Defendant-appellant Julio Ramirez appeals from a June 28, 2002 judgment of conviction entered, following his guilty plea, in the United States District Court for the Southern District of New York (Richard M. Berman, Judge). Ramirez contends that his conviction should be reversed because the district court erroneously denied his motion to withdraw from his plea agreement. The district court analyzed Ramirez's motion as if it were a motion to withdraw a guilty plea, even though Ramirez sought to keep his guilty plea intact. We affirm, but on different grounds than those relied on by the district court.

On June 3, 2000, Ramirez was charged with one count of conspiracy to distribute five or more kilograms of cocaine, in violation of 21 U.S.C. §§ 846 and 841(b)(1)(A). On May 24, 2001, he signed a plea agreement with the Government and entered a plea of guilty in the district court. The plea agreement provided that "[n]o additional understandings, promises, or conditions have been entered into other than those set forth in [the] Agreement, and none will be entered into unless in writing and signed by all parties."

Before accepting Ramirez's plea, the district court examined him in accordance with Rule 11(b) of the Federal Rules of Criminal Procedure.1 In response to the court's inquiry, Ramirez stated, inter alia, that he was satisfied with his attorney's representation and that no one had made any promises to induce him to plead guilty. At the end of the allocution, the court accepted the plea.

In September 2001, Ramirez retained new trial counsel, replacing Roy R. Kulcsar, Esq., with B. Alan Seidler, Esq. On February 18, 2002, Seidler moved for an order "permitting defendant to withdraw his guilty plea pursuant to the Plea Agreement, and plead guilty without such an agreement." In his affidavit in support of the motion, Ramirez sought simply "to withdraw from the restrictions of [his] Plea Agreement," and Seidler clarified during oral argument on the motion that Ramirez sought "to withdraw from the plea agreement, not to withdraw the guilty plea." When the district court inquired about Ramirez's motives for seeking release from his plea agreement, Seidler stated that Ramirez wanted to be free to move for a downward departure from the sentencing range set forth in the agreement, and that he wanted to be able to appeal his sentence.

In support of his motion, Ramirez argued that his decision to enter into the plea agreement had been based on a misrepresentation by his prior counsel. Ramirez claimed that Kulcsar had represented to him that if he signed the agreement, the Government would allow Ramirez's brother to assist in ongoing investigations and would give Ramirez a reduced sentence in return for his brother's cooperation. Ramirez further asserted that those purported promises had not been fulfilled.

The district court held a two-day evidentiary hearing to assess whether Ramirez should be permitted to withdraw from his plea agreement. Ramirez testified that Kulcsar had made the above-described representations to him. Kulcsar testified that he had not done so. The district court then denied Ramirez's motion from the bench.

The district court analyzed the motion as if it were a motion to withdraw a guilty plea, on the assumption that if Ramirez were allowed to withdraw from his plea agreement, his guilty plea would also have to be withdrawn. Crediting Kulcsar's version of the facts, the court concluded that Ramirez had failed to demonstrate any "fair and just reason" warranting withdrawal of the guilty plea. See Fed.R.Crim.P. 11(d)(2)(B).2 On June 27, 2002, the district court sentenced Ramirez to 108 months' imprisonment, in accordance with his plea agreement. Ramirez filed this appeal on July 11, 2002, contending that the district court erred by applying the standard set forth in Rule 11(d)(2)(B) to evaluate his motion.

We agree with Ramirez that neither Rule 11(d)(2)(B) nor the analytical criteria that courts have developed to assist in applying that Rule (and its precursor, Fed.R.Crim.P. 32(e) (effective Dec. 1, 1999)) govern the disposition of a motion to withdraw from a plea agreement where the defendant does not seek to withdraw his guilty plea. We conclude, however, that the "fair and just reason" requirement of Rule 11(d)(2)(B) provides an appropriate benchmark for considering such a motion. Because application of that requirement here warranted denial of Ramirez's motion, we affirm.

Rule 11 of the Federal Rules of Criminal Procedure sets forth the procedures according to which guilty pleas may be entered, accepted by the court, and, in appropriate circumstances, withdrawn.3 It also authorizes the use of plea agreements as vehicles for the entry of guilty pleas. Subsection (c)(1) of the Rule describes three different kinds of promises that may be incorporated into a plea agreement; a plea agreement may specify that "[a]n attorney for the government will: (A) not bring, or will move to dismiss, other charges ["Charge Bargain"];

(B) recommend, or agree not to oppose the defendant's request, that a particular sentence or sentencing range is appropriate or that a particular provision of the Sentencing Guidelines, or policy statement, or sentencing factor does or does not apply (such a recommendation or request does not bind the court) ["Non-binding Sentence Agreement"]; or

(C) agree that a specific sentence or sentencing range is the appropriate disposition of the case, or that a particular provision of the Sentencing Guidelines, or policy statement, or sentencing factor does or does not apply (such a recommendation or request binds the court once the court accepts the plea agreement) ["Binding Sentence Agreement"].

Fed.R.Crim.P. 11(c)(1).4 In the usual course, the court will accept a defendant's guilty plea as soon as it is entered, but will defer acceptance of the plea agreement until it has reviewed the pre-sentence report. See United States Sentencing Guidelines ("U.S.S.G." or "Guidelines") § 6B1.1(c) (directing district court to defer acceptance of any plea agreement until it has had "an opportunity to consider the presentence report"); but see Fed.R.Crim.P. 11(c)(3)(A) (giving court the option, where the agreement is a Charge Bargain or a Binding Sentence Agreement, to "accept the agreement, reject it, or defer a decision until the court has reviewed the presentence report").5

Subsections (d) and (e) of Rule 11 govern the withdrawal of guilty pleas.6 Subsection (d) provides as follows:

A defendant may withdraw a plea of guilty or nolo contendere:

(1) before the court accepts the plea, for any reason or no reason; or

(2) after the court accepts the plea, but before it imposes sentence if:

(A) the court rejects a plea agreement under Rule 11(c)(5); or

(B) the defendant can show a fair and just reason for requesting the withdrawal.

Subsection (e) provides that a defendant may not withdraw a guilty plea once sentence has been imposed; after that point, a plea may be set aside only on direct appeal or through collateral attack. See Fed.R.Crim.P. 11(e).7

In every instance, irrespective of the type of plea agreement involved, a defendant may, as a matter of right, withdraw his guilty plea before it has been accepted by the district court. See Fed.R.Crim.P. 11(d)(1).8 A defendant who has entered into a Charge Bargain or Binding Sentence Agreement also has an unrestricted right to withdraw his guilty plea after its acceptance but before sentence if the district court rejects his plea agreement.9 See Fed.R.Crim.P. 11(d)(2)(A); Fed.R.Crim.P. 11(c)(5).10 But if the defendant has entered into a Non-binding Sentence Agreement, the agreement becomes effective upon entry of the guilty plea; it does not require a formal acceptance or rejection by the district court. See Fed.R.Crim.P. 11 advisory committee's note on 1979 amendments. As a result, in the latter case, the district court's refusal to follow the agreement's sentencing recommendations will not entitle the defendant to withdraw his plea. See Fed.R.Crim.P. 11(c)(3)(B) (requiring court to advise defendant at plea allocution that he "has no right to withdraw the plea if the court does not follow [a Non-binding Sentence Agreement's] recommendation or request").11 Instead, the defendant will be permitted to withdraw his plea only if he can show a "fair and just reason" for the withdrawal — the same standard that applies when the court has already accepted the plea agreement (or, in the case of a Non-binding Sentence Agreement, adopted the agreement's sentencing recommendations). See Fed.R.Crim.P. 11(d)(2)(B).12

Although Rule 11 provides extensive direction concerning the procedures for pleading guilty pursuant to a plea agreement and for subsequent withdrawal of the guilty plea itself, it makes no mention of the rules and procedures attendant on a defendant's withdrawal from a plea agreement. The district court concluded that Ramirez's withdrawal from his plea agreement would require withdrawal of his guilty plea and entry of a new guilty plea. Accordingly, the court treated Ramirez's motion as a motion to withdraw a guilty plea pursuant to Rule 11(d)(2)(B). It then denied the motion after concluding that Ramirez had failed to demonstrate a "fair and just reason" for withdrawing the plea.

We disagree with the district court's analysis in two respects. As an initial matter, nothing in the text of the Federal Rules of Criminal Procedure would have required that Ramirez's guilty plea be withdrawn either following or as a prerequisite to nullification of his plea agreement. Second, we conclude that while Rule 11(d)(2)(B)'s "fair and just reason" standard is useful in considering a motion to withdraw from a plea agreement, neither the Rule itself nor the judicially crafted criteria for applying it govern the disposition of such a motion.

A. Withdrawal of Guilty Plea as Prerequisite to, or Necessary Consequence of, Withdrawal from the Plea Agreement

Ramirez's plea agreement was a Non-binding Sentence Agreement; it included only provisions of the type described in Rule 11(c)(1)(B). As explained above, had the district court refused to follow the sentencing recommendations contained in the agreement, that refusal would not have entitled Ramirez to withdraw his plea. See Fed.R.Crim.P. 11(d)(2)(A) and (c)(3)(B). In light of this fact, we cannot agree with the district court that Ramirez would have been required to withdraw his plea if the district court had granted his motion. As Rule 11(c)(3)(B) makes plain, the validity of a guilty plea made pursuant to a Non-binding Sentence Agreement is wholly independent of the court's acceptance of the recommendations contained in the agreement — a fact of which the defendant, when he enters his plea, is made fully aware. See Fed.R.Crim.P. 11(c)(3)(B); see also Fed.R.Crim.P. 11 advisory committee's note on 1979 amendments (there is no "need for rejection with opportunity for withdrawal [from a Non-binding Sentence Agreement] in light of the fact that the defendant knew the nonbinding character of the recommendation or request"). If a guilty plea is impervious to the court's rejection of a Non-binding Sentence Agreement, then it should also be impervious to a defendant's voluntary withdrawal from a Non-binding Sentence Agreement. (In the latter case, of course, the Government is released from its obligations under the plea agreement and may, therefore, seek a higher sentence than stipulated in the agreement.)

For the foregoing reasons, we hold that where a defendant seeks to withdraw from a Non-binding Sentence Agreement and consents to keeping his guilty plea intact, permission to withdraw from the plea agreement does not necessitate the entry of a new guilty plea. See United States v. Standiford, 148 F.3d 864, 867-68 (7th Cir.1998) (assuming that plea agreement can be withdrawn without withdrawing guilty plea); United States v. Modafferi, 112 F. Supp. 2d 1192, 1196-97 (D.Haw.2000) (following Standiford).13 Accordingly, the district court's assumption that Ramirez would have had to withdraw his plea in order to withdraw from his plea agreement was erroneous.

The question remains, however, whether the requirements for withdrawal of a guilty plea set forth in Rule 11 ought nevertheless to govern motions to withdraw from plea agreements as well. The Seventh Circuit, the only other circuit to have considered the question, has held that they should. See Standiford, 148 F.3d at 868.14 As explained below, although we do not share that court's view that Rule 11(d)(2)(B) applies of its own accord to a defendant's motion to withdraw from his plea agreement without withdrawing his guilty plea, we think the standard incorporated in the Rule serves as an appropriate benchmark for evaluating such a motion.

In Standiford, the defendant argued that because a plea agreement is not accepted by the district court until sentencing, see U.S.S.G. § 6B1.1(c), he had an unrestricted right to withdraw from his plea agreement (while keeping his guilty plea intact) up until that point.15 The court rejected the argument. It reasoned that a plea agreement, unlike a plea, is a contract between a defendant and the Government to which the court is not a party. Standiford, 148 F.3d at 868. Accordingly, the court explained, a plea agreement is binding on a defendant even in the absence of court approval, and thus may not be freely repudiated prior to its acceptance by the district court. Id. The court further held that because, in its view, pleas and plea agreements are "'bound up together,'" id. (quoting United States v. Hyde, 520 U.S. 670, 677, 117 S. Ct. 1630, 137 L. Ed. 2d 935 (1997)), the rule requiring a defendant to show a "fair and just reason" for withdrawing a guilty plea governs a motion to withdraw from a plea agreement as well, even where (as was the case in Standiford) the court has not yet accepted or adopted the plea agreement. Id.

We disagree with the Standiford court that Rule 11(d)(2)(B) applies, of its own accord and with all its attendant jurisprudence, to a motion to withdraw from a plea agreement where the defendant wants to keep his guilty plea intact. As noted above, Rule 11 does not expressly address motions to withdraw from plea agreements; it deals only with motions to withdraw guilty pleas. And the guilty plea and plea agreement are not so intertwined with one another that a rule governing one should be read to govern the other when the language of the rule does not so provide. See Hyde, 520 U.S. at 677, 117 S. Ct. 1630 ("[The Federal Rules of Criminal Procedure] nowhere state that the guilty plea and the plea agreement must be treated identically."); Ellis, 356 F.3d at 1205 (rejecting view that plea and plea agreement are "inextricably bound up together"). Indeed, as the Standiford court recognized, the plea and the plea agreement bind a defendant in different ways and are accepted (or, in the case of Non-binding Sentence Agreements, informally adopted) by the sentencing court at different stages of the litigation.

But our rejection of the view that Rule 11 governs a defendant's motion to withdraw from his plea agreement (but not his guilty plea) does not prevent us from borrowing the "fair and just reason" requirement for use in this situation. And an examination of the competing concerns surrounding withdrawal from a plea agreement compels us to do just that.

On the one hand, it would be anomalous to require a defendant seeking to withdraw only from his plea agreement to show more than a fair and just reason for permitting withdrawal, for a defendant in such a case is asking the court to sanction a lesser breach of the plea agreement than withdrawal of the guilty plea would entail. There is less prejudice to the Government where a defendant seeks to withdraw from his plea agreement but not his guilty plea because such a withdrawal does not raise the prospect of a trial. And withdrawal from a plea agreement, without more, does not "`undermine[ ] confidence in the integrity of our judicial procedures,'" United States v. Maher, 108 F.3d 1513, 1529 (2d Cir.1997) (quoting United States v. Sweeney, 878 F.2d 68, 70 (2d Cir.1989) (per curiam)), in the same way that withdrawal of a guilty plea does because the court is not a party to the abrogated agreement and the withdrawal does not result in a wholesale reexamination of the central issue, viz., the guilt or innocence of the defendant.

On the other hand, permission to withdraw from a plea agreement without withdrawing a guilty plea should not simply be granted upon request. Although the prejudice to the Government may be less than it would be if the guilty plea were withdrawn, it is not negligible; the Government will have to litigate sentencing disputes that had previously been resolved by the plea agreement. Moreover, there are administrative costs imposed on the court as a result of a defendant's withdrawal motion and subsequent sentencing arguments. Finally, a defendant like Ramirez has, in one sense, a less worthy claim to relief than a defendant seeking to withdraw his plea: whereas the latter may harbor a claim that he is innocent, the former has no such claim; he is hoping only to secure a more favorable sentence.

In light of these competing concerns, we think it appropriate to require a defendant who wants to withdraw from his plea agreement, but keep his guilty plea intact, to demonstrate some justification for the withdrawal, and the "fair and just reason" standard articulated in Rule 11(d)(2)(B) satisfactorily captures the proper balance.

We hasten to observe, however, that the criteria developed by courts to determine whether a defendant has presented a "fair and just reason" for withdrawing his guilty plea under Rule 11(d)(2)(B) do not readily apply where a defendant wants to keep his guilty plea intact. A district court deciding whether to permit a defendant to withdraw his guilty plea generally considers, among other things: (1) the amount of time that has elapsed between the plea and the motion to withdraw; (2) whether the defendant is asserting his innocence; and (3) whether — and to what degree — the Government will suffer prejudice as a result of the withdrawal. See, e.g., United States v. Couto, 311 F.3d 179, 185 (2d Cir.2002); United States v. Karro, 257 F.3d 112, 117 (2d Cir.2001). Moreover, "[t]o get permission to withdraw a guilty plea, a defendant must raise a significant question about the voluntariness of the original plea." United States v. Torres, 129 F.3d 710, 715 (2d Cir.1997).

The first factor may be relevant here, if we assume that withdrawal from a plea agreement, like withdrawal of a guilty plea, should be allowed only in those cases where a defendant plausibly maintains that his entry into the plea agreement was procured by nefarious means or was the result of his own confusion. As the First Circuit has noted,

the timing of a defendant's attempted plea withdrawal is highly probative of motive.... While an immediate change of heart may well lend considerable force to a plea withdrawal request, a long interval between the plea and the request often weakens any claim that the plea was entered in confusion or under false pretenses.

United States v. Doyle, 981 F.2d 591, 595 (1st Cir.1992). But the fact of delay may not always cast suspicion on a defendant's reason for seeking withdrawal from his plea agreement; where, as in this case, the assertion is that the Government failed to fulfill its side of the bargain (as the defendant understood it), a delay may be consistent with the defendant's good faith expectation of the Government's performance.

The second factor is even less helpful. A defendant's assertion of innocence normally weighs in favor of permitting plea withdrawal. See, e.g., id. at 596 ("Courts look more hospitably on a motion to withdraw a guilty plea when the motion is coupled with an assertion of innocence."). It is easy to see why; greater injustice is done to a defendant who may actually be innocent yet is denied a trial than to one who admits his guilt but insists on a trial. Cf. United States v. Roberts, 570 F.2d 999, 1009 (D.C.Cir.1977). Where a defendant is seeking only to withdraw from his plea agreement, however, the same rationale does not apply; the defendant necessarily admits his guilt, and he is not seeking a trial.

The third factor, prejudice to the Government, will be essentially the same (and relatively minor) in most cases where a defendant seeks to withdraw from his plea agreement but not to withdraw his guilty plea. "Prejudice" in the context of plea withdrawal typically refers to depriving the Government of the benefit of its bargain by having the burden of trial preparation suddenly thrust upon it, as well as the potential difficulty to the Government in securing evidence against the defendant that would have been easier to secure at an earlier moment in time. See, e.g., United States v. Lineback, 330 F.3d 441, 445 (6th Cir.2003) (Gilman, J., concurring). But neither form of prejudice exists where the defendant simply wants to be released from his plea agreement.

The Government here claims, rather vaguely, that it would have lost "the benefit of the bargain it expected" if Ramirez's motion had been granted. But the primary benefit to the Government under a plea agreement is the defendant's guilty plea. See Hyde, 520 U.S. at 677, 117 S. Ct. 1630 (noting that defendant performs his side of plea bargain when he pleads guilty). This benefit would have been fully realized even if Ramirez's motion had been granted. The only prejudice to the Government in a case like this is that it may be forced to litigate the defendant's sentence (which, under a typical plea agreement, is subject to stipulation) and, perhaps, defend or seek an appeal from whatever sentence is imposed. These added burdens are not trifling, but they are not as onerous or as unpredictable as those associated with withdrawal of a guilty plea.

We can imagine circumstances in which the potential prejudice to the Government is of a different kind and/or magnitude than that presented here. For example, prejudice may be an important factor weighing against withdrawal from a plea agreement in a case where the defendant has agreed, as part of his deal, to provide assistance to the Government, and wishes to renege on that deal.16 Because such possibilities exist, the prejudice factor should never be ignored. As indicated above, however, it will generally have little weight where the defendant is not seeking to withdraw his guilty plea.

Finally, where a defendant seeking to withdraw from his plea agreement does not also seek withdrawal of his guilty plea, the voluntariness of the guilty plea is not at issue, and, therefore, should bear no direct relation to the fairness of granting the defendant's request. Instead, the relevant question is likely to be whether the defendant failed to understand, was misled about, or simply does not like certain subsidiary terms of the plea agreement (e.g., the length of the sentence). Therefore, while voluntariness may be relevant, the focus here ought to be on the voluntariness of the defendant's consent to the terms of the plea agreement, not the voluntariness of the plea itself.

Taken together, the criteria used under Rule 11(d)(2)(B) to determine whether withdrawal of a guilty plea should be permitted are largely unhelpful in deciding whether to grant a defendant's motion to withdraw from his plea agreement without withdrawing his guilty plea. Therefore, although we borrow Rule 11(d)'s "fair and just reason" requirement, we do not borrow its accoutrements. Instead, the factors to be considered in determining whether a defendant has demonstrated a "fair and just reason" should be tailored to the context of plea agreement withdrawal. For example, rather than ask whether the defendant is asserting his innocence (he is not), or whether the guilty plea itself was voluntary (an inquiry that is beside the point), the district court should ask whether there is credible evidence that the defendant did not freely and voluntarily enter into the plea agreement, either because he was coerced or improperly induced to accept its terms, or because he misunderstood them. The court should also seek to determine the extent of any prejudice likely to be suffered by the Government, keeping in mind that, as we have explained, such prejudice will generally be considerably less than would attend withdrawal of a guilty plea. Other factors may be relevant to the analysis, but these two questions ought to be at its center.

In the instant case, the district court believed that permitting Ramirez to withdraw from his plea agreement would have required withdrawal of his guilty plea. Accordingly, it evaluated Ramirez's motion in light of the criteria normally applied to motions to withdraw guilty pleas. The court found that the "absence of a claim of innocence weigh[ed] in favor of allowing [Ramirez's] guilty plea [and plea agreement] to stand." It also concluded that the voluntariness of Ramirez's guilty plea — i.e., the fact that no one had induced him to plead guilty — weighed against granting the motion. As explained above, Ramirez's failure to either claim his innocence or challenge the voluntariness of his guilty plea should not have been considered in examining the merits of his motion.

Nevertheless, based on our review of the facts of this case, we conclude that the district court's error was harmless. Ramirez's only argument in support of his motion was that his counsel had misled him, thus inducing him to sign the plea agreement under false pretenses. After an evidentiary hearing, the district court rejected that claim. Specifically, the court found that Kulcsar had not "made any representation to the defendant prior to the signing of the plea agreement [that] was at odds or variance with the plea agreement." Under any analysis, the district court's crediting of Kulcsar's account of events and its concomitant discrediting of Ramirez's testimony preclude Ramirez from obtaining the relief he seeks. Because Ramirez has not adduced any credible evidence that he either misunderstood the terms of the plea agreement or was improperly induced into signing it, there is no "fair and just reason" to warrant withdrawal from the plea agreement.

For the foregoing reasons, the conviction is AFFIRMED.