United States v. Joseph Smith

United States Court of Appeals

FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued September 7, 2023 Decided July 23, 2024

No. 22-3015

UNITED STATES OF AMERICA,

APPELLEE

v.

JOSEPH SMITH,

APPELLANT

Appeal from the United States District Court

for the District of Columbia

(No. 1:19-cr-00324-1)

Jonathan Zucker, appointed by the court, argued the cause

and filed the briefs for appellant.

David B. Goodhand, Assistant U.S. Attorney, argued the

cause for appellee. With him on the brief were Chrisellen R.

Kolb and John P. Mannarino, Assistant U.S. Attorneys.

Before: SRINIVASAN, Chief Judge, GARCIA, Circuit Judge,

and RANDOLPH, Senior Circuit Judge.

Opinion for the Court filed by Chief Judge SRINIVASAN.

2

SRINIVASAN, Chief Judge: Appellant Joseph Smith was

convicted of child sexual abuse and other related offenses after

sexually abusing his stepdaughter. In this appeal, Smith brings

four challenges to his convictions. First, he contends that an

underrepresentation of Black residents in his jury pool violated

his Sixth Amendment right to a jury drawn from a fair cross-

section of the community. Second, he challenges the district

court’s denial of his motion to suppress evidence discovered on

two cell phones and a personal computer. Third, he asserts that

the government’s case agent should have been excluded from

the courtroom. And fourth, he argues that the case agent

improperly testified as an expert at trial. We are unpersuaded

by any of those arguments and thus affirm Smith’s convictions.

I.

A.

In May 2016, Joseph Smith began sexually abusing A.S.,

his stepdaughter, when she was twelve years old. For eleven

months, Smith forced A.S. to receive oral sex from and perform

oral sex on him. Smith also sent A.S. sexually explicit text

messages and forced her to send nude photos of herself to him.

In April 2017, A.S. and her mother reported Smith’s abuse to

the police.

Police obtained a warrant to search Smith’s residence for

evidence of A.S.’s allegations. The affidavit supporting the

warrant relied on A.S.’s statements describing her text

messages with Smith and the photos she had sent him. The

affiant, a detective specializing in child sex abuse, additionally

averred based on her experience that child sexual abusers often

use their cell phones to take and store pictures of victims and

then save the pictures to their personal computers. The affiant

explained that those images would be “excellent evidence of

3

someone who is engaged in committing sexual offenses against

children.” J.A. 65.

As requested in the affidavit, the warrant authorized a

search for, and seizure of:

Cellular phones, computers, digital storage

devices, thumb drives, removable electronic

devices such as external hard drives, and the

extraction of all electronic data stored inside of

them to take place at the residence or a police or

court facility, mail matter, any material

identifying any resident of the house and to take

photographs and sketches of the entire

premises, and any items or materials relating to

the offense of First Degree Child Sexual Abuse.

J.A. 62.

When executing the warrant, the officers seized three

tablets, an Xbox, an air mattress, a personal computer, and

twelve cell phones. Police discovered substantial amounts of

incriminating evidence on the personal computer and two of

the cell phones.

B.

A grand jury indicted Smith on ten counts related to child

sexual abuse under federal and D.C. law. Four of Smith’s

pretrial and trial motions are at issue in this appeal. The district

court denied all four motions.

First, Smith moved to dismiss the indictment based on his

Sixth Amendment right to a jury drawn from a fair cross-

section of the community. He pointed to statistical evidence

that Black persons were underrepresented in Washington,

4

D.C., jury pools relative to the percentage of Black adults in

the D.C. population. Smith asserted that the disproportionate

impact of the COVID-19 pandemic on racial and ethnic

minorities caused the disparity in the jury pool.

Second, Smith moved to suppress the evidence found on

the computer and two cell phones. He argued that police had

unconstitutionally seized those devices while executing an

invalid warrant to search his home, and that any evidence

discovered on the devices thus should have been excluded at

trial.

Finally, Smith brought two challenges related to the

government’s case agent, a special agent in the Federal Bureau

of Investigation who testified against Smith. Smith first moved

to exclude the agent from the courtroom to prevent her from

hearing the testimony of other witnesses. Smith also separately

objected to a portion of the agent’s trial testimony in which she

reviewed text message exchanges with A.S. found on Smith’s

cell phone. The agent explained which messages Smith sent

and which he received based on her interpretation of a report

from a program called Cellebrite, which is used to extract

information from digital devices. Smith moved to strike the

agent’s testimony as improper expert testimony.

A jury convicted Smith of seven counts of child sexual

abuse, as well as one count each of production of child

pornography, possession of child pornography, and enticement

of a minor. The district court sentenced Smith to two

concurrent terms of life imprisonment.

II.

On appeal, Smith challenges the district court’s denial of

the four motions described above. We reject each of Smith’s

challenges.

5

A.

We first consider Smith’s Sixth Amendment challenge to

the composition of the jury pool. The Sixth Amendment

guarantees a criminal defendant the right to a trial “by an

impartial jury,” U.S. Const. amend. VI, which the Supreme

Court has held must be drawn from a “representative cross-

section of the community,” Taylor v. Louisiana, 419 U.S. 522,

528 (1975) (quoting Williams v. Florida, 399 U.S. 78, 100

(1970)). Smith claims that his Sixth Amendment fair cross-

section right was violated because Black residents were

underrepresented in the jury pool from which his jury was

drawn.

In order to establish a prima facie violation of his fair

cross-section right, Smith must satisfy all three of the prongs

set out by the Supreme Court in Duren v. Missouri, 439 U.S.

357, 364–66 (1979). He must show: (i) that the group

allegedly excluded (here, Black persons) qualifies as a

“‘distinctive’ group in the community”; (ii) that the

representation of the group in jury venires “is not fair and

reasonable in relation to” the group’s representation in the

community; and (iii) that the underrepresentation stems from

“systematic exclusion of the group in the jury-selection

process.” Id. at 364.

The district court held that Smith established the first

Duren prong but not the second or third. We affirm based on

the third prong: we conclude that Smith cannot show that the

jury-selection process systematically excluded Black residents.

We therefore have no need to address the second prong or to

resolve how to determine the baseline population or measure

underrepresentation for purposes of that prong.

To understand why Smith has failed to demonstrate

systematic exclusion of Black residents in the jury-selection

6

process, it is necessary to outline how that process works for

the United States District Court for the District of Columbia.

The District’s Jury Office initially constructs a master jury

wheel from lists of people who: are registered to vote in D.C.;

hold a D.C. driver’s license, D.C. learner’s permit, or other

valid D.C. identification card; or pay D.C. income taxes. From

the master jury wheel, the Office periodically draws sets of

potential jurors for two-week windows of trial start dates. Each

of those potential jurors receives a summons and juror-

qualification questionnaire in the mail. Some share of those

potential jurors responds, and the Office does not follow up

with (or take any action against) those who do not respond.

Based on the responses to the questionnaires, the Office

filters out people who are disqualified or excused from jury

service. The remaining group of eligible jurors is called the

qualified two-week jury pool. When there is a trial, the Office

instructs a portion of the people in the qualified two-week jury

pool to appear at the courthouse for jury selection. From that

group, a venire of the size requested by the presiding judge is

randomly drawn. Voir dire then occurs, yielding a jury of

twelve jurors and two alternates.

Smith contends that, around the time of his trial, Black

residents responded to the Jury Office’s summonses and

questionnaires at lower rates than other groups and thus were

underrepresented in the qualified two-week jury pools. In his

view, because the jury-selection process allows disparate

response rates to affect the composition of the qualified two-

week jury pools, the process systematically excludes Black

jurors. Smith appears to allege both that the COVID-19

pandemic caused the differential response rate and that the fact

of the differential response rate alone suffices regardless of the

reason.

7

Either way, Smith cannot demonstrate the existence of

systematic exclusion within the meaning of Duren’s third

prong. As the Supreme Court explained in Duren, the cause of

underrepresentation is “systematic” when it is “inherent in the

particular jury-selection process utilized.” Id. at 366. Neither

of Smith’s theories involves systematic exclusion of that kind.

Smith’s first theory involves the COVID-19 pandemic,

which of course profoundly affected many aspects of day-to-

day life. According to Smith, one of those effects bore on the

jury-selection process, in that the pandemic depressed response

rates among Black residents, giving rise to nonrepresentative

jury pools.

Even assuming the pandemic brought about differential

response rates, however, that is not “systematic exclusion”

under Duren. The pandemic was an exogenous shock rather

than something “inherent in the . . . jury-selection process.” Id.

Indeed, Smith acknowledges that the jury-selection process

“was carefully calibrated to produce a fair cross-section of the

community” and that the COVID-era data “does not resemble”

the process’s intended results. Defendant’s Motion to Dismiss

the Indictment at 7, United States v. Smith, No. 19-cr-00324

(D.D.C. Oct. 18, 2021).

To the extent Smith’s challenge encompasses differential

response rates more generally, he still has not shown systematic

exclusion in the jury-selection process. Smith alleges that

Black residents respond to jury summonses at lower rates than

other groups. Even if that is so, the resulting

underrepresentation is not “due to [Black residents’] systematic

exclusion in the jury-selection process.” Duren, 439 U.S. at

366. It is instead due to the independent choices of potential

jurors—here, choices about whether to respond to a jury

8

summons. Those sorts of autonomous choices are not

“inherent in the particular jury-selection process utilized.” Id.

In Duren, by contrast, the Supreme Court found systematic

exclusion of women when the jury-selection process offered

certain opportunities to claim exemptions from service only to

women and presumed that women (but not men) who failed to

respond had claimed exemptions. The resulting

underrepresentation was “quite obviously due to the system by

which juries were selected.” Id. at 367. That is untrue when

underrepresentation results from the independent choices of

potential jurors rather than from, as in Duren, “the operation of

[the jury-selection process’s] exemption criteria.” Id.

Smith also asserts that the Jury Office systematically

excludes Black jurors because it fails to follow up on

nonresponses or enforce summonses against nonrespondents.

But Smith does not explain why Black residents respond at

lower rates, why subsequent action by the Office would

ameliorate (rather than cement) the disparity, or how many

additional steps the Office should be required to take to satisfy

the Sixth Amendment. Smith, in other words, has provided

insufficient evidence that the Office in fact could remedy the

disparities in jury representation by following up on

nonresponses or that it would be reasonable to require the

Office to do so. In those circumstances, we have no basis to

impose an obligation on the Office to take further measures that

may or may not mitigate differential response rates or to

conclude that the Office’s failure to take those measures

constitutes systematic exclusion.

For those reasons, we affirm the district court’s conclusion

that Smith has failed to show a violation of his Sixth

Amendment fair cross-section right.

9

B.

We next consider Smith’s challenge to the warrant

authorizing the search of his apartment. The Fourth

Amendment provides that a warrant must “particularly

describ[e] . . . the persons or things to be seized.” U.S. Const.

amend. IV. The particularity requirement “ensures that the

search will be carefully tailored to its justifications, and will

not take on the character of the wide-ranging exploratory

searches the Framers intended to prohibit.” Maryland v.

Garrison, 480 U.S. 79, 84 (1987). To that end, a warrant with

an “indiscriminate sweep” is “constitutionally intolerable.”

Stanford v. Texas, 379 U.S. 476, 486 (1965).

The warrant in this case authorized police to search for and

seize “[c]ellular phones, computers, digital storage devices,

thumb drives, removable electronic devices such as external

hard drives, and the extraction of all electronic data stored

inside of them to take place at the residence or a police or court

facility, mail matter, any material identifying any resident of

the house and to take photographs and sketches of the entire

premises, and any items or materials relating to the offense of

First Degree Child Sexual Abuse.” J.A. 62. Smith contends

that the warrant was unconstitutionally overbroad in violation

of the Fourth Amendment’s particularity requirement.

Smith relies largely on our decision in United States v.

Griffith, 867 F.3d 1265 (D.C. Cir. 2017). In Griffith, we held

that a warrant “to search for and seize all electronic devices”

(including cellular phones and computers) at a residence was

insufficiently particular. Id. at 1276–77. The circumstances in

Griffith, though, differed meaningfully from those here.

In Griffith, the warrant affidavit gave no reason to suppose

that the suspect owned a cell phone (or other electronic device)

at all, and there was also a “limited likelihood that any cell

10

phone discovered in the apartment would contain incriminating

evidence of Griffith’s suspected crime.” Id. at 1272–75. In

those circumstances, we held that it was impermissible to issue

a warrant granting officers unfettered access to every electronic

device in the apartment.

Here, by contrast, police had ample cause to believe that

multiple devices containing incriminating evidence would be

found in Smith’s apartment. Smith’s suspected conduct

included exchanging sexually abusive text messages and

photos with A.S., which undoubtedly involved multiple

electronic devices: namely, the cell phones A.S. and Smith

used to communicate with each other. Contra id. at 1272

(noting that there was “no information about anyone having

received a cell phone call or text message from” the suspect).

A.S. confirmed as much in her statements to investigators,

when she identified multiple phones that she said had been used

to carry out the alleged offenses. And the affidavit supporting

the warrant incorporated the information provided by A.S. to

establish probable cause that Smith’s cell phone and A.S.’s cell

phone would contain evidence of A.S.’s allegations.

Moreover, the affiant relied on her experience investigating

child sexual abuse to provide a detailed account of why and

how a suspected abuser would use his personal computer and

cell phone to perpetrate his offense. See United States v.

Cardoza, 713 F.3d 656, 661 (D.C. Cir. 2013) (finding probable

cause based in part on an affiant’s statements drawn from his

training and experience).

Given that probable cause already existed for multiple

electronic devices in Smith’s apartment, police had reason to

believe that other devices in the apartment might also contain

evidence of the suspected offense. Smith could well have

transferred evidence of his conduct onto multiple devices. He

might have done so in the normal course of cycling through

11

devices, or he might have wanted to make backup copies of

photos or disperse evidence across multiple devices. Viewed

in light of Smith’s suspected conduct, the warrant’s “sweep”

did not “far outstrip[] the police’s proffered justification for

entering the home.” Griffith, 867 F.3d at 1276. Rather, the

warrant reasonably authorized police to seize a broad set of

electronic devices.

Smith, pointing to the fact that A.S. had identified specific

phones in her statements to investigators, contends that the

warrant should have limited the authorized seizure to those

particular phones or that police should have conducted a

reasonable investigation into which devices likely contained

incriminating evidence. We disagree. A.S. was thirteen years

old when she gave her statements to investigators, and she may

have been unable to accurately remember and describe which

particular devices would be relevant. In addition, she would

not have known whether Smith transferred stored photos and

other incriminating evidence to other devices. We decline to

hold that police officers armed with information that Smith

stored evidence of his crimes on phones and personal

computers were obligated to strictly conform the parameters of

their investigation to the precise information recalled and

related by A.S.

In all events, the good-faith exception precludes

suppression of the evidence recovered in the search. Under that

exception, suppression of evidence is appropriate “only if the

officers were dishonest or reckless in preparing their affidavit

or could not have harbored an objectively reasonable belief in

the existence of probable cause.” United States v. Leon, 468

U.S. 897, 926 (1984). To justify suppression, the affidavit

must be “so lacking in indicia of probable cause as to render

official belief in its existence entirely unreasonable.” Id. at

923.

12

The affidavit here does not meet that high bar. The

affidavit included A.S.’s detailed descriptions of Smith’s

sexual abuse, how he used electronic devices to carry out that

abuse, and what evidence would likely be found on those

devices. The affidavit also contained several paragraphs of

information from a detective describing how sex offenders tend

to use multiple electronic devices to carry out their crimes.

There was thus ample cause for police to believe that multiple

electronic devices found in Smith’s residence could contain

evidence of his suspected abuse. The officers’ reliance on the

warrant was reasonable.

Smith also argues the warrant was overbroad in that it

allegedly did not limit the types of data that could be taken from

the seized devices. Without deciding the underlying merits of

the claim, we hold that the good-faith exception also precludes

that argument. Given the information in the affidavit and the

fact that incriminating data was likely to exist in many forms—

including text messages, photos, and internet activity—a

reasonable officer could have concluded that probable cause

existed for the scope of the search.

None of this is to say that the warrant in this case was

necessarily a model of particularity. And when officers can

draft affidavits with greater particularity, they presumably

would do so to avoid a challenge like the one in this case. That

challenge fails here because the warrant was constitutionally

sufficient.

C.

Finally, we consider Smith’s challenges to the courtroom

presence and trial testimony of the government’s case agent.

We begin with Smith’s challenge to the district court’s

ruling permitting the case agent to remain in the courtroom

13

during the trial. Because Smith did not preserve his argument

that the district court failed to recognize its own inherent

authority to exclude the agent, we review for plain error. A

legal error is plain only if it is “clear or obvious, rather than

subject to reasonable dispute”; “affected the appellant’s

substantial rights, which in the ordinary case means . . . that it

‘affected the outcome of the district court proceedings’”; and

“‘seriously affects the fairness, integrity, or public reputation

of judicial proceedings.’” Puckett v. United States, 556 U.S.

129, 135 (2009) (alteration omitted) (quoting United States v.

Olano, 507 U.S. 725, 734, 736 (1993)).

The district court did not clearly or obviously err in

allowing the agent to remain in the courtroom. The district

court relied on Federal Rule of Evidence 615, which generally

requires courts to exclude witnesses from the courtroom at a

party’s request. Fed. R. Evid. 615 (Dec. 1, 2011) (amended

Dec. 1, 2023). (Although Rule 615 was recently amended, we

interpret the version in effect during Smith’s trial.) But the

Rule “does not authorize excluding” “an officer or employee

of a party that is not a natural person, after being designated as

the party’s representative by its attorney.” Id. 615(b) (now

located at Rule 615(a)(2)). Accordingly, the government

designated the agent as its representative at trial and the court

allowed her to remain in the courtroom.

A government case agent fits squarely within the text of

Rule 615(b): she is an “officer or employee” of the

government, which is “not a natural person.” Id.; see S. Rep.

No. 93-1277, at 26 (1974) (“[I]nvestigative agents are within

the group specified under the second exception made in the

rule . . . .”). Our sister circuits uniformly agree that the

government’s case agent in a criminal case falls within Rule

615(b)’s exception. See, e.g., United States v. Dennison, 73

F.4th 70, 73 (1st Cir. 2023); United States v. Rivera, 971 F.2d

14

876, 889 (2d Cir. 1992); United States v. Gonzalez, 918 F.2d

1129, 1137–38 (3d Cir. 1990); United States v. Parodi, 703

F.2d 768, 773 (4th Cir. 1983); United States v. Robles-Pantoja,

887 F.2d 1250, 1256–57 (5th Cir. 1989); United States v.

Pulley, 922 F.2d 1283, 1285 (6th Cir. 1991); United States v.

Edwards, 34 F.4th 570, 585 (7th Cir. 2022); United States v.

Sykes, 977 F.2d 1242, 1245 (8th Cir. 1992); United States v.

Valencia-Riascos, 696 F.3d 938, 941 (9th Cir. 2012); United

States v. Avalos, 506 F.3d 972, 978 (10th Cir. 2007), vacated

on other grounds, 555 U.S. 1132 (2009); United States v.

Butera, 677 F.2d 1376, 1380–81 (11th Cir. 1982). And our

court has already recognized that Rule 615’s exception for

designated representatives “appears to cover” the

government’s case agents. United States v. Cooper, 949 F.3d

744, 749 n.4 (D.C. Cir. 2020).

It is true that, while Rule 615(b) “does not authorize

excluding” a party’s representative, it also does not expressly

prohibit courts from excluding the representative. It appears to

be an open question in this court whether district courts have

discretion to exclude under a source of authority other than

Rule 615. See, e.g., Bradshaw v. Perdue, 319 F. Supp. 3d 286,

288–89 (D.D.C. 2018) (collecting authorities). Because no

binding precedent squarely resolves that question, the district

court did not plainly err in allowing the agent to remain in the

courtroom. See United States v. Vizcaino, 202 F.3d 345, 348

(D.C. Cir. 2000).

Smith also objects to a portion of the agent’s testimony at

trial. In that testimony, the agent interpreted a report from the

program Cellebrite to explain an exchange between A.S. and

Smith that police had discovered on Smith’s cell phone. After

a previous expert witness testified that A.S. sent all the

messages in the conversation, the agent sought to clarify which

text messages were sent to her and which were sent by her.

15

Smith contends that the agent was not qualified as an expert

and improperly gave that testimony based on specialized

knowledge.

Because Smith failed to preserve his argument that the

government laid an inadequate foundation for the case agent’s

expertise, we need not address whether the district court erred

in allowing the agent’s testimony. No such error would have

“affected the outcome of the district court proceedings” in light

of the overwhelming evidence against Smith. See Puckett, 556

U.S. at 135 (quoting Olano, 507 U.S. at 734). If Smith had

successfully blocked the agent from testifying based on her

knowledge of Cellebrite, he would have prevented only an

explanation of who sent and received a handful of text

messages. In those messages, A.S. and Smith discussed A.S.’s

feeling ill, whether she could leave school early, and her

journey home. Regardless of that exchange, there was a vast

amount of incriminating evidence of Smith’s conduct,

including sexually explicit text messages and photos

exchanged with A.S. What is more, the messages about A.S.’s

illness had already been introduced into evidence by the

previous expert witness’s testimony; the agent simply added an

explanation of which messages were sent by A.S. There thus

was no plain error in allowing the government’s case agent to

testify based on her knowledge of Cellebrite.

* * * * *

For the foregoing reasons, we affirm Smith’s convictions.

So ordered.