Opinion
No. 19-cr-00669
08-23-2020
UNITED STATES, v. GREGG SMITH, MICHAEL NOWAK, JEFFREY RUFFO, and CHRISTOPHER JORDAN.
ORDER
In August 2019, Gregg Smith was indicted on criminal charges alleging, among other things, commodity-futures spoofing. R. 1. Before that, Smith also was being investigated for potential parallel civil violations by the Commodity Futures Trading Commission. As part of the civil investigation, the CFTC retained Dr. Henrik Bessembinder, who is a financial-markets expert. Over the course of 2017 and 2018, Bessembinder analyzed Smith's trading activity. Bessembinder ultimately concluded that Smith's trading activity—at least the activity that the expert reviewed—did not appear to be consistent with spoofing. But Bessembinder apparently did not generate any written report memorializing that conclusion. Bessembinder's consulting contract with the CFTC ended in late 2018.
Citations to the record are noted as "R." followed by the docket number and the page or paragraph number.
At some point after this analysis was performed, a member of the CFTC Enforcement Division disclosed Bessembinder's conclusion to the Justice Department in case the prosecution had any disclosure obligations to Smith. R. 103, Cogan Decl. ¶ 2; R. 116-1, Chiang Decl. ¶¶ 4-5. But the prosecution decided not to pry any deeper into the details of Bessembinder's analysis and, as a result, never requested that the CFTC turn over any of the documents relating to the analysis. Cogan Decl. ¶ 3; Chiang Decl. ¶ 5. Instead, the prosecution simply disclosed to the defense that the CFTC's consulting expert, Bessembinder, had reviewed some of Smith's trading activity and found it inconsistent with spoofing. R. 103-1, DOJ Letter at 3. Beyond that bare conclusion, the prosecution explained that it knew nothing else:
The government is not aware of the specific trading activity that served as the basis for Dr. Bessembinder's opinion, nor has the government received any documents or other materials generated as part of Dr. Bessembinder's analysis. Nonetheless, in an abundance of caution, the government is disclosing the extent of its knowledge regarding Dr. Bessembinder's work for the CFTC relating to Mr. Smith's trading activity."Id.
Not surprisingly, Smith now seeks to access Bessembinder's underlying analysis and communications. So Smith went straight to the source, serving the CFTC—which is formally a non-party for purposes of this criminal case—with a subpoena requesting all documents "that refer or relate to Dr. Hendrik Bessembinder's review of certain trading activity by Smith and/or his conclusion whether Smith's trading activity was consistent with spoofing," "that Dr. Bessembinder relied on in preparing or forming his conclusions in the Bessembinder Report," and "that refer to, relate to, or concern the Bessembinder Report, including but not limited to Communications between the CFTC and DOJ." R. 87, Mot. Quash, Exh. A, CFTC Subpoena. The CFTC has filed a motion to quash the subpoena. R. 87. The CFTC also submitted the 38 contested documents to the Court for an ex parte, in camera review. R. 119, 120.
The governing rule is Rule 17 of the Federal Rules of Criminal Procedure. Under Rule 17(c), a district court may quash a subpoena "if compliance would be unreasonable or oppressive." Fed. R. Crim. P. 17(c)(2). The parties naturally dispute what "unreasonable or oppressive" means. According to the CFTC, a subpoena is "unreasonable or oppressive" if it goes beyond seeking relevant, admissible, and specific evidence, as articulated by the Supreme Court in United States v. Nixon, 418 U.S. 683, 700 (1974). Mot. Quash at 5. Thus, argues the CFTC, the subpoena here must be quashed because the evidence it seeks: (1) is irrelevant; (2) constitutes inadmissible hearsay; and (3) is protected by the deliberative-process privilege, attorney-client privilege, and the work-product doctrine. Id. (Specificity does not appear to be in dispute.) In response, Smith maintains that Nixon never explicitly set out a blanket "admissibility" requirement for third-party subpoenas under Rule 17, and in any event, the evidence that he seeks is both admissible and not privileged. R. 102, Def. Resp. Br. at 4.
For the reasons explained below, the CFTC's motion to quash is granted, because the deliberative-process privilege shields the records from disclosure. It is true that, as Smith argues, trial admissibility is not a blanket requirement for third-party subpoenas under Rule 17(c). But the privilege does apply.
I. Rule 17
As an overarching matter, Rule 17 is not meant to "provide an additional means of discovery." Bowman Dairy Co. v. United States, 341 U.S. 214, 220 (1951). Building on that general principle, the Supreme Court has interpreted Rule 17's "unreasonable or oppressive" standard to mean that a subpoena must request evidence that is (1) relevant, (2) admissible, and (3) specific. Nixon, 418 U.S. at 700. Many courts have since applied the Nixon standard to both subpoenas directed at the government and at non-parties, though most without analyzing whether a lesser standard should apply to non-party subpoenas. See United States v. Henry, 482 F.3d 27, 30 (1st Cir. 2007) (noting that under Rule 17, "the defense may use subpoenas before trial to secure admissible evidence but not as a general discovery device" and affirming district court decision to quash third-party subpoena); United States v. Bergstein, 788 F. App'x 742, 746 (2d Cir. 2019) (citing Nixon standard for third-party subpoenas and noting that the Second Circuit has "applied the Nixon standard to Rule 17(c) subpoenas requested by a defendant") (summary order); United States v. Tokash, 282 F.3d 962, 971 (7th Cir. 2002) (citing Nixon standard without analysis as governing third-party subpoena); United States v. Stevenson, 727 F.3d 826, 831 (8th Cir. 2013) (noting that the Eighth Circuit has applied the Nixon standard to third-party subpoenas); United States v. Sleugh, 896 F.3d 1007, 1012 (9th Cir. 2018) (citing Nixon rule in context of Rule 17 subpoenas to phone companies).
There are good reasons to follow the example of the majority of Circuits and interpret the text of Rule 17 to require that third-party subpoenas only seek relevant, admissible, and specific evidence. For one, the text of Rule 17 assumes that there will be a witness at a trial or hearing. Section 17(a) says that the subpoena's content must "command the witness to attend and testify at the time and place the subpoena specifies." Fed. R. Crim. P. 17(a). Section 17(c)(1) similarly refers to "the witness" and early production of documents "before trial or before they are offered in evidence." Fed. R. Crim. P. 17(c)(1). So the rule itself assumes that the subpoena recipient is, at the very least, a potential witness at the trial or hearing. That in turn does refute the scope of a Rule 17 subpoena as a broad discovery or investigation tool. See Bowman Dairy, 341 U.S. at 220.
What's more, the Fourth Circuit in Rand articulated three reasons for why Nixon's requirements should apply to third-party subpoenas. First, Rand reiterated the Supreme Court's general declaration that the subpoena duces tecum "was not intended to provide a means of discovery for criminal cases." United States v. Rand, 835 F.3d 451, 463 (4th Cir. 2016) (cleaned up). This statement, noted the Fourth Circuit, did not limit itself to only discovery from the government. Id. Second, Rand explained that, to the extent that there were any concerns about making it more difficult for criminal defendants to defend themselves, "[t]he right to defend oneself does not extend to using the power of the Court to compel third parties to provide information that may not even be admissible at trial or at a hearing or that is merely 'investigatory.'" Id. (citing United States v. Al-Amin, 2013 WL 3865079, at *7 n.3 (E.D. Tenn. July 25, 2013)). (For what it is worth, however, Al-Amin itself does not cite anything for that proposition.) And finally, the Fourth Circuit explained that the Nixon standard "is not at odds with" the explicit "unreasonable or oppressive" standard laid out in the text of Rule 17. Id.
This Opinion uses (cleaned up) to indicate that internal quotation marks, alterations, and citations have been omitted from quotations. See Jack Metzler, Cleaning Up Quotations, 18 Journal of Appellate Practice and Process 143 (2017).
Having said all that, it is not immediately apparent that the admissibility requirement (which encompasses relevance) should apply to non-party subpoenas. For one, Nixon itself explicitly declined to decide whether the articulated rule (relevance, admissibility, and specificity) was meant to cover subpoenas issued by defendants to non-parties. Specifically, the Supreme Court expressed in a footnote that it "need not decide whether a lower standard exists" in cases where "the subpoena duces tecum is issued to third parties rather than to government prosecutors." Nixon, 418 U.S. at 700 n.12. (Nixon itself involved a third-party subpoena, but the Supreme Court explained in that same footnote that the admissibility and specificity standards had been met either way. Id.) In other words, even if the Supreme Court has not explicitly stated that the Nixon requirement applies only to government subpoenas, it also has not yet stated the opposite, that the Nixon rule absolutely applies to third-party subpoenas as well.
Indeed, this was picked up on by at least one other district court in Tucker, which also dealt with a defense subpoena to a third-party and ultimately chose to impose a "materiality" standard instead of Nixon's admissibility requirement. United States v. Tucker, 249 F.R.D. 58, 66 (S.D.N.Y. 2008). According to the court in Tucker, the Nixon standard certainly made sense in the context of a defendant requesting evidence from the government, but "the standard is inappropriate where production is requested by (A) a criminal defendant; (B) on the eve of trial; (C) from a non-party; (D) where the defendant has an articulable suspicion that the documents may be material to his defense." Tucker, 249 F.R.D. at 66. In that context, Tucker proposed that a defendant "need only show that the request is (1) reasonable, construed as 'material to the defense,' and (2) not unduly oppressive for the producing party to respond." Id. Tucker's rejection of the Nixon standard for non-party subpoenas appears to be animated by fairness concerns—the court remarked that federal criminal defendants have far fewer opportunities to gather evidence compared to the government, even though "defendants have the right to put before a jury evidence that might influence the determination of guilt. To effect this right, a defendant must have the ability to obtain that evidence." Id. at 65.
Caselaw aside (and without specifically endorsing the policy considerations of Tucker), there is also a rule-based reason to distinguish subpoenas issued to the government versus those issued to non-parties. For one, Criminal Rule 16 explicitly governs and limits the scope of discovery as to the government, so it makes sense to apply the Nixon rule (as opposed to a broader catch-all "unreasonable or oppressive" standard) when subpoenas are issued to the government under Rule 17. So a subpoena would be "unreasonable" if it sought from the government more than what other criminal-discovery rules allow. That reading of "unreasonable" does not apply to subpoenas issued to third-parties. See United States v. Tomison, 969 F. Supp. 587, 593 (E.D. Cal. 1997) ("The notion that because Rule 16 provides for discovery, Rule 17(c) has no role in the discovery of documents can, of course, only apply to documents in the government's hands; accordingly, Rule 17(c) may well be a proper device for discovering documents in the hands of third parties."). Indeed, for purposes of this motion, the key words in Rule 17 are "compliance is unreasonable" (because no one says that "oppressive" is at issue here), which as a textual matter is broader than "compliance would require the production of inadmissible evidence."
So the Court rejects a bright-line admissibility requirement for the subpoena issued to the CFTC in this case. Indeed, even Nixon did not require absolute certainty on admissibility or absolute precision in specificity. The opinion characterized the admissibility elements as "a sufficient preliminary showing that each of the subpoenaed tapes contains evidence admissible with respect to the offenses charged in the indictment." Nixon, 418 U.S. at 700 (emphasis added). This makes sense; it is not as if a finding of a subpoena's propriety would automatically result in the allowance of the evidence at trial. On specificity, here, too, absolute precision is not required, as Nixon recognized in saying: "Of course, the contents of the subpoenaed tapes could not at that stage be described fully by the Special Prosecutor," and again only requiring a "sufficient likelihood" that the tapes contained relevant conversations. Id. Consider this example, with the shoe on the other foot: the prosecution routinely invokes Rule 17 to subpoena all non-privileged recordings of jail-phone conversations of a detained defendant. Yet a strict admissibility requirement would render those subpoenas invalid, because it generally cannot be known that the recordings will be admissible before receiving and reviewing them.
At the same time, even if the admissibility requirement is rejected, Bowman Dairy did still say, without making a distinction between the government and non-parties, that "Rule 17(c) was not intended to provide an additional means of discovery." Bowman Dairy, 341 U.S. at 220. Rule 17's "chief innovation was to expedite the trial by providing a time and place before trial for the inspection of the subpoenaed materials." Id. So application of the term "unreasonable" does still consider the need to avoid fishing expeditions, which Nixon emphasized too. See Nixon, 418 U.S. at 700. This limitation protects both the government and the defendant. The key is weighing the likelihood that the subpoena will uncover relevant evidence and the potential probative force of that evidence, against the burdens on complying. Tucker proposed materiality, 249 F.R.D. at 66, which is a useful and well-known way to assess probative force. But on top of materiality, reasonable specificity of the subpoena is still required to avoid fishing expeditions.
In short, the Court is not persuaded that a Rule 17 subpoena to a non-party must, as a blanket rule, seek already-known admissible evidence. Because of that, there is no need to reach the hearsay arguments raised by the CFTC, because even if the evidence constitutes hearsay inadmissible at trial, that alone is not necessarily a reason to quash the subpoena. Instead, as described in more detail below, there is separate reason the subpoena must be quashed: the deliberative-process privilege.
For what it is worth, the CFTC is probably right on the hearsay question. The public-records hearsay exception, Fed. R. Evid. 803(8)(A), does not apply in this case because there is no suggestion that any of Bessembinder's preliminary work product constitutes a report or finding of the CFTC. Separately, the CFTC also argues that the evidence sought is both irrelevant under Federal Rule of Evidence 401 and would be confusing to the jury under Federal Rule of Evidence 403. Mot. Quash at 6. For the same reason that the Court is not basing this decision on the CFTC's hearsay arguments, there is also no need to reach the relevance and Rule 403 arguments for purposes of deciding this motion.
II. Deliberative-Process Privilege
Although there is much dispute over whether Nixon's admissibility standard should apply to defense subpoenas directed at non-parties, the parties do not seriously dispute that the various evidentiary privileges must be considered part of the Rule 17 analysis. Specifically, there is no hint that the "unreasonable and oppressive" language of Rule 17 rejects privilege as a basis for quashing a subpoena. So, if a subpoena seeks privileged information, then the subpoena can be quashed as unreasonable or oppressive. Here, the CFTC asserts three separate privilege arguments: deliberative-process privilege, attorney-client privilege, and work-product protection. Because this motion can be resolved on the first basis—deliberative-process privilege—there is no need to reach the other two.
"The deliberative process privilege protects communications that are part of the decision-making process of a governmental agency." United States v. Farley, 11 F.3d 1385, 1389 (7th Cir. 1993) (citing N.L.R.B. v. Sears, Roebuck & Co., 421 U.S. 132, 150-52 (1975)). The idea behind the deliberative-process privilege is that "frank discussion of legal and policy matters is essential to the decisionmaking process of a governmental agency," which means that "communications made prior to and as a part of an agency determination are protected from disclosure." Farley, 11 F.3d at 1389. On the flip side, "[c]ommunications made subsequent to an agency decision are, however, not similarly protected." Id. In other words, to qualify for protection, a communication must be both predecisional and deliberative." Enviro Tech Int'l, Inc. v. U.S. E.P.A., 371 F.3d 370, 375 (7th Cir. 2004). Here, the relevant agency decision for purposes of the deliberative-process privilege analysis is the CFTC's September 2019 decision to bring a civil action against Smith. See R. 87, McDonald Decl. ¶ 11.
Smith raises two arguments for why the deliberative-process privilege should not apply to the 38 contested documents. First, Smith asserts that the privilege only protects intra-agency communications, whereas here, Bessembinder was an outside consultant. Def. Resp. Br. at 13. Second, Smith argues that the "deliberative" element has not been met because the CFTC only hired Bessembinder for the narrow purpose of assisting staff in analyzing trade data. Id. at 14.
Turning first to the outside-consultant argument, it is true that Bessembinder was not an employee of the CFTC. But courts have extended the deliberative process privilege to apply to reports and communications by outside consultants retained by agencies. See Ryan v. Dep't of Justice, 617 F.2d 781, 790 (D.C. Cir. 1980) ("When an agency record is submitted by outside consultants as part of the deliberative process, and it was solicited by the agency, we find it entirely reasonable to deem the resulting document to be an 'intra-agency' memorandum"); Tigue v. U.S. Dep't of Justice, 312 F.3d 70, 77 (2d Cir. 2002) (noting that "agencies may require assistance from outside consultants in formulating policy" and that "nothing turns on the point that ... reports were prepared by outside consultants ... rather than agency staff") (cleaned up); Pub. Employees for Envtl. Responsibility v. Bloch, 532 F. Supp. 2d 19, 21-22 (D.D.C. 2008). Here, too, the fact that the communications came from an outside consultant is not, standing alone, dispositive.
That brings us to Smith's second point. It is also true that Bessembinder was retained for the ostensibly narrow purpose of analyzing raw trade data. And the emails between Bessembinder and the CFTC staff members do not literally contain discussions about whether or not to bring an enforcement action against Smith. But the record provides no reason to doubt that the CFTC considered and treated the communications with Bessembinder about the trade analyses as anything but a genuine and confidential part of the broader deliberative process meant to inform the CFTC's decision to either pursue claims or not against Smith. See McDonald Decl. ¶¶ 7-10, 12. In a financial-markets case like this, the ultimate decision to bring or not bring an enforcement action against an individual will almost always necessitate some form of data analysis, which will often require the help of an outside consultant. So there is no line between the CFTC's own discussions about whether or not to bring an action and the CFTC's preliminary communications with an outside consultant about the data analyses underlying the broader decision. Indeed, the record (including the Court's in camera review) suggests that even though Bessembinder was technically an outside consultant, the expectation of all the parties was that the documents and communications would be for "internal use only" and should be kept only within the ambit of the CFTC and Dr. Bessembinder himself. McDonald Decl. ¶ 9. Similarly, the privilege log provided by the CFTC labels most of the documents as "confidential." R. 116-1. Thus, Bessembinder was bound by confidentiality agreement as to the work that he performed for the CFTC. To be clear, "materials are not to be withheld on the basis of deliberative process privilege simply because the agency deems them confidential and would prefer not to disclose them." Stinson v. City of New York, 304 F.R.D. 432, 435 (S.D.N.Y. 2015) (cleaned up). But the confidentiality expectation here matters because it further establishes that Bessembinder was fully enveloped in the deliberative process leading up to the CFTC's decision to bring an action against Smith.
Of course, if it turns out that the CFTC's retainer agreement with Bessembinder did not in fact contain any sort of confidentiality expectation (or at least some sort of professional norm of client confidentiality), then the argument for the privilege is much weaker; the lack of a confidentiality agreement could suggest that the CFTC did not actually expect that the analyses and emails would remain "internal." This seems unlikely, though, which is why the Court is operating under an inference (based on the available record) that some type of confidentiality agreement was in place. --------
In short, these documents fall squarely within the deliberative-process privilege. It is true that Bessembinder ultimately came out with a conclusion ostensibly contrary to the CFTC's later decision to pursue a civil enforcement action against Smith. But that illustrates precisely why the deliberative-process privilege is so important in this context—to the extent that the goal of the privilege is to promote full and frank deliberations within government agencies, then society as a whole wants the CFTC to seek out all sorts of expert opinions from outside consultants and to consider as many facets of a decision as possible in order to arrive at the most informed outcome. In contrast, forcing an agency to disclose this type of "contrary" analysis would potentially discourage the agency from even seeking out "contrary" opinions in the first place the next time a decision needed to be made.
If it is any comfort, based on the in camera review, the Court is skeptical that these 38 documents would have been especially helpful to the defense anyway. These emails truly reflect a preliminary back-and-forth, as opposed to a full-fledged report or conclusion, which is precisely why the privilege applies.
III. Conclusion
Reasonable minds can disagree whether it would have been more fair for the prosecution to simply request from the CFTC (and potentially turn over to Smith) the documents underlying Bessembinder's conclusion that Smith's trading activity did not appear to be consistent with spoofing. But a Rule 17 subpoena is simply not enforceable here against the deliberative-process privilege. For that reason, the production of those documents would be "unreasonable" under Rule 17, and as a result, the CFTC's motion to quash must be granted. For the same reasons, Smith's motion to expand the in-camera submission (seeking internal CFTC notes memorializing Dr. Bessembinder's conclusion) is also denied, R. 125, because the deliberative-process privilege also squarely applies in that context.
ENTERED:
s/Edmond E. Chang
Honorable Edmond E. Chang
United States District Judge DATE: August 23, 2020