23-6662-cr United States v. Mensah

Jn the United States Court of Appeals For the Second Circuit

August Term, 2023 No. 23-6662-cr

UNITED STATES OF AMERICA, Appellee,

v.

ENOCK MENSAH, Defendant-Appellant.

On Appeal from a Judgment of the United States District Court for the Eastern District of New York.

ARGUED: MAY 24, 2024 DECIDED: AUGUST 2, 2024

Before: LIVINGSTON, Chief Judge, NARDINI, AND ROBINSON, Circuit Judges.

Defendant-Appellant Enock Mensah was a social worker who billed publicly funded agencies for over 1,600 treatment sessions that never took place. He was convicted, following a jury trial, of theft of public funds, in violation of

(a)(1)(A), and health care fraud, in violation of

(a). The United States District Court for the Eastern District of New York sentenced Mensah to forty- two months of imprisonment, to be followed by one year of supervised release. The court also ordered Mensah to pay $177,345 in restitution. Mensah now appeals, arguing that the district court erred in (1) failing to excuse or conduct further examination of a juror who knew a government witness; (2) denying Mensah’s post-trial motion for a new trial based on the prosecutor’s objection, during the cross- examination of a trial witness, that suggested Mensah had the ability to testify; and (3) applying a ten-level enhancement based on the loss amount in its calculation of the United States Sentencing Guidelines advisory range. We disagree. First, the district court did not abuse its discretion by failing to excuse a juror whose sister’s friend’s husband was a government witness. That connection was too attenuated to give rise to any presumption of bias. Nor did it err in its voir dire of that juror: After the acquaintanceship was disclosed, the district court adequately screened for any actual bias arising from the juror’s knowledge of the government’s witness. Second, the district court did not err in denying Mensah’s motion for a new trial based on the prosecutor’s comment before the jury that implicated Mensah’s Fifth Amendment right not to testify. Any prejudice from the prosecutor’s comment was rendered moot when the defendant elected to testify. And the record does not suggest that the comment somehow compelled or coerced Mensah to testify. Finally, we discern no clear error in the district court’s finding at sentencing that Mensah’s fraud resulted in a loss of $177,345, which was based on a careful reconstruction of Mensah’s whereabouts— using video surveillance, license-plate reading technology, and his

2 cellphone records, among other things—when he claimed to have performed treatment sessions. Accordingly, we AFFIRM the district court’s judgment.

ANTHONY BAGNUOLA (David C. James, on the brief), Assistant United States Attorneys, for Breon Peace, United States Attorney for the Eastern District of New York, Central Islip, NY, for Appellee.

RICHARD WASHINGTON, New York, NY, for Defendant-Appellant.

WILLIAM J. NARDINI, Circuit Judge:

Defendant-Appellant Enock Mensah was a social worker who

participated in a state-run program that provided remedial services

to developmentally delayed children and their families. Mensah

billed publicly funded agencies for over 1,600 sessions that never took

place, and he pocketed those payments. Mensah was convicted,

following a jury trial, of theft of public funds, in violation of

(a)(1)(A), and health care fraud, in violation of

(a). The United States District Court for the Eastern District of

New York sentenced Mensah to forty-two months of imprisonment,

3 to be followed by one year of supervised release. The district court

also ordered Mensah to pay $177,345 in restitution.

Mensah now appeals, arguing that the district court erred by

(1) failing to excuse or conduct further examination of a juror who

knew a government witness; (2) denying Mensah’s post-trial motion

for a new trial based on the prosecutor’s objection, during the cross-

examination of a trial witness, that suggested Mensah had the ability

to testify; and (3) applying a ten-level enhancement for the loss

stemming from Mensah’s fraud in its calculation of the advisory

range under the United States Sentencing Guidelines. We disagree.

First, the district court did not abuse its discretion by failing to

excuse a juror whose “sister’s friend’s husband,” App’x at 345, was a

government witness. That connection was too attenuated to give rise

to any presumption of bias. Nor did the district court plainly err in

its voir dire of that juror. After the acquaintanceship was disclosed,

4 the district court adequately screened for any actual bias arising from

the juror’s knowledge of the government’s witness.

Second, the district court did not err in denying Mensah’s post-

trial motion for a new trial based on the prosecutor’s comment before

the jury that implicated Mensah’s Fifth Amendment right not to

testify. Any prejudice from the prosecutor’s comment was rendered

moot when the defendant elected to testify. And the record does not

suggest that the comment somehow compelled or coerced Mensah to

testify.

Finally, we discern no clear error in the district court’s finding

at sentencing that the loss stemming from Mensah’s fraud was

$177,345, which was based on a careful reconstruction of Mensah’s

whereabouts—using video surveillance, license-plate reading

technology, and his cellphone records, among other things—when he

claimed to have performed treatment sessions.

Accordingly, we AFFIRM the district court’s judgment.

5 I. Background

The New York State Early Intervention Program (“EIP”)

provides to developmentally delayed children under the age of three

years old a broad range of remedial services, such as physical therapy,

occupational therapy, speech therapy, and social work. In order to

provide these services, the New York State Department of Health

(“NYS DOH”), which administers the EIP, contracts with various

agencies, non-profits, and private companies.

Mensah was employed as a social worker by two agencies—the

City Pro Group and All About Kids—contracted by NYS DOH to

provide EIP services. Social workers providing EIP services are

tasked with addressing challenges affecting the children’s caregivers

or environment, which may involve the social worker assisting

families with securing benefits, or counseling caregivers on how to

provide for children with disabilities, among other responsibilities.

6 Agencies contracted by the NYS DOH pay providers for their

EIP services. Those agencies, in turn, are reimbursed by NYS DOH

with funds from Medicaid, the New York City Department of Health

and Mental Hygiene (“NYC DOHMH”), or private insurance. Most

of the reimbursements come from Medicaid or NYS DOHMH.

A. The Charges

On February 4, 2019, a grand jury returned an indictment

charging Mensah with theft of public funds, in violation of

(a)(1)(A), and health care fraud, in violation of

(a). The indictment alleged that between August 2013 and

August 2017, Mensah reported that he provided more than 1,200

therapy sessions for EIP when he, in fact, had not. As a result of those

fraudulently reported sessions, the indictment alleged that Mensah

received more than $105,000 in Medicaid funds and more than

$20,000 in NYC DOHMH funds.

7 On August 30, 2019, a grand jury returned a superseding

indictment with the same charges but that expanded the relevant

timeframe of Mensah’s fraud to October 2018 and additionally

alleged that he fraudulently reported more than 1,500 non-existent

EIP sessions, resulting in payments of more than $130,000 from

Medicaid funds and more than $25,000 from NYC DOHMH funds.

B. Trial

The United States District Court for the Eastern District of New

York (Sterling Johnson, Jr., District Judge) held a jury trial from

December 2 to December 10, 2019.

i. Jury Selection

Before jury selection commenced, the government submitted a

list of names and places that would likely be mentioned during trial,

which included Michael Cecilio, a Special Investigator for the New

York City Department of Investigation (“NYC DOI”). Cecilio later

testified as a government witness. During jury selection, the

8 presiding magistrate judge (James Orenstein, Magistrate Judge)

distributed the list to prospective jurors and asked them to indicate

whether they knew any of the names. None of the prospective jurors

reported any familiarity with Cecilio.

ii. The Government’s Case

At trial, the government presented the following evidence of

the EIP’s requirements and procedures, Mensah’s scheme, and the

loss stemming from Mensah’s scheme.

a. The EIP’s Requirements and Procedures

A NYC DOHMH employee testified about the requirements

that providers must follow when administering services for the EIP,

including that they must administer services in person (and are

prohibited from doing so over the phone). Furthermore, during a

session, an EIP provider must contemporaneously document his

services, as well as the location, date, and time of service, in a session

note. At the end of the session, the EIP provider and the child’s

9 caregiver must sign the session note. EIP providers must then submit

the session notes to the appropriate agencies to obtain payment for

their services. EIP providers are not permitted to complete a session

note before a session occurs, have a caregiver sign a blank session

note, or forge a caregiver’s signature. Many of these requirements

related to the sessions notes are codified in state regulations. See, e.g.,

10 CRR-NY 69-4.26(c). And Mensah agreed contractually to abide by

such regulations when he joined City Pro Group and All About Kids.

A City Pro Group employee testified about the mechanics of

the agency’s payment process for providers. Providers must submit

their session notes to the agency for processing. After the notes are

processed, the agency reviews notes selected at random to ensure that

providers have completed the required components. The agency

does not independently verify whether the sessions occurred but

instead views the session notes as proof that such sessions took place

10 Once the agency completes this process, the providers are paid for

their services.

b. Mensah’s Scheme

The government introduced the following evidence to support

the allegations that Mensah billed agencies for sessions that never

occurred.

1. Victims’ Testimony

Seven parents of children receiving EIP services testified about

their experiences with Mensah. Mensah was assigned to provide

social work services to those parents to support and counsel them on

their children’s disabilities. For example, Mensah was tasked with

assisting one parent with applying for benefits for her son with

autism spectrum disorder and with obtaining furniture that her son

required. For another parent, Mensah was assigned to provide her

with counseling to address her feelings of being “overwhelmed” by

her son’s needs. App’x at 506.

11 Broadly speaking, the parents testified that Mensah had them

sign blank session notes, that session notes submitted by Mensah

contained forged signatures of their names, that Mensah did not

provide services as frequently as he was scheduled, and/or that it was

impossible for Mensah to have provided services on dates specified

on certain session notes submitted for payment. Additionally, one

parent testified that Mensah drove to the sessions, and another

testified that she met Mensah at his car for sessions.

The parents’ testimony that Mensah did not provide services

on dates indicated in the session notes was often corroborated by

other evidence. For instance, Mensah submitted two session notes for

payment for sessions purportedly conducted during the time that a

parent and her son were out of the country. On December 15, 2017,

another parent texted Mensah to inquire when he would next provide

services to her son, noting that he had not done so for five weeks.

However, during the five weeks preceding the text message, Mensah

12 submitted several session notes for the son, some of which contain the

parent’s forged signature. Furthermore, yet another parent texted

Mensah that Mensah “forgot to come” on April 16, 2018,

, but

he nevertheless submitted a session note for a session that allegedly

took place that day.

During the cross-examination of another parent, defense

counsel pressed her on the extent of the assistance that Mensah

provided her. Defense counsel’s questioning led the government to

raise an objection implicating Mensah’s right to remain silent:

Q: All right. Now, at the time you indicated that you lived in a fifth floor walkup?

A: Yes, I did.

Q: And Mr. Mensah was supposed to help you get relocated to another apartment?

A: Yes.

Q: And he told you about the New York City Housing Authority?

A: He did.

13 Q: And you said that that took too long, right?

A: And I applied.

Q: But you told him – you told him that it took too long?

A: I told him it takes long, but I also applied like he asked me to.

Q: And he told you that if it was not the New York City Housing Authority, every program had guidelines, right?

A: Yes.

Q: And he told you the only emergency –

[Prosecutor]: Objection, your honor.

The court: What’s the objection?

[Prosecutor]: Objection, your honor. If he wants the defendant to testify –

The court: No, no, no, no, no. Overruled.

at 712–13. Defense counsel resumed his questioning of the parent

after this exchange. After the jury was dismissed, the district court

admonished the prosecutor for his objection, noting that it was

14 inappropriate. Later, at the charge conference, defense counsel did

not request a curative instruction to address the government’s

objection. Instead, defense counsel stated that the proposed jury

instructions “look[ed] fine” and that the defense had “no complaints.”

. The jury instructions advised, among other things, that

“[t]he defendant in a criminal case never has any duty to testify or

come forward with any evidence.”

. After the district court

read those instructions to the jury, it held a sidebar during which it

asked the parties if there was “[a]nything” further, to which the

parties responded, “[n]o.”

.

2. The ExxonMobil Gas Station

The government introduced evidence that Mensah frequently

spent considerable time at an ExxonMobil gas station in Englewood,

New Jersey. A cashier at the gas station, testified that Mensah was a

regular customer of the gas station and visited it “[a]lmost every day”

for an hour or more.

. Video surveillance captured Mensah

15 “hanging out” at the gas station on August 24, 2018, during the same

time that he allegedly conducted several sessions and billed for them.

. The cashier testified that he never saw Mensah “conduct

business” at the gas station.

.

3. Mensah’s Arrest

On October 4, 2018, law enforcement agents—including

Cecilio—arrested Mensah at his apartment. Agents discovered blank

session notes that had been signed and pre-filled session notes

reflecting future appointments that had not yet occurred.

c. Loss Amount

The government introduced the following evidence regarding

the loss amount, meaning the amount of public funds paid to Mensah

for services that he did not render. The government’s primary

witness on this point, Natalie Lin, estimated the loss amount based,

in part, on license plate recognition (“LPR”) data, which tracked the

movements of Mensah’s car. Because the LPR data relied on the

16 identity of Mensah’s car, the government also produced evidence of

the car that Mensah drove.

1. Mensah’s Car

The government introduced evidence that Mensah owned and

operated a silver 2002 Lexus ES 300 registered in New Jersey with

license plate number D72ETE (the “Silver Lexus”). Records from the

New Jersey Motor Vehicle Commission (“NJMVC”), parking tickets

from the New York City Department of Finance, and moving

violation tickets from the New York Department of Motor Vehicles,

all indicate that Mensah owned and drove the Silver Lexus. A

NJMVC employee testified that the Silver Lexus was registered to

Mensah between January 1, 2015, and October 3, 2018, and that no

other cars were registered to Mensah during that time.

Other witnesses corroborated that Mensah drove the Silver

Lexus. Mensah’s ex-wife testified that she drove the Silver Lexus until

September 2014, after which she sold the car to Mensah and never

17 drove the car again. She was not aware of any time that her son—her

only child who drives—used the Silver Lexus. The ExxonMobil

cashier testified that he saw Mensah driving the Silver Lexus “many

times.”

. Moreover, Cecilio testified that he surveilled

Mensah driving the Silver Lexus at Mensah’s apartment building, and

that he photographed Mensah entering the Silver Lexus and driving

away.

2. Lin’s Analysis

Lin, a data analyst at the New York City Department of

Investigation, testified that, based on her analysis, Mensah submitted

session notes for payment for 1,725 sessions that never happened. Of

those sessions, 1,689 resulted in payment from public funds, totaling

$177,345. Medicaid paid for $148,155 of that total, while NYC

DOHMH paid for $29,190.

She examined six types of evidence to reach that conclusion.

First, she reviewed approximately 3,000 session notes submitted by

18 Mensah for payment to determine the date and time of the alleged

service as well as the child receiving the service.

Second, she reviewed a spreadsheet of data maintained by

NYC DOHMH containing the addresses of the children receiving EIP

services, which were presumably the locations of Mensah’s purported

sessions. All of the children listed on the spreadsheet resided within

the five boroughs of New York City. The spreadsheet also identifies

the payor of Mensah’s sessions and the amounts, if any, that were

paid for those sessions.

Third, she reviewed Mensah’s cellphone records, which totaled

more than 780 pages and indicated the date, time, and duration of

calls involving Mensah’s cellphone. Those records did not indicate

who was using the cellphone. Fourth, she reviewed surveillance

footage of the ExxonMobil gas station in Englewood, New Jersey from

August 24, 2018.

19 Fifth, she reviewed the LPR data generated by New York City

municipal cameras that captured the Silver Lexus’s license plate. That

data came in two forms: directional and static. Directional LPR

reflected instances when cameras captured the license plate crossing

a bridge or a tunnel, indicating whether the car was traveling inbound

to or outbound from New York City—say, on the George Washington

Bridge. Static LPR captured the car at various moments, providing

the latitude and longitude of where the license plate was

photographed, as well as the date and time that the license plate

appeared at that location. Cecilio testified that he reviewed the LPR

data to confirm that each image returned from the cameras matched

the Silver Lexus’s license plate number. The LPR data did not

establish who was driving the Silver Lexus, because the photos did

not show who was in the driver’s seat.

20 Sixth, Lin reviewed Mensah’s personnel records maintained by

City Pro Group and All About Kids to determine that at all relevant

times, Mensah lived in New Jersey.

Lin summarized her findings in a spreadsheet, admitted into

evidence, that used the data from these various sources to identify

instances when evidence of Mensah’s location and activities

conflicted with the services that Mensah purportedly administered.

Using the directional LPR data, Lin recorded a conflict when

Mensah billed for services in New York City either before a camera

photographed the Silver Lexus’s license plate entering New York for

the day, or after the license plate returned to New Jersey for the

evening. Using the static LPR data, Lin recorded a conflict when the

Silver Lexus’s license plate was photographed at a location where

Mensah would have had to drive at a constant speed of more than

forty miles per hour within New York City to provide the services

reflected in a session note. Lin determined the distance between the

21 locations “as the crows flies,”

, meaning she did not consider

the traffic, configuration of the roadways, or traffic lights, among

other things. Relying on the cellphone records, Lin recorded a conflict

when Mensah’s cellphone was in use for at least fifteen minutes

during a purported session and when another source of evidence

independently corroborated that Mensah was likely not providing

services. Finally, using the surveillance footage from the gas station,

Lin recorded a conflict when Mensah was depicted on that footage at

the same time that he purportedly provided a session.

Lin provided several examples of conflicts that she identified

suggesting that Mensah did not provide certain sessions reflected in

his submitted session notes. These conflicts were corroborated by

other evidence, such as contemporaneous text messages sent from

Mensah’s cellphone.

For example, Mensah submitted session notes for six purported

sessions on July 6, 2015. Lin testified, however, that there were no

22 LPR entries on July 6, 2015, which suggested that the Silver Lexus

never traveled into New York City, where all of the children Mensah

served resided, on that specific day. A text message sent from

Mensah’s cellphone on the same day stated, “I did not come to New

York today.”

.

Similarly, Mensah submitted session notes for four purported

sessions on November 26, 2015, which was Thanksgiving Day. Lin

testified again that the lack of LPR entries for that day indicated that

the Silver Lexus did not travel into New York City. This conclusion

is corroborated by a text message sent from Mensah’s cellphone the

day before Thanksgiving Day, which stated: “I’ll be going to

Pittsburgh tomorrow and will be back tomorrow or Friday.”

.

In addition, Mensah submitted session notes for four purported

sessions on January 10, 2016. Several of those sessions allegedly took

place between 3:00 p.m. and 8:55 p.m. Like the previous examples,

23 Lin testified there were no LPR entries for that day, so she concluded

that the Silver Lexus did not travel into New York City that day. A

text message sent from Mensah’s cellphone on the same day

suggested that Mensah entertained company at his home between the

hours of 3:00 p.m. and 7:30 p.m.

Furthermore, Mensah submitted session notes for two alleged

sessions on September 13, 2016, and Lin testified that the LPR data

indicated that the Silver Lexus never traveled into New York City that

day. This conclusion is corroborated by a text message sent from

Mensah’s cellphone the same day, which stated, “I am in

Pennsylvania.”

.

Mensah also submitted session notes for purported sessions on

August 24, 2018. Lin testified that the three sessions occurred before

the Silver Lexus traveled into New York City that day and during

times when Mensah was captured on the surveillance footage from

the gas station. Similarly, Mensah submitted session notes for two

24 purported sessions on September 13, 2018, between 9:30 a.m. and

11:30 a.m. Lin found that both of those sessions happened before the

Silver Lexus traveled into New York City that day. This conclusion is

supported by a text message sent from Mensah’s cellphone at 10:00

a.m. that day, stating, “I am in the gym. Will go to work in a few.”

.

d. Issue with Juror #6

Before Cecilio testified, the government informed the district

court that Cecilio recognized Juror #6. The district court asked Cecilio

if he had a relationship with Juror #6. Cecilio responded that he “went

to college with [Juror #6’s] older sister” and that his “wife and [Juror

#6’s] sister were roommates after college.”

at 342–43. Cecilio

stated that Juror #6 would know him but clarified that they had “[n]o

relationship.”

. Defense counsel stated that “if there’s a

friendly relationship,” he was not sure whether that relationship

25 “would impact the juror’s ability to properly evaluate this witness’

testimony,”

and proposed asking Juror #6 if he recognizes Cecilio.

The district court brought Juror #6 to the courtroom where

Cecilio was sitting at a table and asked Juror #6 if he recognized

anybody seated at the tables. Juror #6 responded that he recognized

Cecilio, identifying Cecilio as his “sister’s friend’s husband.”

. Juror #6 stated that he did not “have a relationship with him”

but that he “know[s] him . . . through seeing him.”

The district

court asked if Juror #6 could still be fair and impartial even though he

recognizes Cecilio, to which Juror #6 replied, “Hundred percent.”

The district court confirmed again that Juror #6 could still be fair and

impartial.

Defense counsel noted on the record that Juror #6 appeared to

be “very happy” to see Cecilio in court,

at 345–46, so defense

counsel was not sure if Juror #6 could actually be fair and impartial as

to Cecilio’s testimony. The district court disagreed with defense

26 counsel’s account of the events, stating that Juror #6 smiled at the

judge, not at the witness. Juror #6 continued serving on the jury.

iii. The Defense’s Case

Mensah testified in his own defense. He stated that he

commuted to work by driving or through public transit, such as a

shuttle from his apartment, a bus, or the subway. He also said that he

drove the Silver Lexus given to him by his ex-wife and that he used

to drive his ex-girlfriend’s car, a Chrysler 300. He denied ever forging

anyone’s signature and testified that in practice, providers have

parents sign blank session notes.

Mensah also attempted to explain the discrepancy regarding

August 24, 2018, the day he was featured on surveillance footage from

the gas station and submitted session notes for sessions that

purportedly occurred during that time. He testified that he “pre-

fill[ed]” the session notes for that day with the scheduled times of the

sessions, which he does “most of the time.”

. He further

27 testified that parents rescheduled the visits for that day, and that he

had inadvertently failed to change the time on the pre-filled session

notes to the rescheduled times during which he actually provided

services.

On cross-examination, the government questioned Mensah on

various aspects of his testimony. For example, addressing Mensah’s

testimony that he works on Thanksgiving Day, the government

produced correspondence dated November 25, 2015, suggesting that

Mensah did not work on Thanksgiving Day in 2015, which was on

November 26, 2015. In that correspondence, someone asked Mensah,

“what are you doing tomorrow?”, to which Mensah responded, “I

will be going to Pittsburgh tomorrow and be back on Friday or

Saturday[.]”

. Mensah testified that the person involved in

the correspondence “is a model friend of [his] ex-girlfriend” who had

been “bugging [him] . . . [to] help her out to establish [an] LLC.”

. Mensah denied traveling to Pittsburgh on November 26.

28 Resisting the government’s characterization of his

correspondence as dishonest or false, Mensah testified that he was

trying “to distract [his] friend” because he did not “want to see her.”

. He explained that he “[did not] know anybody in

Pennsylvania.”

. The government then asked Mensah if his

son attended Carnegie Mellon University, which is located in

Pittsburgh, Pennsylvania. Mensah responded that his son did. And

when asked if his previous statement that he had no connection to

Pennsylvania was therefore a mistake, Mensah testified that his son

did not live there but merely went to school there. Mensah could not

recall whether his son attended Carnegie Mellon in 2015.

iv. The Jury’s Verdict

The jury found Mensah guilty of both Count One (theft of

funds) and Count Two (health care fraud).

29 C. Post-Trial Motions

On January 30, 2020, Mensah moved for a judgment of acquittal

under Federal Rule of Criminal Procedure 29 or, in the alternative, for

a new trial under Federal Rule of Criminal Procedure 33. He argued

that relief was warranted because (1) the testimony of the

government’s witnesses was incredible and inconsistent; (2) he acted

in good faith, which is a complete defense to the fraud count; and (3)

the prosecutor improperly commented on his right to remain silent,

which prejudiced him.

The district court denied Mensah’s motion. The district court

found that the prosecutor’s comment was improper, but nevertheless

concluded that vacatur of the conviction or a new trial was not

necessary because the court could not identify any prejudice to

Mensah as a result of that comment. The court reasoned that (1) the

prosecutor did not finish the sentence before the district court

interrupted him; (2) the district court advised the jury at the

30 beginning and end of the trial that Mensah had no obligation to

testify; and (3) Mensah testified, “render[ing] moot” “any negative

inference the jury may have drawn from a decision to remain silent,”

and “Mensah does not argue that his decision to testify was in any

way motivated by the prosecutor’s comment or that his testimony

ultimately harmed his case,”

.

D. Sentencing

The Presentence Investigation Report (“PSR”) stated that

Mensah received payment for 1,689 sessions that never happened,

totaling $148,155 in Medicaid funds and $29,190 in NYC DOHMH

funds, which is consistent with Lin’s trial testimony. The PSR

additionally reported that restitution was statutorily required—

$20,000 to NYC DOHMH and $105,000 to Medicaid.

The PSR, which was amended after receiving objections from

the government, calculated a total offense level of 24, which, coupled

with Mensah’s criminal history category of I, resulted in an advisory

31 Guidelines range of 51 to 63 months of imprisonment. In calculating

the total offense level, the Probation Office applied enhancements

based on Mensah’s abuse of his position of public trust in committing

the offenses, his obstruction of justice as a result of his false testimony

during trial, and the fact that his crimes involved vulnerable victims.

The U.S. Probation Office also found that the estimated total loss

stemming from Mensah’s scheme was $177,345, which is more than

$150,000 but less than $250,000, resulting in a ten-level enhancement

under U.S.S.G § 2B1.1(b)(1)(F).

In Mensah’s sentencing memorandum, he raised numerous

“objections” to the PSR, including that the table in the Guidelines for

calculating the enhancement for loss amount “overstates the

seriousness of the offense” and therefore “the impact of the loss table

should be limited.” Id. at 1386–87. Although he generally objected to

the impact of the loss table, he did not object to the application of the

loss enhancement. He did object to the application of the obstruction

32 of justice enhancement, as well as the restitution, arguing that the

government had failed to prove the loss amount by a preponderance

of the evidence.

On June 8, 2023, the district court (Ann M. Donnelly, District

Judge) 1 sentenced Mensah principally to a below-Guidelines sentence

of forty-two months of imprisonment, to be followed by one year of

supervised release. The district court also imposed a special

assessment of $200 and ordered restitution in the amount of $177,345.

The district court adopted the PSR in its entirety, including the

Guidelines calculation.

The district court entered judgment on June 13, 2023, which

listed the restitution amount as $177,745. On June 20, 2023, Mensah

filed a notice of appeal. On June 28, 2023, the government requested

an amended judgment because the restitution amount listed in the

written judgment did not match the restitution amount orally

1After the trial and post-trial ruling, the case was reassigned to Judge Donnelly on September 13, 2022. Judge Johnson passed away on October 10, 2022.

33 imposed by the district court at sentencing. On July 11, 2023, the

district court entered an amended judgment that updated the total

restitution amount to $177,345. Finally, on July 18, 2023, the district

court entered a separate order of restitution again providing that the

total restitution amount is $177,345.

II. Discussion

On appeal, Mensah contends that the district court erred in:

(1) failing to excuse or conduct further examination of Juror #6 for

bias; (2) denying Mensah’s request for post-trial relief based on the

prosecutor’s comment implicating Mensah’s right to testify; and

(3) applying the ten-level enhancement for the loss amount. For the

reasons below, we affirm the district court’s judgment.

A. Juror #6

Mensah argues that Juror #6 “was partial to the prosecution . . .

based on his prior acquaintance with Mr. Cecilio,” Appellant’s Br. at

14, and accordingly, the district court should have excused Juror #6

or conducted additional inquiry into the duration of Juror #6’s

34 relationship with Cecilio and why Juror #6 did not disclose that he

knew Cecilio during jury selection to ensure that Juror #6 held no bias.

“We review for abuse of discretion a district court’s handling

of juror dismissal,” United States v. Walker,

(2d Cir.

2020), “revers[ing] only if there is ‘clear abuse’ of the district court’s

discretion,” Cruz v. Jordan,

(2d Cir. 2004) (quoting

United States v. Nelson,

(2d Cir. 2002)). “[A]

reviewing court will only find abuse of that discretion where there is

bias or prejudice to the defendant.” United States v. Breen,

(2d Cir. 2001) (internal quotation marks and citation omitted).

Because Mensah did not object to the district court’s

questioning of Juror #6, we review his challenge to the court’s failure

to further question the juror for plain error. See Fed. R. Crim. P. 52(b);

United States v. Parse,

(2d Cir. 2015) (applying plain-

error review when the defendant failed to object to alleged juror

misconduct before the district court). Under that standard, the

35 defendant must show that “(1) there is an error; (2) the error is clear

or obvious, rather than subject to reasonable dispute; (3) the error

affected the appellant’s substantial rights; and (4) the error seriously

affects the fairness, integrity or public reputation of judicial

proceedings.” United States v. Chaires,

(2d Cir. 2023)

(quoting United States v. Miller,

, 557–58 (2d Cir. 2020)).

We start by reviewing the longstanding principles regarding

juror bias. Under the Sixth Amendment, “[a] criminal defendant is

guaranteed a trial ‘by an impartial jury.’” United States v. Torres,

(2d Cir. 1997) (quoting U.S. Const. amend. VI). An

impartial jury is one “capable and willing to decide the case solely on

the evidence before it.”

(quoting McDonough Power Equip., Inc. v.

Greenwood,

(1984)).

“[I]t is uncontroversial that ‘part of the guarantee of a

defendant’s right to an impartial jury is an adequate voir dire to

identify unqualified jurors.’” United States v. Nieves,

, 631

36 (2d Cir. 2023) (quoting Morgan v. Illinois,

(1992)). A

trial court is afforded broad discretion in the way it conducts voir dire,

and specifically, in “deciding what questions to ask prospective

jurors,” United States v. Tsarnaev,

(2022). Indeed, for

determining whether a juror is impartial, “the Constitution lays down

no particular tests and procedure is not chained to any ancient and

artificial formula.” Torres,

(quoting United States v.

Wood,

, 145–46 (1936)).

The broad discretion afforded to a trial court in this realm is for

good reason because “the trial court . . . is best positioned to ‘reach

conclusions as to impartiality and credibility by relying on [its] own

evaluations of demeanor evidence and of responses to questions.’”

Nieves,

(quoting Rosales-Lopez v. United States,

(1981)). It is for that very same reason that “the adequacy of

voir dire is not easily subject to appellate review.”

(quoting Rosales-

Lopez,

); see also United States v. Greer,

, 171

37 (2d Cir. 2002) (“The district court, which observ[es] the jury on a day

to day basis . . . is in the best position to sense the atmosphere of the

courtroom as no appellate court can on a printed record.” (internal

quotation marks and citation omitted)).

The trial court’s discretion, of course, is not limitless: it “must

be exercised consistent with ‘the essential demands of fairness.’”

Nieves,

(quoting United States v. Barnes,

,

137–38 (2d Cir. 1979)). And fairness demands that the defense be

provided “a full and fair opportunity to expose bias or prejudice on

the part of veniremen.”

(quoting United States v. Colombo,

(2d Cir. 1989)). “Where the court finds that such bias or

prejudice exists, a juror is subject to dismissal for cause.”

There are generally three forms of bias that may give rise to a

for-cause challenge for dismissal: (1) actual bias, (2) implied bias, and

(3) inferred bias.

“[A]ll three forms of bias must be ‘grounded in

facts developed at voir dire.’”

(quoting Torres,

).

38 Mensah argues that Juror #6’s connection with Cecilio gave rise

to implied or inferable bias. Implied bias “is ‘presumed as a matter

of law’ where a typical person in the juror’s position would be biased,

irrespective of whether actual bias exists.” Nieves,

(quoting Greer,

). The test for implied bias “is whether

an average person in the position of the juror in controversy would

be prejudiced.” Torres,

. For this inquiry, a juror’s

statements regarding his ability to be impartial hold no weight.

A

finding of implied bias “is reserved for exceptional situations,”

(internal quotation marks and citation omitted), such as when

“jurors . . . are related to the parties or . . . were victims of the alleged

crime,”

. And when such exceptional situations are present,

disqualification of the juror “is mandatory.”

“[I]nferable bias . . . arises ‘when a juror discloses a fact that

bespeaks a risk of partiality sufficiently significant to warrant

granting the trial judge discretion to excuse the juror for cause, but

39 not so great as to make mandatory a presumption of bias.’” Nieves,

(quoting Greer,

). In this inquiry, a juror’s

statements about his ability to be impartial are irrelevant “once facts

are elicited that permit a finding of inferable bias.” Torres,

. But, “particularly when considering whether some marginal

types of disclosed facts are enough to show inferable bias,” a judge

may “ask about a juror’s impartiality and might be persuaded by the

force of the juror’s assurance.”

at 47 n.12.

Applying those principles here, we conclude that the district

court did not abuse its discretion by permitting Juror #6 to remain on

the jury. The connection between Juror #6 and Cecilio is not one of

the “exceptional situations,”

(internal quotation marks and

citation omitted), that give rise to implied bias and therefore warrants

mandatory dismissal. The evidence in the record shows that Cecilio

went to college with Juror #6’s sister and that Cecilio’s wife and Juror

#6’s sister were roommates after college. Although Cecilio and Juror

40 #6 knew each other, both unequivocally confirmed that they had “[n]o

relationship.” App’x at 343; id. at 345. Such a non-relationship—more

precisely, a mere acquaintanceship—was too attenuated to compel a

presumption of bias. 2 See United States v. Ferri,

, 991–94

(3d Cir. 1985) (finding no implied bias where the juror’s husband

knew a government witness); United States v. Shaoul,

, 816–

17 (2d Cir. 1994) (finding no implied bias where a juror was the uncle-

in-law of a prosecutor in the same district, but the prosecutor was not

involved in the case at hand). The attenuated nature of the association

between Cecilio and Juror #6 is also, without more, fatal to Mensah’s

claim of inferred bias. See Torres,

(“[A]lthough not

rising to a per se implied bias category, relatives of the prosecutor

might, in a particular case, be excludable.”). In sum, the district court

2 That defense counsel thought that Juror #6 smiled at Cecilio, suggesting a higher level of familiarity, does not alter our conclusion. The district judge’s view of the events differed from defense counsel’s: he concluded that Juror #6 smiled at the judge rather than at the witness. As with demeanor findings generally, we defer to the district judge’s first-hand assessment of the facts on this point. See Greer,

(“[A] district court’s evaluation of the juror’s impartiality is accorded deference.”).

41 did not abuse its discretion in concluding that the acquaintanceship

between Juror #6 and Cecilio gave rise to implied or inferred bias.

Furthermore, we conclude that the district court did not plainly

err by failing to ask Juror #6 additional questions. After Cecilio

disclosed that he knew Juror #6, the district court brought Juror #6 to

the courtroom and asked him if he recognized anyone. Juror #6

replied that he knew Cecilio, prompting the district court to ask

several questions regarding the nature of Juror #6’s connection with

Cecilio. Specifically, the court asked Juror #6 how he recognized

Cecilio and if he had a relationship with Cecilio. The district court

finished by asking Juror #6 twice if he could be impartial, to which

Juror #6 answered affirmatively and unequivocally. This colloquy

was sufficient to afford “a full and fair opportunity to expose bias or

prejudice on the part of [Juror #6],” Nieves,

(internal

quotation marks and citation omitted), especially since defense

counsel did not request any further questioning.

42 Although Mensah admits that the district court asked “seven

substantive questions” during its voir dire, Appellant’s Reply Br. at 1,

he insists that the court’s inquiry “should have been more in-depth,”

Appellant’s Br. at 18. But the district court’s failure to ask additional

questions newly proposed by Mensah on appeal does not constitute

error. See United States v. Lawes,

(2d Cir. 2002)

(“[F]ederal trial judges are not required to ask every question that

counsel—even all counsel—believes is appropriate.”). As we have

previously explained, “the district court must provide some

opportunity for prospective jurors to be meaningfully screened for

biases relevant to a particular defendant.” Nieves,

. But

a district court’s refusal to ask particular questions “will not be

grounds for reversal, provided the voir dire ‘cover[s] the subject[s]’

that may arise in the case to ensure that jurors will be impartial.”

United States v. Rahman,

(2d Cir. 1999) (quoting

Aldridge v. United States,

(1931)). “Deciding precisely

43 how to [screen for bias] remains within the district court’s expansive

voir dire discretion.” Nieves,

; see also United States v.

Treacy,

(2d Cir. 2011) (“[W]e ‘appear[ ] never to have

reversed a conviction for the failure to ask a particular question on the

voir dire of prospective jurors.’” (quoting Lawes,

).

Here, the district court’s questioning was sufficient to apprise

the defense of the connection between Cecilio and Juror #6 and to

flush out any possible bias stemming from that association. That ends

our inquiry. It is not for us to second-guess the precise manner in

which the district court conducted its voir dire.

Mensah nonetheless contends that his case is like Nieves, where

we concluded that a district court’s failure to ask a certain category of

questions aimed at discovering juror bias was an abuse of discretion.

In that case, the defendant was charged with multiple counts

stemming from his conduct as a gang member. See Nieves,

. Because of the nature of the charges, the defense and government

44 proposed gang-related voir dire questions, see

, which the

district court declined to ask during jury selection,

. We held

that the district court abused its discretion based on its failure to

adequately screen for any biases against gangs.

. Notably,

we did not fault the district court for failing to ask the specific

questions proposed by the parties. See

at 638–39. The district

court’s mistake, rather, was its failure to ask any questions that would

permit the defendant to “ferret” out directly or indirectly any gang-

related biases, which both parties viewed as central to the case,

at

638–40 (internal quotation marks and citation omitted), and which we

identified as a sufficiently “pervasive” potential bias in light of an

unusual and “frenetically-publicized reported spike in violent crime

that local and federal law enforcement had broadly attributed to gang

violence, including in large part to violence committed by members

of traditionally Latin-American gangs” like the one involved in the

charged case,

at 633–34.

45 Mensah’s attempt to liken his case to Nieves is inapt. Nieves

represents an extreme case where our Court concluded that the

district court had failed to conduct any inquiry at all that might have

enabled the parties to explore (even indirectly) a possible bias against

a specific type of gang that was identified by both parties, highly

publicized at the time in the community from which the jury pool was

drawn, and closely related to issues that were almost certain to arise

during trial. Putting aside any differences between the types of

potential bias at issue here and in Nieves, the district court in the

present case did inquire into the potential bias—and it did so directly.

The district court asked several questions to ascertain the nature of

Juror #6’s connection with Cecilio, permitting Mensah to discover any

biases that might have arisen as a result.

Accordingly, the district court did not abuse its discretion in

failing to dismiss Juror #6 or plainly err in its voir dire of Juror #6.

46 B. Prosecutor’s Comment

Mensah next argues that the prosecutor’s objection during

Hamilton’s cross-examination violated his Fifth Amendment right to

remain silent, depriving him of a fair trial, and that the district court

abused its discretion by denying his post-trial motion for a new trial

based on this comment.

Rule 33 provides that a district court “may vacate any judgment

and grant a new trial if the interest of justice so requires.” Fed. R.

Crim. P. 33(a). The district court “has broad discretion to decide Rule

33 motions based upon its evaluation of the proof produced.” United

States v. McPartland,

(2d Cir. 2023) (quoting United

States v. Gambino,

(2d Cir. 1995)). The crucial inquiry

“is whether letting a guilty verdict stand would be a manifest

injustice.”

(quoting United States v. Ferguson,

(2d

Cir. 2001)).

47 “We review the district court’s denial of a motion under Rule

33 for abuse of discretion.”

“A district court abuses its discretion

in denying a Rule 33 motion when (1) its decision rests on an error of

law or a clearly erroneous factual finding, or (2) its decision—though

not necessarily the product of a legal error or a clearly erroneous

factual finding—cannot be located within the range of permissible

decisions.”

(internal quotation marks and citation omitted).

“It is settled that prosecutors may not comment adversely on a

defendant’s invocation of his Fifth Amendment privilege not to

testify.” United States v. Whitten,

(2d Cir. 2010). To

determine “whether a prosecutor’s statements amount to an

improper comment on the accused’s silence in violation of the Fifth

Amendment,” we must “look[] at the statements in context and

examine[] whether they naturally and necessarily would be

interpreted by the jury as a comment on the defendant’s failure to

48 testify.” United States v. Knoll,

(2d Cir. 1994)

(internal quotation marks and citation omitted).

The district court correctly interrupted the prosecutor’s

objection before it could be completed. During Hamilton’s cross-

examination, defense counsel asked her twice to confirm what

Mensah had said to her. See, e.g., App’x at 712 (“And [Mensah] told

you that if it was not the New York City Housing Authority, every

program had guidelines, right?”). The prosecutor then objected,

asserting, “If he wants the defendant to testify –.” Id. at 713. Before

the prosecutor could finish, the district court hastily intervened and

stated, “No, no, no, no, no. Overruled.” Id. It appears that the

prosecutor’s interjection was the beginning of an inartfully phrased

hearsay objection. Nevertheless, the prosecutor’s incomplete

objection could have been interpreted by the jury as a comment on

Mensah’s failure to testify at that point in the trial. See Fox v. Mann,

(2d Cir. 1995) (stating that the prosecutor’s remark that

49 “if the defendant wants to speak, he can take the stand” was

“certainly improper”).

A finding that a prosecutor’s remark was improper is not

enough to warrant a new trial. We will grant one only when “the

misconduct alleged [is] so severe and significant as to result in the

denial of [the defendant’s] right to a fair trial.” United States v. Coplan,

(2d Cir. 2012) (internal quotation marks and citation

omitted); see also Whitten,

(“We will reverse on the

ground of prosecutorial misconduct only if that misconduct caused

substantial prejudice by so infecting the trial with unfairness as to

make the resulting conviction a denial of due process.” (internal

quotation marks and citation omitted)). And when assessing whether

a prosecutor’s remark “rise[s] to the level of prejudicial error, we

examine the severity of the misconduct, the measures adopted to cure

the misconduct, and the certainty of conviction absent the

50 misconduct.” Coplan,

(internal quotation marks and

citation omitted).

We conclude that the prosecutor’s remark did not cause

prejudice so substantial that it deprived Mensah of a fair trial. After

the prosecutor’s objection, Mensah took the stand to testify in his own

defense. We have not yet addressed the impact of a defendant’s

testimony on the analysis for prejudice stemming from a prosecutor’s

comment on a defendant’s right to testify. The Tenth Circuit,

however, resolved this question in United States v. Carleo,

(10th Cir. 1978). In that case, the court assumed the prosecutor’s

remark violated the defendant’s rights under the Fifth Amendment

and held that the issue was “mooted by defendant’s election to testify

in his own defense.”

. The court reasoned that “[t]he danger

that a jury might infer guilt from an accused’s failure to testify simply

is not present when he actually does take the stand and testify.”

And “[w]hen comments relating to an accused’s opportunity to testify

51 are followed by his actual testimony,” the court stated that “the

relevant inquiry [becomes] whether his testimony in effect was

coerced or compelled by the prior comments.”

We are persuaded

by and adopt our sister circuit’s reasoning.

Applied here, any possible prejudice from the prosecutor’s

statement dissipated when Mensah testified. As explained above, the

core issue with a prosecutor commenting on a defendant’s silence is

that a jury might infer guilt from that silence. See Griffin v. California,

, 614–15 (1965) (“[T]he Fifth Amendment . . . forbids either

comment by the prosecution on the accused’s silence or instructions

by the court that such silence is evidence of guilt.”). A jury might be

particularly tempted to make such an inference when a defendant

“fail[s] to testify as to facts peculiarly within [his] knowledge.”

. Ultimately, it is this inference of guilt drawn from such a

comment that can undermine a defendant’s Fifth Amendment right

to remain silent. See

at 614–15. But a jury cannot infer guilt from

52 silence if a defendant does not remain silent. Accordingly, there is no

danger of an improper inference where (as here) a defendant elects to

testify.

The question then becomes whether Mensah was compelled to

testify by virtue of the prosecutor’s comment. Mensah argues for the

first time on appeal that he was “forced” to take the stand and waive

his Fifth Amendment right to remain silent because of the

prosecutor’s remark. Appellant’s Br. at 26. Mensah did not raise this

argument at any time before the district court. See App’x at 1371

(district court noting in its ruling on the post-trial motion that

“Mensah [did] not argue that his decision to testify was in any way

motivated by the prosecutor’s comment or that his testimony

ultimately harmed his case”). We thus review Mensah’s claim of

compulsion for plain error. See Fed. R. Crim. P. 52(b).

We find no error, let alone plain error, on the issue of

compulsion. The prosecutor’s remark occurred on a Friday, and

53 Mensah testified the following Monday, so there is no suggestion that

Mensah was pressured to decide whether to testify without adequate

time to consider his options. Indeed, the record is devoid of any

indications whatsoever that Mensah was somehow coerced to testify.

At oral argument, Mensah conceded that the sole evidence of

compulsion was the prosecutor’s improper comment. See Oral Arg.

Audio Recording at 09:38–09:42. However, the prosecutor’s remark

alone does not constitute adequate evidence of compulsion. Offering

his own testimony was not the only available way for Mensah to

redress the prosecutor’s comment. For example, Mensah could have

requested a curative jury instruction. See Fox, 71 F.3d at 72–73

(finding that a district court’s remedial instruction to the jury—

directly addressing the prosecutor’s improper remark and given the

morning after the remark was made—was sufficient “to overcome

whatever limited adverse effect the prosecutor’s [remark] may have

had on the jury”); see also United States v. Bubar,

(2d

54 Cir. 1977) (holding that “any conceivable misunderstanding on the

part of the jury [of the prosecutor’s comments] was nipped in the bud

by Judge Newman’s emphatic curative instructions”). Mensah could

have also offered the testimony of additional witnesses to discredit

the government’s case. See United States v. Jones,

(2d

Cir. 2002) (finding defendant was not compelled to testify where he

could have “chosen to attack the government’s case by offering the

testimony of other individuals”). He did neither. Given that other

options for redress were available, Mensah cannot convincingly argue

that he was forced to testify as a result of the prosecutor’s comment.

Furthermore, there are a variety of reasons, often related to trial

strategy, that can motivate a defendant’s decision to testify. But the

“cho[ice] between asserting his right to silence and pursuing what he

believed to be the most effective defense” does not mean Mensah

“faced the kind and intensity of coercion that would deprive him of

[his] right.”

. Thus, without more than the prosecutor’s

55 statement, we cannot conclude that the district court erred by failing

to find compulsion based on the record before us.

In sum, we discern no abuse of discretion or plain error in the

district court’s denial of Mensah’s motion for a new trial.

C. Loss Amount Enhancement

Finally, Mensah challenges his sentence as procedurally

unreasonable because, he argues, the district court erred by applying

the ten-level enhancement under U.S.S.G. § 2B1.1 based on a loss

amount that is not a reasonable estimate.

“Under U.S.S.G. § 2B1.1, the sentencing court increases the

defendant’s offense level based on the ‘loss’ to victims caused by his

acts.” United States v. Moseley,

(2d Cir. 2020) (quoting

U.S.S.G. § 2B1.1(b)(1)). Here, the district court adopted the loss

amount from the PSR, which assessed the loss associated with

Mensah’s scheme to be $177,345. Because the loss was more than

$150,000 but less than $250,000, the district court applied a ten-level

56 enhancement to Mensah’s total offense level under U.S.S.G §

2B1.1(b)(1)(F). The district court therefore calculated Mensah’s total

offense level to be 24 and Mensah’s criminal history category to be I,

resulting in an advisory Guidelines range of 51 to 63 months of

imprisonment. The district court sentenced Mensah to a below-

Guidelines sentence of forty-two months of imprisonment, to be

followed by one year of supervised release.

The parties dispute the appropriate standard of review. We

generally apply a “particularly deferential form of abuse-of-

discretion review . . . to the procedures used to arrive at the sentence”

or, in other words, the procedural reasonableness of a sentence.

United States v. Davis,

, 195–96 (2d Cir. 2023) (internal

quotation marks and citation omitted). Procedural error occurs when

a district court “fails to properly calculate the Guidelines range or

rests its sentence on a clearly erroneous finding of fact,” among other

things. United States v. Osuba,

(2d Cir. 2023). “The

57 district court must find facts relevant to a sentencing enhancement by

a preponderance of the evidence.”

Where a defendant fails to raise

a particular challenge to the sentence before the district court,

however, we review that challenge for plain error.

Mensah argues that he objected to the calculation of the loss

amount before the district court, albeit in the restitution context.

Thus, in his view, the abuse-of-discretion standard applies (with

subsidiary factual findings reviewed for clear error). In contrast, the

government argues that the more rigorous plain-error standard

applies because Mensah did not specifically object to the loss amount

enhancement. We need not decide which party’s view is correct,

because even under the standard advocated by Mensah, his argument

fails.

The district court did not clearly err as a factual matter when it

found the loss amount to be $177,345, which was consistent with Lin’s

trial testimony. As a preliminary matter, evidence at trial showed that

58 during all relevant times, Mensah lived in New Jersey, and that all the

children he served lived in New York City. Since the EIP required

providers to administer services in person, Mensah would have had

to travel into New York City to provide services.

During trial, Lin comprehensively testified about her analysis

underpinning the loss amount. She testified that she reviewed data

indicative of Mensah’s activities and locations, including the LPR data

for the license plate associated with the Silver Lexus and Mensah’s

cellphone records. Cross-referencing those records with the

purported sessions reflected in Mensah’s session notes, Lin identified

conflicts and catalogued each one in a spreadsheet.

Lin’s conflicts indicated that it was highly unlikely that Mensah

was providing services at the time and place he reported he was in

his session notes. She described specific examples of conflicts she

identified, which were corroborated by other evidence, such as text

messages from Mensah’s cellphone. The text messages often left no

59 room for imagination as to where Mensah was for the day. For

instance, Lin recorded conflicts for Mensah’s purported sessions on

July 6, 2015, based on her analysis that the Silver Lexus did not enter

New York City that day. A text message sent from Mensah’s

cellphone on the same day provided, “I did not come to New York

today.” App’x at 685.

Mensah attacks Lin’s analysis as speculative, arguing that the

LPR data did not show who was driving the Silver Lexus. But there

was ample other evidence in the record that Mensah owned and

drove the Silver Lexus. According to NJMVC records, the Silver

Lexus was the only car registered to Mensah during the relevant time.

Mensah’s ex-wife testified that she sold him the Silver Lexus, and that

she and her son do not use that car. A gas station attendant and

Cecilio both observed Mensah driving the Silver Lexus. And

although Mensah testified that he took other transportation to New

York City, he did not suggest that anyone else drove the Silver Lexus.

60 Furthermore, at least two of the parents who testified indicated that

Mensah drove to sessions. It was permissible for the court to infer

based on this cumulative evidence that Mensah drove his Silver Lexus

to New York City for his sessions.

Mensah makes a similar argument with respect to the cellphone

records, arguing that there is no direct evidence that he was the one

using his cellphone. But circumstantial evidence will do. See United

States v. Coppola,

(2d Cir. 2012) (finding no

procedural error in the trial court’s loss calculation where the

evidence “permit[ted] the district court to make a reasonable estimate

of the loss given the available information” (internal quotation marks

and citation omitted)). The court could permissibly infer that it was

likely the owner of the phone—Mensah—who was using the phone,

especially since Mensah does not point to anyone else who used his

cellphone. Mensah takes issue with the fact that Lin recorded

conflicts based on the use of the phone for more than fifteen minutes

61 when a purported session was ongoing. He argues that such

conversations might have been related to social work services. Lin,

however, did not rely solely on the duration of the phone usage to

record a conflict. She only recorded conflicts with respect to phone

usage when there was also independent evidence suggesting that

Mensah was not providing services.

Accordingly, the district court did not err by applying the ten-

level loss amount enhancement under U.S.S.G. § 2B1.1.

III. Conclusion

In sum, we hold as follows:

1. The district court did not abuse its discretion by permitting a juror whose sister’s friend’s husband was a government witness to remain on the jury. That connection was too attenuated to give rise to a presumption of bias. Nor did the district court plainly err in its voir dire of that juror, as it adequately screened for any actual bias arising from his acquaintance with the government’s witness.

2. The district court did not err in denying Mensah’s motion for a new trial based on the prosecutor’s remark before the jury, during cross-examination of a trial witness, that Mensah had the ability to testify. Any prejudice to Mensah’s

62 Fifth Amendment right not to testify was rendered moot when Mensah later elected to testify on his own behalf. And the record does not suggest that the comment somehow compelled or coerced Mensah to testify.

3. The district court did not clearly err in finding that Mensah billed $177,345 to the agencies for treatment sessions that he had not actually provided, and therefore did not err in applying the ten-level loss amount enhancement under U.S.S.G. § 2B1.1.

Accordingly, we AFFIRM the district court’s judgment.

63