United States of America v. Albert William Upshur

PRECEDENTIAL
UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT
__________
No. 21-3281
__________
UNITED STATES OF AMERICA
v.
ALBERT WILLIAM UPSHUR,
Appellant
__________
On Appeal from the United States District Court
for the Eastern District of Pennsylvania
(District Court No. 2-18-cr-00124-002)
Honorable Wendy Beetlestone, U.S. District Judge
__________
Submitted Under Third Circuit L.A.R. 34.1(a)
on May 5, 2022
Before: KRAUSE, BIBAS, and AMBRO, Circuit Judges
(Filed: May 8, 2023)
Thomas A. Dreyer
30 Running Brook Road
Glen Mills, PA 19342
Counsel for Appellant

-- 1 of 6 --

2
Katie Bagley
Hannah Cook
Samuel R. Lyons
Joseph B. Syverson
United States Department of Justice
Tax Division
P.O. Box 972
Washington, DC 20044
Counsel for Appellee
__________
OPINION OF THE COURT
__________
KRAUSE, Circuit Judge.
When it comes to punishing fraud crimes under the
United States Sentencing Guidelines, it is typically the case
that the greater the monetary loss is, the harsher the penalty.
But what is the rule if the loss is merely intended, not actual?
We recently answered this question for theft offenses in United
States v. Banks, holding that in the absence of Guideline text
extending “loss” to intended loss, U.S.S.G. § 2B1.1’s loss table
was properly interpreted to reach only actual loss. 55 F.4th
246, 258 (3d Cir. 2022).
From this precedent, Appellant Albert William Upshur
urges that we likewise limit the tax loss table at U.S.S.G.
§ 2T4.1 to actual loss, which would require his resentencing at
a lower offense level. But the texts of U.S.S.G. §§ 2T1.1 and
2T1.4, which prescribe the Guidelines calculation for tax fraud,
make plain that § 2T4.1’s loss table covers not only the actual
loss to the United States Treasury, but also the loss the
perpetrator intended. Thus, the District Court relied on the

-- 2 of 6 --

3
proper base offense level, and we will affirm the sentence it
imposed.
I. BACKGROUND
Upshur and co-defendant Yolanda Thompson were
indicted in 2018 for engaging in two fraudulent tax schemes
over the course of a decade.
In one, they operated a trust and told people that they
could pay off their debts if they wired certain fees to Upshur
and allowed the defendants to file tax forms representing that
the Trust had withheld significant amounts of income tax on
their behalf, hopefully yielding sizable refunds. This scheme
required participants to pay $500 to join and $250 for each debt
they wanted to pay off, as well as 20% of any payout they
received from the IRS. Upshur and Thompson participated in
it themselves as well. Though this scheme was largely
unsuccessful, the IRS did issue one participant a $1.5 million
refund in 2011. Even then, the IRS realized its mistake and
was able to freeze the payment.
In a second scheme, they made large fraudulent tax
overpayments to the IRS, hoping to generate refunds. This
scheme apparently did not generate any payments from the
IRS, but together with the first scheme, Upshur’s and
Thompson’s conspiracy resulted in over $325 million in
fraudulent tax claims.
Proceeding pro se, Upshur waived his right to a jury
trial and was convicted at a bench trial of conspiracy to defraud
the United States and eight counts of aiding and assisting in the
preparation of false tax returns, in violation of 18 U.S.C. § 371
and 26 U.S.C. § 7206(2), respectively. When it came to
sentencing Upshur, the District Court recognized there was no
actual loss to the United States Treasury, and calculated
Upshur’s base offense level under U.S.S.G. § 2T1.4, the
Guideline for aiding and assisting tax fraud, using the
intended-loss figure of $325 million. Applying the tax loss
table at U.S.S.G. § 2T4.1, that loss figure produced a base
offense level of thirty-four, which, with a two-point upward
adjustment for obstruction under U.S.S.G. § 3C1.1 and
Upshur’s Criminal History Category of VI, resulted in a

-- 3 of 6 --

4
Guidelines range of 324 to 348 months’ imprisonment. The
Court, however, imposed a below-Guidelines sentence of
eighty-four months.
Upshur filed a notice of appeal, but originally did not
file an appellate brief. Instead, his standby counsel moved to
withdraw, pursuant to Anders v. California, 386 U.S. 738
(1967), and Third Circuit L.A.R. 109.2(a). After we published
our opinion in Banks, however, Upshur’s counsel diligently
identified a non-frivolous issue it offered for appeal and sent a
letter to the Court pursuant to Federal Rule of Appellate
Procedure 28(j), arguing that Banks required reversal of his
client’s sentence. We treated that letter as a motion to
withdraw his Anders brief, which we granted, and merits
briefing ensued.
II. JURISDICTION AND STANDARD OF REVIEW
The District Court had jurisdiction under 18 U.S.C.
§ 3231. We have jurisdiction under 28 U.S.C. § 1291.
We review a district court’s interpretation of the
Sentencing Guidelines de novo. United States v. Seibert, 971
F.3d 396, 399 (3d Cir. 2020) (citation omitted). But where a
defendant fails to preserve the issue in the district court, we
review only for plain error. United States v. Scarfo, 41 F.4th
136, 192 (3d Cir. 2022) (citation omitted). We review a district
court’s factual findings for clear error. Seibert, 971 F.3d at 399
(citation omitted).
III. DISCUSSION
Upshur’s argument on appeal is that the “spirit” of our
intervening decision in Banks requires resentencing because
the District Court erred in relying on intended loss in
calculating his base offense level. Opening Br. at 19. Below,
we first summarize our decision in Banks and the reasons we
concluded that the theft loss table at § 2B1.1 was limited to
actual loss. We then consider the implications of Banks for the
tax loss table at § 2T4.1.

-- 4 of 6 --

5
A. Banks and the Theft Loss Table
Banks involved the sentencing enhancement table for
theft offenses at § 2B1.1. The defendant had attempted to
make hundreds of thousands of dollars’ worth of fraudulent
deposits into and withdrawals from an international exchange
system, but he was unsuccessful. Banks, 55 F.4th at 251. The
district court, in reliance on the application notes to § 2B1.1,
concluded that the table covered both a scheme’s “actual loss”
and its “intended loss,” and therefore applied a 12-point
enhancement based on the amount of money Banks had
attempted to withdraw from the exchange. Id. at 253.
In view of Kisor v. Wilkie, 139 S. Ct. 2400 (2019), we
reversed, holding that the application notes were not entitled to
the deference the district court accorded them and that the plain
text of § 2B1.1 controlled. See Banks, 55 F.4th at 255–57.
Because the “ordinary meaning of ‘loss’ in the context of
§ 2B1.1 is ‘actual loss,’” we remanded for resentencing. Id. at
257–58; see also United States v. Kousisis, Nos. 19-3679 and
19-3774, Slip Op. at 22 & n.97 (3d Cir. Apr. 21, 2023)
(applying Banks to the § 2B1.1 loss calculation in the context
of benefits-program fraud).
B. The Application of Banks to Upshur’s Case
Because Upshur did not argue a similar interpretation of
§ 2T4.1 in the District Court, we review that argument on
appeal only for plain error. See Fed. R. Crim. P. 52(b). Thus,
Upshur can only prevail if (1) there was an error, (2) which was
plain, and (3) affected his substantial rights. United States v.
Johnson, 899 F.3d 191, 200 (3d Cir. 2018) (citation omitted).
In making these assessments, we consider the implication of
Banks because plain error means error under controlling law at
the time of direct appeal. Id. at 199–200.
Despite the superficial similarities between the loss
tables at § 2B1.1 and § 2T4.1, Upshur’s argument founders at
the first step of the plain error test because the District Court
did not err at all. Unlike § 2B1.1, § 2T1.4—and § 2T1.1, to
which § 2T1.4 refers for its definition of “tax loss”—
encompass both actual and intended monetary losses. In
Banks, our reasoning was that the text of § 2B1.1 did not

-- 5 of 6 --

6
differentiate between actual and intended monetary loss, and,
in the absence of a specialized definition, “loss” must carry its
“ordinary meaning” of “actual loss.” 55 F.4th at 257–58. But
§ 2T1.1, in contrast, provides a detailed, capacious definition
of “tax loss”: In the context of tax crimes, “loss is the total
amount of loss that was the object of the offense (i.e., the loss
that would have resulted had the offense been successfully
completed).” U.S.S.G. § 2T1.1(c)(1). That definition by its
terms encompasses both actual and intended losses from tax
fraud schemes, and § 2T1.4 expressly applies that definition to
“the tax loss . . . resulting from the defendant’s aid, assistance,
procurance or advice.” Id. § 2T1.4(a).
Upshur attempts to circumvent this clear textual
directive by arguing that he and Banks were similarly situated
in that they both “planned and executed a scheme to steal
money.” Opening Br. at 10. In the broadest sense, that may
be true, but Upshur’s crimes of conviction are governed by a
different Sentencing Guideline than Banks’s, and, in contrast
to § 2B1.1, which covered Banks’s offense, § 2T1.4 uses a
definition of “loss” that unambiguously includes both actual
and intended losses.1
In sum, the District Court did not err in relying on
Upshur’s $325 million in intended losses to calculate his base
offense level, and resentencing is not required.
IV. CONCLUSION
For the foregoing reasons, we will affirm the judgment
of the District Court.
1 Because we conclude that the text of § 2T1.1(c)(1) is
unambiguous, we need not go further and examine its
“structure, history, and purpose” or determine if the relevant
Guidelines Commentary merits Auer deference. See Kisor,
139 S. Ct. at 2415.

-- 6 of 6 --