United States Court of Appeals For the First Circuit

No. 19-1584

UNITED STATES OF AMERICA,

Appellee,

v.

LOUIS GARDNER,

Defendant, Appellant.

APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF NEW HAMPSHIRE

[Hon. Joseph A. DiClerico, Jr., U.S. District Judge]

Before

Lynch, Lipez, and Kayatta, Circuit Judges.

William S. Maddox for appellant. Seth R. Aframe, Assistant United States Attorney, with whom Scott W. Murray, United States Attorney, was on brief, for appellee.

July 20, 2021 LIPEZ, Circuit Judge. Pursuant to a plea agreement,

Louis Gardner pled guilty to a variety of drug and firearm offenses

in exchange for a 120-month sentence. While in custody prior to

sentencing, he assaulted a fellow inmate, a breach of the plea

agreement. In light of that breach, the government withdrew from

the agreement. When Gardner then moved to withdraw his guilty

plea, the government opposed, insisting that Gardner was still

bound by the plea. The district court denied the motion to

withdraw and sentenced Gardner to 160 months' imprisonment -- 40

months above the agreed-upon sentence. Gardner now appeals the

denial of his motion to withdraw the plea, as well as the length

of his sentence. Because we agree that Gardner should have been

allowed to withdraw his plea, we do not reach the sentencing issue.

I.

Louis Gardner was charged with six related drug and

firearm offenses. He and the government negotiated a plea

agreement. The agreement explained that "[i]n exchange for the

defendant's guilty pleas" on three of the six counts, the

government "agrees" to certain sentencing stipulations and to

dismiss the remaining counts of the indictment.1 It also stated

that the parties "stipulate and agree that 120 months' imprisonment

1 Although the plea agreement (and the district court, at times) referred to Gardner's "guilty pleas" (i.e., the pleas to each of the three separate counts), we use "guilty plea" to refer to all three pleas collectively.

- 2 - is an appropriate disposition of this case," and that the parties

intended this sentencing stipulation to be "binding" under Federal

Rule of Criminal Procedure 11(c)(1)(C),2 meaning that "if the Court

will not accept the plea agreement under Fed. R. Crim. P.

11(c)(3)(A),3 the plea agreement is null and void and the defendant

will be allowed the opportunity to withdraw his guilty pleas." In

addition to other recitals, waivers, and stipulations, the

agreement included a breach provision, which specified that if,

"before sentencing," Gardner "violates any term or condition of

this Plea Agreement, engages in any criminal activity, or fails to

appear for sentencing," the government "may consider such conduct

to be a breach of the Plea Agreement and may withdraw therefrom."

The district court duly held a change of plea hearing.

The court went over the agreement and confirmed that Gardner

understood it and was entering into the plea knowingly and

voluntarily. As part of its review, the district court also

explained the significance of the stipulated sentence:

THE COURT: All right. So Mr. Gardner, you and the government have agreed that the total sentence that's to be imposed in this . . .

Rule 11(c)(1)(C) allows the parties to "agree that a 2

specific sentence or sentencing range is the appropriate disposition of the case," and "binds the court [to the recommended sentence] once [it] accepts the plea agreement." 3Rule 11(c)(3)(A) provides that, as to plea agreements containing an agreed-upon sentence pursuant to Rule 11(c)(1)(C), "the court may accept the agreement, reject it, or defer a decision until the court has reviewed the presentence report."

- 3 - case is 120 months in prison. That's a binding agreement which means if the Court accepts that agreement and imposes that sentence, do you understand you cannot withdraw your guilty plea?

THE DEFENDANT: Yes.

THE COURT: If the court does not accept that binding agreement and does not impose that sentence, do you understand that you would have the opportunity then to withdraw your guilty plea?

THE DEFENDANT: Yes.

In accordance with the agreement, the district court then accepted

Gardner's guilty plea on the three counts and set a date for

sentencing.

While in custody awaiting sentencing, Gardner assaulted

a fellow inmate. Citing the breach provision that allowed it to

withdraw from the plea agreement if the defendant committed

criminal activity before sentencing, the government moved to

withdraw. After a hearing, the district court found, by a

preponderance of the evidence, that Gardner did indeed commit the

assault, granted the government's motion to withdraw from the plea

agreement, and rescheduled sentencing on the previously-entered

plea.

Sixteen days after the government's motion to withdraw

from the plea agreement was granted, Gardner moved to withdraw the

underlying guilty plea. At the hearing on the motion, his counsel

argued that, given that "the government has withdrawn from the

- 4 - agreement," and assuming that "the Court is not going to accept

the 120 month[]" stipulated sentence, Gardner "should be allowed

the opportunity to withdraw his plea." The district court denied

the motion, stating that, as a matter of fairness, Gardner's loss

of the stipulated sentence was "a consequence of his own actions,"

and "the government's withdrawal from the plea agreement because

of Gardner's breach is not a sufficient reason to permit him to

withdraw his guilty pleas." Then, addressing the fact that the

agreement allowed Gardner to withdraw his plea "if the Court will

not accept the plea agreement under Fed. R. Crim. P. 11(c)(3)(A),"

the court reasoned that it had not actually rejected the plea

agreement, a process which entails its own, somewhat elaborate

procedures for rejection under Rule 11(c)(5).4 Instead, the court

explained that it had simply allowed the government to withdraw

(as permitted by the breach provision), meaning that Gardner's

right to withdraw the plea was never triggered.

After denying Gardner's motion to withdraw his plea, the

district court proceeded to sentencing. Now unbound by the

4 Specifically, in order to reject a plea agreement under Rule 11(c)(5), a court must, "on the record and in open court (or, for good cause, in camera)": "(A) inform the parties that the court rejects the plea agreement; (B) advise the defendant personally that the court is not required to follow the plea agreement and give the defendant an opportunity to withdraw the plea; and (C) advise the defendant personally that if the plea is not withdrawn, the court may dispose of the case less favorably toward the defendant than the plea agreement contemplated." Fed. R. Crim. P. 11(c)(5).

- 5 - agreement's 120-month stipulation, it applied the Sentencing

Guidelines and ultimately imposed a sentence of 160 months. On

appeal, Gardner challenges the denial of his motion to withdraw

his plea.

II.

Guilty pleas and plea agreements are distinct, governed

by different parts of Rule 11. See Fed. R. Crim. P. 11(a), (b)

(guilty pleas); Fed. R. Crim. P. 11(c) (plea agreements). Although

a defendant usually pleads guilty pursuant to a plea agreement,

"[g]uilty pleas can be accepted while plea agreements are deferred,

and the acceptance of the two can be separated in time." United

States v. Hyde,

(1997).

In general, we review a district court's denial of a

pre-sentencing motion to withdraw a guilty plea for abuse of

discretion. United States v. Rodríguez-Morales,

(1st Cir. 2011). When presented with such a motion, a district

court must determine whether there is a "fair and just reason for

requesting the withdrawal." Fed. R. Crim. P. 11(d)(2)(B); see

also United States v. Dunfee,

(1st Cir. 2016).

In applying that standard, there is not an "exclusive list of

reasons that might allow withdrawal of a plea." United States v.

Aker,

(1st Cir. 1999). However, according to

case law, relevant considerations include: (1) whether the

original plea was knowing, intelligent, and voluntary and in

- 6 - compliance with Rule 11, (2) the strength of the reason for

withdrawal, (3) the timing of the motion to withdraw, (4) whether

the defendant has a serious claim of actual innocence, (5) whether

the parties had reached (or breached) a plea agreement, and (6)

whether the government would suffer prejudice if withdrawal is

permitted. See Dunfee,

; United States v. Tilley,

(1st Cir. 1992). The overall standard is

"liberal," United States v. Kobrosky,

(1st Cir.

1983), and "permissive," United States v. Merritt,

(1st Cir. 2014).5

Gardner's argument for allowing his withdrawal turns on

the language of his plea agreement. When interpreting a plea

agreement, we apply "[b]asic contract principles." United States

v. Newbert,

(1st Cir. 2007). The touchstone is

the "defendant's reasonable understanding" of the agreement.

United States v. Conway,

(1st Cir. 1996).6

5 To be sure, we have questioned whether the standard is quite as liberal as some of our older cases suggest. See United States v. Torres-Rosario,

(1st Cir. 2006) ("Although older case law endorses a liberal approach to pre- sentence plea withdrawals, it is questionable how far this view has survived the pressure of growing dockets and an increasing appreciation of the grim dynamics of plea bargaining, including the prevalence of 'buyer's remorse' among those who have pled." (citations omitted)). But we have not actually abdicated the "liberal" standard. See, e.g., Merritt,

(noting the prevailing liberal standard while cautioning that "liberal allowance is not to be confused with automatic allowance"). 6See also United States v. Gregory,

(2d Cir. 2001) (considering "the 'reasonable meaning' of the parties'

- 7 - III.

Our first task, then, is to determine the parties'

reasonable understanding of the agreement, specifically as to

whether Gardner would be permitted to withdraw his guilty plea if

the government withdrew from the agreement because of Gardner's

breach.

In our view, that question is resolved by the agreement

itself, which states: "[I]f the Court will not accept the plea

agreement under Fed. R. Crim. P. 11(c)(3)(A), the plea agreement

is null and void and the defendant will be allowed the opportunity

to withdraw his guilty pleas." Although this "opportunity to

withdraw" provision refers to the court not accepting the agreement

under Rule 11(c)(3)(A), it does not say that the right to withdraw

is limited to any particular reason for non-acceptance. Rather,

it says that Gardner will have the option to withdraw his plea if

the event contemplated by the language -- the acceptance of the

plea agreement under Rule 11(c)(3)(A) and the imposition of the

stipulated sentence -- did not occur for any reason. Here,

undisputedly, that event never occurred.

overall agreement"); United States v. Cruz-Romero,

(5th Cir. 2017) ("In examining the government's compliance with its promises in the plea agreement, we ask 'whether the Government's conduct was consistent with the parties' reasonable understanding of the agreement'" (quoting United States v. Harper,

(5th Cir. 2011))).

- 8 - In denying Gardner's motion to withdraw his plea,

therefore, the district court misread the "opportunity to

withdraw" provision. It first determined that the agreement

allowed Gardner to withdraw the plea only if the court

"'reject[ed]' the plea agreement under Rule 11(c)(3)(A) and

(c)(5)." And then it reasoned that:

In this case . . . the court did not "reject" the plea agreement under Rule 11(c)(3)(A) and (c)(5). In granting the government's motion to withdraw from the agreement, the court enforced . . . the plea agreement, which permitted the government to withdraw if Gardner engaged in criminal activity. . . . Gardner cannot transform his breach of the plea agreement, which cost him his right to receive the agreed-upon sentence, into a rejection of the agreement by the court.

We see two difficulties with the court's approach.

First, while the district court was undoubtedly "enforcing" the

agreement by allowing the government to withdraw in light of

Gardner's breach, that enforcement is only one half of the

equation. Both parties to the agreement have a claim to its

enforcement. The second question is what impact the government's

withdrawal had on Gardner's rights -- and specifically, whether

Gardner remained bound by his guilty plea or had a right to

withdraw it.

On that question, the district court, in evaluating

whether it "rejected" the agreement, focused on a word that does

not appear in the plea agreement. As we have seen, the

- 9 - "opportunity to withdraw" provision is worded differently. It

permits withdrawal of the plea if the district court "will not

accept the plea agreement under Fed. R. Crim. P. 11(c)(3)(A)."

And, as we have noted, the court indisputably did not accept the

agreement. Said differently, the agreement's reference to "not

accept" is a broader term than "reject." "Not accepting"

encompasses every situation in which the court does not actually

impose the stipulated sentence; "rejecting" the agreement under

Rule 11 is just one of those situations.

While not necessarily dispositive, it is telling that

the district court seemed to endorse the correct reading of the

"opportunity to withdraw" provision at the change of plea hearing:

THE COURT: All right. So Mr. Gardner, you and the government have agreed that the total sentence that's to be imposed in this . . . case is 120 months in prison. That's a binding agreement which means if the Court accepts that agreement and imposes that sentence, do you understand you cannot withdraw your guilty plea?

THE DEFENDANT: Yes.

THE COURT: If the court does not accept that binding agreement and does not impose that sentence, do you understand that you would have the opportunity then to withdraw your guilty plea?

THE DEFENDANT: Yes.

These statements suggest that, consistent with the agreement,

there were two -- and only two -- mutually exclusive outcomes: if

- 10 - the court "accept[ed] the agreement and impose[d] that sentence,"

Gardner would not be able to withdraw his guilty plea, whereas if

it "d[id] not accept that binding agreement and d[id] not impose

that sentence," Gardner would have the opportunity to withdraw his

guilty plea.

The dissent disagrees with our reading of the agreement.

It suggests that the agreement "explicitly requires the district

court to indicate that it will not accept the agreement" in order

to trigger Gardner's withdrawal right. However, nothing in the

operative phrase ("if the Court will not accept the plea agreement

under Fed. R. Crim. P. 11(c)(3)(A)") requires any kind of

affirmative statement. Rather, the language simply establishes a

future condition which, if not satisfied, will trigger a specified

consequence. Therefore, once it was clear that the district court

would not accept the agreement, Gardner had a right to withdraw

his plea. The dissent, in effect, is introducing a further

requirement: that the district court had to affirmatively reject

the agreement. While the agreement could have been written that

way, it was not.7

7 We note that, if the agreement had been written to give Gardner the right to withdraw only if the district court affirmatively rejected the agreement, it would simply be reciting what Rule 11 already guarantees. See Fed. R. Crim. P. 11(c)(5) (requiring the court, if it "rejects a plea agreement" specifying a sentence pursuant to Rule 11(c)(1)(C), to "give the defendant an opportunity to withdraw the plea").

- 11 - The dissent also suggests that "[t]he district court's

actions demonstrate that it accepted the plea agreement." That

view of what occurred is unsupportable. "Accepting a plea

agreement" is a formal act under the Rules and triggers serious

consequences and obligations. Rule 11(c)(1)(C) "binds the court

[to the recommended sentence] once [it] accepts the plea

agreement." Obviously, the court never imposed the recommended

sentence of 120 months. It imposed a sentence of 160 months.

Additionally, Rule 11(c)(5) provides that, "[i]f the court accepts

the plea agreement" proposing a recommended sentence, it must

inform the defendant that "the agreed disposition will be included

in the judgment." Here, at the change of plea hearing, the court

accepted Gardner's guilty plea but deferred a decision on the

recommended sentence. It did not inform Gardner that the

recommended sentence would be included in the judgment or state

that he would be sentenced as provided in the agreement. Under no

plausible construction of the rules or the proceedings, then, did

the district court accept the agreement, in fact or constructively.

Additionally, nothing else in the agreement suggests

that the "opportunity to withdraw" provision should not be taken

to mean what it says. The breach provision itself says nothing

about the defendant's right of withdrawal after a breach,

explaining only that "if, before sentencing, [Gardner] violates

any term or condition of this Plea Agreement, engages in any

- 12 - criminal activity, or fails to appear for sentencing, the United

States may consider such conduct to be a breach of the Plea

Agreement and may withdraw therefrom." That is, the breach

provision is silent about whether Gardner would be held to his

guilty plea, despite the government's withdrawal, or whether, in

light of the government's withdrawal, he would be permitted to

withdraw his plea.

Similarly, nothing in the general law of plea bargaining

or our case law precludes Gardner from withdrawing his guilty plea

in these circumstances. To the contrary, "it is generally accepted

that 'when a defendant breaches his plea agreement, the Government

has the option to either seek specific performance of the agreement

or treat it as unenforceable' (at least absent language in the

plea agreement specifying fewer or other remedies)." 5 Wayne R.

LaFave et al., Crim. Proc. § 21.2(e) (4th ed. 2020) (quoting United

States v. Cimino,

(2d Cir. 2004)). In this

context, "specific performance" means that the government can

enforce the remaining provisions of the agreement and hold the

defendant to the guilty plea. See United States v. Alexander,

(2d Cir. 1989). If the government instead chooses

to treat the entire agreement as unenforceable (sometimes referred

to as "cancellation"), the presumption is that the defendant may

withdraw his plea, unless "the plea agreement itself . . .

describe[s] the government's remedies in such a fashion as to

- 13 - foreclose plea withdrawal by the defendant in these

circumstances." 5 Wayne R. LaFave et al., Crim. Proc. § 21.2(e).

Here, as we have explained, the plea agreement did not foreclose

plea withdrawal by the defendant in the event of his own breach.

The government cites our decision in United States v.

Tilley,

(1st Cir. 1992), for the proposition that a

defendant should not be able to withdraw a plea because of his own

breach of a plea agreement. Like Gardner, Tilley breached his

plea agreement by committing a crime (in Tilley's case, by

perjuring himself before a grand jury and at a criminal trial).

. Facing the loss of certain benefits under the

agreement, Tilley then moved to withdraw his plea.

The

district court denied the motion, and we affirmed.

. We

first determined that the court did not err in finding that Tilley

breached the agreement.

. Then, after analyzing the

appropriate factors, we concluded that the court did not abuse its

discretion in finding that there was not a "fair and just reason"

to permit withdrawal.

.

Some of our language in the opinion suggests that Tilley

was foreclosed from withdrawing his plea simply because he breached

his agreement by committing a crime. See

(in arguing

for the opportunity to withdraw his plea, we said, Tilley

"overlooks the fact that he did indeed violate the plea agreement

by engaging in behavior which constitutes a crime"). But as the

- 14 - rest of the opinion makes clear, Tilley's holding relies on the

specific language of Tilley's plea agreement, which provided that,

if Tilley violated the agreement, "the Government had the option

to declare the agreement null and void, or to bring the failure to

fully cooperate to the attention of the court."

at 71 n.17.

In response to Tilley's breach, the government simply chose the

latter, specifically-provided remedy, which did not trigger any

right to withdraw the plea. See

("In view of a violation

by appellant of the plea agreement, the Government was entitled to

bring all factors related to said violation to the attention of

the court."). Here, there is no analogous remedy specified in the

plea agreement. The plea agreement does not make clear that the

defendant would be bound to the plea if the government opted to

withdraw from the agreement.

In sum, this is not a case where the plea agreement

specifies that "if [the defendant] committed new crimes, he would

be bound to his guilty plea even if the Government exercised its

right to rescind the [plea] agreement." United States v. Gregory,

(2d Cir. 2001). To the contrary, the agreement

indicated (and the district court confirmed at the change of plea

hearing) that Gardner would be able to withdraw his plea if the

court did not accept the agreement -- and the court did not accept

the agreement.

- 15 - With that understanding of the agreement in mind, we

proceed to consider whether there is a fair and just reason to

permit withdrawal of Gardner's plea.

IV.

As we have noted, the substantive standard for

evaluating a motion to withdraw a guilty plea in the trial court

(and the standard that informs our review of the trial court's

exercise of discretion in denying a motion to withdraw a guilty

plea) is whether there was a "fair and just reason" for withdrawal.

The relevant considerations include: (1) whether the plea was

knowing and voluntary and in compliance with Rule 11, (2) the

strength of the reason for withdrawal, (3) the timing of the motion

to withdraw, (4) whether the defendant has a serious claim of

actual innocence, (5) whether the parties had reached (or breached)

a plea agreement, and (6) whether the government would suffer

prejudice if withdrawal is allowed. See Tilley,

.

Three of the considerations favor withdrawal. First,

the reason for withdrawal is highly compelling: the agreement

explicitly guaranteed Gardner the opportunity to withdraw his plea

in these circumstances. This reason is far from a mere "second

thought[] about some fact or point of law, or about the wisdom of

his earlier decision." United States v. Parrilla-Tirado,

(1st Cir. 1994) (citations omitted). Indeed, it goes to

the heart of the bargain that Gardner struck with the government.

- 16 - Second, Gardner moved to withdraw his guilty plea

promptly. He did so only sixteen days after the court granted the

government's motion to withdraw from the plea agreement, when

Gardner first learned that the court would not be accepting the

agreement. Although we typically measure any delay from the entry

of the guilty plea, see Ramos, 810 F.2d at 313, we have recognized

that a different approach may be warranted in certain

circumstances. For example, we observed that, when a motion to

withdraw is motivated by post-plea developments in a separate case,

it may not be correct to "rel[y] on the . . . passage of time

between the plea and the motion to withdraw it as an indication

that [the defendant] was simply employing sharp tactics," at least

when the motion to withdraw "was filed so soon after the events in

the [other] case." United States v. Isom,

, 53 n.15

(1st Cir. 2009). In the present circumstances, we consider the

sixteen-day delay to be the appropriate measure for evaluating

promptness, as Gardner had no reason to move to withdraw earlier.

And even though we have held a delay as brief as thirteen days

against a defendant, see Ramos, 810 F.2d at 313, the delay here is

certainly on the shorter side, and we have countenanced much longer

delays when other considerations favor withdrawal, see United

States v. Daniels,

(1st Cir. 1987) (allowing

withdrawal after ten-week delay).

- 17 - Third, there does not appear to be any prejudice to the

government, beyond the burdens that inevitably accompany any

withdrawal -- namely, those of processing the withdrawal and

negotiating a second plea or proceeding to trial. Indeed, there

is no suggestion by the government of any kind of prejudice, such

as any difficulty in tracking down witnesses or otherwise preparing

for a potential trial. See Kobrosky,

(explaining

that "[t]he most common form of prejudice is the difficulty that

the government would encounter in reassembling its witnesses").

The dissent, on the government's behalf, speculates about possible

kinds of prejudice, but it is not our role to make an argument

that the government never makes.

To be sure, on the negative side of the balance, there

are also three factors. Gardner has not advanced any plausible

theory of innocence.8 But this deficiency is not fatal; it just

"counsels against" allowing withdrawal. United States v. Mercedes

Mercedes,

(1st Cir. 2005). Additionally, he

unquestionably breached the agreement by committing an assault, a

significant breach to be sure. And, finally, there is no

8 According to the plea agreement, Gardner was arrested while driving home, armed with a gun, from a drug purchase. Also in the vehicle were a confidential informant and an eventual co- defendant. Gardner had previously sold drugs to the informant on two occasions. Gardner argues that if his guilty plea were withdrawn, he would have the opportunity to move to suppress evidence and raise an entrapment defense.

- 18 - indication that the guilty plea was not "knowing, intelligent, and

voluntary" at the time it was made. United States v. Adams,

(1st Cir. 2020). The district court scrupulously

followed the technical requirements of Rule 11, and Gardner

voluntarily pled guilty in accordance with the agreement as written

and as explained by the district court.

Despite these countervailing considerations, we conclude

that the strength of Gardner's reason for withdrawal so far

outweighs the offsetting factors that the inquiry tilts in his

favor. The government failed to recognize that the plea agreement,

by its terms, gave Gardner the right to withdraw his plea under

the circumstances of this case. That failure, sanctioned by the

court with its denial of Gardner's motion to withdraw his plea,

was tantamount to a breach of the plea agreement by the government.

Given the importance of contract principles to the enforcement of

plea agreements, see Newbert,

, and the other

factors cited in Gardner's favor, there was a fair and just reason

for the withdrawal of Gardner's plea, and the district court abused

its discretion in concluding otherwise. 9

9 The dissent also maintains that any prejudice to the government should be considered only after Gardner has met his burden on the initial five factors (and thereby established some good reason for withdrawal). Some of our cases do suggest such a bifurcated analysis. See, e.g., United States v. Pellerito,

(1st Cir. 1989) ("If a defendant advances a plausible reason, the court should also weigh the prejudice, if any, to the government."). But other cases of ours simply list

- 19 - V.

We close with two additional points. The government and

the dissent argue that allowing a plea withdrawal in these

circumstances would, effectively, reward Gardner for his own

breach. Even worse, the government and dissent suggest, such a

decision would encourage future defendants to deliberately breach

their plea agreements in the hope of getting out of their pleas.10

We are unpersuaded. In the face of a defendant's

strategic breach, the government will not be obligated to cancel

the agreement and concede to the withdrawal of the plea. Rather,

as the non-breaching party, the government will have the option of

how to respond. The government might elect specific performance:

that is, keep the rest of agreement in place, in which case the

defendant would be stuck with the same plea, plus, potentially,

additional exposure for a new crime. See Cimino,

n.3. Alternatively, the government might choose to void the

agreement, concede to withdrawal of the original plea, and

prejudice as one factor among the others. See, e.g., United States v. Dunfee,

(1st Cir. 2016); United States v. Isom,

(1st Cir. 2009); United States v. Kobrosky,

(1st Cir. 1983). Here, even if we adopted the dissent's preferred approach, we would still find that Gardner had established a plausible reason for withdrawal based on the initial five factors -- in particular, the strength of the reason for withdrawal. To be clear, there is no argument or indication here that 10

Gardner committed the assault in order to facilitate his plea withdrawal.

- 20 - "demand[] that [the defendant] either plead guilty a second time

or go to trial."

In the latter case, the price is high: a

defendant would lose the benefits of the existing plea agreement

(including any agreed-upon sentence and benefit for acceptance of

responsibility) and, if he breached by committing a crime, face

the possibility of a new prosecution.

Additionally, of course, the government can avoid a

repetition of the scenario here by being clearer in future plea

agreements about the consequences of a defendant's breach (i.e.,

by explicitly specifying in the agreement that the defendant will

still be held to the guilty plea even if the government exercises

its right to withdraw). See United States v. Rivera,

(2d Cir. 1992) (instructing that "[t]he government should

make it absolutely clear in a plea agreement that a breach by the

defendant releases the government from its obligation to recommend

leniency but does not release the defendant from the plea of

guilty"). The prosecution, defendants, and the courts would all

benefit from this additional clarity.

VI.

For the reasons set forth herein, Gardner must be

permitted to withdraw his guilty plea. We vacate the judgment.

So ordered.

- Dissenting Opinion Follows -

- 21 - LYNCH, Circuit Judge, dissenting. With respect, I

believe the majority got this wrong and departed from controlling

law at every key point in its analysis. After entering into a

plea agreement, Gardner assaulted a codefendant and possible

witness against him. This assault triggered a provision in his

plea agreement giving the government, but not Gardner, the right

to "consider [Gardner's] conduct to be a breach of the Plea

Agreement and . . . withdraw" from it. The government exercised

this right. The majority opinion allows Gardner to also withdraw

from the plea agreement due to his own breach. In doing so, it

concludes that the district court never accepted the plea agreement

under Rule 11, directly contradicting the fact that the court

enforced the plea agreement. Further, the majority erroneously

reads into the plea agreement terms not bargained for by the

defendant, disadvantaging the government. Beyond that, the

majority erroneously holds that the district court abused its

discretion when it did not allow Gardner to withdraw his plea and

posits that no harm will come from its decision. I disagree on

all of these key points.

The majority begins its analysis by saying that the plea

agreement's "opportunity to withdraw" provision gives Gardner the

right to withdraw his plea. That provision reads: "[I]f the Court

will not accept the plea agreement under Fed. R. Crim. P.

- 22 - 11(c)(3)(A), the plea agreement is null and void and [Gardner]

will be allowed the opportunity to withdraw his guilty pleas."

When presented with a plea agreement like Gardner's,

Rule 11 says that a court "may accept the agreement, reject it, or

defer a decision until the court has reviewed the presentence

report [('PSR')]." Fed. R. Crim. P. 11(c)(3)(A). The text is

clear that, even if the court chooses to defer its decision pending

review of the PSR, the court has to choose between two options:

accepting or rejecting the plea agreement. See United States v.

Hyde,

(1997) (discussing the district court's

ability to "defer its decision about whether to accept [a] [Rule

11(c)(1)(C)] agreement (emphasis added)); United States v. Fokker

Servs. B.V.,

(D.C. Cir. 2016) (describing "a

district court's authority to 'accept' or 'reject' a proposed plea

agreement under Rule 11"); see also Fed R. Crim. Proc. 11(c)

(outlining the procedures for "[a]ccepting a [p]lea [a]greement"

in Rule 11(c)(4) and "[r]ejecting a [p]lea [a]greement" in Rule

11(c)(5)).

The majority concedes that the district court never

rejected the plea agreement.11 It should have then concluded that

11 As the majority recognizes, a district court must follow the procedures in Rule 11(c)(5) to reject a plea agreement, which it did not do here. If the district court had rejected the agreement, there would be no need to interpret it. Gardner would have been permitted to withdraw his plea pursuant to Rule 11(c)(5)(B).

- 23 - the court either deferred acceptance of the agreement or accepted

it outright. In either case, that would mean that the "opportunity

to withdraw" provision does not apply to Gardner. Instead, the

majority erroneously holds that the provision applies (and Gardner

can withdraw his plea) because, even though the district court did

not reject the agreement, the majority says that the district court

did not accept the agreement.12

This holding is inconsistent with the text of the plea

agreement. The "opportunity to withdraw" provision says that if

the court "will not accept" the agreement, then the agreement

becomes null and void. (Emphasis added). The majority's reading

ignores the word "will." Non-acceptance alone would not be

enough.13 The agreement explicitly requires the district court to

indicate that it will not accept the agreement, something it did

not do.

12 The majority erroneously reads the change of plea hearing transcript to say that the district court had "two -- and only two -- mutually exclusive" options. In the majority's view, the district court could either (1) accept the plea agreement and impose the sentence in the agreement or (2) not accept the agreement and not impose that sentence. This reading ties acceptance of the agreement to sentencing, making it impossible for the court to accept the plea agreement until a defendant is sentenced. That is flatly inconsistent with Rule 11. 13 If non-acceptance alone were enough, then the plea agreement would have become null and void when the district court, consistent with Rule 11, chose to defer acceptance of the agreement at Gardner's change of plea hearing. Fed. R. Crim. P. 11(c)(3)(A).

- 24 - Indeed, it did the opposite. The district court's

actions demonstrate that it accepted the plea agreement. The

district court said it relied on the "facts . . . set forth in the

offense conduct paragraph of the plea agreement" when it accepted

Gardner's guilty plea during his change of plea hearing and later

said it was enforcing, not rejecting, the agreement when it allowed

the government to withdraw from the agreement due to Gardner's

breach. See United States v. Soloff,

(4th Cir.

2021) ("Where the record furnishes sufficient evidence to conclude

that a district court constructively accepted the plea agreement,

the court's failure to explicitly accept the agreement will not

undo the parties' bargain."); United States v. Leyva-Matos,

, 1216 n.1 (10th Cir. 2010) (concluding that "the district

court . . . constructively accepted the plea agreement by working

within its terms and accepting certain stipulations while

rejecting others"); United States v. Brown,

(7th

Cir. 2009) (finding that a district court accepted a plea agreement

when "every aspect of the court's disposition . . . was consistent

with an acceptance of the plea agreement"); United States v.

Skidmore,

(6th Cir. 1993) ("[T]he court's failure

to elect clearly one of the options specified in [Rule 11(c)'s

predecessor] amounted to an acceptance of the plea agreement.").

The majority uses the "opportunity to withdraw" provision (which,

if it had actually been triggered, would have voided the agreement

- 25 - entirely) to imply a new and unbargained-for term into the

agreement.14 Doing so upsets the bargain the parties struck. I

would conclude that because the district court never said it would

not accept the agreement, did not treat the agreement as null and

void, and in fact enforced the plea agreement's breach provision,

Gardner has no opportunity to withdraw.

Even if Gardner did have an opportunity to withdraw his

guilty plea under the plea agreement, Gardner cannot withdraw this

plea unless he can "show a fair and just reason for requesting the

withdrawal." Fed. R. Crim. Proc. 11(d)(2)(B). We review for abuse

of discretion the district court's determination that Gardner has

not shown a fair and just reason. See United States v. Merritt,

(1st Cir. 2014). There was no abuse of discretion

here.

To determine if a fair and just reason for withdrawal

exists, district courts consider five factors: "(1) the timing of

defendant's change of heart; (2) the force and plausibility of the

14 At times, the majority implies that the government is choosing to treat the agreement as either unenforceable or null and void. This implication misunderstands the government's position. The government wants to treat the agreement as enforceable, enforce the provision allowing it to withdraw from the agreement, and hold Gardner to his guilty plea pursuant to the agreement. The majority's discussion of the "general law of plea bargaining" governing what happens when the defendant breaches a plea agreement is not relevant here because the plea agreement explicitly says what happens: if Gardner breaches, the government (but not Gardner) has the right to withdraw from the plea agreement.

- 26 - reason; (3) whether the defendant has asserted his legal innocence;

(4) whether the parties had reached (or breached) a plea agreement;

and (5) most importantly, whether the defendant's guilty plea can

still be regarded as voluntary, intelligent, and otherwise in

conformity with Rule 11 . . . in light of the proffered reason and

the disclosed circumstances." United States v. Tilley,

(1st Cir. 1992). If, and only if, a fair and just reason

exists, then the district court considers whether the government

will be prejudiced by the withdrawal of the plea. See United

States v. Adams,

(1st Cir. 2020) ("If the totality

of [the other five factors] militates in favor of allowing the

plea to be withdrawn, the court should then consider whether, and

to what extent, withdrawal would prejudice the government.");

United States v. Flete-Garcia,

(1st Cir. 2019);

United States v. Marrero-Rivera,

(1st Cir. 1997);

United States v. Muriel,

(1st Cir. 1997); United

States v. Desmarais,

(1st Cir. 1992); United States

v. Ramos,

(1st Cir. 1987). Here, the district

court correctly concluded that no fair and just reason for

withdrawal exists. All of the factors cut against Gardner.

On the first factor, Gardner pleaded guilty in October

2018 and moved to withdraw his plea almost four months later, in

- 27 - February 2019.15 We have held that waiting only thirteen or

fourteen days from the date of the original plea cuts against the

defendant. See Ramos,

(1st Cir. 1987); Nunez

Cordero v. United States,

(1st Cir. 1976). That

is because a quick withdrawal might indicate that the original

plea was "made in haste," Nunez Cordero,

, and that

the request for withdrawal is not being made "to gain personal

advantage," Ramos,

; see also United States v.

Fernández-Santos,

(1st Cir. 2017) ("The timing of

a motion to withdraw a guilty plea is important . . . because it

is 'highly probative of motive.'") (quoting United States v. Doyle,

(1st Cir. 1992)); United States v. Sanchez-

Barreto,

(1st Cir. 1996) ("[Defendants] belated

plea-withdrawal motions substantially weakened [their] claims that

their guilty pleas resulted from confusion or coercion."). The

four-month delay here shows that Gardner's request to withdraw was

15 The majority measures the timing of the plea withdrawal as "sixteen days after the court granted the government's motion to withdraw from the plea agreement." The majority ignores the fact that the reason we consider timing at all is because it informs whether the defendant's original plea was knowing and voluntary and whether the defendant is trying to strategically withdraw the plea. Even in United States v. Isom,

, 53 n.15 (1st Cir. 2009), on which the majority relies, the Court noted that the delay in filing a withdrawal motion after the original plea was relevant and cut against the defendant. See

(explaining that the "delay works to [the defendant's] detriment" when "his claim of innocence did not depend on the events in the companion case, but could have been raised earlier").

- 28 - to gain personal advantage, not because he made his original plea

too hastily.

As to the second factor, Gardner wants to withdraw his

guilty plea because he chose to assault his codefendant. The

majority's holding "would allow the defendant to withdraw his

guilty plea simply on a lark" after he "has sworn in open court

that he actually committed the crimes, after he has stated that he

is pleading guilty because he is guilty, after the court has found

a factual basis for the plea, and after the court has explicitly

announced that it accepts the plea." Hyde,

. It

"debases the judicial proceeding at which a defendant pleads and

the court accepts his plea" by "degrad[ing] the otherwise serious

act of pleading guilty into something akin to a move in a game of

chess."

. Allowing defendants to benefit by

breaching their plea agreements makes it harder for the government

to enforce plea agreements. The majority opinion forces the

government to either excuse breaches or void plea agreements and

take on the costs and risks associated with allowing defendants to

withdraw their guilty pleas. If defendants can escape their plea

agreements through intentional breaches, they may choose to gamble

on future acquittals by breaching. Encouraging breaches in this

way will cause defendants to treat plea agreements as disposable

and to take their terms less seriously. The "force and

- 29 - plausibility" of Gardner's reason therefore does not support

withdrawal. Tilley,

.

The majority agrees that the remaining three factors cut

against Gardner. Gardner has made no claim of innocence, "an

'important factor' in determining whether there is fair and just

reason to allow him to withdraw his plea." United States v.

Daniels,

(1st Cir. 1987) (quoting United States v.

Kobrosky,

(1st Cir. 1983)). He admits that he

is responsible for breaching the agreement by assaulting his

codefendant. See Tilley,

. And his plea was knowing

and voluntary, which is the most important factor in the analysis.

See United States v. Muriel,

(1st Cir. 1997).

By the majority's own count, three factors, including the most

important one, do not support withdrawal. Only two support it.16

I do not see how the majority can conclude that the district court

committed a "demonstrable abuse of discretion" here in holding

that no fair and just reason existed for Gardner to withdraw his

plea. See Marrero-Rivera,

("[W]e accord

considerable deference to the firsthand assessment ultimately made

16 The majority says there is no prejudice to the government and that this fact helps Gardner. Even if this were true, prejudice is only relevant if the totality of the other factors supports withdrawal. See United States v. Adams,

(1st Cir. 2020).

- 30 - by the district court, which must be affirmed absent a demonstrable

abuse of discretion.").

Finally, even if a fair and just reason for withdrawal

did exist, I strongly disagree that the government's mere assertion

of prejudice is insufficient. The majority's decision will

obviously cause harm in general and clearly, quite specifically on

the facts of this case. The majority's result is not a simple

matter of the government reinstating the original charges. The

events leading to Gardner's prosecution occurred in April 2017,

and it is self-evident that the government will be prejudiced by

having to prosecute Gardner more than four years after the fact.

The government will likely have difficulty securing

witnesses.17 After so much time has passed, it is likely that some

witnesses could not reliably testify against Gardner. See United

States v. Allard,

(1st Cir. 1991) (holding

that the fact that a "witness against [the defendant] is no longer

available to testify" "clearly . . . constitutes the kind of

prejudice that may be considered under . . . the 'fair and just

reason' standard"); Kobrosky,

("The most common

17 Gardner pleaded guilty while the government was negotiating plea agreements with his codefendants and the government did not require any of them to testify against Gardner as part of their plea deals. Gardner also assaulted one of his codefendants, who was in a wheelchair at the time, for being a "rat" and "ratt[ing] on his case," which could discourage his other codefendants from testifying against him.

- 31 - form of prejudice is the difficulty that the government would

encounter in reassembling its witnesses; and the longer the delay

in moving for a plea withdrawal, the greater this prejudice is

likely to be."). Further, a confidential informant was important

to the government's case. There is no evidence that this informant

is still alive or able to testify against Gardner.

Finally, much of the delay prejudicing the government

here is attributable to Gardner, who deliberately prolonged this

appeal (perhaps for tactical reasons) by seeking briefing

extensions for almost eight months. See Allard,

(explaining that "a delay that prejudices the government's case is

a factor weighing against withdrawal" when "the defendant can be

blamed for the delay"). Gardner's brief was supposed to be filed

in December 2019, before the COVID-19 pandemic. His counsel

repeatedly requested extensions, many of which were totally

unrelated to the pandemic. In all, he received ten extensions and

had to be ordered to file a brief by August 14, 2020. Such delay

tactics should not be rewarded.

I respectfully dissent.

- 32 -