United States v. Sanchez‑Zurita: Enforceability of Informal Email-Based Sentencing Promises in Open Pleas

I. Introduction

In United States v. Sanchez‑Zurita, No. 24‑60602 (5th Cir. Dec. 17, 2025), the United States Court of Appeals for the Fifth Circuit vacated a 240‑month sentence for illegal re‑entry and remanded for resentencing before a different district judge. The core of the decision is not about immigration law or the substantive reasonableness of the twenty‑year sentence, but about the enforceability of an informal prosecutorial promise made by email in the context of an “open plea” (a guilty plea without a formal written plea agreement).

The panel, in an opinion by Judge Priscilla Richman, holds that:

A prosecutor’s clear promise—communicated by email—to state at sentencing that the government would be “satisfied with a guideline sentence” can constitute an enforceable plea-related agreement even when:
- there is no formal, written plea agreement;
- the parties and the court all state on the record that there is “no plea agreement”; and
- the plea is characterized as “open.”
Where the defendant reasonably relies on that promise in deciding to plead guilty, the government must strictly perform it under Santobello v. New York.
The government breached the promise by taking “no position” on an upward variance instead of stating it would be satisfied with a guideline sentence.
The appropriate remedy, at the defendant’s election, is specific performance: resentencing before a different judge, with the government stating that it is satisfied with a guideline sentence.

The decision significantly clarifies that, in the Fifth Circuit, plea-related promises can arise and be enforced even outside the conventional framework of Rule 11(c) written plea agreements. It also underscores the binding character of prosecutorial communications in email negotiations, and the continued vitality of contract-law and promissory-estoppel principles in the plea-bargaining context.

II. Summary of the Opinion

Luis Javier Sanchez‑Zurita, previously removed after an aggravated felony conviction, was indicted under 8 U.S.C. § 1326(a)(1), (b)(2) for illegal re‑entry. The government and defense counsel exchanged emails about a potential plea agreement. The government initially proposed an agreement that would include a recommendation of a “within guideline sentence (lower 50%).” Defense counsel, wanting to preserve appellate rights, asked what would happen if the defendant entered an “open plea” (a guilty plea without a written agreement).

In response, the prosecutor wrote (paraphrasing the key language emphasized by the court):

“I stand by the recommendation, however, when a defendant pleads open, there is no recommendation.”
“If the court asks me during sentencing, I will state the government would be satisfied with a guideline sentence.”
“But an open plea is without a recommendation, correct.”

Defense counsel met with Sanchez‑Zurita, who opted to plead guilty without a written agreement. At the plea hearing, all participants—including the defendant—told the court there was “no plea agreement.” The Presentence Investigation Report (PSR) calculated an offense level of 10 and a guideline range of 8–14 months, and noted that there was no plea agreement. The district court adopted the PSR.

At sentencing, however, the court announced it was considering an upward variance and asked the government about its position. The government:

confirmed it was not requesting a variance;
confirmed there was “no plea agreement”; and
told the court it took “no position” on a variance and left sentencing “to the Court’s discretion.”

Despite defense counsel’s oral objection—explicitly referencing his understanding of an agreement that the government would recommend the guidelines and the lower 50%—the court imposed the statutory maximum of 240 months (20 years), a massive upward variance from the 8–14 month guideline range. Defense counsel renewed his objection.

On appeal, and before merits briefing, both sides moved to remand. The government, after reviewing the emails, acknowledged that Sanchez‑Zurita could reasonably have understood that it would tell the court it was satisfied with a guideline sentence and sought remand “out of an abundance of caution” to fulfill any inadvertent promise. A motions panel denied the government’s motion, stating that “there was no plea agreement,” but did so without having the emails before it. Later, the record on appeal was supplemented with those emails.

The merits panel:

Re-examines whether any enforceable agreement existed and holds that the emails created an informal but binding plea-related promise.
Finds that Sanchez‑Zurita reasonably relied on that promise in deciding to plead guilty.
Holds that the government breached the agreement by taking “no position” instead of stating it would be satisfied with a guideline sentence.
Vacates the sentence and grants the defendant’s chosen remedy: specific performance of the agreement before a different sentencing judge.

III. Detailed Analysis

A. Factual and Procedural Background in Greater Detail

1. The charge and statutory framework

Sanchez‑Zurita was charged under 8 U.S.C. § 1326(a)(1), (b)(2) for illegal re‑entry after removal following an aggravated felony conviction. Section 1326(b)(2) authorizes a maximum term of imprisonment of 20 years. The district court advised him of that maximum at the plea hearing.

2. Email negotiations: from proposed plea agreement to “open plea”

The email exchange unfolded in several stages:

Initial proposal. The government floated a written plea agreement that would include a recommendation of a “within guideline sentence (lower 50%).”
Defense concern about appellate rights. Defense counsel sought to preserve an appeal right for any sentence above the guidelines and asked whether that could be included.
Shift toward an open plea. When the government declined that appeal carve‑out, defense counsel inquired what the government’s stance would be if Sanchez‑Zurita pled “open” (without a written agreement).
The key promise. The prosecutor replied that although an open plea does not involve a formal recommendation:
- the government would still “stand by” its view; and
- “If the court asks me during sentencing, I will state the government would be satisfied with a guideline sentence.”
Defense follow‑up. Defense counsel indicated he would discuss the options with his client and then respond. Subsequently, Sanchez‑Zurita chose to plead open.

3. The plea hearing and the “no plea agreement” statements

At the Rule 11 plea colloquy:

Defense counsel characterized plea negotiations as “not necessarily fruitful” and explained that this was why they were proceeding with an open plea.
The prosecutor confirmed there was no plea agreement.
The defendant, under oath, agreed there was no plea agreement.
The district court similarly stated that the defendant had “not … agree[d] to a plea agreement with the Government.”

The Fifth Circuit later treats these statements as directed to the absence of a formal written plea agreement, not necessarily to any informal or oral/email-based understanding about the government’s sentencing stance.

4. Sentencing: guideline range vs. statutory maximum

The PSR fixed an offense level of 10, incorporating prior felony conviction and acceptance of responsibility. The resulting guideline imprisonment range was 8 to 14 months, and the PSR stated that there was no plea agreement. Neither party objected, and the district court adopted the PSR.

At sentencing, the critical exchange went as follows:

The court announced it was considering an upward variance.
The court asked whether the prosecution was asking for a variance; the government said no.
The court confirmed with the government that there was no plea agreement.
The court then said it would “imagine” the prosecution took “no position” on whether a variance should be imposed; the prosecutor agreed and stated the government left sentencing to the court’s discretion.

Defense counsel:

argued against an upward variance; and
explicitly “note[d] for the record” that in discussions with the government, when the defendant opted to plead open, counsel “understood” there to be an agreement that the government would recommend the guidelines, including the lower 50%.

The court nonetheless imposed 240 months (20 years)—the statutory maximum—representing a staggering upward variance from 8–14 months.

5. Appeal and the motions panel

On appeal:

Both parties moved to remand prior to merits briefing.
The government, after reviewing the email chain, acknowledged that Sanchez‑Zurita could have reasonably read the emails as committing the government to inform the court it would be satisfied with a guideline sentence. To “fulfil any agreement it inadvertently made,” it sought remand for resentencing.
However, in its reply, the government disclaimed any “plea agreement” and framed the issue as a misunderstanding in pre‑plea email exchanges, not a breach of a binding agreement.
A motions panel denied the government’s motion, stating “there was no plea agreement” and noting that no emails or other evidence had been submitted. That panel’s order did not decide the defendant’s separate request for reassignment on remand.
Subsequently, Sanchez‑Zurita successfully moved to supplement the record with the emails.

The merits panel later underscores that decisions by a motions panel are not binding on the merits panel—citing Texas Democratic Party v. Abbott—and re‑examines the plea issue in light of the full record.

B. Precedents and Authorities Cited

1. Santobello v. New York and the basic rule

The controlling Supreme Court principle comes from Santobello v. New York, 404 U.S. 257 (1971). The Fifth Circuit quotes the key passage:

“[W]hen a plea rests in any significant degree on a promise or agreement of the prosecutor, so that it can be said to be part of the inducement or consideration, such promise must be fulfilled.”

Santobello established that due process requires enforcement of prosecutorial promises that induce a guilty plea: the state may not secure a plea on one set of assurances and then disregard them at sentencing.

2. United States v. Harper and standards for breach and remedy

The panel relies on United States v. Harper, 643 F.3d 135 (5th Cir. 2011), for several propositions:

The defendant bears the burden of proving the factual basis for an alleged plea‑agreement breach by a preponderance of the evidence.
The government must “strictly adhere” to the terms of any promises it makes in a plea agreement.
If the government breaches, the defendant is entitled to relief “even if the Government's breach did not ultimately influence the defendant's sentence.”

Harper also frames the two classic remedies for breach:

Specific performance (typically resentencing before a different judge, with the government required to perform the promise), or
Withdrawal of the guilty plea.

3. United States v. Williams and the defendant’s choice of remedy

In United States v. Williams, 821 F.3d 656 (5th Cir. 2016), the court recognized that when the government breaches a plea agreement, the defendant has the right to have his chosen remedy accepted—either specific performance and resentencing before a different judge or withdrawal of the plea—subject to a potential “material breach” limitation when rescission is sought.

Citing Williams, the panel here honors Sanchez‑Zurita’s selected remedy: specific performance, not plea withdrawal. The court does not engage in a separate “materiality” analysis, implicitly treating the breach as sufficiently serious to warrant full enforcement.

4. United States v. Malmquist: standard of review

The panel cites United States v. Malmquist, 92 F.4th 555 (5th Cir. 2024), for the proposition that whether the government breached a plea agreement is reviewed de novo. This standard reflects that the determination largely turns on legal interpretation of the agreement’s scope and meaning, rather than purely factual disputes.

5. United States v. Long: contract-law lens

The Fifth Circuit reiterates its longstanding view that “this court applies general principles of contract law in interpreting the terms of a plea agreement,” citing United States v. Long, 722 F.3d 257, 262 (5th Cir. 2013).

In practice, that approach means:

Looking for offer, acceptance, consideration, and clear terms;
Favoring objective manifestations of assent over subjective understandings; and
Constricting prosecutorial discretion where the government has bargained away some of its flexibility in exchange for the plea.

6. Johnson v. Beto and promissory estoppel concepts

The panel draws explicitly on Johnson v. Beto, 466 F.2d 478 (5th Cir. 1972), which analogized plea bargaining to promissory estoppel:

“Analogous to promissory estoppel, plea bargaining must have more substantiality than mere expectation and hope. It must have explicit expression and reliance and is measured by objective, not subjective, standards.”

Johnson provides the doctrinal framework that the panel uses to decide whether the emails rose above “mere expectation” and became an enforceable promise on which the defendant could reasonably rely.

7. United States v. Melton and misrepresentation

The panel also quotes United States v. Melton, 930 F.2d 1096 (5th Cir. 1991):

“[T]he government may neither misrepresent its intentions nor renege on representations reasonably relied and acted upon by defendants and their counsel.”

This principle is directly applied to the government’s email promise here: once the prosecutor represented that he would state the government was satisfied with a guideline sentence if asked, it was impermissible to renounce that representation at sentencing after the defendant had already pled guilty in reliance on it.

8. United States v. McDaniels and United States v. Cervantes:

a. General rule about plea-hearing statements

The government invoked United States v. McDaniels, 907 F.3d 366 (5th Cir. 2018), which holds that the court “generally will not allow a defendant to contradict his testimony given under oath at a plea hearing.” This rule promotes finality and integrity in plea proceedings: defendants cannot easily undo sworn statements that there were no promises or that they understand the consequences.

b. The exception: precise proof of a contrary promise

However, McDaniels, relying on United States v. Cervantes, 132 F.3d 1106 (5th Cir. 1998), recognizes an exception in collateral proceedings: a defendant may challenge plea‑hearing statements if he can prove:

“the exact terms of the alleged promise,”
“exactly when, where, and by whom the promise was made,” and
“the precise identity of an eyewitness to the promise.”

While this case arises on direct appeal (not collateral review), the panel notes that the emails effectively supply (1) and (2) with precision: they document the exact language, timing, and authorship of the government’s promise, and because the communication is written, there is no need for a live eyewitness (3).

The panel thereby harmonizes the general rule of McDaniels with enforcement of the emails: the defendant is not attempting to contradict his plea colloquy in an unsubstantiated way; he brings documentary proof that a real, specific promise existed.

9. United States v. Harrier (unpublished): “no plea agreement” statements

The panel cites United States v. Harrier, 229 F. App’x 299 (5th Cir. 2007) (unpublished) for the proposition that district court statements that there is “no plea agreement” are not necessarily dispositive of whether a non‑written plea agreement exists. In Harrier, the court interpreted such statements as referring to the absence of a written agreement, not to any oral or informal understanding; it found an oral agreement existed when one count was dismissed in exchange for a plea to another.

Harrier bolsters the conclusion here that repeated references to “no plea agreement” at the plea and sentencing hearings do not defeat the existence of an informal promise documented in emails.

10. Gibson v. Blackburn and supplementation of the record

The panel quotes Gibson v. Blackburn, 744 F.2d 403, 405 n.3 (5th Cir. 1984), to note that appellate courts do have authority (though they generally refrain) to enlarge the appellate record to include material not before the district court. Here, that authority was used when the court granted Sanchez‑Zurita’s motion to supplement the record with the plea‑negotiation emails—evidence critical to resolving whether a promise existed.

11. Texas Democratic Party v. Abbott: motions panel vs. merits panel

Relying on Texas Democratic Party v. Abbott, 978 F.3d 168, 176 (5th Cir. 2020), the panel underscores that “[o]pinions and orders of a panel with initial responsibility for resolving motions filed in an appeal are not binding on the later panel that is assigned the appeal for resolution.” This allows the merits panel to revisit the earlier motions-panel statement that “there was no plea agreement,” especially in light of the now-complete record.

12. Villa v. United States (6th Cir. 2025): comparison on vague email offers

The panel cites Villa v. United States, 135 F.4th 513 (6th Cir. 2025), as a contrast. In Villa, the Sixth Circuit found that an email did not constitute a plea offer where it:

required additional cooperation by the defendant, and
used tentative language that the prosecutor “would consider” changing his recommendation.

By invoking Villa, the panel highlights that unlike in Villa, the promise to Sanchez‑Zurita:

did not require any further cooperation beyond pleading open;
was not tentative (“would consider”); and
was “not deliberately guarded and vague,” but concrete: “I will state the government would be satisfied with a guideline sentence.”

13. Missouri v. Frye: defense counsel’s duties

The panel references Missouri v. Frye, 566 U.S. 134 (2012), in a parenthetical noting defense counsel’s duty to promptly inform clients of plea offers. This supports the inference that counsel met with Sanchez‑Zurita promptly after the key email because the email was, from counsel’s perspective, materially important to the decision whether to plead.

C. The Court’s Legal Reasoning

1. Existence of an informal but enforceable agreement

The government conceded there was no formal written plea agreement, and the defendant admitted as much at the plea hearing. The central question was whether the email exchange nonetheless created a binding, plea-related promise subject to Santobello enforcement.

The panel concludes yes, emphasizing:

The emails reveal the exact terms of the promise: “If the court asks me during sentencing, I will state the government would be satisfied with a guideline sentence.”
The promise was specific, unconditional (beyond the condition that the defendant plead open), and operationally clear.
The promise was “unusual” for an open plea, but that does not make it any less enforceable once made.
Nothing in the email made the statement contingent on additional cooperation or discretionary reconsideration by the prosecutor.

By treating this email as an enforceable promise, the court holds that “plea agreement” for Santobello purposes is a functional concept: it includes informal promises and assurances that form part of the inducement for the guilty plea, not just formally titled “Plea Agreements” signed and filed under Rule 11(c).

2. Reasonable reliance by the defendant

The government argued there was no evidence that Sanchez‑Zurita actually relied on the prosecutor’s email when deciding to plead guilty, pointing to:

the surrounding language stressing that “an open plea is without a recommendation”; and
the lack of an express “acceptance” email saying, “we accept your offer.”

The panel disagrees, finding reliance reasonable and objectively manifested:

Defense counsel explicitly asked what sentence recommendation the government would give if the defendant pled open—showing the question was important to the decision.
The prosecutor responded with the clear promise about stating satisfaction with a guideline sentence if asked.
Counsel then immediately met with the defendant, and again a week later, consistent with advising him about the import of the government’s position.
The defendant then chose to plead open, the very course contemplated by the email promise.
At sentencing, defense counsel objected specifically on the ground that he understood there to be an agreement that the government would recommend the guidelines, “including the lower 50 percent.”

Under the promissory‑estoppel/contract framework of Johnson v. Beto, the court concludes:

There was “explicit expression” of a sentencing stance by the government; and
The defendant’s actions in pleading open were objectively consistent with relying on that expression.

The government’s own pre‑briefing motion to remand—acknowledging that the emails could reasonably be read as promising to tell the court the government would be “satisfied with a guideline sentence”—effectively corroborates this conclusion.

3. The effect of the “no plea agreement” statements

The government also relied on the defendant’s sworn statement at the plea hearing that there was no plea agreement, invoking McDaniels’ general rule that defendants may not contradict their plea‑hearing testimony.

The panel sidesteps any rigid reading of that rule by:

treating the “no plea agreement” colloquy as referring to the absence of a formal written agreement (consistent with Harrier), not to all possible informal promises;
emphasizing that the emails supply the exact terms, timing, and author of the promise; and
recognizing that such documented proof satisfies the rationale behind the Cervantes / McDaniels exception, even though this is a direct appeal rather than collateral review.

Thus, the plea colloquy does not bar enforcement of the email promise.

4. The breach

Once the panel finds that there was a promise and reasonable reliance, the breach analysis is straightforward:

The government promised: If the court asks at sentencing, the prosecutor will state that the government is satisfied with a guideline sentence.
At sentencing, the court asked about the government’s position in the face of a contemplated upward variance—effectively asking whether the government supported, opposed, or was content with a guidelines sentence.
Instead of stating that it would be satisfied with a guideline sentence, the government told the court it took “no position” on a variance and left the matter to the court.

The panel emphasizes that even though the court did not recite the magic words “Would the government be satisfied with a guideline sentence?”, its inquiry about whether the government asked for a variance, coupled with its assumption that the government took no position, “essentially asked for the Government's position on a within‑guideline sentence.”

Under Harper and related cases, the government must “strictly adhere” to the promise; “no position” is not equivalent to “we would be satisfied with a guideline sentence.”

The government had itself effectively recognized this as a failure to honor its representation when it moved for remand “so that Sanchez‑Zurita can be re‑sentenced with the benefit of a statement from the government that it would be ‘satisfied with a guideline sentence.’” The panel, now having found a binding agreement rather than a mere misunderstanding, holds that this was a legally cognizable breach.

5. Remedy: specific performance and reassignment to another judge

Under Harper and Williams, a defendant whose plea-related promise has been breached can:

seek specific performance (usually resentencing with the government required to comply, before a different judge), or
seek to withdraw the plea.

Sanchez‑Zurita asked for specific performance and resentencing before a different judge. The panel grants that remedy, reasoning that:

Specific performance best ensures that the defendant receives the benefit of the bargain (the government’s declared satisfaction with a guideline sentence);
Resentencing before a different judge mitigates any risk—real or perceived—that the original judge might feel constrained or influenced by the prior maximum sentence and breach context; and
Relief does not depend on showing that the breach “influenced the defendant’s sentence”—breach alone is sufficient for relief under Harper.

The court therefore:

VACATES the 240‑month sentence;
REMANDS for resentencing; and
GRANTS the motion for reassignment to a different district judge on remand.

D. Impact and Implications

1. Broadened understanding of “plea agreement” in the Fifth Circuit

This decision reinforces and clarifies that, in the Fifth Circuit, the concept of a “plea agreement” or plea‑related promise is functional and not limited to:

formal written agreements filed under Rule 11(c); or
terms recited verbatim at the plea colloquy.

Informal agreements—including email exchanges—can create binding obligations if:

they contain clear and specific prosecutorial promises;
they are communicated in the context of plea negotiations; and
the defendant reasonably relies on them in electing to plead guilty.

2. Practical consequences for prosecutors

For prosecutors, Sanchez‑Zurita is a cautionary precedent:

Email communications in plea negotiations are not mere “informal” or “off the record” chatter; they can form enforceable commitments.
Disclaimers such as “an open plea is without a recommendation” may not suffice to avoid binding obligations if, in the same communication, the prosecutor makes a more concrete promise (“I will state the government would be satisfied with a guideline sentence”).
Attempts to correct or downplay the promise on appeal (“misunderstanding,” “abundance of caution”) will not prevent the court from characterizing it as a binding agreement if the objective language supports that conclusion.

Prosecutors in the Fifth Circuit will likely respond by:

Using more standardized, carefully vetted language in email negotiations;
Avoiding any definitive representation about future sentencing positions unless they intend to be bound; and
Ensuring that any limited commitments are fully and accurately disclosed to the district court at the plea hearing under Rule 11.

3. Practical consequences for defense counsel

For defense counsel, the case offers several lessons:

Any written communication from the government that appears to commit to a sentencing recommendation or stance should be preserved and, where appropriate, treated as a binding promise.
If a client pleads in reliance on such a promise, counsel must be prepared to:
- explicitly raise the existence of the promise at sentencing, as counsel did here;
- move for remedy if the government does not adhere to it; and
- ensure that the emails or other documentation are made part of the record through proper motions, especially if the district court has never seen them.
Even if the plea is described as “open,” counsel should recognize that informal sentencing commitments may still be enforceable under Santobello.

4. Guidance for district courts

The decision implicitly sends guidance to district judges:

When parties state there is “no plea agreement,” it may still be prudent to ask whether any informal or unwritten promises or understandings exist regarding sentences or recommendations.
Where defense counsel suggests at sentencing that such an informal agreement exists, the court may need to explore the issue more fully or invite formal motions to develop the record, rather than dismissing the claim out of hand.
If it later emerges that the government made an unfulfilled sentencing promise, a remand and reassignment may be ordered, as here.

5. Appellate practice and motions panels

The case also has implications for appellate practice:

It confirms that motions‑panel rulings—particularly when made on an incomplete record—are not binding on the merits panel.
It illustrates the importance of supplementing the record promptly with documentary evidence (such as emails) that bears on plea‑agreement disputes.
Appellate counsel should not assume that an early motions‑panel characterization (“no plea agreement”) will control where new evidence emerges or is admitted later.

6. Sentencing dynamics and variance practice

Finally, the case touches on the dynamics of sentencing and variances:

The defendant’s guideline range (8–14 months) versus the imposed sentence (240 months) highlights the extraordinary effect a sentencing judge’s variance can have.
A prosecutor’s statement that the government is “satisfied with a guideline sentence” can play a meaningful role in a court’s variance decision, especially when balanced against concerns about recidivism, prior convictions, and deterrence.
By requiring the government to perform its promised satisfaction with a guideline sentence on remand, the court ensures a structurally fair process even if the district judge ultimately retains discretion to vary upward (subject to reasonableness review).

IV. Complex Concepts Simplified

1. “Open plea”

An “open plea” is a guilty plea entered without a formal written plea agreement. The defendant pleads guilty to the charged offense, and:

the government does not agree to dismiss other counts (if any);
there is typically no agreed sentence or recommendation; and
the sentencing judge retains full discretion within the statutory limits, guided by the Sentencing Guidelines and 18 U.S.C. § 3553(a) factors.

In Sanchez‑Zurita, despite the plea being “open,” the court found that an email promise about the government’s stance at sentencing was still enforceable.

2. Sentencing guidelines and “within‑guidelines sentence”

The United States Sentencing Guidelines provide an advisory range of months based on:

offense level; and
criminal history category.

A “within‑guidelines sentence” is any sentence that falls inside that calculated range. Here:

The guideline range was 8–14 months.
The government’s informal promise was to be “satisfied with a guideline sentence”—that is, any sentence in the 8–14‑month range.

3. “Variance”

A “variance” is a sentence that departs from the guideline range, either:

upward (higher than the range), or
downward (lower than the range).

It is distinct from a “departure” under the Guidelines, although the terms are sometimes loosely interchanged. A variance is based on the multi‑factor analysis in 18 U.S.C. § 3553(a), giving the court discretion to tailor the sentence to the individual circumstances.

In this case, the district court contemplated—and then imposed—an extraordinary upward variance from 8–14 months to 240 months.

4. “Specific performance” vs. withdrawal of plea

Specific performance in this context means enforcing the government’s promise by requiring it to do what it said it would do—here, to state at resentencing that it would be satisfied with a guideline sentence—usually before a new judge.
Withdrawal of the plea would allow the defendant to undo the guilty plea and return the case to a pre‑plea posture, subject to re‑prosecution on the original charge(s).

The defendant in Sanchez‑Zurita chose specific performance, and the court granted that remedy.

5. “Material breach”

A “material breach” is a breach of such significance that it defeats the purpose of the agreement, potentially justifying rescission (undoing the contract). Williams suggests that materiality may be required for a defendant to withdraw his plea as a remedy, but not necessarily for specific performance.

Here, the Fifth Circuit did not explicitly label the breach “material,” but, given the gravity of the variance and the government’s failure to honor a core sentencing promise, it effectively treated the breach as sufficiently serious to merit full relief.

6. Motions panel vs. merits panel

In appellate practice:

A motions panel handles preliminary matters (such as motions to remand or stay) early in the appeal.
A merits panel later hears the full appeal and issues the final opinion.

Under Fifth Circuit law (Texas Democratic Party v. Abbott), the merits panel is not bound by the motions panel’s earlier decisions. That allowed the panel in Sanchez‑Zurita to revisit the “no plea agreement” conclusion of the motions panel once the email evidence was in the record.

V. Conclusion

United States v. Sanchez‑Zurita stands as an important Fifth Circuit precedent on the enforceability of informal, email-based prosecutorial promises in the plea‑bargaining context. The decision confirms that:

Plea agreements, for Santobello purposes, are not confined to formal written contracts; informal but specific promises of sentencing conduct—such as a commitment to state satisfaction with a guideline sentence—can be binding.
Defendants may reasonably rely on such promises when deciding to plead guilty, even in “open plea” situations.
The government must strictly honor its promises; taking “no position” when it promised to endorse the guideline range constitutes a breach.
When a breach is found, the defendant is entitled to meaningful relief—here, specific performance and resentencing before a different judge—regardless of whether the breach can be shown to have altered the original sentence.
Plea‑hearing statements disclaiming a “plea agreement” do not categorically foreclose later proof of an informal agreement, particularly where contemporaneous written communications supply the exact terms of the promise.

In the broader legal landscape, the case will likely influence how prosecutors structure and phrase their email negotiations, how defense lawyers document and litigate plea‑related promises, and how district courts inquire about and handle informal agreements at plea and sentencing hearings. Above all, it reinforces a core constitutional and contractual principle: when the government secures a guilty plea based in any significant part on its assurances, it must keep its word.