United States v. Aurelio Cano-Flores

United States Court of Appeals

FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued January 20, 2015 Decided August 7, 2015

No. 13-3051

UNITED STATES OF AMERICA,

APPELLEE

v.

AURELIO CANO-FLORES, ALSO KNOWN AS YANKEE, ALSO

KNOWN AS YEYO,

APPELLANT

Appeal from the United States District Court

for the District of Columbia

(No. 1:08-cr-00057-16)

Richard K. Gilbert, appointed by the court, argued the

cause for appellant. With him on the briefs was Kristen Grim

Hughes, appointed by the court.

Nina S. Goodman, Attorney, U.S. Department of Justice,

argued the cause and filed the brief for appellee.

Before: ROGERS and SRINIVASAN, Circuit Judges, and

WILLIAMS, Senior Circuit Judge.

2

Opinion for the Court filed by Senior Circuit Judge

WILLIAMS.

WILLIAMS, Senior Circuit Judge: Appellant Aurelio

Cano-Flores appeals from his conviction for conspiring to

manufacture and distribute cocaine and marijuana for

importation into the United States, in violation of 21 U.S.C.

§§ 959, 960, and 963. He raises a number of challenges to his

conviction and sentence. We reject all, with one exception:

we conclude that the $15 billion forfeiture assessed by the

district court must be recalculated under the terms of 21

U.S.C. § 853(a)(1), and we remand to the district court for that

purpose. Accordingly we do not reach Cano-Flores’s

argument that the forfeiture constituted an “excessive” fine in

violation of the Eighth Amendment, or his contentions that the

court miscalculated the forfeiture under its understanding of

§ 853(a)(1).

* * *

Cano-Flores was a member of the Gulf Cartel, one of the

largest and most infamous drug cartels in Mexico. A former

Mexican state police officer, Cano-Flores participated in the

cartel’s takeover of Miguel Alemán, a Mexican border city or

“plaza” across the Rio Grande from Roma, Texas. The

officials and police in the town turned a blind eye to the

cartel’s drug trafficking, which took in roughly between $1

million and $2 million in weekly profits in Miguel Alemán

alone. Cano-Flores was responsible for guarding shipments of

marijuana and cocaine, and he several times completed drug

sales. In late 2005 or early 2006, Cano-Flores became a

“plaza commander” in Los Guerra, a town near Miguel

Alemán that also borders Texas. As a “trusted man” in the

cartel, he was in charge of transporting, storing, and

distributing drugs in his territory, as well as accounting for the

drugs and money that moved across the border.

3

Using wiretaps that are the subject of several claims on

appeal, the Drug Enforcement Administration gathered

evidence of the cartel’s activities, leading to a 2008

indictment of Cano-Flores along with other cartel members.

A warrant was issued for his arrest, and he was extradited to

the United States in August 2011.

Cano-Flores argues that the wiretap authorization orders

exceeded the jurisdiction of the issuing court, that the

listening agents failed to properly minimize their overhearing

of the intercepted conversations, and that the transcripts of

those conversations were improperly sent into the jury

deliberation room. As to sentencing, Cano-Flores argues that

his below-Guidelines 35-year sentence was substantively

unreasonable and that his $15 billion criminal forfeiture

assessment was incorrectly calculated and in violation of the

Eighth Amendment.

* * *

The DEA conducted its wiretaps under authorizations

from various federal district judges in the U.S. District Court

for the Southern District of Texas. For each targeted

telephone number, the telephone service provider (evidently

always Nextel) directed the calls’ content to a DEA “wire

room” in Houston, where Spanish-speaking DEA contractors

monitored the calls. So far as appears, the process intercepted

only calls made near the border; when the cell phones were in

roaming mode, they would seek the strongest signal, which

was very commonly a cellphone tower in the United States.

Before trial, Cano-Flores moved to suppress evidence

from the wiretap, arguing that the district court in Texas

lacked jurisdiction to issue wiretap authorization orders

targeting the calls because the devices were located in Mexico

and the authorizing statute grants no authority to intercept

4

communications outside the United States. The statutory

basis for the interceptions was Title III of the Omnibus Crime

Control and Safe Streets Act of 1968, Pub. L. No. 90-351, 82

Stat. 211, codified at 18 U.S.C. §§ 2510-2520, which permits

district judges to issue orders authorizing “interception of

wire, oral, or electronic communications within the territorial

jurisdiction of the court in which the judge is sitting.” 18

U.S.C. § 2518(3). Section 2510(4) in turn defines

“intercept[ion]” as “aural or other acquisition of the contents

of any wire, electronic, or oral communication.”

Without addressing Cano-Flores’s arguments about the

ultimate reach of Title III, the district court found the

interceptions lawful: they had taken place not in Mexico, but

“in the DEA wire room located in Houston, Texas (a location

within the Southern District of Texas) after they had been

accessed by cellular towers located in the United States.”

Although the statute does not supply an explicit rule for

determining where interception occurs, courts have integrated

the language allowing “interception . . . within the territorial

jurisdiction of the court in which the judge is sitting” with the

language that defines “intercept” as the “aural or other

acquisition of the contents of any . . . communication.” On

the basis of these provisions, for example, United States v.

Rodriguez, 968 F.2d 130 (2d Cir. 1992), held that besides

occurring at the site of the telephone, an interception “must

also be considered to occur at the place where the redirected

contents are first heard.” Id. at 136. In a separate opinion,

Judge Meskill, though rejecting this reasoning, gave it its

name—the “listening post” theory. Id. at 144. The basic

reasoning has been accepted in all courts of appeals to address

the issue. See United States v. Henley, 766 F.3d 893, 911-12

(8th Cir. 2014); United States v. Luong, 471 F.3d 1107, 1109-

10 (9th Cir. 2006); United States v. Jackson, 207 F.3d 910,

914-15 (7th Cir.), vacated on other grounds, 531 U.S. 953

(2000); United States v. Denman, 100 F.3d 399, 402-03 (5th

5

Cir. 1996); United States v. Tavarez, 40 F.3d 1136, 1138 (10th

Cir. 1994).

Cano-Flores points out that in United States v. Glover,

736 F.3d 509 (D.C. Cir. 2013), we observed that the statute

“does not refer to a ‘listening post,’” id. at 514, which is of

course true. But all we held there was that the statute did not

authorize a magistrate in one jurisdiction to authorize the

planting of a physical bug on private property in another

jurisdiction, and we distinguished Rodriguez and similar cases

as dealing with telephone intercepts. Id. at 514-15. We didn’t

mention and had no occasion to construe the “aural . . .

acquisition” language of § 2510(4).

Cano-Flores argues that none of the cases applying the

“listening post” theory involved taps of conversations

occurring abroad. True enough, but we don’t see how that

alters the force of the general principle, which turns on the

statutory language.

He also argues that the listening post theory, which

predates the wireless era, should be inapplicable in cases

involving wireless communications. But he points to no

distinction between the two eras that calls for a different

result. Of course it is true that the primary means by which

end users interface with the telephone system has significantly

changed. But that change alone is not what accounts for the

expansiveness of the listening post theory, which Cano-Flores

suggests is boundless. Whatever boundlessness the theory

may imply is due to the fact that phones used in one location

can be tapped in a way that allows agents to first hear them

somewhere else, and he points to no special change in this

characteristic. In fact courts have applied the principle

equally to landlines, see Rodriguez, 968 F.2d at 135 (applying

listening post theory to support jurisdiction in the Southern

District of New York for evidence from four landline

6

telephones located in a New Jersey café), and cell phones, see

Henley, 766 F.3d at 911-12 (upholding the Eastern District of

Missouri’s authorization of a wiretap on communications

from a cellular phone located in Illinois).

The alleged boundlessness of which Cano-Flores

complains stems from the statutory language, especially the

definition of “intercept,” which Cano-Flores does not try to

parse. Moreover, whatever the force of the effects to which

he points, there are opposing concerns. On his view

government officials would be required to obtain a wiretap

order in every district where they thought a target could make

calls. Such a scheme seems unworkable. Moreover, by

diffusing oversight responsibilities, it might weaken the

courts’ ability to protect citizens’ privacy by monitoring the

wiretap process. As Rodriguez suggested, “If all of the

authorizations are sought from the same court, there is a better

chance that unnecessary or unnecessarily long interceptions

will be avoided.” 968 F.2d at 136.

* * *

Cano-Flores also sought suppression on the grounds that

the agents listening to the calls failed to heed the Federal Wire

Tap Statute’s requirement that officials “conduct[] [the

wiretap] in such a way as to minimize the interception of

communications not otherwise subject to interception under

this chapter.” 18 U.S.C. § 2518(5). Officials must make

“reasonable” efforts to minimize the interception of non-

relevant conversations. United States v. Carter, 449 F.3d

1287, 1295 (D.C. Cir. 2006) (citing Scott v. United States, 436

U.S. 128, 139-40 (1978)).

Cano-Flores proposes that we adopt a burden-shifting

approach for determining whether the government adequately

minimized. Under his proposal, Cano-Flores would need to

7

make a prima facie case of improper minimization, at which

point the government would be required to “provide

permissible explanations for the failure to minimize,

especially explanations derived from the facts of its

investigation.” Of course the plausibility of the proposal turns

largely on what constitutes a prima facie case. He argues that

he made such a case when he provided the district court with a

long list of calls that were longer than two minutes and were

ultimately deemed “non-pertinent” in their entirety; non-

minimization of such calls (continued listening by the agents,

beyond the two minutes), he argues, is presumptively

unreasonable.

But such an approach grossly oversimplifies the

interception process. We’ve rejected the idea that a high

percentage of nonpertinent non-minimized calls is, or is even

likely to be, inconsistent with reasonable minimization efforts.

Carter, 449 F.3d at 1295. As the Court made clear in Scott, a

host of factors determines the reasonableness of interceptors’

treatment of particular calls. A call may have been “very

short,” Scott, 436 U.S. at 140, a concern perhaps answered by

Cano-Flores’s two-minute dividing line. Calls may have been

onetime, id., a matter Cano-Flores doesn’t try to address. The

Court also pointed to special problems with a wide-ranging

conspiracy, such as the one here, where an initial wide cast of

the net may be necessary to trace the conspiracy’s scope. Id.

at 140-41. As a consequence, we and the Supreme Court

require defendants to “identify particular conversations so that

the government can explain their non-minimization.” Carter,

449 F.3d at 1295.

Cano-Flores’s list of calls essentially mirrors the

approach rejected in Carter and Scott. While the list identifies

a large subset of calls, it does so primarily on the basis of

length; it does not explain why specific call characteristics

(such as recipients, content, or context) should have caused

8

the agents to recognize early on that the remainder of the call

would not be pertinent. Once a defendant is provided with the

list of intercepted calls and the tapes of those calls (as Cano-

Flores was), he has not only the incentive to make a case on

inadequate minimization but the information needed to do so.

Whatever merit a burden-shifting scheme might have, Cano-

Flores’s proposal is unsound and his challenge to the

minimization efforts is clearly insufficient under established

law. The district court correctly rejected his motion to

suppress the evidence.

* * *

Cano-Flores raises a third issue related to the wiretap

evidence: the district court’s decision to allow translated

transcripts from the wiretap recordings to go back into the

jury room during deliberations. The parties spent several

months negotiating over the transcription and translation of

the calls, which of course were originally in Spanish. Cano-

Flores contends that various stipulations made as to

unintelligible and ambiguous portions of the wiretap

recordings were made under the explicit understanding that

the transcript binders would not go back to the jury room, and

that the district court’s reversal on that front (allowing the

binders to go back) constituted error.

At the end of the trial, the district judge asked the parties

whether they thought the transcripts should be sent back to the

jury; over a defense objection, she eventually ruled that they

would. She introduced the evidence in the recordings and

transcripts to the jury as follows:

During this trial you have been given transcripts of

translations from Spanish into English of the

conversations that could be heard on the wiretap

recordings received in evidence. I have admitted the

9

transcripts for the purpose of aiding you and [sic; in]

following the content of the conversations as you listen to

the wiretaps which were spoken in Spanish and also to

aid you in identifying the speakers. The transcripts are

evidence just like any other evidence in this case.

However, the wiretap recordings are the actual evidence

of what was said and should you find it necessary during

your deliberations I can arrange to have them played back

to you while you follow along with the transcripts. The

parties have stipulated that the transcripts accurately

translate the conversation between the speakers in all

material respects.

This court has previously warned against the dangers of

the indiscriminate use of transcripts, noting that the “the jurors

may . . . transform the transcript into independent evidence of

the recorded statements.” United States v. Law, 528 F.3d 888,

901 (D.C. Cir. 2008) (quoting United States v. Holton, 116

F.3d 1536, 1540 (D.C. Cir. 1997)). Holton held that “[t]he

jury must be instructed that they should disregard anything in

the transcript that they do not hear on the recording itself.

Moreover, the court must ensure that the transcript is used

only in conjunction with the tape recording.” 116 F.3d at

1543.

But Holton’s general rule favoring the use of recordings

over transcripts did not categorically prohibit the use of

transcripts. Here, the recordings were in a foreign language

and the jurors could only understand the evidence through the

translated transcripts. It would be redundant to require the

jury to go through the pretense of rehearing the recordings,

when its real need was an ability to refer back to the

transcripts. Emphasis and vocal inflection may of course be

critical, but jurors dealing with calls made in a foreign

language are likely to take the vast majority of their

10

understanding from the translations, turning to the recordings

only for special issues.

Cano-Flores argues that the judge’s statement that “[t]he

transcripts are evidence just like any other evidence in this

case” was error, emphasizing that certain inaccuracies in the

transcripts had been established during the trial. But he

doesn’t point to any inaccuracies material enough to have

affected the outcome. Furthermore, we agree with other

circuits that when the tapes are in a foreign language, it

generally makes little sense to say that accurate transcriptions

do not qualify as evidence. Absent unusual circumstances,

there was no error in instructing the jurors that they could

“consider those transcripts like any other evidence during

[their] deliberations.” United States v. Placensia, 352 F.3d

1157, 1165 (8th Cir. 2003); see also United States v. Franco,

136 F.3d 622, 626 (9th Cir. 1998).

* * *

Cano-Flores challenges three aspects of his sentencing,

claiming that his 35-year imprisonment term is substantively

unreasonable, that the court improperly calculated the $15

billion forfeiture, and that assessment of that forfeiture

violates the Eighth Amendment’s prohibition against

excessive fines.

While the Sentencing Guidelines are advisory, the first

step for the sentencing court is to calculate the range they

prescribe. Gall v. United States, 552 U.S. 38, 49, 51 (2007).

The district court determined (and Cano-Flores does not

challenge) that the Guidelines recommended a sentence of life

imprisonment. Explaining the actual 35-year sentence by

reference to the factors named in 18 U.S.C. § 3553(a)(2), the

court emphasized the enormity of the threat posed by cartels

and the drug trade between Mexico and the United States, as

11

well as § 3553(a)(2)(B)’s directive to adequately deter such

conduct. Nonetheless, the court imposed a below-Guidelines

sentence in light of the need to avoid “unwarranted sentencing

disparities between [Cano-Flores] and defendants found guilty

of similar crimes.”

When reviewing a sentencing court’s application of the

Guidelines to facts, we grant the court “due deference,” which

we have said lies “somewhere between de novo and ‘clearly

erroneous.’” United States v. Kim, 23 F.3d 513, 517 (D.C.

Cir. 1994); see also 18 U.S.C. § 3742(e).

Cano-Flores argues that his sentence constituted too great

a “trial penalty,” severely punishing him for his decision to go

to trial rather than accept a plea bargain, and thus violates 18

U.S.C. § 3553(a)(6)’s requirement that the court consider “the

need to avoid unwarranted sentence disparities among

defendants with similar records who have been found guilty of

similar conduct.” He cites the sentences of a number of cartel

members whose roles were greater than his but who, having

pleaded guilty, received lesser sentences. His best but not

altogether atypical example is the cartel’s “supreme leader,”

Osiel Cardenas-Guillen, whom the judge in the Southern

District of Texas sentenced to 25 years in prison.

The argument runs into two difficulties. First, the

sentencing judge clearly took into account the need to avoid

unwarranted disparities, and indeed offered that concern as the

primary reason to give Cano-Flores a below-Guidelines

sentence. Second, “[b]ecause it is well established that

sentences that fall within the Guidelines range are entitled to a

presumption of reasonableness, it is hard to imagine how we

could find [a] below-Guidelines sentence[] to be unreasonably

high.” United States v. Jones, 744 F.3d 1362, 1368 (D.C. Cir.

2014) (internal citations and quotations omitted). In light of

the deferential standard that we apply to sentencing

12

determinations—combined with the sentencing judge’s

explicit acknowledgment of the relevant factors and a below-

Guidelines sentence—we conclude that Cano-Flores’s

sentence is not substantively unreasonable.

* * *

The district judge ordered a $15 billion forfeiture against

Cano-Flores pursuant to 21 U.S.C. § 853(a)(1). That

provision reads:

(a) Property subject to criminal forfeiture

Any person convicted of a violation of this subchapter or

subchapter II of this chapter punishable by imprisonment

for more than one year shall forfeit to the United States,

irrespective of any provision of State law—

(1) any property constituting, or derived from, any

proceeds the person obtained, directly or indirectly,

as the result of such violation;

The district court arrived at the $15 billion figure by

relying on the attribution principles set out in Pinkerton v.

United States, 328 U.S. 640 (1946), essentially accepting what

the government claimed was a “conservative” interpretation of

evidence on gross cartel proceeds that were reasonably

foreseeable to Cano-Flores. The cartel employed tens of

thousands of people, and Cano-Flores argues that to impose a

forfeiture so calculated on him violates the Eighth

Amendment’s prohibition against excessive fines.

At oral argument we asked whether such an expansive

approach to forfeiture was consistent with the statutory text:

13

Court: How do you get from the statute that refers to

what a “person obtained” to assigning to the person $15

billion based on what the entire cartel obtained?

Government Counsel: Courts have held . . . in a

drug conspiracy case, specifically, a defendant is jointly

and severally liable for the reasonably foreseeable

proceeds of a conspiracy, and . . . that’s consistent with

general conspiracy law, subject, of course, to an Eighth

Amendment constraint . . . .

Oral Argument Recording at 44:30-45:30.

Although Cano-Flores did not raise the question whether

§ 853(a)(1) authorized a forfeiture based on the attribution

principles of Pinkerton, we ordered supplemental briefing in

order to determine whether a correct interpretation of the

statute would allow us to avoid Cano-Flores’s constitutional

challenge. Order for Supplemental Briefing, May 13, 2015

(citing U.S. National Bank of Oregon v. Independent

Insurance Agents of America, 508 U.S. 439, 445-48 (1993);

Meredith Corp. v. FCC, 809 F.2d 863, 872 (D.C. Cir. 1987)).

We now conclude that the calculation procedure employed by

the district court was inconsistent with the language of

§ 853(a)(1), which in our view does not authorize imposition

of a forfeiture based on the total revenues of a conspiracy

simply because they may have been reasonably foreseeable.

We acknowledge at the outset that government counsel’s

summary of the views of the circuit courts that have spoken to

the issue is essentially correct. Under both 21 U.S.C. § 853

and 18 U.S.C. § 1963, a similarly-worded forfeiture provision

also enacted as part of the Comprehensive Forfeiture Act of

1984, courts have applied Pinkerton principles and

characterized the resulting forfeiture calculation as one of

“joint and several liability.” See, e.g., United States v. McHan,

14

101 F.3d 1027, 1042-43 (4th Cir. 1996) (noting that the court

“generally construe[s] the drug and RICO forfeiture statutes

similarly”); United States v. Edwards, 303 F.3d 606, 643-44

(5th Cir. 2002); United States v. Corrado, 227 F.3d 543, 553

(6th Cir. 2000); United States v. Pitt, 193 F.3d 751, 765 (3d.

Cir. 1999); United States v. Simmons, 154 F.3d 765, 769 (8th

Cir. 1998); United States v. Hurley, 63 F.3d 1, 22 (1st Cir.

1995); United States v. Masters, 924 F.2d 1362, 1369 (7th Cir.

1991); United States v. Benevento, 663 F. Supp. 1115, 1118

(S.D.N.Y. 1987), aff’d, 836 F.2d 129 (2d Cir. 1988)

(expressly adopting the district court opinion); United States

v. Caporale, 806 F.2d 1487, 1506 (11th Cir. 1986). We

respectfully disagree, reading the statutory language as

providing for forfeiture only of amounts “obtained” by the

defendant on whom the forfeiture is imposed.

We begin our analysis with the statutory text itself, which

appears, on its face, to embrace only property that a defendant

has “obtained.” The government’s dispute of that position, to

the extent it goes beyond stating that other courts have applied

Pinkerton, appears to rest on the word “indirectly,” arguing

that a co-conspirator “‘indirectly’ obtains proceeds [that are]

foreseeably acquired by his co-conspirators in furtherance of

the conspiracy.” Gov. Supp. Brief at 6.

But the government’s view reads the word “obtained” out

of the statute. In ordinary English a person cannot be said to

have “obtained” an item of property merely because someone

else (even someone else in cahoots with the defendant)

foreseeably obtained it. And there is no need to read

“obtained” in such a strained way, given that “indirectly” can

be meaningfully understood in ways completely consistent

with giving “obtained” its ordinary meaning. Most obviously,

“indirect” naturally covers any situation where funds are

transferred by a victim (or purchaser) to a defendant through

an intermediary. That understanding reconciles “indirectly”

15

and “obtained” by giving power to the word “indirect” while

still encompassing only funds that actually reach the

defendant. While it might be argued that the final stage of the

transaction is the only one which “counts,” so that any such

multi-party transaction would include one “direct” step (the

last), such a formulation drains “indirectly” of its most

obvious and natural meanings.

There are also cases where the flow of funds is a good

deal more subtle. For example, an employee engineering a

fraud for his or her firm may receive increased compensation

as an indirect benefit of the fraud. See SEC v. Stoker, 865 F.

Supp. 2d 457, 463-64 (S.D.N.Y. 2012) (finding that the

complaint adequately pled that the defendant “personally

obtained money indirectly” after a “doubling of his yearly

bonus” which was plausibly “at least partly the fruit of his

fraud”).

“Indirectly” might also be seen as embracing property

received by persons or entities that are under the defendant’s

control (such as a closely held corporation, or an employee or

other subordinate of the defendant), or property applied to the

benefit of persons for whom that defendant has a legal or

moral obligation of support (such as his children). Thus, in

United States v. Peters, 732 F.3d 93, 102-04 (2d Cir. 2013),

the Second Circuit held that, under a similarly worded

forfeiture provision, an individual defendant indirectly

obtained proceeds received by a corporation 98% owned by

the defendant and his wife. See also United States v. Stolee,

172 F.3d 630, 631 (8th Cir. 1999) (applying the bank fraud

enhancement from the Sentencing Guidelines and holding that

the defendant indirectly obtained funds deposited into a

corporation solely owned by the defendant).

In all these cases the defendant would normally be seen,

as a matter of ordinary language, as having obtained the

16

amount in question. Forfeiture amounts calculated under the

government’s view, by contrast, may consist almost entirely

of amounts that the defendant has never obtained.

The Sentencing Guidelines further confirm the oddity of

the government’s assumption by adopting rules under which

proceeds “indirectly” obtained by a violator refer exclusively

to proceeds actually obtained by him individually. For

example, § 2B4.1 of the Guidelines, “Bribery in Procurement

of Bank Loan and Other Commercial Bribery,” provides for a

two-level increase if “the defendant derived more than

$1,000,000 in gross receipts from one or more financial

institutions as a result of the offense . . . .” § 2B4.1(b)(2)(A).

The notes explain that § 2B4.1(b)(2)(A) encompasses all

property obtained “directly or indirectly” and also that the

defendant is deemed to have “derived” only sums received

“individually”:

(A) In General. For purposes of subsection

(b)(2)(A), the defendant shall be considered to have

derived more than $1,000,000 in gross receipts if the

gross receipts to the defendant individually, rather

than to all participants, exceeded $1,000,000.

(B) Definition. “Gross receipts from the offense”

includes all property, real or personal, tangible or

intangible, which is obtained directly or indirectly as a

result of such offense. See 18 U.S.C. § 982(a)(4).

USSG § 2B4.1 Application Note 4 (emphasis added); see also

USSG § 2B1.1 (treating “Theft, Embezzlement, Receipt of

Stolen Property, Property Destruction, and Offense Involving

Fraud or Deceit” similarly). Thus the Commission plainly

recognizes that there is no inconsistency in saying that, under

language clearly imputing to a person property received

“indirectly,” the court is to exclude property received by other

17

people with whom he or she is in cahoots, and to include only

property received by the defendant in question.

We now turn to the reasoning of the decisions that, as we

noted, adopt a view equivalent to the government’s. First,

courts using the concept of joint and several liability often

rely on 21 U.S.C. § 853(o)’s instruction to construe the

provisions in the statute “liberally” in order “to effectuate [the

statute’s] remedial purposes.” See Benevento, 663 F. Supp. at

1118; see also Simmons, 154 F.3d at 771; McHan, 101 F.3d at

1043; United States v. Saccoccia, 823 F. Supp. 994, 1003

(D.R.I. 1993).

We put aside for a moment some general problems with

reliance on the “be liberally construed” clause and focus

instead on the remedial purposes of the legislation. See

Caporale, 806 F.2d at 1507 (“[T]he legislative history of the

forfeiture provision indicate that joint and several liability is

not only consistent with the statutory scheme, but in some

cases will be necessary to achieve the aims of the

legislation.”). The essence of the theory appears to be that

since Congress undoubtedly wanted to improve forced

disgorgement as a tool for dissuading people from embarking

on drug (or RICO) crimes, Congress sought basically to

expand the amounts forfeitable, and application of Pinkerton

has that effect. There are at least two flaws in the reasoning.

First, neither the statutory language nor the legislative

history suggests any such general expansive purpose. The

Senate Report explained: “For the most part, [these] forfeiture

amendments do not focus on significant expansion of the

scope of property subject to forfeiture . . . [i]nstead, they focus

primarily on improving the procedures applicable in forfeiture

cases.” S. Rep. No. 98-225, at 192 (Sept. 14, 1983). Instead

of intending some sort of generalized expansion, Congress

appeared to be intent on specific improvements aimed at

18

preventing transfer or concealment of assets before

conviction, id. at 195, and at creating (in the drug forfeiture

sections applicable here) a more efficient forfeiture process

which would no longer require a wasteful “separate civil

forfeiture proceeding[] against property of the defendant . . . .”

Id. at 210 (emphasis added).

Second, even if Congress explicitly asserts a particular

purpose, the courts do not assume that it intended to pursue

that purpose to the exclusion of all others. See, e.g.,

Rodriguez v. United States, 480 U.S. 522, 525-26 (1987)

(“[N]o legislation pursues its purposes at all costs.”). Here,

for example, there is nothing to suggest that Congress

intended to rank forfeiture maximization above all normal

principles, such as the idea that the punishment should fit the

crime.

Reliance on the “be liberally construed” provision also

presents more general problems. First, the Supreme Court has

been clear that identical language (the “provisions of this title

shall be liberally construed to effectuate its remedial

purposes”) cannot be used to apply a statute “to new purposes

that Congress never intended”; the instruction “only serves as

an aid for resolving an ambiguity; it is not to be used to beget

one.” Reves v. Ernst & Young, 507 U.S. 170, 183-84 (1993)

(internal quotations and citation omitted).

Second, even if the statute were ambiguous in the sense

of permitting the government’s construction, “[t]he rule of

lenity requires ambiguous criminal laws to be interpreted in

favor of the defendants subjected to them.” United States v.

Santos, 553 U.S. 507, 514 (2008). In the context of the RICO

forfeiture provision, which has both civil and criminal

application, we held that any ambiguity in the statute would

need to be narrowly construed, as the rule of lenity prevails

over the explicit instruction to construe the statute liberally.

19

Yellow Bus Lines, Inc. v. Drivers, Chauffeurs & Helpers Local

Union 639, 913 F.2d 948, 955-56 (D.C. Cir. 1990). (A

footnote in Reves suggests disagreement with a different

aspect of Yellow Bus Lines, 507 U.S. at 179 n.4, but doesn’t

address this principle.) There may be little clash here. The

rule of lenity, which “applies to sentencing as well as

substantive provisions,” United States v. Batchelder, 442 U.S.

114, 121 (1979), requires Congress to speak clearly so that

courts need “not play the part of a mindreader” “[w]hen

interpreting a criminal statute.” Santos, 553 U.S. at 515.

Whatever the force of “liberally construed,” it cannot support

interpretations that require us to play mindreader to Congress,

which did not manifest any decision, so far as we can tell

(much less a clear one), that forfeitures be calculated on a

theory of joint and several liability.

Finally, in addition to the rule of lenity, the canon of

constitutional avoidance requires that if one of two

linguistically permissible interpretations raises “serious

constitutional problems” and the other does not, we are to

choose the second unless it is “plainly contrary to the intent of

Congress.” See Solid Waste Agency of N. Cook County v.

U.S. Army Corps of Engineers, 531 U.S. 159, 173 (2001)

(quoting Edward J. DeBartolo Corp. v. Florida Gulf Coast

Building & Construction Trades Council, 485 U.S. 568, 575

(1988)). A forfeiture equal to a cartel’s gross take of $15

billion, imposed on a mid-level manager such as Cano-Flores

(or even a trivial courier) within a conspiracy—a result which

appears to be commanded under the government’s

interpretation of § 853(a)(1)—poses serious Eighth

Amendment concerns. See United States v. Bajakajian, 524

U.S. 321, 334-44 (1998) (outlining inquiry for determining

whether a fine is unconstitutionally excessive). Even if the

government’s view of the statute were a plausible

interpretation—which we question—the canon counsels us to

go with the narrower reading.

20

The final theme invoked by our sister circuits is the

thought that imposition of vicarious liability under § 853

“resonates with established criminal law principles.” McHan,

101 F.3d at 1043. Some courts have argued that the

imposition of joint and several liability in forfeiture is “even

less theoretically problematic than vicarious liability for a

substantive conviction might be because it goes only to the

penalty imposed rather than to the individual’s criminal

liability.” Caporale, 806 F.2d at 1508.

But vicarious liability’s supposed “resonance” with

Pinkerton seems a woefully inadequate reason for

disregarding the normal meaning of the word “obtained.”

First, as Congress made no mention of the case or the

principle in either the statute or in the legislative history, the

fact that it is and was “established” would seem to weaken the

case for its implicit incorporation.

Further, Pinkerton, even on its own terms, is a doctrine

which speaks only to a defendant’s substantive liability—not

to the consequences of such liability. Applying Pinkerton of

course tends to increase consequences (i.e., imprisonment) for

criminal defendants, but applying vicarious liability principles

to forfeiture under § 853(a)(1) yields a growth in forfeitures

that doesn’t parallel the growth in imprisonment lengths. At

least in the case of drug convictions, the Sentencing

Guidelines do not link imprisonment with drug quantities by a

linear formula under which imprisonment time increases in

direct proportion to increases in the quantity of drugs

attributed to the defendant. Although the ranges of

recommended imprisonments increase, they do so at generally

declining rates. For example, a defendant who possessed

(with intent to distribute) 100 grams of cocaine, and to whom

1900 additional grams are attributed under Pinkerton (a 20-

fold increase), would be subject to only a three-fold increase

in minimum imprisonment (63 months compared to 21

21

months). See Drug Quantity Table, USSG § 2D1.1 (outlining

a “base level” of 16 for possession with intent to distribute of

100-200 grams of cocaine, and a base level of 26 for

possession with intent to distribute of 2-3.5 kilograms of

cocaine). In contrast, application of Pinkerton to the

computation of forfeiture would increase a defendant’s

monetary liability in relation to quantities handled by the

entire conspiracy at a steady 1:1 rate—a much larger

increment in monetary punishment than the Guidelines-

directed increment in imprisonment. While imprisonment and

forfeiture are both subject to ceilings (life for the first, and

maximum lifetime wealth for the second), calculating

forfeitures via the joint and several theory does not truly align

the growth paths of the two types of criminal consequences.

Moreover, the language of “joint and several liability” is

derived from torts, but the courts invoking it have not deeply

considered where there is a sound analogy between forfeiture

and tort law. We doubt there is one. In torts, the doctrine of

joint and several liability rests on a serious policy rationale:

the judgment that it is better that the risk of an insolvent co-

defendant should fall on a partially guilty defendant than on a

completely innocent victim. See Paul Bargren, Joint and

Several Liability: Protection for Plaintiffs, 1994 Wis. L. Rev.

453, 464 (1994). This suggests that the tort analogy might

well apply to restitution in a criminal case, and 18 U.S.C.

§ 3664(h) does indeed authorize (but does not require)

application of joint and several liability as a means of

protecting victims. See Jonathan R. Hornok, A Right to

Contribution and Federal Restitution Orders, 2013 Utah L.

Rev. 661, 678 (discussing joint and several liability under

§ 3664(h)). But the reasoning doesn’t extend to forfeitures,

which are collected by the government. Moreover, in the

normal tort case a defendant who is jointly and severally

liable has at least a chance of securing contribution from co-

defendants, see id. at 670-71, but there appears to be no

22

suggestion by any court imposing joint and several liability

that defendants would have a right of contribution among

themselves.

Because we conclude that the “joint and several”

calculation procedure erroneously included amounts not

obtained by Cano-Flores, we need not reach any of Cano-

Flores’s other arguments against the forfeiture imposed,

including his constitutional claim and his dispute of specific

aspects of the calculation. We vacate the $15 billion

forfeiture assessment against him and remand to the district

court for determination of the proper amount to be forfeited

under § 853(a)(1). We otherwise affirm the judgment of the

district court.

So ordered.