United States v. Russell

United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued January 12, 2010 Decided April 2, 2010

No. 08-3120

UNITED STATES OF AMERICA,
APPELLANT

v.

MARK WAYNE RUSSELL,
APPELLANT

Appeal from the United States District Court
for the District of Columbia
(No. 1:06-cr-00176-RBW-1)

Tony Axam Jr., Assistant Federal Public Defender, argued
the cause for appellant. With him on the briefs was A. J.
Kramer, Federal Public Defender. Neil H. Jaffee, Assistant
Federal Public Defender, entered an appearance.

Peter S. Smith, Assistant U.S. Attorney, argued the cause
for appellee. With him on the brief were Roy W. McLeese II1,
Mary B. McCord, and Julieanne Himelstein, Assistant U.S.
Attorney.

Before: HENDERSON and TATEL, Circuit Judges, and
WILLIAMS, Senior Circuit Judge.
 2

Opinion for the Court filed by Senior Circuit Judge
WILLIAMS.

Concurring opinion filed by Circuit Judge HENDERSON.

Concurring opinion filed by Senior Circuit Judge
WILLIAMS.

WILLIAMS, Senior Circuit Judge: Defendant Mark
Russell pleaded guilty to one count of travel with intent to
engage in illicit sexual conduct, in violation of 18 U.S.C.
§ 2423(b) (2006). The district court sentenced him to 46
months of imprisonment and 30 years of supervised release.
A special condition of his supervised release specifies that
Russell may not “possess or use a computer for any reason.”
Russell challenges the duration of his supervised release and
the computer restriction, arguing that each is substantively
unreasonable. See Gall v. United States, 552 U.S. 38, 51
(2007). We affirm the length of the supervised release, but
vacate the computer restriction and remand for resentencing.

* * *

In June 2006, using a computer at his home in Columbia,
Maryland, Russell entered an internet chat room and initiated
a conversation with an individual identifying herself as a 13-
year old girl; “she” was actually a member of the District of
Columbia Metropolitan Police Department. Three days later,
Russell again engaged the “child” in an online chat. Over the
course of their second chat, Russell performed a solo sex act
live via webcam and invited the “child” to have sex with him.
The purported child, in response, provided her address in
Washington, D.C., and said that her mother would not be
home until seven or eight that evening. Russell drove to the
address, parked his car, and e-mailed the “child” to say he had
 3

arrived. After a period of waiting, he began to drive away, at
which point he was arrested.

Russell was 46 when he was sentenced, and
approximately 50 at the time of his release. He had worked as
an applied systems engineer at Johns Hopkins University for
ten years before becoming unemployed at the end of April
2006. Before this arrest, he had had no contact with the law.
According to his wife of 23 years, the mother of their three
children, he had been depressed in the period just before his
arrest.

30-Year Term of Supervised Release. Russell challenges
the 30-year term of his supervised release as substantively
unreasonable. The parties initially spar over the proper
standard of review. The government, though acknowledging
that counsel posed an adequate objection to the conditions of
supervised release, contends that it did not embrace the term.
We assume in the government’s favor that Russell’s objection
in fact went only to the conditions.

We held in United States v. Bras, 483 F.3d 103 (D.C. Cir.
2007), that we review claims of substantive unreasonableness
for abuse of discretion, regardless of whether an objection on
those terms was made. Id. at 113. Noting that reasonableness
is simply “the standard of appellate review,” we quoted the
7th Circuit’s discussion of the point:

To insist that defendants object at sentencing to preserve
appellate review for reasonableness would create a trap
for unwary defendants and saddle busy district courts
with the burden of sitting through an objection—probably
formulaic—in every criminal case. Since the district
court will already have heard argument and allocution
from the parties and weighed the relevant § 3553(a)
factors before pronouncing sentence, we fail to see how
 4

requiring the defendant to then protest the term handed
down as unreasonable will further the sentencing process
in any meaningful way.

Id. (quoting United States v. Castro-Juarez, 425 F.3d 430,
433-34 (7th Cir. 2005)); see also United States v. Vonner, 516
F.3d 382, 389-90 (6th Cir. 2008) (no duty to object to
sentence on grounds of substantive unreasonableness). Such a
requirement would indeed yield a “formulaic” statement.
Substantive reasonableness is the catch-all criterion under
which the reviewing court monitors (deferentially—for abuse
of discretion) whether the district court has given reasonable
weight to all the factors required to be considered.
Presumably all (or virtually all) defendants would prefer a
shorter sentence, a shorter period of supervised release, and
less restrictive conditions. It would hardly alert the district
court to anything new for defense counsel to say that
defendant sought a more favorable sentence. By contrast,
unnoticed errors of the sort characterized by the Supreme
Court in Gall as procedural, “such as failing to calculate (or
improperly calculating) the Guidelines range, treating the
Guidelines as mandatory, failing to consider the § 3553(a)
factors, selecting a sentence based on clearly erroneous facts,
or failing to adequately explain the chosen sentence—
including an explanation for any deviation from the
Guidelines range,” 552 U.S. at 51, would normally be
reviewed for plain error. In re Sealed Case, 527 F.3d 188,
191-93 (D.C. Cir. 2008) (applying plain error review, in the
absence of objection at trial, to a sentencing judge’s failure to
provide an explanation of the sentence); see also Vonner, 516
F.3d at 386-88.

The government argues that in the absence of objection
we should apply plain error in reviewing a substantive
reasonableness challenge to the duration of supervised
release, citing United States v. Sullivan, 451 F.3d 884 (D.C.
 5

Cir. 2006). In Sullivan, we reviewed a challenge to the
substantive reasonableness of a condition of supervised
release for plain error. Id. at 894. See United States v. Love,
593 F.3d 1, 8 (D.C. Cir. 2010) (like Sullivan, applying plain
error review to substantive reasonableness challenge of a
condition of supervised release). Whatever the precedential
effect of Sullivan and Love may be with respect to
discretionary conditions, they do not address the applicable
standard with respect to the duration of supervised release.
On this question, we find the reasoning in Bras to be
persuasive and thus proceed under an abuse of discretion
standard.

Appellate review of the duration of supervised release
parallels review of a term of imprisonment. While in the
latter we inquire whether the district court abused its
discretion in applying the factors mandated by 18 U.S.C.
§ 3553(a), Gall, 552 U.S. at 56, here we ask whether the court
abused its discretion in applying the factors specified by 18
U.S.C. § 3583(c) for fixing a term of supervised release.
These are in fact simply a subset of those specified in
§ 3553(a), namely:

(1) the nature and circumstances of the offense and
the history and characteristics of the defendant;

(2) the need for the sentence imposed-- . . .

(B) to afford adequate deterrence to criminal
conduct;

(C) to protect the public from further crimes of the
defendant;

(D) to provide the defendant with needed
educational or vocational training, medical care, or
 6

other correctional treatment in the most effective
manner; . . .

(4) [the applicable Sentencing Guidelines range
based on the defendant’s offense and criminal
history];

(5) [pertinent policy statements issued by the
Sentencing Commission]; and

(6) the need to avoid unwarranted sentence
disparities among defendants with similar records
who have been found guilty of similar conduct; and

(7) the need to provide restitution to any victims of
the offense.

18 U.S.C. § 3553(a), cross-referenced in 18 U.S.C. § 3583(c).

The Sentencing Guidelines provide that violators of
§ 2423(b) (such as Russell) should receive a term of
supervised release ranging from three years to life. U.S.S.G.
§ 5D1.2(b)(2). In United States v. Law, 528 F.3d 888 (D.C.
Cir. 2008), we held that sentences within the applicable
Guidelines range are presumed reasonable. Id. at 902; see
Gall, 552 U.S. at 51 (allowing courts of appeal to apply such a
presumption); Rita v. United States, 551 U.S. 338, 350-51
(2007) (same).

Russell does not challenge Law directly, but argues that
Rita, in saying that an appellate presumption of
reasonableness for within-Guidelines sentences was
permissible under the Sixth Amendment, relied on logic that
is inapplicable here. He says that Rita is premised on an
assumption “that the Guidelines, insofar as practicable, reflect
a rough approximation of sentences that might achieve
§ 3553(a)’s objectives.” Appellant’s Br. at 15 (quoting Rita,
 7

551 U.S. at 350). Here that condition is absent, he thinks,
because the broad range of three years to life does not serve
§ 3553(a)(6)’s goal of “avoid[ing] unwarranted sentencing
disparities among defendants with similar conduct.” Id.

In embracing a presumption in favor of within-Guidelines
sentences, the Rita Court faced—and rejected—the objection
that such a presumption might so constrain sentencing judges’
discretion on the basis of judge-found facts as to deny
defendants their Sixth Amendment rights as construed by
Apprendi v. New Jersey, 530 U.S. 466 (2000), and its
progeny. See 551 U.S. at 350 (alluding to Justice Souter’s
dissent), id. at 391 (Souter, J., dissenting). Such an objection
would be at its minimum, perhaps nil, where the presumption
is applied to so broad a sentencing range (and one based
entirely on the offense of conviction, quite independent of
judge-found facts).

Affirmatively, the Rita Court found a presumption of
reasonableness to be proper because the trial judge and the
Sentencing Commission are both directed to apply the same
statutory factors, “the one, at retail, the other, at wholesale.”
551 U.S. at 348. If the trial judge’s judgment falls within “the
Commission’s view of the appropriate application of
§ 3553(a) in the mine run of cases, it is probable that the
sentence is reasonable.” Id. at 351.

When the Commission specifies a broad sentencing
range, that justification does not disappear—though it is of
course weakened. And the breadth in § 5D1.2(b) can be
reconciled with § 3553(a)(6)’s interest in avoiding
unwarranted disparities by supposing, not unreasonably, that
the Commission believed either that variability among
violators of § 2423(b) made a narrower range unsuitable, or
that allowing considerable subjectivity among sentencing
judges might (for now at least) be reasonable, or both. See
 8

Rita, 551 U.S. at 349 (acknowledging that uniformity may
conflict with the goals of “proportionality in sentencing
through a system that imposes appropriately different
sentences for criminal conduct of different severity”
(emphasis omitted)).

Russell notes the existence of a Commission Policy
Statement recommending that sex offenders receive the
maximum statutory term of supervised release, U.S.S.G.
§ 5D1.2(b)(2) (Policy Statement), but does not challenge the
effect the statement may have had (if any) in causing the
district court’s selection of a term close to the high end of the
range. To the extent that courts, both trial and appellate, are
obliged to give at least Chevron deference to such statements,
see United States v. Anderson, 82 F.3d 436, 439, 442-43, 446
(D.C. Cir. 1996) (treating particular policy statement as
worthy of at least Chevron deference), this statement’s very
narrow range—indeed, not a range but a point—would give
Russell a stronger Apprendi argument vis-à-vis Rita’s
willingness to allow courts to presume an in-Guidelines
sentence to be reasonable, at least it would if the statement
were triggered by judge-found facts. But he makes no
arguments with reference to the Policy Statement. We thus do
not address its possible vulnerability vis-à-vis Rita’s reasoning
or the Guideline itself.

Analysis of the term of supervised release under the
remaining § 3553(a) factors made applicable by § 3583(c)
does not rebut the presumption established by the within-
Guidelines sentence. The Policy Statement appended to
U.S.S.G. § 5D1.2(b)(2), as we just noted, recommends that
violators of § 2423(b) be subject to supervised release for life.
Coupled with the computer restriction, to be sure, the duration
of the sentence might conflict with § 3553(a)(2)(D)’s
rehabilitative goals, but we analyze that restriction separately
under 18 U.S.C. § 3583(d)’s criteria, which are designed
 9

specially to govern conditions; for the reasons given below,
we vacate the restriction. Given the policy statement and the
relaxation of the computer restriction, the remaining § 3553(a)
factors leave intact the presumption in favor of the within-
Guidelines sentence. See, e.g., Gall, 552 U.S. at 51-52.

Computer Restriction. Russell’s second challenge is to
the special condition of his supervised release, providing that
he “shall not possess or use a computer for any reason.” The
government and Russell agree that the 30-year prohibition on
the possession and use of computers—a prohibition not
subject to modification by the probation office—is
substantively unreasonable. Brief for Appellee 19; Brief for
Appellant 18. Despite the government’s concession, the
Supreme Court tells us that we are to conduct an independent
review of the issue:

The considered judgment of the law enforcement officers
that reversible error has been committed is entitled to
great weight, but our judicial obligations compel us to
examine independently the errors confessed.

Young v. United States, 315 U.S. 257, 258-59 (1942); see also
Roberts v. Galen of Virginia, Inc., 525 U.S. 249, 253 (1999)
(per curiam) (“Although the concession of a point on appeal
by respondent is by no means dispositive of a legal issue, we
take it as further indication of the correctness of our
decision.”). While Young and Roberts involved issues of law,
courts of appeals have also applied this doctrine to issues of
fact. See, e.g., United States v. Cooke, 110 F.3d 1288, 1294
(7th Cir. 1997). We can see no basis for drawing the line at a
mixed issue of law and fact. Thus, despite the tension with
Article III’s limitation of our jurisdiction to “cases” and
“controversies,” and the government’s de facto power to yield
in a defendant’s favor by a timely exercise of its prosecutorial
discretion (a tension that of course may also inhere in rules
 10

such as those giving district courts authority to reject pleas),
we proceed with our review.

Sections 3583(c) and 3583(d) govern the conditions
attached to a term of supervisory release. As we have seen,
§ 3583(c) requires that a sentencing court, in deciding the
duration and conditions of the supervised release, “consider
the factors set forth in section 3553(a)(1), (a)(2)(B), (a)(2)(C),
(a)(2)(D), (a)(4), (a)(5), and (a)(6).” And, where such a term
is imposed, 18 U.S.C. § 3583(d) specifies certain conditions
and grants the district court discretion to impose additional
ones, so long as each such condition:

(1) is reasonably related to the factors set forth in section
3553(a)(1), (a)(2)(B), (a)(2)(C) and (a)(2)(D);

(2) involves no greater deprivation of liberty than is
reasonably necessary for the purposes set forth in section
3553(a)(2)(B), (a)(2)(C), and (a)(2)(D); and

(3) is consistent with any pertinent policy statements
issued by the Sentencing Commission pursuant to 28
U.S.C. 994 (a).

18 U.S.C. § 3583(d). Because independent review shows that
the challenged condition deprives the defendant of
substantially more liberty than is “reasonably necessary for
the purposes set forth in section 3553(a)(2)(B), (a)(2)(C), and
(a)(2)(D),” it must be vacated.

Subsections 3553(a)(2)(B) and (C) codify the penal goals
of general and specific deterrence, requiring disincentives to
match the severity of punishment to the harmfulness of the
crime. Had the “child” been an actual minor and had Russell
proceeded to have sex with her, the harm would have been
great. Thus a stiff condition that deters would-be molesters
and prevents Russell from reoffending may be appropriate.
 11

This court has observed that the harm inflicted when someone
uses a computer or the internet to arrange for sex with a minor
is generally greater than the harm inflicted when someone
uses the internet to trade child pornography. Love, 593 F.3d
at 12. Because of the relative harm, computer and internet
restrictions “while perhaps unreasonably broad for defendants
who possess or distribute child pornography, may be
appropriate for those who use the Internet to ‘initiate or
facilitate the victimization of children.’” Id.

Here, however, the question is not the appropriateness of
an internet restriction but its form and severity. A condition
that takes into account only the magnitude of the harm from
defendant’s offense may serve the general deterrence goals of
§ 3553(a)(2)(B), but may be more or less severe than required
to serve § 3553(a)(2)(C)’s goal of preventing this defendant
from repeating his offense. The sentence already achieves
considerable severity by its 30-year term and several other
conditions, e.g., requirements that Russell register as a sex
offender in any jurisdiction where he resides and not be in the
presence of anyone under the age of 18 in a private setting
without another adult present.

More important for analysis under § 3583(d)(2), the
computer restriction affirmatively and aggressively interferes
with the goal of rehabilitation. See, e.g., United States v.
Holm, 326 F.3d 872, 878 (7th Cir. 2003). It is hard to imagine
white collar work in 2010 not requiring access to computers,
just as white collar work 100 years ago would almost
invariably have required the use of pens and pencils. In fact
Russell’s training and experience mark him not only as a
white collar worker but as one at the most technically
sophisticated end of the white collar distribution. He holds a
Bachelors of Science degree in engineering and a Masters
degree in Strategic Intelligence, see Presentence Investigation
Report at 8, and his 10 years as an applied systems engineer at
 12

Johns Hopkins suggest a work life fitted to the skills so
acquired. Even a lot of blue collar work requires some
computer use. Although we cannot rely on evidence
developed since the sentence, it is totally unsurprising in the
realities of the modern world that in his post-release search for
employment Russell has evidently found that computer use is
required for filling out most job applications, including those
at McDonald’s, as well as discharging the duties of even low
tech occupations, such as keeping inventory at PETCO, and
producing frames at A.C. Moore. Oral Argument Recording
at 30:45-31:40. See United States v. Voelker, 489 F.3d 139,
148-49 (3rd Cir. 2007) (explaining that a restriction
prohibiting defendant, who worked as a respiratory therapist
before his arrest, from using computer equipment was overly
burdensome because his employment “‘necessarily entails
access to and the use of computers and computer equipment
for record keeping [and] patient care’”). Because the
computer restriction prevents Russell from continuing in a
field in which he has decades of accumulated academic and
professional experience, it directly conflicts with the
rehabilitative goal of sentencing. It also, of course, places a
substantial burden on Russell’s liberty, which under 18 U.S.C.
§ 3583(d)(2) must be no greater than reasonably necessary to
achieve the goals of deterrence as well as rehabilitation.

The district court’s restriction is scheduled to elapse more
than three decades after sentencing. A provision for
modification by the probation department—a minimum
change suitable on remand—would allow the restriction to
adjust to ongoing developments in technology and to secure a
reasonable balance between the statute’s rehabilitative and
deterrence goals. See Love, 593 F.3d at 12. Given the ample
room for adjusting the sentence to enable a better balance
among those goals, the computer restriction in its current form
is substantively unreasonable.
 13

The government’s concession, of course, is a further
indication, under Young and Roberts, supporting our
conclusion that the district court abused its discretion in
imposing an unqualified prohibition on Russell’s use of
computers.

The experience of other courts also supports our
conclusion. We have found only one case, United States v.
Paul, 274 F.3d 155 (5th Cir. 2001), that upholds, against
proper challenge, a categorical prohibition on computer
possession or use without provision for probation office
modification. Not only is Paul an outlier, but in key respects
it represents a far stronger case for blanket restriction. Most
obviously, Paul’s computer restriction was to last three years,
or one tenth of the duration of Russell’s—a difference that
makes Paul’s restriction both less burdensome and less likely
to become a still poorer fit over time. Moreover, Paul had
suggested no way in which the computer and internet ban
would adversely affect his occupational prospects, id. at 170,
whereas Russell made clear that his rehabilitation would
depend at least in part on his ability to apply his professional
training, which in turn would depend on his being able to
access computers. Although Paul had pled guilty only to
possessing child pornography that traveled through interstate
commerce, in violation of 18 U.S.C. § 2252A(a)(5)(B), id. at
158, the record showed that Paul had engaged in a sustained,
extensive and sophisticated pattern of sexual predation over
many years, continuously aided and magnified by the internet.
He had, for example, given e-mail advice (presumably
reflecting experience) on how to “find ‘young friends’ by
scouting single, dysfunctional parents through Alcoholics
Anonymous or local welfare offices and winning their
friendship, thereby securing access to their young sons.” Id.
at 158. And he had a prior child pornography conviction. Id.
Finally, of course, the computer and internet have permeated
 14

everyday life in ways that make a restriction on their use far
more burdensome than when Paul was decided.

We have found no instance other than Paul where a court
has upheld a prohibition on the defendant’s use of computers
or the internet that was not subject to relaxation by the
probation office. Many cases have upheld restrictions subject
to such relaxation. Love, 593 F.3d at 11-13; Sullivan, 451
F.3d at 892-896; United States v. Bender, 566 F.3d 748, 751-
52 (8th Cir. 2009); United States v. Lay, 583 F.3d 436, 449-50
(6th Cir. 2009); United States v. Thielemann, 575 F.3d 265,
278 (3rd Cir. 2009); United States v. Alvarez, 478 F.3d 864,
866-68 (8th Cir. 2007); United States v. Johnson, 446 F.3d
272 (2nd Cir. 2006); United States v. Crandon, 173 F.3d 122,
125 (3rd Cir. 1999). In several of these cases defendants’
conduct was more egregious than Russell’s—he had either
completed sex acts with a child (Bender, Alvarez, Johnson,
and Crandon), or caused another to do so in order to obtain
images of the conduct (Thielemann), or took more drastic
steps toward completion of the acts than did Russell (Lay
(defendant flew to site of intended rendezvous after
developing a ruse to separate the minor from his or her mother
for the weekend)). In contrast, the courts have generated a
large universe of decisions rejecting such unmodifiable
restrictions, typically invoking (as here) the public interest in
the defendant’s rehabilitation through the productive use of
his or her skills. See, e.g., Voelker, 489 F.3d at 144-50;
United States v. Mark, 425 F.3d 505, 508-11 (8th Cir. 2005);
Holm, 326 F.3d at 877-78; United States v. Sofsky, 287 F.3d
122, 124 (2nd Cir. 2002); United States v. White, 244 F.3d
1199, 1206 (10th Cir. 2001). In at least one such case, White,
the defendant’s conduct, unlike Russell’s, manifested a course
of completed child molestations. Cf. United States v.
Perazza-Mercado, 553 F.3d 65, 69-74 (1st Cir. 2009)
(vacating a restriction prohibiting defendant from using the
internet at his home for the fifteen year period of his
 15

supervised release, where his molestation of a nine-year old
with special needs whom he supervised had not involved
internet use).

Conclusion

While we affirm the 30-year term of supervised release,
we vacate the computer restriction and remand for
proceedings consistent with this opinion.

So ordered.
KAREN LECRAFT HENDERSON, Circuit Judge, concurring:
Although I agree with my colleagues’ decisions to affirm
the duration of Russell’s supervised release and to remand the
computer ban for refinement, I write separately to distance
myself on various points. First, I do not share the majority’s
doubt regarding “the precedential effect,” Maj. Op. at 5, of
United States v. Sullivan, 451 F.3d 884 (D.C. Cir. 2006), and
United States v. Love, 593 F.3d 1 (D.C. Cir. 2010). In Sullivan,
the “Appellant urge[d] us to review the substantive validity of
the [unobjected-to] terms of supervised release for abuse of
discretion” but “[w]e reject[ed] this argument,” holding that
“[t]he proper standard of review [t]here [was] plain error.” 451
F.3d at 894. Likewise, Love applied plain error review to four
unobjected-to conditions of supervised release, twice explaining
its application of that standard. 593 F.3d at 11, 14. It first relied
on United States v. Breedlove, 204 F.3d 267 (D.C. Cir. 2000),
and Fed. R. Crim. P. 52(b) in support thereof and later cited
Sullivan to the same effect. Id. I do not know what else is
needed to decide the scope-of-review question. Nor should
United States v. Bras, 483 F.3d 103 (D.C. Cir. 2007), eclipse the
precedential effect of Sullivan and Love. Williams Concurrence
at 1-2. Bras did not address Sullivan, perhaps because Bras
dealt with a term of imprisonment rather than conditions of
supervised release. 483 F.3d at 104. And Bras preceded Love.
Thus, while Sullivan and Love are distinguishable because “they
do not address the applicable standard with respect to the
duration of supervised release,” Maj. Op. at 5 (emphasis added),
they nevertheless remain good law regarding unobjected-to
conditions of supervised release. And on the latter point,
Sullivan and Love are of course distinguishable because Russell
did challenge the disputed condition of supervised release at
sentencing.1

1
Regarding that condition, whether a “tension” results, Maj. Op.
at 9, when, notwithstanding the government’s concession, we proceed
with our review is irrelevant—as the Supreme Court reminded us long
 2

Second, I do not share my concurring colleague’s faith in
the isolated Justice Department data he cites. Williams
Concurrence at 2-3. Those figures do not sway me. Assuming
their accuracy, lower recidivism rates for sex offenders can just
as easily be explained by the tight leash judges often specify for
sex offenders once released from incarceration, as my colleague
himself recognizes. Id. at 3-4.2
And finally, I am unwilling to subscribe to the notion that
a restriction (or ban) on a criminal defendant’s computer use, at
least where the computer enables the crime, constitutes “a

ago, our duty of review is independent and mandatory, Young v.
United States, 315 U.S. 257, 258-59 (1942); see also United States v.
Escobar Noble, 653 F.2d 34, 36 (1st Cir. 1981) (“It is the particular
function of the court, not the prosecutor, to say the last word about the
justice of a sentence.”).
2
In Love, for example, we recognized a nation-wide move
towards broad restrictions on child predators:
Consensus is emerging among our sister circuits that
Internet bans, while perhaps unreasonably broad for
defendants who possess or distribute child pornography,
may be appropriate for those who use the Internet to
“initiate or facilitate the victimization of children.” [United
States v.] Holm, 326 F.3d [872, 878 (7th Cir. 2003)]; see
United States v. Thielemann, 575 F.3d 265, 278 (3d Cir.
2009); United States v. Johnson, 446 F.3d 272, 283 (2d Cir.
2006); United States v. Boston, 494 F.3d 660, 668 (8th Cir.
2007); United States v. Paul, 274 F.3d 155, 169 (5th Cir.
2001). The distinction is grounded in the simple proposition
that when a defendant has used the Internet to solicit sex
with minors, “the hazard presented by recidivism” is greater
than when the defendant has traded child pornography.
Johnson, 446 F.3d at 283.
593 F.3d at 12.
 3

substantial burden” on liberty, Maj. Op. at 12. A defendant
convicted of vehicular homicide can permanently lose his
driving privilege and the resulting ban on his use of an
automobile—which, like Russell’s computer, enabled the
crime—does not deprive him of his liberty. That Russell’s
white collar career may be adversely affected by the computer
ban—a result the majority supports with anecdotal predictions,
Maj. Op. at 12—does not ipso facto translate into a deprivation
of liberty. We can judicially note that millions of Americans
every day perform jobs without using (or even seeing) a
computer. If Russell cannot find a job, it is more likely because
of his criminal record than the computer ban. While I do not
believe the thirty-year computer ban implicates Russell’s liberty,
I nevertheless recognize that the weight of authority is to the
contrary. For that reason, I join in the remand for the district
court to again exercise its discretion in refining3 the computer
ban condition of Russell’s supervised release.

3
I note that, in the event the district court permits circumscribed
computer use, 18 U.S.C. § 3583(d) expressly provides:
The court may order, as an explicit condition of supervised
release . . . that [Russell] submit his . . . computer, other
electronic communications or data storage devices or media,
and effects to search at any time, with or without a warrant,
by any law enforcement or probation officer with reasonable
suspicion . . . and by any probation officer in the lawful
discharge of the officer’s supervision functions.
 WILLIAMS, Senior Circuit Judge, concurring: I write
separately to elaborate on two issues: the scope of review and
the probability of recidivism.

Scope of review. Our opinion refers to two cases, United
States v. Sullivan, 451 F.3d 884 (D.C. Cir. 2006), and United
States v. Love, 593 F.3d 1 (D.C. Cir. 2010), which applied a
plain error standard of review to a defendant’s challenge of a
discretionary condition of supervised release after he had
failed to lodge an objection at sentencing. But in neither of
these cases did the defendant argue or the court address
whether a plain error standard of review is appropriate in the
absence of an objection. The defendant in Sullivan
“acknowledge[d] that, ‘generally, where a sentencing court
affords the defendant an opportunity to object to the special
conditions but the defendant remains silent, an appellate court
reviews for plain error’.” 451 F.3d at 894 (internal brackets
omitted). Of course he could not have cited our decision in
United States v. Bras, 483 F.3d 103 (D.C. Cir. 2007), which
hadn’t yet issued; nor did he anticipate Bras’s argument (or
any others) on the background question of the need for an
objection vel non. Rather, he argued that the circumstances
under which the challenged condition was imposed—the trial
court announced the discretionary condition “after resolving
objections to the PSR and after permitting defendant to make
a statement,” he “ha[d] no meaningful opportunity to
comment,” 451 F.3d at 894 (emphasis in the original)—
prevented him from making an objection. We rejected that
specific argument, observing that “once appellant was made
aware of the conditions . . . his counsel was in a position to
respond or to seek additional time.” Id. But we did not
consider the logically anterior question whether plain error
review was suitable for a substantive reasonableness challenge
not posed at sentencing.
 2

Love, appealing after Bras had issued, failed to invoke it;
nor did he in any way raise the threshold question of whether
plain error review is appropriate absent an objection. Rather,
he argued that he had made an adequate objection. We
rejected the argument, finding that counsel’s words fell short
of the specificity requirement expressed in United States v.
Breedlove, 204 F.3d 267 (D.C. Cir. 2000). Love, 593 F.3d at
11.

Thus the decisions in Sullivan and Love both lacked the
benefit of adversarial briefing on the issue. Neither case
considered whether, as substantive reasonableness is a
standard of appellate review rather than a direct mandate to
trial courts, plain error review would be appropriate. Bras,
483 F.3d at 113. Neither case considered other arguments as
to the likely impracticality of requiring an objection. See Maj.
Op. at 4.

As the Sullivan and Love courts never addressed the
assumption that plain error review was appropriate in the
absence of an objection, the courts’ implicit assumptions,
under standard principles, lack precedential effect. Brecht v.
Abrahamson, 507 U.S. 619, 630-31 (1993). It remains for a
future case to determine whether the standard of review
applicable to a challenge of a discretionary condition as
substantively unreasonable should differ from the standard
applicable to terms of incarceration under Bras.

Risk of recidivism. At sentencing Russell argued that he
was unlikely to repeat his offense, pointing to Department of
Justice data on the recidivism rates for various types of
offenders, including child molesters. The data, which Russell
never mentioned in his appellate briefs, are quite interesting
and seem to place child molesters at the low end of the
distribution. One report, U.S. Department of Justice: Bureau
of Justice Statistics, Recidivism of Prisoners Released in 1994
 3

Table 10 (June 2002), available at http://bjs.ojp.usdoj.gov/
content/pub/pdf/rpr94.pdf, provides the percentage of
approximately 272,111 inmates released in 1994 in 15 states
who were rearrested within three years of release for the type
of crime for which he/she was imprisoned, namely, homicide,
rape, robbery, assault, burglary, larceny/theft, motor vehicle
theft, fraud, drug offenses, and public order offenses. The
numbers for those categories are, respectively, 1.2, 2.5, 13.4,
22.0, 23.4, 33.9, 11.5, 19.0 , 41.2 and 31.2 percent. But
another report, Bureau of Justice Statistics, Recidivism of Sex
Offenders Released from Prison in 1994 Table 22 (Nov.
2003), available at http://bjs.ojp.usdoj.gov/content/
pub/pdf/rsorp94.pdf, shows the analogous rates of recidivism
for child molesters to be lower than all but two categories of
that set of offenders, and radically lower than most. It reports
that of the 4,295 inmates convicted of child molestation who
were released in 1994, only 5.1 percent were rearrested for
any sex crime within three years.

The DOJ statistics also have implications for Russell’s
age and his lack of prior offenses. The rearrest rate for child
molesters with no arrests prior to the one leading to their
imprisonment was about half that for those with a prior arrest
for any crime. See id. at Table 28. And the rearrest rate for
persons 45 or over was only about 60 percent of the average
rate. See id. at Table 25.

On appeal Russell does not argue that the sentencing
court committed the procedural error of relying on erroneous
facts when assessing the risk of recidivism, Gall v. United
States, 552 U.S. 38, 51 (2007), let alone raise the Justice
Department statistics, thereby depriving the government of a
chance to argue why the data may not be as telling as they
appear. It may be, for example, that assiduous supervision
under judgments such as the one under review, as well as
general legislation constraining the offenders’ movement and
 4

sensitizing neighbors (e.g., the various versions of Megan’s
Law adopted by every state and by Congress by 1996, see
Smith v. Doe, 538 U.S. 84, 89-90 (2003)), have played a key
role in producing the low rates shown in DOJ’s statistics.

In any event, given Russell’s failure to press the Justice
Department figures, the other applicable § 3553(a) factors and
the deference owed to the trial court, see, e.g., Gall, 552 U.S.
at 51-52, I am unable to find the 30-year term of supervised
release to be substantively unreasonable.