United States v. Williams

[PUBLISH]

IN THE UNITED STATES COURT OF APPEALS

FOR THE ELEVENTH CIRCUIT FILED
U.S. COURT OF APPEALS
____________________________ ELEVENTH CIRCUIT
AUGUST 5, 2003
No. 02-12234 THOMAS K. KAHN
CLERK
____________________________

D.C. Docket No. 00-06312-CR-NCR

UNITED STATES OF AMERICA,

Respondent-Appellee,

versus

JERMAINE C. WILLIAMS,
a.k.a. Jermaine Williams,

Petitioner-Appellant.

__________________________

Appeal from the United States District Court
Southern District of Florida
__________________________

(August 5, 2003)

Before TJOFLAT, ANDERSON and CUDAHY*, Circuit Judges.
________________
* Honorable Richard D. Cudahy, United States Circuit Judge for the Seventh Circuit, sitting by
designation.
TJOFLAT, Circuit Judge:

I. STATEMENT OF THE CASE

A. Factual Background

On October 20, 2000, Jermaine Williams and Lowen Espinueva attempted

to rob an armored car in Coral Springs, Florida. As the armored car’s drivers

(Frank Granja and Eshaman Ruiz) were restocking an ATM with cash, Williams

and Espinueva drove up and started shooting at them. Espineuva used a 12-gauge

shotgun, while Williams was firing a 9-millimeter pistol; both guards were

seriously wounded. When the guards returned fire, Williams and Espinueva fled

without stealing any money.

Williams and Espinueva were later arrested and indicted on three counts:

conspiring to obstruct interstate commerce through robbery (Count I),1 attempting

1
Federal law provides,
Whoever in any way or degree obstructs, delays, or affects
commerce or the movement of any article or commodity in
commerce, by robbery or extortion or attempts or conspires so to
do, or commits or threatens physical violence to any person or
property in furtherance of a plan or purpose to do anything in
violation of this section shall be fined under this title or imprisoned
not more than twenty years, or both.
18 U.S.C. § 1951(a) (1994) (emphasis added). The statute defines robbery as “the unlawful
taking or obtaining of personal property from the person or in the presence of another, against his
will, by means of actual or threatened force, or violence, or fear of injury, immediate or future, to
his person or property, or property in his custody or possession . . . .” Id. § 1951(b)(1).

2
to obstruct interstate commerce through robbery (Count II),2 and discharging a

firearm in connection with a crime of violence (Count III).3 Williams pled guilty

to all three counts and was sentenced to 200 months in prison by the United States

District Court for the Southern District of Florida.4 This includes 80 months for

Counts I and II, and a mandatory consecutive 120-month sentence for Count III.5

This appeal concerns Williams’s prison sentence for Counts I and II.

B. Williams’s Sentence

Under the sentencing guidelines, the various charges of which a defendant

is convicted are sorted into different “groups” based on the rules set forth in

United States Sentencing Commission, Guidelines Manual, § 3D (Nov. 2002). In

general, related charges are supposed to be grouped together, while charges arising

from separate incidents are supposed to be grouped apart from each other. “In

2
Id. § 1951(a) (establishing penalties for “[w]hoever . . . attempts . . . [to] obstruct[],
delay[], or affect[] commerce or the movement of any article or commodity in commerce, by
robbery or extortion”).
3
Federal law provides, “[A]ny person who, during and in relation to any [federal] crime
of violence . . . uses or carries a firearm, or who, in furtherance of any such crime, possesses a
firearm, shall, in addition to the punishment provided for such crime of violence . . . if the
firearm is discharged, be sentenced to a term of imprisonment of not less than 10 years.” Id. §
924(c)(1)(A)(iii) (1994).
4
The court also imposed a $300 special assessment against Williams, and ordered
restitution of $404,080.29.
5
See supra note 3.

3
essence, counts that are grouped together are treated as constituting a single

offense for purposes of the guidelines.” Id.,ch.3, pt.D, introductory cmt. (2002).

Each group is assigned a numerical “offense level,” which is determined by the

most serious offense in that group. See id. § 3D1.3(a). Based on the number of

groups the defendant has, as well as each group’s offense level, the defendant is

assigned a “combined offense level,” which is used to determine his sentence. See

id. § 3D1.4. As a result of this system, a defendant will receive a much higher

sentence if two crimes are grouped separately than if they are grouped together.6

Section 3D1.2 of the Sentencing Guidelines states that multiple offenses may be

grouped together only “[w]hen counts involve the same victim and the same act or

transaction,” U.S.S.G. § 3D1.2(a), or “[w]hen counts involve the same victim and

two or more acts or transactions connected by a common criminal objective or

constituting part of a common scheme or plan,” id. § 3D1.2(b).7

6
If two or more offenses are grouped separately, each group is assigned an offense level,
and these individual offense levels will be considered in conjunction with each other to
determine the defendant’s “combined offense level.” This combined offense level will typically
be much higher than any of the groups’ individual offense levels.
If, in contrast, multiple offenses are grouped together, that group is assigned an offense
level based only on the most serious charge it contains. The other offenses in that group are
essentially ignored and do not cause the defendant to have a higher “combined offense level.”
Thus, it is to a defendant’s advantage to have as many charges as possible squeezed into as few
groups as possible.
7
The other circumstances under which multiple counts may be grouped together are not
pertinent to the instant case.

4
 The Pre-Sentence Investigation Report (“PSI”) completed by Williams’s

probation officer recommended that Counts I and II be assessed separately under

the guidelines rather than grouped together. The report claimed, “Counts One and

Two represent separate harms and are specifically excluded from grouping rules in

§ 3D1.2. . . . Group One will represent the harm caused to Frank Granja and

Group Two will represent the harm caused to Eshaman Ruiz.” Williams filed an

objection to the report, stating “Counts One and Two of the indictment, which

charge a conspiracy to commit a robbery and an attempt to commit a robbery, are

part of a single criminal episode and should be grouped together, pursuant to §

3D1.2(b).” The probation officer again declined to group Counts I and II together,

emphasizing that “[t]he counts cannot be grouped under either § 3D1.2(a) or (b)

since neither count involved the same ‘victim.’ ”

Williams then filed an objection to the amended report with the district

court, arguing yet again that Counts I and II (conspiracy to commit robbery and

the attempted robbery) should have been grouped together. The district court,

without explanation, overruled Williams’s objection and adopted the

recommendation of the PSI, putting each charge in a separate group in calculating

Williams’s sentence. Because Counts I and II were grouped separately,

Williams’s sentencing range for these counts was 78-97 months; the judge

5
sentenced him to 80 months (plus a mandatory consecutive 120-month sentence

for Count III). Had Counts I and II been grouped together, Williams would be

eligible for a sentence of between 63-78 months, in addition to his mandatory

sentence for Count III.

II. STANDARD OF REVIEW

Before turning to the substance of Williams’s claims, it is first necessary to

determine the appropriate standard of review. Federal law states, “The court of

appeals . . . shall accept the findings of fact of the district court unless they are

clearly erroneous and, . . . shall give due deference to the district court's

application of the guidelines to the facts.” 18 U.S.C. § 3742(e). Our precedent

clearly follows the first part of this statute, requiring us to review district courts’

factual findings under a “clear error” (or “clearly erroneous”) standard.8 See

United States v. Maung, 267 F.3d 1113, 1118 (11th Cir. 2001) (“When a

defendant challenges the district court’s application of the sentencing guidelines,

we review the district court’s underlying findings of fact for clear error . . . .”).

8
Because the statute is silent as to the standard of review for “pure” questions of law, we
review them under our customary de novo standard. See, e.g., United States v. Calderon, 127
F.3d 1314, 1339 (11th Cir. 1997) (“In reviewing challenges to the application of U.S.S.G. §
3B1.3, we review the legal meaning of the term ‘special skills’ de novo.”) (internal quotations
omitted); United States v. Brenson, 104 F.3d 1267, 1287 (11th Cir. 1997) (“[T]he question of
whether conduct by a grand juror justifies the ‘abuse of trust’ enhancement is a legal conclusion
requiring a de novo review.”).

6
For example, “[a] district judge’s attribution of drugs to a particular defendant

under the Sentencing Guidelines is subject to clearly erroneous review.” United

States v. Alred, 144 F.3d 1405, 1416 (11th Cir. 1998). Similarly,“[p]ossession of

a firearm for sentencing purposes is a factual finding . . . reviewed under a clearly

erroneous standard.” United States v. Geffrard, 87 F.3d 448, 452 (11th Cir. 1996).

The proper standard of review for a district court’s application of the

sentencing guidelines to the facts of a particular case, also referred to as “mixed

questions of fact and law” involving the guidelines, is much less settled. Section

3742's “due deference” language has proven to be a source of great ambiguity

within this circuit, and our caselaw is, to say the least, muddled. Subpart A begins

by explaining the various approaches this circuit has taken in considering this

matter. Subpart B goes on to explain how “due deference” should not be

interpreted as establishing a fixed quantum of review, but instead requires varying

degrees of deference to the lower courts based on the precise nature of the

guideline provision at issue. Subpart C applies these principles to determine the

appropriate standard of review in the instant case.

A. Prior Interpretations of § 3742(e)’s “Due Deference” Language

18 U.S.C. § 3742(e) requires the court of appeals to accord “due deference”

7
to a district court’s ruling on how the sentencing guidelines should be applied to

the facts before it. Our precedents have not implemented this provision in a

wholly consistent manner. This Subpart attempts to explain our circuit’s various

interpretations of the proper standard of review for guidelines cases.

1. Ignoring 18 U.S.C. § 3742(e) - We have characterized “question[s]

about whether a particular guideline applies to a given set of facts” both as

questions of law, see United States v. Shriver, 967 F.2d 572, 574 (11th Cir. 1992),

and (more appropriately) as mixed questions of fact and law, see United States v.

Scroggins, 880 F.2d 1204, 1206 n.5 (11th Cir. 1989). In general, de novo review

is appropriate for both types of issues. See, e.g., Parker v. Sec’y for the Dep’t of

Corr., 331 F.3d 764, 765 (11th Cir. 2003) (“We review the district court’s findings

of fact for clear error and its legal conclusions and mixed questions of law and fact

de novo.”).

It is possible that prior panels simply applied these general principles and,

as a result, overlooked § 3742(e) and its “due deference” standard. Consequently,

many of our cases simply state, “This court reviews the district court’s sentencing

hearing findings of fact for clear error and its application of the sentencing

guidelines to those facts de novo.” United States v. Renick, 273 F.3d 1009, 1021

8
(11th Cir. 2001); accord United States v. De La Mata, 266 F.3d 1275, 1302 (11th

Cir. 2001); Maung, 267 F.3d at 1118; United States v. Jamieson, 202 F.3d 1293,

1295 (11th Cir. 2000); United States v. Gallo, 195 F.3d 1278, 1280 (11th Cir.

1999); United States v. Herrera, 931 F.2d 761, 762 (11th Cir. 1991). Even if, in

these cases, § 3742's due deference standard was interpreted to mean de novo

review, it is both confusing and improper for a court of appeals, or the parties

appearing before it, to fail to cite this statute (or a case interpreting this statute) in

a sentencing guidelines appeal.

By its very terms, § 3742 clearly applies to sentencing guidelines cases.

Moreover, we reject the notion that Congress intended § 3742's due deference

standard to leave undisturbed our general de novo standard for reviewing issues

concerning the application of the law to particular sets of facts. This “due

deference” language was undoubtedly intended to require, at least in some cases,

greater respect for district court holdings than is traditionally accorded under de

novo review. See United States v. Malone, 78 F.3d 518, 520-21 n.2 (11th Cir.

1996) (“The ‘due deference’ standard in 18 U.S.C. § 3742 ‘serves as an additional

caution against overly intense judicial review.’ ” (quoting United States v. Mejia-

Orosco, 868 F.2d 807, 808 (5th Cir. 1989))).

9
 2. “Due deference” always means de novo review - Many of our cases,

despite quoting the due deference standard from § 3742(e), nevertheless proclaim

in sweeping terms that “we review the district court’s application of the

Sentencing Guidelines de novo.” United States v. Ortiz, 318 F.3d 1030, 1036

(11th Cir. 2003) (internal citation omitted); see also United States v. Hall, 312

F.3d 1250, 1260 n.12 (11th Cir. 2002); United States v. Singh, 291 F.3d 756, 763

(11th Cir. 2002); United States v. Smith, 231 F.2d 800, 806-07 (11th Cir. 2000).

As noted above, we refuse to interpret the phrase “due deference” as used in §

3742 as establishing a blanket de novo standard. While, for reasons explained later

in this opinion, due deference may have required some panels to apply de novo

review in certain cases (given their facts and the particular guidelines at issue),

those cases should not be read as establishing a general proposition that due

deference always means de novo.

3. Due deference always means clear error review - At least one of our

cases has also strayed to the other extreme, holding “[W]e review for clear error

the district judge’s application of the guidelines to the facts of the case.”

Calderon, 127 F.3d at 1339. We also arguably equated the due deference and clear

error standards in United States v. De Varon, 175 F.3d 930 (11th Cir. 1999) (en

10
banc), where we held, “[A] district court’s determination of a defendant’s role in

the offense is a finding of fact to be reviewed only for clear error. . . . [It is] a

fundamentally factual determination entitled to due deference and not a legal

conclusion subject to de novo review.” Id. at 937-38. These quotes might suggest

that the clear error and due deference standards were interchangeable for

reviewing a district court’s findings of fact.

It is clearly inconsistent with the overwhelming weight of the precedents

discussed above to conclude that due deference always means clear error, or that

clear error is always the appropriate standard for reviewing a district court’s

application of the guidelines to a particular case. Moreover, such conclusions are

contrary to the clear text of § 3742, which states that although a district court’s

factual determinations should be reviewed under a “clear error” standard, its

application of the law should be given only “due deference.” This deliberate

variation in terminology within the same sentence of a statute suggests that

Congress did not interpret the two terms as being equivalent. See United States v.

Bean, 531 U.S. 71, 123 S. Ct. 584, 587 n.4, 154 L. Ed. 2d 483, 489 n.4 (2002)

(“The use of different words within related statutes generally implies that different

meanings were intended.” (quoting 2A N. Singer, Sutherland on Statutes and

Statutory Construction § 46.06, at 194 (6th ed. 2000))). While clear error review

11
may have been warranted for the guideline at issue in Calderon, our precedents

should not be read as establishing a uniform rule of always equating the due

deference standard with clear error review.

4. Due deference is either a fixed standard of review independent of the de

novo and clear error standards, or a type of de novo review - Many of our cases set

forth our general de novo standard for reviewing mixed questions of fact and law,

then cite § 3742's “due deference” standard for mixed questions involving

application of the sentencing guidelines. For example, United States v. Yount

states, “[W]e give due deference to the district court’s application of the guidelines

to the facts. . . . The district court’s application of law to the facts is reviewed de

novo.” 960 F.2d 955, 956 (11th Cir. 1992); accord United States v. Geffrard, 87

F.3d 448, 452 (11th Cir. 1996) (“[T]he application of the law to the facts found is

reviewed de novo. The application of the guidelines to the facts, however, is

entitled to ‘due deference.’ ” (citation omitted)); Singh, 291 F.3d at 763.9

These passages are misleading on several grounds. First, in distinguishing

between application of “the guidelines” and application of “the law,” they obscure

9
This is essentially the approach for which the Government advocated in the instant case.
See Brief for the United States, at *8 (“This Court reviews the district court’s . . . application of
the sentencing guidelines de novo. Furthermore, this Court reviews the district court’s refusal to
group multiple counts under USSG § 3D1.2 [sic] with due deference.”) (internal citations
omitted).

12
the fact that the guidelines are the law. Second, the passages can be interpreted in

two different ways, both of which are erroneous. On the one hand, one might

interpret this passage to mean that we review de novo all mixed questions of fact

and law, including questions involving application of the guidelines, but in the

course of doing so we somehow accord the district court’s ruling some special

degree of deference. This is apparently the approach we took in United States v.

Hall, where we held, “Although this Court, as part of its de novo review, gives due

deference to the district court's application of the guidelines to the facts, this Court

is not bound by the district court's application of the guidelines to the facts and our

review remains de novo.” 312 F.3d 1250, 1262 n.16 (11th Cir. 2002) (emphasis

added). This understanding is also conveyed by cases such as United States v.

McIntosh wherein we held, “We review the district court’s application of the

sentencing guidelines de novo, giving due deference to the district court’s refusal

to group multiple counts under U.S.S.G. § 3D1.2.” 216 F.3d 1251, 1253 (11th

Cir. 2000).

It is unclear how we can give due deference to a district court’s conclusions

when applying de novo review, since de novo review requires us to look at a

question as if we are the first court to consider it. Put simply, it is definitionally

impossible to give deference of any sort to a decision being reviewed de novo.

13
Moreover, none of these cases explain how a de novo review involving “due

deference” to the district court would differ from a de novo review not involving

such deference. Thus, the Hall/McIntosh interpretation of the above-quoted

passage from Yount is both unenlightening and internally contradictory.

The only other possible interpretation is that while in general we review

mixed questions of fact and law de novo, we apply a special, independent “due

deference” standard when the mixed question involves application of the

sentencing guidelines. This would imply that due deference is a standard of

intermediate scrutiny falling somewhere between de novo and clearly erroneous.

This interpretation is not only a stretch, given the text of those opinions, but is also

erroneous. For the reasons discussed in the next Subpart, we do not believe that

Congress intended the phrase “due deference” to establish a new, independent

standard of review.

B. The Correct Interpretation of § 3742(e)’s “Due Deference” Language

This Subpart attempts to harmonize our precedents concerning the proper

standard of review for a district court’s application of the sentencing guidelines to

particular factual scenarios (“mixed questions of fact and law” involving the

guidelines). We begin with Supreme Court decisions interpreting § 3742(e) that

14
most of our precedents have overlooked. We then turn to a handful of opinions

within this circuit that have interpreted § 3742(e) in accordance with both the

statute’s clear language as well as these Supreme Court precedents. Finally, in the

absence of a case that can clearly be identified as our panel’s earliest definitive

ruling on this subject, we set forth general principles to guide future panels in

resolving these complex issues.

1. Supreme Court precedent - Two Supreme Court cases shed light on §

3742(e). In Koon v. United States, 518 U.S. 81, 98, 116 S. Ct. 2035, 2046, 135 L.

Ed. 2d 392 (1996), the Court held, while interpreting § 3742(e)’s “due deference”

requirement,

[t]he deference that is due depends on the nature of the question
presented. The district court may be owed no deference, for instance,
when the claim on appeal is that it made some sort of mathematical
error in applying the Guidelines; under these circumstances, the
appellate court will be in as good a position to consider the question
as the district court was in the first instance.

Five years later, in Buford v. United States, 532 U.S. 59, 121 S. Ct. 1276, 149 L.

Ed. 2d 197 (2001), the Court applied Koon’s “sliding scale” understanding of “due

deference.” The Buford Court considered the degree of deference due a district

court’s holding that a defendant’s prior convictions had not been “functionally

consolidated,” to determine whether a § 4B1 enhancement for repeat offenders

15
was appropriate. Id. at 63, 121 S. Ct. at 1279.

The Buford Court upheld the Seventh Circuit’s decision to apply a

particularly deferential standard of review, akin to “clearly erroneous,” to this type

of decision. It stated, “[T]he district court is in a better position than the appellate

court to decide whether a particular set of individual circumstances demonstrates

‘functional consolidation’ [of prior sentences, because] . . . [a]s a trial judge, a

district judge is likely to be more familiar with trial and sentencing practices in

general, including consolidation procedures.” Id. at 64, 121 S. Ct. at 1280. The

Court also emphasized:

The legal question at issue is a minor, detailed, interstitial question of
sentencing law . . . . That question is not a generally recurring, purely
legal matter, such as interpreting a set of legal words, say, those of an
individual guideline . . . . Nor is that question readily resolved by
reference to general legal principles and standards alone. Rather, the
question at issue grows out of, and is bounded by, case-specific
detailed factual circumstances.

Id. at 65, 121 S. Ct. at 1280-81.

These cases lead to several conclusions. Koon tells us that “due deference”

is not a fixed standard of scrutiny akin to de novo or “clear error” review, but

instead establishes a “sliding scale.” When reviewing decisions involving the

application of certain types of guidelines, no deference will be due a district

court’s rulings, effectively leading to de novo review. With other guidelines, an

16
appellate court should accord the trial court’s decisions a high degree of deference,

reviewing those rulings under the substantive equivalent of a “clear error”

standard.

Buford shows us how to distinguish between guidelines requiring

significant deference and those requiring no deference. The two most important

factors are: (1) whether the ruling depends on a wide range of facts that are more

readily available to the district court than to the court of appeals, due to the district

court’s first-hand experience trying the case, and (2) whether the issue primarily

involves a legal interpretation of a guideline (suggesting de novo review), or

instead involves the application of a clearly-established, well-understood legal

standard or principle to a detailed fact pattern (indicating “clear error” review).

“Clear error” review is particularly appropriate in the latter type of cases because

“the fact-bound nature of the decision limits the value of appellate court

precedent.” Buford, 532 U.S. at 65-66, 121 S. Ct. at 1281. These principles are

more fully fleshed out with reference to our own precedents, to which we now

turn.

2. The Sliding Scale Interpretation of § 3742 in the Eleventh Circuit -

Despite the wide range of our circuit’s interpretations of § 3742's “due deference”

17
language, discussed in the previous Subpart, our precedents are not entirely

without support for the Supreme Court’s sliding scale approach. See, e.g., United

States v. Taylor, 88 F.3d 938, 942 (11th Cir. 1996) (recognizing that the meaning

of the term “due deference” depends on the particular guidelines at issue). One of

our early guidelines cases, United States v. Malgoza, 2 F.3d 1107 (11th Cir. 1993),

best exemplifies this understanding of § 3742. Citing the Fourth Circuit, we held,

“To say that we review a finding with due deference means that we give the

district court the deference that is due in regard to that finding. The deference due

will depend upon whether the determination is primarily factual or legal.” Id. at

1109 (citing United States v. Daughtrey, 874 F.2d 213, 217 (4th Cir. 1989)). We

explained,

When the district court's application of sentencing guidelines to facts
involves primarily a legal decision, such as the interpretation of a
statutory term, less deference is due to the district court than when the
determination is primarily factual. . . . In contrast, when an
application of the guidelines to facts is closer to a pure question of
fact, the appellate court will review the application with the deference
due under the clearly erroneous standard.

Id. at 1109-10 (citations omitted). This holding reflects the principles articulated

above. Where a determination turns primarily on the evaluation of facts (such as a

witness’s credibility, intonation, and demeanor) that are more accessible to the

18
district court than to the court of appeals, we will defer to the district court’s

application of the law to those facts and apply “clear error” review. Similarly, a

case involving application of a fairly well-understood legal standard to a complex

factual scenario will be considered “primarily factual,” and be reviewed for clear

error. Most other cases, in contrast, will be reviewed de novo.

a. Specific types of cases in which de novo review is particularly

appropriate - Unless a case falls into one of the categories specified above, de

novo review of the district court’s application of the law to the facts of the case is

generally appropriate. But see infra Part I.B.3. Many (but certainly not all) cases

in which we apply de novo review involve legal issues for which most or all of the

relevant facts are contained in the counts in the indictment for which the defendant

has been convicted or pled guilty. For example, in United States v. Smith, 231

F.3d 800 (11th Cir. 2000), one of the defendants was the county’s deputy registrar

of voters. She was convicted of, among other things, fraudulently submitting an

absentee ballot on behalf of a voter without that voter’s knowledge or consent.

We did not have to go far beyond the allegations contained in the indictment to

affirm the district court’s ruling that this offense, under the circumstances,

involved an “abuse[] [of] a position of public or private trust.” Id. at 819 (quoting

U.S.S.G. § 3B1.3). Consequently, there was no special need to defer to the district

19
court’s assessments, and de novo review was appropriate. Id. at 806 (“We review

the district court’s application of the Sentencing Guidelines de novo . . . .”).

Similarly, in United States v. Singh, 291 F.3d 756, 762 (11th Cir. 2002), we

reviewed the district court’s interpretation of U.S.S.G. § 2F1.1(b)(6), which

provides for an offense-level enhancement if part of a fraud is committed outside

of the United States. Because it was undoubtedly apparent from the indictment

whether part of the fraud for which the defendant was convicted had occurred in a

different country, de novo review was appropriate. See Singh, 291 F.3d at 762

(citing United States v. Bradford, 277 F.3d 1311, 1312 (11th Cir. 2002)). Another

reason de novo review was appropriate in Singh was because our opinion focused

more on interpreting the guideline’s language than on digging through a complex

factual record.

De novo review is also warranted in cases where we must determine

whether the district court applied the correct sentencing guideline (or subsection

of a sentencing guideline) for the defendant’s underlying conduct. See, e.g.,

United States v. De La Mata, 266 F.3d 1275, 1302 (11th Cir. 2001) (reviewing de

novo the district court’s decision as to whether the “money laundering” or “fraud”

sentencing guideline was more appropriate, given the facts of the case); Smith,

231 F.3d at 806-07, 818-19 (applying de novo review and upholding district

20
court’s decision to apply U.S.S.G. § 2H2.1(a)(2) instead of § 2H2.1(a)(3) because

“[t]he offenses for which [defendants] were convicted involved . . . the forging of

other voters’ names on applications of absentee ballot and affidavits of absentee

voter”). Such matters turn primarily on the contents of the indictment and

interpretation of the sentencing guidelines, areas for which we are not especially

dependent on district courts.

b. Specific types of cases in which “clear error” review is particularly

appropriate - A key factor in determining whether “due deference” requires us to

review a district court’s application of the guidelines in a particular case is the

degree to which its ruling is based upon a variety of “intangible” factors that a

district judge is in a better position than we are to assess. With certain types of

issues, a variety of considerations may have a significant influence on the judge’s

determination, yet cannot be adequately reflected in the cold, barren wasteland of

the record. For example, U.S.S.G. § 3C1.1 provides for a two-level enhancement

if the defendant engaged in obstruction of justice during the course of the

investigation, prosecution, or sentencing of his crime. One way in which a

defendant may obstruct justice is by perjuring himself at trial.10 See U.S.S.G. §

10
Perjury involves making a false statement “with the willful intent to provide false
testimony, rather than as a result of confusion, mistake, or faulty memory.” Singh, 291 F.3d at

21
3C1.1, cmt. n.4(b) (“The following is a non-exhaustive list of examples of the

types of conduct to which this adjustment applies: . . . committing, suborning, or

attempting to suborn perjury.”). An appellate court is not in a position to assess a

defendant’s demeanor, apparent sincerity, intonation, expression, gesticulations,

and a wide range of other considerations that are pertinent in determining whether

he has perjured himself. See also Owens v. Wainwright, 698 F.2d 1111, 1113

(11th Cir. 1983) (“Appellate courts reviewing a cold record give particular

deference to credibility determinations of a fact-finder who had the opportunity to

see live testimony.”); United States v. De Varon, 175 F.3d 930, 938 (11th Cir.

1999) (en banc) (“Intensely factual inquiries . . . are properly consigned to the

experienced discretion of the district judge.”); see also United States v. Abbell,

271 F.3d 1286, 1303 (11th Cir. 2001) (“[B]ecause the demeanor of the pertinent

juror is important to juror misconduct determinations, the district court is uniquely

situated to make the credibility determinations that must be made in cases like this

one: where a juror’s motivations and intentions are at issue.”).

Consequently, it is not only appropriate, but necessary, for us to accord

special deference to district court rulings under guidelines such as § 3C1.1 when

763 (quoting United States v. Dunnigan, 507 U.S. 87, 94, 113 S. Ct. 1111, 122 L. Ed. 2d 445
(1993)).

22
matters such as perjury are at issue (by reviewing the district court’s

determinations only for “clear error”). E.g., Singh, 291 F.3d at 764 (“[W]e cannot

say that the district court’s decision to enhance [defendant’s] adjusted offense

level pursuant to U.S.S.G. § 3C1.1 was clearly erroneous. Therefore, we affirm

the district court on this issue.”). However, because § 3C1.1 encompasses a wide

range of conduct which can obstruct justice, we need not examine all § 3C1.1

enhancements with special deference. For example, in United States v. Taylor, we

reviewed de novo a court’s two-level enhancement for obstruction of justice under

§ 3C1.1 for a defendant who had “repeatedly refused to provide the government

with handwriting exemplars, and when he finally did, h[ad] attempted to disguise

his writing.” 88 F.3d at 941-42. Unlike the perjury at issue in Singh, there was no

need for a particularized assessment of the credibility or demeanor of the

defendant in Taylor. The pertinent conduct (failing to provide the handwriting

exemplars) could clearly be set forth in detailed, non-conclusory findings by the

district court, which were accorded great deference. Consequently, we were in a

much better position to review de novo the district court’s application of the

obstruction of justice guideline to those facts than in a typical perjury case. Id. at

942. Thus, even with regard to application of a particular guideline, the degree of

deference “due” a district court can vary dramatically based on the allegations at

23
issue.

Because of the importance of determining a defendant’s apparent sincerity

through firsthand observation, we have also accorded special deference to district

court rulings concerning offense-level reductions under § 3E1.1 for a defendant’s

acceptance of responsibility, reviewing them under a “clear error” standard as

well. See, e.g., United States v. Brenson, 104 F.3d 1267, 1288 (11th Cir. 1997)

(“We review the district court’s decision as to acceptance of responsibility only for

clear error.”); United States v. Pritchett, 908 F.2d 816, 824 (11th Cir. 1990) (“The

district court is in a unique position to evaluate whether a defendant has accepted

responsibility for his acts, and the determination is entitled to great deference on

review. Unless a court’s determination is without foundation, it should not be

overturned on appeal.”).

Of course, these are only some of the most common types of cases where §

3742's “due deference” requirement will require us to apply a “clear error”

standard to district courts’ resolutions of mixed questions of fact and law

involving the guidelines. As discussed earlier, the Supreme Court’s ruling in

Buford, as well as our own holding in Malgoza, remind us that whenever a district

court’s determination is best characterized as the application of a clear-cut

standard to a complex set of facts, rather than a legal interpretation that can have

24
broader value beyond the instant case, “clear error” review is generally

appropriate.

3. Exception to the Two-Step Process of Review: “Dispositive” Factual

Findings - Some guidelines are fairly cut-and-dry; after reviewing the district

court’s factual findings for clear error, it is rather apparent whether or not a

particular enhancement applies. In these cases, the factual findings tend to be

rather concrete and specific (e.g., the defendant either had a firearm, or he didn’t),

or involve “lay” judgment calls that rely more on common-sense, everyday

assessments than on specialized legal knowledge (e.g., did the defendant play a

“major” or “minor” role in the criminal scheme).

Technically, of course, we must review a district court’s application of the

guidelines to the facts, even in these cases. As a practical matter, however, once

we come to a conclusion as to the district court’s factual findings, the question of

the applicability of a guideline of this type practically answers itself. We

frequently decline to specify a level of deference to accord the district court’s

application of these guidelines simply because the issue is essentially moot – the

factual findings all but determine the outcome. Thus, as a type of shorthand, this

court often characterizes questions of these guidelines’ applicability as reviewable

25
under a “clearly erroneous” standard. While the “clearly erroneous” standard

technically applies only to the district court’s factual findings, those findings tend

to be absolutely dispositive of issues involving these types of guidelines.

For example, in United States v. Alred, 144 F.3d 1405 (11th Cir. 1998), this

court reviewed a district court’s ruling imposing an offense-level enhancement

under U.S.S.G. § 2D1.1(b)(1) because the defendant possessed a firearm in

conjunction with his offense. We noted, “We review a sentencing judge’s factual

findings . . . for clear error.” Id. at 1420. After upholding the district court’s

finding that the defendant possessed a firearm in connection with his drug

distribution, we had no choice but to affirm its imposition of a two-level

enhancement under § 2D1.1(b). Id. (“We conclude that the facts that the district

judge used as the basis for the section 2D1.1(b)(1) enhancement were not clearly

erroneous. Thus, [the defendant’s] enhancement for possession of firearms was

appropriate.”); see also United States v. Geffrard, 87 F.3d 448, 452 (11th Cir.

1996) (“Possession of a firearm for sentencing purposes is a factual finding.”).

There was no need to get into the requirements of “due deference” or to review the

district court’s application of the law to those clear-cut facts. The court did not

have to make any independent legal determinations or draw any further

conclusions.

26
 Similarly, in United States v. De Varon, we held that “a district court’s

determination of whether a defendant qualifies for a minor role adjustment under

the Guidelines is a finding of fact that will be reviewed only for clear error.” 175

F.3d at 934. We explained, “[T]he ultimate determination of [the size of a

defendant’s] role in the offense is . . . a fundamentally factual determination

entitled to due deference and not a legal conclusion subject to de novo review.”

Id. at 938; see also United States v. Calderon, 127 F.3d 1314, 1341 (11th Cir.

1997) (“[W]e review the district judge’s factual determination [concerning the size

of the defendant’s role] for clear error.”). Of course, once we affirmed this factual

finding (under the “clear error” standard), the question of a downward adjustment

under § 3B1.2 (for defendants who play only a minor role) was essentially settled.

Again, there was no need for either us or the district court to draw further legal

conclusions.

This reasoning has also been applied on the other end of the spectrum.

Determinations as to whether defendants played “managerial” roles have been

treated as questions of pure fact that do not raise questions about the pertinent

guideline’s application and are reviewed under the clearly erroneous standard.

United States v. Glinton, 154 F.3d 1245, 1260 (11th Cir. 1998) (“The district

court’s determination is reviewed under the clearly erroneous rule.”).

27
 It is critical to emphasize that this type of analysis – essentially folding our

legal conclusions into the district court’s factual findings – is nothing more than a

shorthand we use when we feel there is nothing to be gained by looking at a

district court’s application of a clear-cut principle to a particular set of facts. In

the majority of cases, however, there is usually some sort of legal determination to

be made even after the fact-finding is done. As a result, our usual practice is to

examine the district court’s application of the law to the facts separately,

according this determination the level of deference it is due under § 3742 (as

discussed in the previous Subpart). See, e.g., United States v. Hall, 312 F.3d

1250, 1260-61 (11th Cir. 2002) (applying a de novo standard of review in

determining whether “the district court erred in failing to apply a four-level

sentence enhancement under U.S.S.G. § 2G2.2(b)(3) (2001)” for trafficking child

pornography that portrayed “sadistic conduct”); Singh, 291 F.3d at 762 (reviewing

de novo a district court’s decision to apply a two-level enhancement under

U.S.S.G. § 2F1.1(b)(6)(B) “because a substantial part of the fraudulent scheme of

which [the defendant] was convicted was committed from outside of the United

States”); Taylor, 88 F.3d at 942 (applying a de novo standard in reviewing the

district court’s application of U.S.S.G. § 2A6.1(b)(1) for defendants who engage

in “any conduct evidencing an intent to carry out [a] threat”). This line is

28
admittedly indistinct and uncertain, but we are bound by our precedents and so

must adhere to this doctrine.

4. A Concluding Note about Upward and Downward Departures - Of

course, nothing in this opinion is meant to affect the manner in which we review

district court rulings concerning upward or downward departures. “We review the

district court’s discretionary refusal to depart downward only if the district court

erroneously believed it did not have the statutory authority to do so.” De La Mata,

266 F.3d at 1303; see also Calderon, 127 F.3d at 1342 (“[D]ownward departure is

a matter of discretion that rests with the district judge and is ordinarily not

appealable. . . . [W]e simply have no jurisdiction to review the district judge’s

discretionary decision unless it was made based upon the belief that he did not

possess such discretion.”). In contrast, “[w]e review a district court’s decision to

[upwardly] depart from the Guidelines for an abuse of discretion.” United States v.

Melvin, 187 F.3d 1316, 1320 (11th Cir. 1999).

C. Standard of Review for This Case

Williams asks us to review the district court’s grouping of his offenses.

Since this question clearly involves the application of the guidelines to Williams’s

case, § 3742 requires us to review this question “with due deference” to the district

29
court’s ruling. See United States v. Bradford, 277 F.3d 1311, 1312 (11th Cir.

2002) (“[T]his Court views the district court’s refusal to group multiple counts

under § 3D1.2 with due deference.”); accord United States v. Tillmon, 195 F.3d

640, 642 (11th Cir. 1999); United States v. McIntosh, 216 F.3d 1251, 1253 (11th

Cir. 2000); United States v. Bonner, 85 F.3d 522, 525 (11th Cir. 1996).

Notwithstanding the implication of these cases – that “due deference” is either a

substantive standard in its own right, or a sort of modification on de novo review –

the preceding discussion makes clear that “due deference” requires us to assess the

type of question we are reviewing in order to determine the degree of respect to

accord the district court’s findings.

Applying our Malgoza standard, the question of whether charges should

have been grouped differently seems “primarily” one of law rather than of fact.

We need not rely on the district court’s factfinding, because the charges of which

the defendant was convicted are clearly set forth in the indictment. To resolve this

issue, we merely must interpret and apply § 3D1.2's standards to these charges.

Moreover, unlike Buford, this case does not involve an overly complex or unique

fact pattern that will render our decision of little value in future cases. Thus, we

need not afford the district court’s ruling any special deference, and we review its

ruling de novo.

30
 III.

Having established the proper standard of review, we now turn to the

substance of Williams’s claims. He argues that Counts I and II – for conspiring to

rob the armored car and attempting to rob the armored car, respectively – should

have been grouped together under either §§ 3D1.2(a) or (b). We consider each

section in turn.

A. Grouping Multiple Counts Under § 3D1.2(a)

U.S.S.G. § 3D1.2(a) is not applicable because multiple crimes must be

committed in the same general area and at approximately the same time to be

considered part of the “same act or transaction.” The guidelines provide the

following example:

The defendant is convicted of two counts of assault on a federal
officer for shooting at the same officer twice while attempting to
prevent apprehension as part of a single criminal episode. The counts
are to be grouped together. . . . But: The defendant is convicted of two
counts of assault on a federal officer for shooting at the officer on two
separate days. The counts are not to be grouped together.

Id. § 3D1.2, cmt. n.3 (emphasis in original). In virtually all cases, a conspiracy is

formed prior to the underlying offense it concerns, and in a place other than where

that offense is to be committed. Thus, the formation of the conspiracy and the

commission of the underlying crime are more akin to shooting the federal officer

31
on two separate days than shooting him twice “as part of a single criminal

episode.” Id. Only where the conspiracy and the underlying offense are roughly

contemporaneous with each other – that is, when they are in both temporal and

physical proximity with each other – are they part of the “same act or transaction”

as contemplated by § 3D1.2(a). As we held in United States v. Beard, 960 F.2d

965, 969 (11th Cir. 1992), “the passage of time between offenses, differences in

place of the offense, the nature of the offense, and the intervening circumstances

taken together may justify separate grouping of the offenses.”

B. Grouping Multiple Counts Under § 3D1.2(b)

Consequently, if Counts I and II are to be grouped together, it must be under

U.S.S.G. § 3D1.2(b) (requiring counts to be grouped when they “involve the same

victim and two or more acts or transactions connected by a common criminal

objective or constituting part of a common scheme or plan”). The conspiracy to

rob the armored car and the attempted robbery of that car were clearly parts of the

same “common plan or scheme.” In fact, the attempted robbery of the armored car

was the sole objective of the conspiracy.

Thus, the only real issue is whether these acts involved the same victim.

Inexplicably, the PSI concluded that these counts involved separate victims,

32
stating “Group One [including only Count I] will represent the harm caused to

Frank Granja and Group Two [including only Count II] will represent the harm

caused to Eshaman Ruiz.” This makes no sense. Count I was a conspiracy to rob

the armored car; Count II was the attempted robbery of that car. It is impossible to

claim that only Frank Granja was harmed by the conspiracy, while only Eshaman

Ruiz was harmed by the attempted robbery. Both men were attacked during the

robbery, and both were put in jeopardy by the formation of the conspiracy. The

only reasonable explanation is that both men were victims of both offenses.

Although as the government claims, multiple “counts arising out of a single

transaction or occurrence [may not be] grouped together . . . when there are

distinct victims,” see Brief for the United States, at *10 (quoting U.S.S.G. §

3D1.2, cmt. background), the victims of the two offenses at issue here are the

same. Consequently, the offenses are suitable for grouping under § 3D1.2.11 As a

11
Williams’s attorney made this same point during Williams’s sentencing hearing:
The probation officer’s analysis fails because the probation officer
concluded that one guard was the victim of the conspiracy and
another guard is the victim of the robbery. That is clearly not what
happened here. And as far as the conspiracy, either both guards
were also the victim of the conspiracy or no one was the victim of
the conspiracy. But [it cannot be argued that] one guard was . . .
the victim of the conspiracy and one guard was . . . the victim of
the robbery. That doesn’t make common sense. The indictment
charges one conspiracy with 2 victims. And the robbery charge,
the indictment charges one robbery with two victims.
Trans. Apr. 3, 2002, at *4.

33
matter of law, a conspiracy to commit a substantive offense will almost always

have the same victims as the commission (or attempted commission) of that

substantive offense, and so the two should almost always be grouped together

under § 3D1.2. This result is essentially mandated by the commentary to the

guidelines, which states, “When one count charges a conspiracy or solicitation and

the other charges a substantive offense that was the sole object of the conspiracy

or solicitation, the counts will be grouped together under subsection (b).”

U.S.S.G. § 3D1.2, cmt. n.4.

Thus, we do not contest the government’s claim that Williams’s conduct

“resulted in ‘two separate victims.’ ” Brief for the United States, at *6. However,

this observation does not allow the Government to separate two offenses that

would otherwise be grouped together by arbitrarily “assigning” one victim to each

offense where there is no clear basis for doing so. As discussed above, it is

precisely because there were the same “two separate victims” for each of

Williams’s offenses that we must group them together. If the Government wanted

to punish Williams separately for the harm he inflicted upon each individual

driver, it should have included two aggravated assault or attempted murder counts

34
in the indictment.12

Our conclusion compels us to reject the notion, implicit in the government’s

argument, that each offense can have only one “victim.” The Commentary to the

Sentencing Guidelines states, “Generally, there will be one person who is directly

and most seriously affected by the offense and is therefore identifiable as the

victim.” U.S.S.G. § 3D1.2, cmt. n.2. While such commentary is binding on us,

see Stinson v. United States, 508 U.S. 36, 38, 113 S. Ct. 1913, 1915, 123 L. Ed. 2d

598 (1993) (“[C]ommentary in the Guidelines Manual that interprets or explains a

guideline is authoritative unless it violates the Constitution or a federal statute, or

is inconsistent with, or a plainly erroneous reading of, that guideline.”), this

particular statement is a broad generalization rather than an inflexible rule that

must be applied in every single case. Where offenses such as the robbery and

conspiracy at issue here equally harm multiple people, the guidelines do not force

us to designate the “one real victim.” The main purpose of this provision is simply

to emphasize that “secondary victims,” such as bystanders who may be

12
It is possible, however, that aggravated assault charges would have been grouped
together with the robbery. See U.S.S.G. ch.3, pt.D, introductory cmt. (“Other offenses, such as an
assault causing bodily injury to a teller during a bank robbery, are so closely related to the more
serious offense that it would be appropriate to treat them as part of the more serious
offense . . . .”).

35
traumatized at the sight of a crime, should not be counted as victims when

grouping offenses under § 3D1.2. See U.S.S.G. § 3D1.2, cmt. n.2 (“The term

‘victim’ is not intended to include indirect or secondary victims.”).

Consequently, Appellant’s sentences on Counts I and II are VACATED,

and the case is REMANDED for resentencing.

SO ORDERED.

36