework in mind, the court found that
“[a]lthough appellant’s injuries may be grievous, ‘[of]
paramount concern is the detrimental effect a finding of
liability would have on prisoner release and rehabilitation
programs. Were we to find a cause of action stated we would
in effect be encouraging the detention of prisoners in
disregard of their rights and society’s needs.’” Id. at 348
(emphasis added) (quoting Whitcombe v. County of Yolo,
73 Cal. App. 3d 698, 716 (1977)). In order to address the
paramount concern about the effect that liability would have
on the development of innovative prisoner release and
rehabilitation programs, the Court reached for the immunity
cloak provided by section 845.8 to public entities and
employees discussed and distinguished in Johnson, and
applied it to private facilities as if they were public:

Respondent concededly is not a “public entity
or public employee” within the meaning of
section 845.8. But the same public policy that
moved the Legislature to immunize public
release and rehabilitation programs from
liability—to encourage such innovations in
the interests of criminal justice—compels the
conclusion that respondent’s private release
and rehabilitation program owed no legal duty
 DUGARD V. UNITED STATES 21

to this appellant. In light of the purpose
behind the governmental immunity, it would
be incongruous to hold that, while the state is
immune from liability for its decision to
assign Bentley to, and his unauthorized
departure from, the Synanon program, the
program itself owed appellant a duty not to
accept Bentley or to prevent his unauthorized
departure.

Id. (emphasis in original). Thus, the Beauchene decision was
driven by the court’s conclusion that “[t]o hold respondent
civilly liable would deter the development of innovative
criminal offender release and rehabilitation programs, in
contravention of public policy.” Id.

Following Beauchene, the Supreme Court of California
decided Thompson v. County of Alameda, 27 Cal. 3d 741
(1980). There, a juvenile offender who was released from a
county detention facility into the custody of his mother had
threatened that, if released, he would kill a young child in the
neighborhood. Id. at 746. The court held that the county of
Alameda had no duty to warn the neighborhood parents, the
police, or the child’s mother. Id. at 758. The court explained
that “[i]n deciding whether a duty to warn should be imposed,
we inquire under our Rowland v. Christian, [69 Cal. 2d 108,
113 (1968)], formulation concerning the probable beneficial
effect if such warnings were routinely and generally given.”
Id. at 755. The court then reasoned:

We are skeptical of any net benefit which
might flow from a duty to issue a generalized
warning of the probationary release of
offenders. In our view, the generalized
22 DUGARD V. UNITED STATES

warnings sought to be required here would do
little to increase the precautions of any
particular members of the public who already
may have become conditioned to locking their
doors, avoiding dark and deserted streets,
instructing their children to beware of
strangers and taking other precautions.

Id. The court distinguished Johnson and Tarasoff because in
those cases there was a specifically identifiable victim, and
therefore a warning would not have been futile or against
public policy:

[I]t is fair to conclude that warnings given
discreetly and to a limited number of persons
would have a greater effect because they
would alert those particular targeted
individuals of the possibility of a specific
threat pointed at them. In contrast, the
warnings sought by plaintiffs would of
necessity have to be made to a broad segment
of the population and would be only general
in nature. In addition to the likelihood that
such generalized warnings when frequently
repeated would do little as a practical matter
to stimulate increased safety measures, as we
develop below, such extensive warnings
would be difficult to give.

Id. at 755. The court acknowledged that, “[i]n those instances
in which the released offender poses a predictable threat of
harm to a named or readily identifiable victim or group of
victims who can be effectively warned of the danger, a
releasing agent may well be liable for failure to warn such
 DUGARD V. UNITED STATES 23

persons.” Id. at 758 (emphasis added). With respect to the
police, the court noted that “warnings to the police as urged
by plaintiffs ordinarily would be of little benefit in preventing
assaults upon members of the public by dangerous persons
unless we were simultaneously and additionally to impose a
concurrent duty on the police to act upon such warnings. As
we noted in Tarasoff, supra, no such duty to act exists.” Id.
at 756. Finally, regarding the juvenile offender’s mother, the
court reasoned that:

notification to the offender’s mother of
James’ threat in our opinion would not have
the desired effect of warning potential
victims, at least in a case such as that herein
presented. . . . The imposition of an
affirmative duty on the County to warn a
parent of generalized threats without
additionally requiring, in turn, some
affirmative action by the parent would prove
ineffective.

Id. at 757. As with the police, because the mother did not
have a duty to take some affirmative action upon receiving
the information about her son’s threat, the court found that
the warning would not have been effective.

Notably, Justice Tobriner, the author of Johnson and
Tarasoff, dissented from the majority’s formulation, finding
that the failure to warn the mother did state a cause of action.
Justice Tobriner explained that “the absence of an identifiable
victim does not postulate the absence of a duty of reasonable
care.” Id. at 760. “Instead, [Johnson and Tarasoff] stand for
the principle that a special relationship, such as that between
the state and a person in its custody, establishes a duty to use
24 DUGARD V. UNITED STATES

reasonable care to avert danger to foreseeable victims. If the
victim can be identified in advance, a warning to him may
discharge that duty; if he cannot be identified, reasonable care
may require other action.” Id. The fact that the specific
victim was not identifiable in advance “cannot absolve the
state from its failure to warn James’ mother so that she could
exercise proper care in observing and supervising James and
thereby preventing the harm that ensued.” Id. at 761. The
majority countered this argument—not by disputing the
dissent’s formulation of the law concerning an identifiable
victim—but instead by noting that “[t]he dissent[’s]
speculat[ion] that the mother ‘might’ have taken special care
to control her son had she been warned of James’ threats”
was “attenuated conjecture,” which “cannot alone support the
imposition of civil liability.” Id. at 757. Moreover, the
majority explained, “it is contrary to the very purpose of such
a release to speculate that a mother in whose care a nearly 18-
year-old offender has been temporarily placed would thereby
assume the constant minute-to-minute supervision that would
have been required to prevent the tragedy.” Id. at 758.2

At the heart of the majority’s decision in Thompson was
a deep concern about expanding the reach of Johnson—which
held that statutory immunity under section 845.8 did not
protect a county from liability for failure to warn of danger
disconnected from the decision to parole—and Tarasoff—
which cemented the probability of liability for failure to warn

2
The Court’s conclusion on this point is dubious to be sure: a warning
to the mother might very well have been effective; for example, reporting
a threatening presence to the police and community is a mainstay of sex
offender notification laws, both state and federal. See Sex Offender
Registration and Notification Act, Pub. L. No. 109-248, Title I (2006).
But that debate is beyond the scope of the question here.
 DUGARD V. UNITED STATES 25

in special relationship situations—to require that the
government issue “generalized warnings” to the public every
time a parolee is released.3 Indeed, the question before the
court in Thompson was actually quite narrow: “[W]e
examine the propriety of imposing on those responsible for
releasing or confining criminal offenders a duty to warn of
the release of a potentially dangerous offender who, as here,
has made a generalized threat to a segment of the
population.” Id. at 750 (emphasis added). Looking to
Beauchene for support, the court found that these generalized
warnings would be against public policy:

Furthermore, such notice might substantially
jeopardize rehabilitative effort both by
stigmatizing released offenders and by
inhibiting their release. It is also possible that,
in addition, parole or probation authorities
would be far less likely to authorize release
given the substantial drain on their resources
which such warnings might require. A stated
public policy favoring innovative release
programs would be thwarted.

Id. at 757. The majority found further support in the fact that
“the County’s original decision to release James . . . is a

3
Although Thompson considers the duty of a public entity, its holding
provides an important basis for the private rehabilitation cases that follow.
Accordingly, the justifications for the holding in Thompson also form the
justifications for exempting private rehabilitation facilities from their
special duty—public policy concerns and the futility of any warning. As
detailed below, neither of these justifications apply to federal parole
officers, a difference that warrants looking to other analogous private
entities.
26 DUGARD V. UNITED STATES

decision we already hold is immunized from liability.” Id. at
757–58.

On the heels of Thompson, this Circuit upheld the finding
of no duty in a case where a Job Corps program failed to
control its residents. Vu v. Singer Co., 706 F.2d 1027, 1030
(9th Cir. 1983). The Court came to this conclusion for two
reasons. First, it found that California law with respect to the
duty to warn mandates that the “victim must be foreseeable
and specifically identifiable.” Id. (citing Thompson, 27 Cal.
3d 741). Second, the Court pointed to the same policy
concerns voiced in Beauchene:

To impose on the operator of a center a duty
to prevent the tortious acts of corps members
and to impose liability to the victims of such
acts for having failed to do so would place in
some degree of jeopardy the Job Corps
program and its efforts towards the
rehabilitation of disadvantaged young people.
Faced with such potential liability an operator
with any concern for its economic survival
could be expected to terminate from the corps
any person whose conduct suggests that he
might pose a risk, whether it otherwise
justifies termination or not. This would
deprive of the program’s benefits those most
in need of rehabilitation.

Id.4

4
Judge Rothstein filed a concurrence in Vu, which recognized that the
Court was bound by Thompson, but criticized that decision (relying on
Justice Tobriner’s dissent) as “enunciat[ing] a myopic view of
 DUGARD V. UNITED STATES 27

One month later, however, Myers v. Quesenberry clarified
the holding of Thompson, stating that “the fact [that the
plaintiff] was a foreseeable but not a readily identifiable
victim . . . does not preclude him from stating an action.”
144 Cal. App. 3d 888, 892 (Ct. App. 1983). In Myers, a
doctor had failed to warn a patient not to drive due to her
uncontrolled diabetic condition; she struck the plaintiff, who
sued the doctor. The court’s analysis centered on the fact that
“[a]s a practical matter, the doctors here could not have
effectively warned [the plaintiff victim] Myers of the danger
presented by [the patient] Hansen’s driving. . . . However,
they could easily have warned Hansen not to drive because of
her irrational and uncontrolled diabetic condition.” Id. at
892–93 (citations omitted). Of significance was the fact that
“this probably would not have been a futile act. Having
otherwise complied with her doctors’ professional
recommendations, Hansen presumably would have
continued to follow their advice had they warned her not
to drive.” Id. at 893. The court distinguished this holding
from Thompson—not because it involved a doctor-patient
relationship, as the majority argues—but because the warning
would not, as a matter of law, have been futile: “Hansen is
unlike the homicidal actors in Thompson and Tarasoff . . . .
On these pleadings, we cannot factually presume Hansen
would have ignored the doctors’ warning.” Id. Therefore the
court concluded that “under these circumstances where
warning the actor is a reasonable step to take in the exercise
of the standard of care applicable to physicians . . . , liability
is not conditioned on potential victims being readily
identifiable as well as foreseeable. Id. (emphasis added)
(citing Thompson, 27 Cal. 3d at 752–53, 758; Mavroudis v.

foreseeability in the context of the duty to warn and to supervise.” Vu,
706 F.2d at 1031 (Rothstein, J., concurring).
28 DUGARD V. UNITED STATES

Superior Court, 102 Cal. App. 3d 594, 599–601 (Ct. App.
1980)).

In 1987, Cardenas v. Eggleston Youth Center, 193 Cal.
App. 3d 331 (Ct. App. 1987), again shielded a private
rehabilitation facility from liability. As with Beauchene, the
result was wholly driven by public policy:

We are persuaded that the public policy
concerns that led the Beauchene court to find
no duty on the part of that respondent should
lead us to the same conclusion in the case at
bar. We recognize that plaintiff has been
grievously injured, but to permit him recourse
against defendant would have a potentially
devastating impact on private rehabilitation
programs. Such programs, particularly in
these times of overcrowded penal facilities,
serve an indispensable public function.
Forced to choose between the competing
interests presented in this case, we must, like
the Beauchene court, without in the least
minimizing the seriousness of the injury to the
individual, defer to the imperative policy
objective of encouraging innovative release
and rehabilitation programs for criminal
offenders.

Id. at 335–36. The fact that the victims were not specifically
identifiable again did not factor into the analysis.

In 1995, a California Court of Appeals reached a similar
decision to that in Myers in Reisner v. Regents of University
of California, 31 Cal. App. 4th 1195 (Ct. App. 1995). There,
 DUGARD V. UNITED STATES 29

a doctor failed to inform a patient, Jennifer, that she had
contracted HIV from a blood transfusion. Id. at 1198. She
later infected her partner, Daniel, who sued the doctor. Id.
As in Myers, the court held that “[f]or several reasons, it is
immaterial that, in Tarasoff, the therapist knew the identity of
his patient’s intended victim whereas, in this case, Defendants
did not know Daniel or even that he existed.” Id. at 1199. In
particular, “warning Jennifer would have been a reasonable
step to take in the exercise of the standard of care applicable
to physicians . . . and we cannot factually presume Jennifer or
her parents would have ignored Defendants’ warning.” Id.
at 1200 (emphasis added).

In 2003, Bragg v. Valdez, 111 Cal. App. 4th 421 (Ct. App.
2003), held a psychiatrist liable for releasing a dangerous
patient from a psychiatric hospital because the patient did not
have insurance. The court noted that “without receiving any
therapy, and without any notification to any other mental
health care provider or the police, Lee was turned loose into
an unsuspecting community. . . . It is foreseeable that
someone who is dangerous to others is liable to act out and be
dangerous.” Id. at 431. The court explained:

This rule is in accordance with Tarasoff and
subsequent case law that holds, “When the
avoidance of foreseeable harm to a third
person requires a defendant to control the
conduct of a person with whom the defendant
has a special relationship (such as physician
and patient) or to warn the person of the risks
involved in certain conduct, the defendant’s
30 DUGARD V. UNITED STATES

duty extends to a third person with whom the
defendant does not have a special
relationship.”

Id. at 432 (quoting Reisner, 31 Cal. App. 4th at 1198).

The court also found that the psychiatrist was not immune
under two state statutes: one granting immunity to
psychiatrists who release patients prior to the end of the
initial 72-hour commitment, and another, enacted after
Tarasoff, that mandated that there could be no monetary
liability to warn of a patient’s threats “except where the
patient has communicated to the psychotherapist a serious
threat of physical violence against a reasonably identifiable
victim or victims.” Id. at 432–34. The court reasoned that
“[t]his case is not dealing with judgment calls, but a
dereliction of judgment.” Id. at 435. Thus, it did not fall
under the purview of either of these statutes, the first of which
was enacted to prevent the “inappropriate, indefinite, and
involuntary commitment of mentally disordered persons” by
protecting a doctor’s assessment of whether or not the person
is a danger; and the second, which “sought to protect [] the
privileged communications between a psychotherapist and his
or her patient in a treatment setting.” Id. at 432–35.
Moreover, the court found that “the trial court erred [] in
analyzing this case as a Tarasoff case.” Id. at 434.
Concluding that the “complaint [did] not allege a failure to
warn a known victim,” but instead concerned “a negligent
release based upon factors other than the professional
judgment that is required when a psychiatrist treats an
individual who is a danger to himself or to others,” the court
held that “Tarasoff [was] inapplicable.” Id.
 DUGARD V. UNITED STATES 31

In 2007, a California Court of Appeals again found that a
private rehabilitation center did not owe a duty to the general
public in its vicinity. Rice v. Ctr. Point, Inc., 154 Cal. App.
4th 949, 959 (Ct. App. 2007). In Rice, the court implicitly
acknowledged that the fact that the plaintiffs were not
identifiable victims was not fatal to their claim: the court first
noted that “[h]ere, it is clear that defendants did not owe a
duty to the plaintiffs as the foreseeable victims of the
residents’ criminal conduct,” but then went on to state that
“[t]he more difficult question is whether a special
relationship existed between defendants and their residents
giving rise to a duty owed to the public to exercise reasonable
care to control the criminal conduct of their residents.” Id.
at 955–56 (emphasis added). On this second question, Rice’s
outcome, like Beauchene and Cardenas, was driven almost
exclusively by public policy. The court first noted that, “[i]n
Beauchene, . . . the court concluded that public policy
considerations preclude the imposition of such a duty on the
operators of a rehabilitation facility.” Id. at 956 (emphasis
added). The court then rejected the plaintiff’s attempts to
distinguish Beauchene and Cardenas, based on the fact that
the Rice defendants had violated mandatory rules, reasoning
that “[the facility]’s failure to follow applicable operational
policies or procedures did not give rise to a duty where none
otherwise existed. Defendants’ failure to comply with
applicable safety regulations would at most demonstrate a
lack of reasonable care, i.e. breach of the duty of care if such
a duty were first determined to exist.” Id. at 958 (emphasis
added).

Reading these cases as a whole makes clear that rather
than two separate lines of cases—one dealing with doctors
and the other dealing with rehabilitation facilities—this is a
single line of cases dealing with the fundamental duty to
32 DUGARD V. UNITED STATES

control or warn, based on a special relationship, which has
developed (sometimes awkwardly) over nearly 50 years.
This line of cases begins with the general rule that a
defendant has no duty to warn third parties who are harmed
by a tortfeaser. An exception to this rule based on “special
relationships” was recognized by the California courts in
Poncher (1966) and Johnson (1968). In these cases, the court
unequivocally said there was a duty to warn and/or control a
person who posed a risk of harm to third persons which was
reasonably foreseeable. (Tarasoff further refined this duty
with respect to specifically identifiable victims in certain
circumstances.) In the decisions that followed, the courts
enacted a judicially created immunity for private
rehabilitation facilities and similar programs based on
important public policy considerations (an exception to the
exception, so to speak). And where these policy
considerations have not been present (mostly in the physician
and psychiatrist context), the Court has continued to apply the
Poncher/Johnson duty analysis.

Moreover, a slight but important refinement of the case
law occurred with Myers, which clarified that a victim must
be “foreseeable,” but need not be “readily identifiable.”
Myers, 144 Cal. App. 3d at 892.5 Instead, Myers and
subsequent cases dealing with victims who were not readily
identifiable, focus on whether the proposed action that the
defendant failed to take would have been “reasonable” given

5
The district court here relied on Vu for the proposition that “one must
know that the target of the risk is an identifiable and foreseeable victim.”
Dugard v. United States, No. 3:11-cv-04718-CTB, 2013 WL 6228625, at
*9 (N.D. Cal. Nov. 27, 2013). But Vu was decided before Myers, which
clearly states that “the fact [that the plaintiff] was a foreseeable but not a
readily identifiable victim . . . does not preclude him from stating an
action.” Myers, 144 Cal. App. 3d at 892.
 DUGARD V. UNITED STATES 33

the facts of the case and public policy, and whether or not it
would have been “futile.” Id. at 892–93. The only exception
is Rice, which did not reach the issue of whether a warning
would have been futile because it found that “public policy
considerations preclude[d] the imposition of such a duty on
the operators of a rehabilitation facility.” 154 Cal. App. 4th
at 956 (emphasis added). This is the same public policy
mandate that drove the decisions in the earlier rehabilitation
cases. See Cardenas, 193 Cal. App. 3d at 335–36 (“Such
programs, particularly in these times of overcrowded penal
facilities, serve an indispensable public function. Forced to
choose between the competing interests presented in this
case, we must, like the Beauchene court, without in the least
minimizing the seriousness of the injury to the individual,
defer to the imperative policy objective of encouraging
innovative release and rehabilitation programs for criminal
offenders.” (emphasis added)); Vu, 706 F.2d at 1030 (“[T]he
same public policy that moved the Legislature to immunize
public release and rehabilitation programs from liability—to
encourage such innovations in the interests of criminal
justice—compels the conclusion that respondent’s private
release and rehabilitation program owed no legal duty to this
appellant.” (second emphasis added; first and third in
original)).

Critically, nothing in Myers limits its holding to a doctor-
patient relationship. And the court had that opportunity: in
distinguishing Thompson, the court could have said that the
duties that exist in the doctor-patient context do not exist in
other special relationships. Instead, the Myers court found
that, unlike in Thompson, it could not “factually presume
Hansen would have ignored the doctors’ warning.” Myers,
144 Cal. App. 3d at 893. In Thompson, by contrast, the
majority found that all of the suggested warnings—including
34 DUGARD V. UNITED STATES

warning the mother, about which the dissent disagreed—
would have been futile, and further, would have created bad
public policy. 27 Cal. 3d at 756. However, in cases where
there is a third party who can be “effectively warned of the
danger,” id. at 758, or another action that could easily be
taken to prevent the harm without creating bad policy, there
is a duty. See Bragg, 111 Cal. App. 4th at 431–34; Reisner,
31 Cal. App. 4th at 1198–1203; Myers, 144 Cal. App. 3d at
892–94.

II. Most Analogous Private Party

With this backdrop in mind, I turn to determining whether
“a private individual under like circumstances,” 28 U.S.C.
§ 2674, to Defendant would be held liable under California
law.

The two most recent cases—Bragg and Rice—are both
progeny of Poncher and Johnson, and both present
potentially compelling analogies: a psychiatrist who released
a mentally ill and dangerous patient into society, and a private
rehabilitation facility from which several inmates escaped. In
both of these cases, like in the instant case, the question was
whether a party responsible for a dangerous person having the
opportunity to commit a crime could be held liable for the
harm caused. Yet the California courts came to opposite
conclusions: while the psychiatrist was held liable, the private
rehabilitation facility was not. The question is why the
opposite results? The answer to that question reveals which
case is more analogous to this one.

Taking a close look at these cases, it becomes clear that
Bragg, not Rice, presents the “like circumstances” to this
case under California law. The majority simplistically
 DUGARD V. UNITED STATES 35

characterizes my argument as preferring “a line of California
cases, particularly involving medical professionals,” Majority
at 12, to the private rehabilitation cases. But as my analysis
above demonstrates, the majority clearly misses the
point—we are dealing here with one line of cases, typified by
the physician and psychologist examples, to which there is an
exception for private rehabilitation facilities based on policy
grounds. Although I concede that it is tempting to conclude,
as the majority has, that because this case deals with the
prison system rather than the medical system, Rice seems
more analogous; but that is a red herring that distracts from
the real comparison. The public policy considerations that
dictated the result in private rehabilitation cases like Rice (the
need for immunity and the futility of the putative warning)
simply do not exist in this case. Instead, the policy
motivations and the futility question in the present case align
much more closely to those in Bragg and demonstrate that a
private person under circumstances similar to those here
would be liable under California tort law.6

6
A recent district court decision with analogous facts applied a similar
approach, using “a person responsible for controlling the acts of a third
party as described in Section 319 of the Restatement (Second) of Torts”
as “the closest private analog.” Ben v. United States, No. 5:14-CV-0370
(CJS), 2016 WL 447713, at *11 (N.D.N.Y. Feb. 4, 2016). In that case, a
criminal defendant, Renz, who had been charged with receiving and
possessing child pornography, was released from custody prior to trial and
placed on electronic monitoring under the supervision of U.S. Probation
and Pretrial Services Office for the Northern District of New York. Id. at
*1. However, “Probation did not follow [its] procedures in Renz’s case.
More specifically, Probation did not develop a supervision plan for Renz,
did not inspect Renz’s electronic monitoring equipment (except perhaps
on one occasion), and did not monitor Renz’s tamper alerts as required
even though tampering with ‘location monitoring’ equipment is
considered a ‘higher-risk violation.’” Id. at *3. While on supervised
release, Renz committed crimes including kidnapping, rape, and murder.
36 DUGARD V. UNITED STATES

In Bragg, the court determined that “[p]reventing future
incidents is [] furthered by imposing a duty to the injured
party.” 111 Cal. App. 4th at 431. Simply put, public policy
favored imposition of a duty on hospitals to refrain from
releasing dangerous patients simply because they do not have
insurance. Moreover, the action required—not releasing the
patient—clearly would not have been futile; it would have
prevented the patient from committing the tort.7 By contrast,
in Rice, the court found that “public policy considerations
preclude the imposition of such a duty on the operators of a
rehabilitation facility.” 154 Cal. App. 4th at 956. The court
in Rice explained that “[w]ere we to find a cause of action

The court denied the government’s motion to dismiss the FTCA claims
against the government, finding that Probation “had sufficient control over
Renz to support a duty under § 319.” Id. at *14. In doing so, the court
rejected the government’s argument that someone on supervised release
is more akin to a voluntary outpatient of a medical/psychiatric facility
(where New York courts had found no duty), as opposed to someone
involuntarily committed to a psychiatric facility (where courts had found
a duty), because supervised release is not “in custody.” Id. Like in the
instant case, on the surface, the outpatient cases looked more analogous
because Probation did not have “actual physical custody of the dangerous
third person”; however, the court looked past the seemingly similar
circumstances in the outpatient cases and interpreted them “as holding that
because the medical/psychiatric facilities had no other way of controlling
the third parties’ actions when they were away from the facilities,” there
was no duty. Id. Probation, by contrast, “had a relationship of control
over Renz that did not require actual physical custody.” Id.
7
It is important to keep in mind that Bragg’s outcome did not hinge on
the existence of a doctor-patient relationship: The psychiatrist was liable
because the negligence did not occur in the course of the doctor-patient
relationship; had the inverse been the case, the psychiatrist would have
had statutory immunity. 111 Cal. App. 4th at 432. Thus, it cannot be said
that the particular contours of the doctor-patient relationship, as opposed
to the duties in other “special relationship” cases, drove the result.
 DUGARD V. UNITED STATES 37

stated we would in effect be encouraging the detention of
prisoners in disregard of their rights and society’s needs.”
154 Cal. App. 4th at 956 (emphasis added). As a practical
matter, liability for private rehabilitation facilities would
result in those facilities going out of business and a worsening
of the already daunting prison crisis in California.

Here, the public policy concerns align much more closely
to Bragg than Rice and the other rehabilitation center cases.
The private rehabilitation facility cases turned on the
assumption that imposing liability would be the death knell
of such facilities and that it would be “incongruous” to attach
liability to private facilities for actions for which their public
counterparts were immune. These concerns simply do not
exist in this case.

The majority suggests that the policy concerns that
animated the private rehabilitation cases are equally present
in the federal probation system, drawing support from a
publication of the Administrative Office of the United States
Courts, which contains some generic statements about
flexible and innovative approaches to offender management
and supervision. Noting that the goals of the parole and
probation systems “rang[e] from protecting the public from
further crime to providing the offender with needed
education, vocational training, medical care, and other
correctional treatment,” Majority at 12 (quoting Probation
Division, Administrative Office of the United States Courts,
The Supervision Process, Publication 106, at 21 (1983), the
majority posits that, “[i]mposition of broader liability
logically forces this range of goals to shift to the former
(protecting the public), and away from the latter (education,
vocational training, medical, and correctional treatment).” Id.
For the reasons outlined below, the majority’s suggestion just
38 DUGARD V. UNITED STATES

does not hold water: the policy concerns of federal probation
and parole are not at all aligned with the concerns described
in the private rehabilitation cases.

To begin, the publication cited by the majority makes
clear that parole and probation officers’ primary
responsibility is to keep apprised of and report on the
behavior of the individuals they supervise, and the “flexible
approach” must be consistent with those mandatory
responsibilities:

The supervision responsibilities of
probation officers begin with the statutory
obligation set forth at title 18, USC, 3665, “to
keep informed concerning the conduct and
condition of each probationer under
supervision and report thereon to the court
placing such person on probation.”

This function is also performed for
persons on parole on behalf of the United
States Parole Commission. For offenders who
do not abide by the conditions of supervision,
it is the responsibility of the probation officer
to promptly report this information to the
court or Parole Commission. The probation
officer is also mandated by statute to “use all
suitable methods, not inconsistent with the
conditions imposed by the court, to aid
probationers to bring about improvements in
their conduct and condition.”

Publication 106, at 2 (emphasis added).
 DUGARD V. UNITED STATES 39

Second, there is virtually no risk that imposing an
obligation on federal parole officers to follow their
mandatory reporting duties will result in fewer people being
released on parole. For starters, the U.S. Parole Commission
effectively went out of business in the 1980s with the passage
of the Sentencing Reform Act of 1984, Pub. L. No. 98-473,
98 Stat. 1988 (codified at 18 U.S.C. § 3551, et seq.). No
offender has been put on parole in nearly 30 years. The only
function of the Parole Commission since 1988 has been the
management of defendants who were paroled prior to that
date. Since the late 1980s, sentencing decisions lie
exclusively with federal district judges, who prescribe both
the length of incarceration and supervision, or, alternatively,
non-incarcerative probation. These decisions, of course, all
enjoy judicial immunity, and no district judge worth his or
her salt would ever decide to incarcerate a defendant for a
longer period than required to meet the objectives of
18 U.S.C. § 3553 out of fear of creating liability for federal
probation. The idea is unthinkable. To be sure, federal
probation officers supervise defendants during their terms of
supervised release or probation, but probation officials are not
responsible for deciding when a prisoner gets released.
Consequently, there is virtually no risk that holding such
officials responsible for negligence for their failure to
perform nondiscretionary functions will in any way
undermine new or creative rehabilitation programs, or result
in defendants being kept in prison for longer than necessary.
Rather, imposing the duty of reasonable care called for by
Poncher, Johnson, Tarasoff, and Bragg, will better ensure
that U.S. Probation officers perform their duties carefully,
and follow protocols dutifully, as they should, bringing
violations to the attention of U.S. District Judges (or, in the
present case, the U.S. Parole Commission). Simply put, the
concern of the California courts was that third party liability
40 DUGARD V. UNITED STATES

to foreseeable (but not specifically identifiable) victims
would make opening and operating private rehabilitation
facilities financially untenable. No such concern exists in the
federal probation system. Regardless of what liability exists
for federal probation officers’ misdeeds, U.S. Probation will
continue to supervise defendants as directed by federal
judges.

Third, the California courts’ concern about holding
private and public facilities to different standards is not at
play here. The Supreme Court of California recognized in
Johnson that the policy considerations underlying the
immunity provided by Section 845.8, did not apply in a
situation like this one where the decision concerned
“ministerial implementation” rather than a “policy
decision[]”:

Once an official reaches the decision to parole
to a given family, however, the determination
as to whether to warn the foster parents of
latent dangers facing them presents no such
reasons for immunity; to the extent that a
parole officer consciously considers pros and
cons in deciding what information, if any,
should be given, he makes such a
determination at the lowest, ministerial rung
of official action. Judicial abstinence from
ruling upon whether negligence contributed to
this decision would therefore be unjustified;
coupled with the administrative laxness that
caused the loss in the first instance, it would
only result in the failure of governmental
institutions to serve the injured individual.
 DUGARD V. UNITED STATES 41

Johnson, 69 Cal. 2d at 795–96 (emphasis added);8 see also
Vu, 706 F.2d at 1029 (“California courts have recognized that
[section 845.8] provides an exception to the rule that special
relationships can result in creation of duty.” (emphasis
added)). Here, the challenged action (or in this case inaction)
is not the policy decision that was made to release Garrido on
parole, nor is it the policy decision that would have been in
the Parole Commission’s discretion to revoke his parole; it is
simply the parole officer’s ministerial task of reporting
Garrido’s numerous drug violations to the Commission.
Thus, the evaluation in the rehabilitation facility cases that
“the same public policy that moved the Legislature to
immunize public release and rehabilitation programs from
liability [under section 845.8]—to encourage such
innovations in the interests of criminal justice—compels the
conclusion that respondent’s private release and rehabilitation
program owed no legal duty to this appellant,” Cardenas,
193 Cal. App. 3d at 335 (emphasis in original) (quoting
Beauchene, 88 Cal. App. 3d at 348), does not apply to these
facts. Moreover, the innovation currently taking place in the
federal system is being driven primarily by the courts. To be
sure, probation offices in innovative courts play a critical
programmatic role, but the policy decision of whether to
adopt a re-entry program or drug court or diversion program
is a judicial function, not a probation function, so it cannot be
said that imposing liability for probation officers’ misdeeds
will stifle innovators and increase incarceration.

8
As Johnson and Thompson were both public entity cases, they cannot
be used as specific examples of a private person in like circumstances.
However, as the majority notes, they nonetheless provide the foundation
for the California tort law concepts and public policy issues at play in this
case.
42 DUGARD V. UNITED STATES

Fourth, this is not a situation where imposing liability
would result in “frequently repeated [warnings that] would do
little as a practical matter to stimulate increased safety
measures” and “would be difficult to give.” Thompson,
27 Cal. 3d at 755. Indeed, unlike the police and the mother
in Thompson, the Parole Commission did have an affirmative
duty to act upon receipt of the warning, and presumably
would have acted in this case. So, futility too is not a
concern.

In contrast, similarities between the public official actor
in this case and the private actor in Bragg are numerous.
Like Bragg, “[t]his case is not dealing with judgment calls,
but a dereliction of judgment.” Bragg, 111 Cal. App. 4th at
435. Here, a parole officer failed to report over 70 drug
violations for a parolee who was known to be extremely
violent when under the influence of drugs. And in this case,
like in Bragg—which analyzed the contours of the duty
because the immunity and public policy issues that are
peculiar to the private rehabilitation facilities were not
applicable—“[p]reventing future incidents is [] furthered by
imposing a duty to the injured party.” 111 Cal. App. 4th at
431. Indeed, if the prevailing public policy in California is to
encourage alternative rehabilitation programs, applying the
same duties to parole officers as those applied to physicians
in Bragg and Myers would advance that policy. As many of
the private rehabilitation cases note, there is an inherent risk
involved in paroling criminals, a risk that is considerably
exacerbated by parole (or probation) officers disregarding
their responsibilities to the degree they were disregarded in
this case. Mitigating the risk of danger associated with parole
or probation/supervision by holding parole and probation
officers responsible for carrying out their mandatory tasks,
makes it more likely, not less, that the public and government
 DUGARD V. UNITED STATES 43

officials, and more importantly judges and prosecutors, will
support expansion of alternative rehabilitation programs.
And moreover, imposing a duty on parole officers to
conscientiously do their jobs has the added benefit of
protecting the public during the expansion of these programs.
In essence, the same public policy that drove the result in
Bragg applies in the context of federal parole and
supervision.

Indeed, just as in Bragg “the burden to the
defendant”—here, that parole and probation officers follow
mandatory reporting guidelines that they should be following
anyway—is low; and, unlike in the private rehabilitation
cases, the “consequences to the community of imposing a
duty to exercise care with resulting liability for breach” are
positive. See Myers, 144 Cal. App. 3d at 894 (“There also are
no significant practical problems involved here in requiring
physicians to warn patients not to engage in foreseeably
dangerous conduct. . . . Our holding does not require the
physician to do anything other than what he was already
obligated to do for the protection of the patient.”). As
described above, imposing liability would increase the
likelihood that officers will perform their duties, which in
turn would enhance public safety, which in turn would
increase the support for alternative rehabilitation programs.

Finally, as if more were needed, there is in this case a
specific party to warn which had control of the wrongdoer by
virtue of his status as a parolee: the Parole Commission.
And the Court here cannot say that, as a matter of law, such
a warning would have been futile.

To sum up then, the FTCA requires this Court to look to
the closest private sector analogue to determine if a private
44 DUGARD V. UNITED STATES

party would have a duty in like circumstances. The line of
cases we should look to is the “duty to control/duty to warn”
cases which represent the special relationship exception to the
general rule that there is no duty to third parties harmed by
the wrongdoer. That line of cases has its roots in Poncher
and Johnson. And as the cases have developed, an exception
to the special relationship exception has evolved where public
policy considerations demand parity for private rehabilitation
facilities with the immunity afforded to public facilities,
and/or where the warning suggested would be futile. Where
neither of these policy-based concerns is present, the courts
have continued to find a duty to control or warn, evidenced
by Bragg (control), Tarasoff, Myers, and Reisner (warn).
Here, the policy considerations undergirding the
rehabilitation center exception cases are plainly not present,
and the warning/control that Plaintiff claims should have been
made both would likely have been effective and would have
promoted exactly the behavior that tort law is meant to
promote: greater care, vigilance, and concern for the safety
of foreseeable victims.9

For all of these reasons, I would reverse the district
court’s grant of summary judgment.

9
The government also argues that Dugard’s claims are barred by the
discretionary function exception of the FTCA and that there is insufficient
evidence for a factfinder to determine that the parole officers’ alleged
failure to report Garrido’s drug violations to the Parole Commission
proximately caused Dugard’s harm. I agree with the district court’s
analysis and decision on these two issues and would find that neither
provides grounds to bar Dugard’s claims.