New York Federal Criminal Practice Blog

Recently in the Plea Bargaining category:


Plea-bargaining is critical to the administration of federal criminal justice, since over 95% of federal criminal cases end in a guilty plea.  In turn, good faith – especially on the part of the most powerful party in that process – is key to maintaining that high non-trial disposition rate.  Why would thousands of defendants plead guilty to plea agreements, like the standard plea agreement in the Eastern District of New York, that permit the government to advocate a higher guideline at sentencing than the one carefully calculated in the plea agreement?  Well, because there is a good faith understanding that the prosecutor will not use this provision to support a higher guideline at sentencing absent extraordinary circumstances, such as new facts coming to light, or the application of some arcane, unanticipated guideline.  When, as in United States v. Habbas (discussed here) and United States v. MacPherson, 590 F.3d 215 (2d Cir., December 30, 2009), the prosecutor does advocate a higher guideline at sentencing on basic, predictable, nuts and bolts issues like drug type and quantity or managerial role, the entire plea bargaining process, and the manageability of judges’ trial calendars, is in jeopardy.  Luckily, MacPherson can be distinguished on its facts, but it highlights the troubling problems in the standard EDNY plea agreement, the need to avoid pleading under a plea agreement unless it genuinely gives the defendant a benefit, and the importance of obtaining some additional assurances that the prosecutor will stand by the guideline estimate in the agreement. 


MacPherson pled guilty to conspiracy to import heroin and cocaine, under a plea agreement that estimated his guideline range to be 120 to 135 months, and in which he stipulated that his sentence would be calculated using 15 kilograms of cocaine.  The agreement provided that the guideline estimate was not binding on the government, and that the defendant could not withdraw his plea if the government argued a different guideline at sentencing.  During his allocution, the defendant admitted to “repeatedly traveling to Peru to make arrangements to purchase heroin and meeting a courier at [JFK] who was carrying ten kilograms of cocaine.”

The Probation Department estimated an offense level that was 6 levels higher than the one in the plea agreement – two additional levels arising from the inclusion of 7 kilograms of heroin in the offense calculation, and 4 levels for his managerial role in the scheme.  At sentencing, the defendant objected to these enhancements, but not on the basis of the inconsistency with his plea agreement.  The government advocated the Probation Department’s estimate, which the district judge adopted.  She imposed a sentence of 262 months (at the high end of the new guideline range).  On appeal, MacPherson argued that the government had violated his plea agreement in advocating the higher guideline at sentencing. 


Reviewing the claim on a plain error standard (as required by the Supreme Court’s recent ruling in Puckett), and noting its conflicting rulings on this issue in Habbas and Palladino, the Court held that the prosecutor’s conduct here (which was similar to that upheld in Habbas) “cannot have precipitated plain error, if any error at all.”  In particular, there was no plain error despite the defendant’s stipulation in his plea agreement that his sentence would be based on 15 kilograms of cocaine:

[W]hereas typical contract stipulations state that the parties stipulate to some agreed upon terms, the agreement in this case states only that the defendant stipulates to a sentence based on the cocaine quantity. In any event, the agreement and the plea colloquy put the defendant on notice that the Pimentel estimate was not binding on the prosecutor and that if the estimate was wrong, the plea could not be withdrawn. In such circumstances, there was no plain error.

In his concurrence, Judge Newman went one step further, finding that “there [was] no error at all,” citing clear language in the plea agreement and during the plea colloquy that the government could advocate a higher offense level at sentencing than the estimate in the plea agreement. 

The Court also found that the district court’s sentencing at the high end of the new guideline was not unreasonable. 


MacPherson can be distinguished, both on the grounds that the defendant did not claim a violation of the plea agreement at the district court level, and also because the defendant allocuted at his plea proceeding to much of the facts underlying the enhancements advocated at his sentencing.  But the bottom line is that the defense should not put too much stock in a plea agreement that permits the government to repudiate its own guideline estimate, and in which the government does not stipulate to the facts underlying that estimate. 

Richard Willstatter adds:

In light of MacPherson and Habbas, defense lawyers should be sure to ask for a written agreement in which the US Attorney’s Office agrees to advocate for the guideline range estimated in the plea agreement absent new facts, or, if that is not forthcoming, to make an oral statement on the record before the plea is entered to the effect that the government will not later advocate a higher sentencing range unless the government learns of a factual basis for such a higher range after the plea is entered.  If there is reluctance to agree to either of these, the client should be counseled that the government may do an about-face at sentencing, against which the client may have little recourse.

MacPherson underlines the importance of figuring out whether a plea agreement (as opposed to a plea without an agreement) actually gets you anything.  If the standard agreement does not bind the government and if there is a possibility that a higher range could be found, as was readily apparent in MacPherson’s case, reliance on the agreement is risky and dependent on the integrity of the assigned prosecutor (or her supervisor).   It is in those cases where the government might advocate for a higher range – but agrees to forego such arguments in exchange for a plea – that a binding plea agreement is most useful.


Lawyers: Todd Merer (not lawyer in district court) (defendant); AUSAs Susan Corkery, Licha Nyiendo

The NYSACDL is holding its Fall Federal Criminal Practice Seminar tomorrow at the Southern District courthouse from 2:00 to 5:00 p.m.  With presentations on proffers, restitution and forfeiture, and innovative approaches to sentencing (not to mention 3 CLE credits), it promises to be an afternoon well-spent.  More information is available here.  To tie in with that program, I wrote an article for the NYSACDL's newsletter on the pitfalls of proffers.  You can access that here. 

The conviction last Friday of handbag magnate Frederic Bourke for bribing the government of Azerbaijan is probably enough on its own to strike fear into those doing business in the developing world – especially in countries with a high score on the Corruption Perceptions Index.  But an evidentiary ruling in this high profile FCPA case may generate some serious cost-benefit analysis in boardrooms.  In United States v. Kozeny, 2009 WL 1514369 (S.D.N.Y. May 29, 2009), SDNY Judge Scheindlin ruled that the government may introduce background evidence of corruption as part of its proof that the defendant consciously avoided knowing about the bribes at issue.  (Another interesting decision in the case is discussed here.)  The May 29 decision includes a detailed primer on the conscious avoidance doctrine, and is notable as the first reported application of the doctrine in an FCPA prosecution.  It is also notable as an illustration of how a defendant can score an “own goal” by doing innocence or cooperation proffers in a triable case.  


The defendant, Frederic Bourke, was charged with violations of the FCPA for allegedly making payments to Azeri officials to encourage the privatization of the State Oil Company of the Azerbaijan Republic and to permit him and others to participate in that privatization.  He moved in limine to preclude the government from introducing at trial evidence of corruption in Azerbaijan to prove his knowledge of the bribes at issue.  Under the FCPA, knowledge of a circumstance may be established by evidence that “a person is aware of a high probability of the existence of such circumstance, unless the person actually believes that such circumstance does not exist.”  The evidence the government sought to introduce included “that Azerbaijan was known to be a corrupt nation, that the post-Communist privatization processes in other countries have been tainted by corrupt practices, that SOCAR was a strategic asset of Azerbaijan, and that Kozeny [Bourke’s co-defendant] was notorious as the ‘Pirate of Prague.’”  

Conscious Avoidance and the FCPA

Agreeing with the defendant that the government may not introduce background evidence of corruption to show that he “should have known” that Azeri officials were being bribed, Judge Scheindlin held that here, however, the purpose of the evidence was to show that “a person of Bourke’s means, who was considering making a large investment in a venture in Azerbaijan, would have at least been aware of the high probability that bribes were being paid.”  As such, it was relevant and admissible.  

Defendant and Attorney Proffers

Critically, the court found no prejudice because the government would be able to establish the necessary factual basis for a conscious avoidance instruction.  In fact, the government’s factual basis rested on the defendant’s own admissions in proffers to the government.  Upon learning that he was the subject of a government investigation, Mr. Bourke appeared for a proffer session with the prosecutor, and admitted that he had been “warned by his counsel that Azerbaijan was the ‘Wild West’ and that doing business in Azerbaijan was like the movie ‘Chinatown,’ where there are ‘no rules.’”  For good measure, one of his lawyers had turned over a recording of a conversation between Bourke, another investor and their attorneys in which Bourke revealed his knowledge of the dark side of business in Central Asia, e.g. “Do you think business is done at arm’s length in this part of the world?” (There are some intriguing waiver issues here, but the decision doesn’t explore them, only to note that everyone appeared to agree that the defendant had waiver his attorney-client privilege in his proffer.)

Imputing Others’ Knowledge to Defendant

Finally, the decision addresses another unusual and important issue: the circumstances under which the government may introduce and impute to the defendant evidence of a third party’s knowledge.  Here, this knowledge could be imputed to Bourke based on his travels in a private jet with Viktor Kozeny (the alleged mastermind behind the SOCAR investment)and Bourke’s friendliness with Kozeny's “inner circle.”  “Based on these [close business] relationships the jury has a fair basis to infer that the knowledge of these individuals can be imputed to Bourke.”


Much ink will be spilled on the conscious avoidance doctrine in an FCPA prosecution.  But a lesson we can all (re)learn from this case is that proffers to the government in criminal cases can have a huge downside and should not be entered into lightly.  There may have been compelling strategic reasons for Bourke's proffer in this case, but it always astonishes me how often and easily lawyers will escort the clients in for meetings with the prosecutors, instead of proceeding by attorney proffer or asserting the right to remain silent.  As a wily criminal defense lawyer once said to me, “nothing is often the perfect thing to say and an excellent thing to do.”  

Lawyers: Harold Haddon and  Saskia  Jordan, Haddon Morgan Mueller Jordan Mackey & Foreman P.C., John Cline and K.C. Maxwell, Jones Day LLP, Dan Webb, James Reich, Jr. and Christopher Paolella, Winston & Strawn LLP (defendant); AUSA Harry Chernoff


Acknowledging that “[a] consensus has developed that the federal cocaine sentencing laws should be reassessed,” the Department of Justice has announced plans to establish a working group to formulate a complete elimination of the crack-cocaine disparity in the sentencing laws.  This disparity is “difficult to justify based on the facts and science, including evidence that crack is not an inherently more addictive substance than powder cocaine,” Lanny A. Breuer, the new Chief of DOJ’s Criminal Division, explains in his prepared remarks to Congress.  He elaborates:

[W]e think that the best way to address drug-related violence is to ensure the most severe sentences are meted out to those who commit violent offenses.  However, increased penalties for this conduct should generally be imposed on a case-by-case basis, not on a class of offenders, the majority of whom do not use any violence or possess a weapon.

For the time being, nothing will change: “Until a comprehensive solution – one that embodies new quantity thresholds and perhaps new sentencing enhancements – can be developed and enacted as legislation by Congress and as amended guidelines by the Sentencing Commission, federal prosecutors will adhere to existing law.”  Recognizing, however, “federal courts have the authority to sentence outside the guidelines in crack cases or even to create their own quantity ratio,” federal prosecutors “will inform courts that they should act within their discretion to fashion a sentence that is consistent with the objectives of 18 U.S.C. § 3553(a) and . . .  will bring the relevant case-specific facts to the courts’ attention.”

So, in the meantime, what can practitioners do?  Well, for one thing, try to get plea agreements based on a 1:1 ratio.  DOJ’s new policy involves achieving crack-cocaine parity with regard to both mandatory minimums and the guidelines.  While the Department has not yet formally implemented any changes, its plans to try to eliminate the disparity support a less hard-line approach in plea-bargaining crack cases, especially cases involving non-violent offenders.  And of course, DOJ’s commitment to an elimination of any disparity, coupled with its acknowledgement of a sentencing court’s powers to develop whatever ratio it chooses, are additional ammunition in the arsenal of reasons why a sentencing court should not wait for Congress to act, but should utilize a 1:1 crack-cocaine ratio right now.

In an interesting decision from Judge Larimer in the Western District of New York, United States v. Nix, 07 CR 60221 (DGL), 2008 WL 4757301 (W.D.N.Y. October 28, 2008), he rejected the government's effort to limit his power to depart below both mandatory minimum sentences called for by the defendant's two counts of conviction.  The government had purported to move for a substantial assistance departure under Count One, and not Count Two (which charged a violation of 924(c) and carried a mandatory consecutive five year sentence). 

Citing the defendant's "extensive and extraordinary" cooperation, the court sentenced Nix to consecutive sentences of 24 months on each count, for a total of 48 months. 

Interpreting the government's departure motion as applying to both counts, the court found the plea agreement to be ambiguous so that Nix could not anticipate that his cooperation would only be rewarded with a downward departure motion as to one count only.  (The language of the plea agreement is not quoted in the decision.)  He points out: "The Plea Agreement is, of course, a contractual obligation and the Government is obligated to comply with the terms of that contract. Since the Government is the principal author of the document and has virtually complete control over its content, any confusion or ambiguity must be held against it."

Lawyers: Donald Thompson and Lawrence Kasperek for the defendant; AUSA Robert Marangola

The government’s obligation to disclose Brady material encompasses not only oral, unrecorded statements of a cooperating witness that are favorable to the accused, as held in Rodriguez, previously discussed here, but also exculpatory statements communicated to the government by the witness’s lawyer, the Second Circuit held in United States v. Triumph Capital Group, Inc., 2008 WL 4349318 (2d Cir. September 25, 2008).

Brady Material Includes Notes of Witness’s Attorney’s Proffer

In Triumph, the defendant had been convicted after trial of racketeering, bribery, fraud and obstruction of justice, arising out a scheme to bribe the Deputy Treasurer of the State of Connecticut.  The bribee was the key cooperating witness at trial, but after the trial, he cooperated with the defense, and, no doubt to the dismay of the prosecutors, provided copies of notes he had made that had been communicated to the government in an attorney proffer as part of his initial plea negotiations.  These notes differed from his trial testimony in key respects.  While the government never had custody of the witness’s actual notes, it did have in its possession the notes an agent took of the attorney proffer.  It withheld these from the defense, however, until the defense made its motion for a new trial arguing suppression of exculpatory evidence.  The district court denied the motion, finding that the notes were not materially different from the witness’s trial testimony.

The Second Circuit disagreed.  In a decision authored by EDNY Judge Gleeson sitting by designation, the Court carefully analyzed the differences between the notes and the trial testimony, demonstrating that the notes “provided scant if any support for the inference that [the defendant] possessed the requisite intent to bribe or defraud.”  Finding the withholding of the agent’s notes of the attorney proffer “inexplicable,” the Court concluded that “the government deprived [the defendant] of exculpatory evidence going to the core of its bribery case against him,” as well as impeachment material that had a “real enough possibility to undermine confidence in the verdict.”  Accordingly, the Court reversed the racketeering, bribery and fraud convictions.

Destruction of Documents Likely to Be Subpoenaed by Grand Jury is Obstruction

Also of note in this decision is the Court’s analysis of the evidence supporting the obstruction of justice conviction.  The defendant had deleted certain documents from his laptop when he became aware that a grand jury investigation had commenced, albeit documents that were, at the time, not the subject of an outstanding subpoena.  On appeal, the defendant argued that there was insufficient evidence to prove that he knew the documents would later be, or were likely later to be, requested by the grand jury, analogizing himself to the defendant in United States v. Aguilar, 515 U.S. 593 (1995), whose obstruction conviction was reversed because he could not be expected to have known that the false statements he made to an F.B.I. agent would be communicated to a grand jury. 

This time, the Second Circuit disagreed with the defendant.  There was a “crucial distinction” between Aguilar and this case: statements made to investigating agents may not necessarily be communicated to a grand jury, but grand jury subpoenas for documents are necessarily broad and sweeping.  Here, the Court concluded that “the inference that the grand jury would issue a subpoena for [the deleted documents] was quite strong, perhaps inescapable.”  Moreover, there was evidence of the defendant’s “awareness of the comprehensive nature” of subpoenas typically issued in grand jury investigations, that his company’s lawyers anticipated future subpoenas, and that the defendant had received advice from a former prosecutor that the grand jury would be likely to inspect the data on his laptop.  The obstruction conviction was therefore affirmed, although the Court remanded for resentencing, in the event the convictions on the reversed counts had influenced the length of time imposed on that count of conviction.

“Queen for a day” (proffer) agreements – bare your soul to prosecutors in exchange for some limited protections – usually benefit the government more than the defendant.  For one thing, in these agreements, the defendant gives up any additional protections he may be entitled to under Fed.R.Evid. 410, which precludes admission at trial of ”statements made in the course of plea discussions with a [prosecutor].”
Not that defendants get much choice about whether to sign the agreement.  Most prosecutors will not agree to listen to a proffer without one.  But what if a proffer does in fact proceed without any proffer agreement: does the defendant automatically get the benefit of Rule 410?  Or should the defense lawyer do something more to invoke the rule’s protections?  That is the interesting question presented but not answered in United States v. Galestro, 06-CR-285 (ARR), 2008 WL 2783360 (E.D.N.Y. July 15, 2008), where the defendant, in his lawyer’s presence, spoke without a proffer agreement to prosecutors several days prior to the unveiling of an indictment charging him with death-eligible murder. 

The Scope of Fed.R.Evid 410

The Second Circuit has held that statements made by a defendant to prosecutors are not “plea discussions” under Rule 410, unless the defendant, “in some way, express[es] the hope that a concession to reduce the punishment will come to pass.”  United States v. Levy, 578 F.2d 896, 901 (2d Cir.1978).  The Levy Court expressly left open the question of whether statements made in a less formal “de facto process of plea bargaining” can be “plea discussions” for purposes of Rule 410 protections. 

In Galestro, the defense proposed that Levy’s void be filled with a two-tier analysis that focuses on whether at the time of the discussion, the defendant showed a subjective expectation to negotiate a plea, and whether that expectation was reasonable under the circumstances.  The government, argued alternatively, that there are no “plea discussions” if the meeting was at the defendant’s behest, he “sought to avoid indictment altogether rather than to plead,” and no plea deal was offered or ever made. 

Invoking Rule 410 Protection

In the end, the Galestro court did not have to decide the issue, because here, it was undisputed that Galestro’s attorney announced at the beginning of the proffer meeting that he considered it to be “in furtherance of settlement negotiations, pursuant to Fed.R.Evid. 410 and Fed.R.Crim.P. 11,” and the prosecutors present said nothing in response.  As the court concluded: “permitting the government to frustrate a defendant’s reasonable, explicit understanding of the nature of a discussion by simply remaining silent would not only be inconsistent with the rationale in Levy, but would undermine the very rationale of Rule 410 – to ‘promote plea negotiations by permitting defendants to talk with prosecutors without sacrificing their ability to defend themselves if no disposition agreement is reached’” (citation omitted).


Here, Galestro's lawyer's prescient statement at the beginning of the proffer meeting saved the day.  Without it, the government might have prevailed on the argument that this wasn’t a plea negotiation at all - it was a (misleading) innocence proffer, in which the defendant wasn't seeking to reduce his punishment, he was seeking to eliminate it.  Which begs another interesting question: whether innocence proffers are exempt from Rule 410 protection, since their goal is not a plea bargain but a dismissal. 

Innocence proffers, however, as the Second Circuit has pointed out elsewhere, are often preludes to plea negotiations.  In other words, they are part of the de facto process of plea bargaining, which runs the gamut of mindsets from denial to acceptance.  It makes no sense, and surely undermines the rationale of Rule 410, to carve out of the plea bargaining process (also known as the coming-to-terms process)  any discussions where the defendant professes innocence.  Moreover, the government, with its oft-described “awesome advantages in bargaining power,” knows full well how to obtain a waiver of Rule 410 protections prior to an innocence proffer, and it should not be rewarded for its failure to do so.

See Archives for all posts since September 2007.