Opinion
CR. S-10-0061 LKK (GGH).
August 2, 2010
ORDER ON RECONSIDERATION
The United States (government) has moved the undersigned for reconsideration of that part of his discovery order dated July 9, 2010 (Docket 132) in which the government was tasked withBrady/Giglio identification of already disclosed, voluminous documents and other materials, identification/disclosure of Brady exculpatory information which had not yet been disclosed, and identification with an eye towards later disclosure of Giglio information which had not yet been disclosed. See order at 13. The undersigned finds reconsideration warranted based on the developed arguments of the United States. After close review of the government's request, especially the burdens arguments, and defendant Salyer's opposition, the court affirms its previous order with modifications to the logistics on implementation. The undersigned reviews each of the government's arguments in turn.
The government first argues that an order for it to identifyBrady/Giglio information compels a revelation of its core work product. The government mistakes statement of a position in a criminal litigation for compelled revelation of work product. The government's position is misguided, albeit commonly argued. For example, the indictment itself is the formalized statement of the government's legal analysis that Facts 1, 2 and 3 comprise the crime of X. Yet the government could not be logically heard to argue that the process of requiring an indictment, and its disclosure in open court, somehow constitutes an unwarranted disclosure of its core work product. Similarly, Fed.R. Ev. 404(b) requires the government, upon request, to provide pretrial the general nature of the "bad acts" evidence it intends to use concerning motive, intent, etc. Someone in the U.S. Attorney's office makes the legal analysis that a certain act would qualify as a "bad act" and is relevant to motive, for example, and then identifies/discloses the nature of the evidence. The identification and analysis of the evidence is surely an attorney's thought process, but no one would consider Rule 404(b) to tread on work product. Further examples abound, even on the defense side — a defendant is required to assess the evidence, come to a determination, and then give notice of an insanity defense if appropriate, Fed.R.Crim.P. 12.2. Surely, such a determination is at the pinnacle of attorney thought process — yet it is required to be identified and disclosed.
A very analogous example of the work product principle exists in civil litigation. Fed.R.Civ.P. 26(b)(5) is a self-executing directive for a party, through his counsel, to make a legal determination as to what information, otherwise disclosable, is privileged. This attorney analysis, i.e., thought process, must be memorialized and served on the other side. This procedure is identical in general principle to what is required of the prosecution in the Brady/Giglio context and would not be considered an invasion of work product.
Rather, work product is classically defined as those in-house, attorney or investigator created reports, memoranda and so forth which discuss a legal issue, a planned course of action in the investigation/litigation, and the like. See Fed.R.Crim.P. 16(a)(2). Thus, the prosecutor's memorandum to the U.S. Attorney as to why a prosecution should be brought is protected work product — the ultimate memorialization of the approval of prosecution, i.e., the indictment, is not, regardless of whether the legal position in the indictment precisely tracks that of the initial memorandum. The cases cited by the government, e.g.,Morris v. Ylst, 447 F.3d 735, 742 (9th Cir. 2006) (involving a prosecution created status report), discuss situations which fall into the former, protected category, and are inapposite to the identification/disclosure requirements of the order.
If the government's position were to prevail, nothing containing a legal determination or analysis could ever be filed. This would lead to an absurd result. Indeed, a court could not order the production of heretofore undisclosed Brady/Giglio material because the AUSA would first have to make a "legal determination" that some information in his or her file constituted Brady/Giglio, and disclosure would then constitute revelation of core work product. There would be a duty to disclose Brady/Giglio, but work product would prevent its disclosure in any meaningful fashion. Catch 22's are disfavored in litigation. Accordingly, when a court compels the government to identify and/or disclose Brady/Giglio material from documents acquired from third parties or entities, that identification, much like every other compelled identification or disclosure in the Rules, is not the revelation of protected work product.
To the extent that underlying third party documentation exists, the court does not at this time require the production of anyBrady memoranda created by the U.S. Attorney's office or its agencies for purposes of this litigation to determine a course of action with respect to identification/disclosure of document(s), assuming that any such documents exist. That would be work product.
Next the government argues that the Ninth Circuit has determined that a Brady violation has not been found for exculpatory or impeaching material that was disclosed, but not identified as favorable to the defense, citing Rhoades v. Henry, 596 F.3d 1170, 1181 (9th Cir. 2010). This is again a correct legal argument in the circumstances of that case, but inapposite to what the undersigned is doing. The case law is established, as a general rule, that if Brady/Giglio were disclosed to the defense, the fact that the government did not identify it as such will not cause the conviction to be reversed. However, that case law does not preclude the undersigned as a matter of case management (and fairness) in ordering identification to be done. For example, as a general rule, a conviction will not be reversed for the government failing to identify its witnesses pretrial. However, that fact does not preclude a court from ordering such identification and disclosure prior to trial as a matter of case management where appropriate. United States v. W R Grace, 526 F.3d 499, 509-510 (9th Cir. 2008) (en banc). The undersigned is doing nothing that the Grace case does not approve in general principle as a matter of case management in this particular case. See also Advisory Committee Notes, Rule 16, 1974 amendment: "The rule is intended to prescribe the minimum amount of discovery to which the parties are entitled. It is not intended to limit the judge's discretion to order broader discovery in appropriate cases." See also Fed. Crim. P. 2. In sum there is no authority which prohibits the court's exercise of discretion here.
Moreover, the government argument that it has not violatedBrady/Giglio by its activities to date in simply disclosing everything to Salyer could, or could not, ultimately be found to be true after trial given the circumstances of this case, but it is not conclusive to the question before the court. As explained by Strickler v. Greene, 527 U.S. at 281, 119 S.Ct. at 1948, there is a distinction between the ultimate finding of a Brady violation and the initial "broad duty of disclosure" that is at issue here at this stage of the case.
Finally, the government arrives at its strongest argument: that nothing in the rules of our adversarial system should require the government to prepare the case for the defense, and specifically, that a trial court should not require the government to identify pretrial, even as a matter of case management, materials which are exculpatory or impeaching. See United States v. Skilling, 554 F.3d 529, 576-77 (5th Cir. 2009, aff'd in part, rvs'd in part,Skilling v. United States, ___ U.S. ___, 2010 WL 2518587 (June 24, 2010)). "As a general rule, the government is under no duty to direct a defendant to exculpatory evidence within a larger mass of disclosed evidence." Id. at 576. But see United States v. Hsia, 24 F. Supp. 2d 14, 29-30 (D.D.C. 1998) rvs'd in part on other grounds, 176 F.3d 517 (D.C. Cir. 1999) ("the government cannot meet its Brady obligations by providing [the defendant] with access to 600,000 documents and then claiming that she should have been able to find the exculpatory information in the haystack").
Despite the fact that the government's arguments have much color for the run of the main case, the undersigned finds for the purposes of this case that Salyer has the better argument.
However, before commencing the analysis of the government's colorable argument, a bit more of the factual circumstances of the information disclosed thus far must be made. Over the course of approximately five years before indictment, the government collected, i.e., subpoenaed, acquired through warrant, or other means, a massive amount of documentary information. Not only does the electronic information consist of multiple gigabytes, pages numbering into the millions, the hard paper information fills more than two "pods" (storage containers), and there is further documentation stored in F.B.I. offices. When asked at hearing why the government would seek so much information, the response, pared to its essence, was "because we can." In most cases, after the acquisition dust settles, a vast amount of acquired documentary material is often useless, if even reviewed at all. It remains unclear precisely how much of the acquired information in this case has actually been reviewed by the government.
One gigabyte of text files will average over 677,000 text file pages and over 100,000 pages of e-mail files. How Many Pages in a Gigabyte, www.lexisnexis.com/applieddiscovery/lawlibrary/whitePapers/ADI_FS_PagesInAGigabyte.pdf.
Of course, the AUSA did not use those words, but his response indicated that grand jury subpoenas are purposefully, broadly worded, and that in a business case, the documents to be seized via search warrant are very voluminous as well.
The government represented on the record, that with the exception of internal documents created by government personnel, for the purposes of the investigation/litigation, every document acquired or seized by the government has been made available to the defense.
Thus, commencing the analysis here, the government's arguments center on the general assertions that it should not have to "crawl in to the mind of defense counsel" to prepare the defense in this case by identifying Brady/Giglio information, and in any event, that the burden of doing so in this case would impose an impossible requirement.
First, upon close inspection, the government's argument is not supported in the "big documents" case. As pointed out by Salyer in referencing the DOJ Brady/Giglio guidelines, and as well established by case law, the prosecution has the duty to affirmatively scour those records of the agencies considered the "government" for purposes of the criminal case in order to determine and acquire those materials which would be considered Brady exculpatory and Giglio impeaching. Kyles v. Whitley, 514 U.S. 419, 437, 115 S.Ct. 1555 (1995); United States v. Price, 566 F.3d 900, 908-909 (9th Cir. 2009). Indeed, the most recent Department of Justice guidelines, fashioned after some embarrassing non-disclosures in other cases, command an actual review of the materials acquired during investigation of a criminal case for the purpose of disclosing Brady/Giglio materials. See Memorandum for Department Prosecutors, dated January 4, 2010, Section B, e.g., items, 1 ("The prosecutor can personally review the file or documents or may choose to request production of . . . materials from the case agent"), 3 ("[g]enerally, all evidence and information gathered during the investigation should be reviewed, including anything obtained during searches or via subpoenas, etc."). Tellingly, nowhere in the memorandum does it even suggest that in lieu of affirmatively looking for Brady/Giglio the prosecutor may determine not to look at all and simply disclose the entire discovery file. The prosecutor's argument that his duty to affirmatively search forBrady/Giglio information is performed by not searching is oxymoronic in nature.
The government's "impossible" burden objections to the court's order (and DOJ Guidelines) are somewhat colorable, although as observed below, these arguments conflict with the inconsistent argument that with a modicum of diligence, the defense can ferret out all of the Brady/Giglio material itself. The objections also are hyperbole. And while the undersigned is sympathetic, he is not too sympathetic. During the course of any large case investigation, investigators and attorneys are going to have to review and organize the sometimes immense amount of evidence that is acquired. During the course of the years long investigation in this case, the government personnel seemed to be able to segregate that evidence which would be useful in the prosecution in terms of guilt, but apparently made no efforts to segregate that evidence which runs counter to the charges. Assuming for the moment that some Brady/Giglio evidence, as the court has defined it below, exists, the reviewing personnel apparently made no note of the evidence, or merely having noted it, "stuck it back" in the ever-increasing pile to be an inevitably hidden part of the mass disclosure. The obligations imposed by Brady et al. have been well established for years, and should be anticipated in every case during the investigation phase. If the government argues that it is now "impossible" to comply with the burden of reviewing evidence for identification purposes, the government more or less made its own bed in this matter by making it impossible.
At hearing, the government argued that it performs itsBrady/Giglio duties as set forth above, by not looking for the required information as the cases seem to direct, but by just disclosing everything. And, there is some general, but ultimately inconsequential, support for that proposition. In Strickler v. Greene, 527 U.S. 263, 119 S.Ct. 1936 (1999) the Supreme Court observed in a footnote on the acts of its case:
We certainly do not criticize the prosecution's use of an open file policy. We recognize that this process may increase the efficiency and the fairness of the criminal process. We merely note that, if a prosecutor asserts that he complies with Brady through an open file policy, defense counsel may reasonably rely on that file to contain all the material the State is constitutionally obligated to disclose under Brady.Id. at 283 (n. 23), 119 S.Ct. at 1949 (emphasis added).
While it could be argued inferentially that the Supreme Court has endorsed an open file Brady disclosure policy, at the very least the Court did not condemn it, the footnote which simply sets a cautionary note to the prosecution to not misrepresent an open file policy cannot stand for authority that one'sBrady/Giglio duties are satisfied in every case by an open file policy no matter what the circumstances. The issue of whether disclosure can be hidden by too large a file was not by any means an issue confronted by the Strickler court.
The government's argument at hearing on the burden of requiringBrady/Giglio identification in the voluminous documents case actually is the argument which conclusively proves Salyer's point in this case. The government argued at length that if it were required to presently, actually review the documentation it acquired, in order to identify Brady/Giglio materials, the burden of doing so would be impossible, even to the point of having to dismiss the case. If the government professes this inability to identify the required information after five years of pre-indictment investigation, its argument that the defense can "easily" identify the materials buried within the mass of documents within months of post-indictment activity is meritless. Obviously, under the government's reasoning, the defense burden is even more impossible. What the government is actually arguing, in effect and for practical purposes, is — that logistics in the "big documents" case render Brady/Giglio a dead letter no matter who has the burden of ascertaining the information. There is no authority to support this evisceration of constitutional rights just because the case has voluminous documentation.
The undersigned is not being naive in this analysis. To be sure, Salyer knew something was up when the records of his company and others were subpoenaed. There has been discussion in various pleadings that Salyer was in discussion with the government pre-indictment concerning possible charges. Thus, he had some ability to review records to determine what might rebut the potential charges. However, the actual charges are finally determined with the indictment, with an explanation of the government's theories, and the real search for exonerating or impeaching materials by the defense begins at that time. Moreover, the undersigned is aware of no authority conditioning the government's Brady/Giglio duties depending on pre-indictment involvement of the defendant. Again, the Supreme Court has placed the initial Brady/Giglio duty on the government, and the undersigned is not free to assign it to Salyer.
Next argued by the government is its protest that it cannot possibly look into defense counsel's mind in attempting to figure out what fanciful theory the defense may advance, or what stretched factual presentation may be presented. Therefore, the argument goes, the prosecution should not be tasked with anticipation of such with the ultimate conclusion that the prosecution should never be tasked with identifying what the prosecution deems Brady/Giglio information. However, Brady/Giglio requires no such clairvoyance. Although the Brady rule often is phrased in terms of evidence "favorable to the defense," the case law has made clear that such evidence is that which tends to exculpate guilt of the offense charged by the government. See Brady v. Maryland, 378 U.S. 83, 87, 83 S.Ct. 1194-1196-97 (1963) (exculpatory evidence is that evidence material to guilt or punishment). See also Cone v. Bell, ___ U.S. ___, 129 S.Ct. 1782-83 (2009). Thus, the prosecution knows, as any litigator would know, what evidence, on its face, significantly detracts from the factual elements which must be proven in a particular case. Indeed, one cannot prepare a case without thinking about how to counteract, i.e., explain away, evidence inconsistent with proving the elements of the charged crime. That determination is made from the prosecution's vantage point, not a speculative insight into defense counsel's theory of the defense. That is why it is the prosecution's initial, unquestionable, and affirmative duty to "learn of any favorable [defense] evidence known to the others acting on the government's behalf in the case." Kyles v. Whitley, supra. The Supreme Court most assuredly did not hold that because it is not possible to look into defense counsel's mind about possible defenses, there is no duty to look for exculpatory evidence. When the prosecution, in good faith, determines that a piece of evidence, on its face, significantly tends to controvert what it is attempting to prove, disclosure (and in this case, identification as well) is mandated. Similarly, for Giglio information, the prosecution knows, from its vantage point, what information is significantly inconsistent with the testimony it expects its potential witnesses to present or with their credibility generally. The Supreme Court has directed prosecutor's to err on the side of disclosure, but it has not held that exculpatory or impeaching evidence is to be speculatively gauged by what defense counsel may desire to argue.
The undersigned rejects the argument of the government that its counsel may be subject to sanctions if it inadvertently mis-identifies disclosed information as Brady/Giglio. The case cited by the government, United States v. Shaffer, 789 F.2d 682, 690 (9th Cir. 1986), involved a situation where the government lawyer had purposefully labeled Brady information as information not meeting that category. Such a situation is the polar opposite of what is being ordered here — an identification that certain material may be Brady/Giglio. The Supreme Court in Kyle did not order prosecutors to be generous in their interpretation ofBrady/Giglio, not to "tack too closely to the wind" in their disclosure obligations, only to punish the prosecutors if they followed the admonition of the Supreme Court.
In light of the above, and to return to the real problem here, it bears repetition to emphasize that the ultimate issue is whether there is "disclosure" in the letter and spirit ofBrady/Giglio simply by turning over a mountain of "everything" acquired over half a decade, and telling defense counsel nothing about where exculpatory/impeaching information can be found. Again, "the government cannot meet its Brady obligations by providing [the defendant] with access to 600,000 documents and then claiming that she should have been able to find the exculpatory information in the haystack." Hsia, supra. Or, as the undersigned put it in the initial order: "[A]t some point (long since passed in this case) a duty to disclose may be unfulfilled by disclosing too much; at some point, "disclosure," in order to be meaningful, requires "identification" as well.
The government can point to cases which disagree with the undersigned — at least in the circumstances of their facts. However, upon close review, at least the in-circuit district court's W.R. Grace opinion supports the conclusions of the undersigned. In United States v. W.R. Grace, 401 F. Supp. 2d 1069, 1080 (D. Mont. 2005), the district court refused to order the government to conduct a Brady/Giglio identification review of the massive amounts of information acquired in that case. However, the district court did not find that such a request was invariably to be rejected; it found in the circumstances of its case, that such a review was not warranted. First noting that much of the information was text searchable, as is the case here, the court found that "most importantly" the fact that the individual defendants had access to corporate assistance in the search for exculpatory evidence, the vast majority of documents at issue were corporate documents, and the fact that a civil litigation had been ongoing for years, making the search for documents that much easier, were dispositive in the court's mind. The undersigned would add that there were multiple defendants in that criminal case with some overlapping discovery needs, and that the defendants were apparently not detained pending trial making their search for Brady materials much easier.
In the present case, there is a singular, individual defendant, who is detained in jail pending trial, and who is represented by a relatively small defense team. There is no parallel civil litigation, and Salyer does not have access to voluntary corporate assistance in attempting to find the documents needed by the defense. Salyer has little practical ability in the jail setting to be of much assistance in the search as he can do no document review, either hard copy or electronically, absent the physical presence of counsel in the jail. The most important factual underpinnings in the district court W.R. Grace case which convinced the court not to order identification are absent here.
Interestingly, a scant few months later, the district court inW.R. Grace grew skeptical of the government's compliance with itsBrady obligations, apparently concerning documents which had been withheld from the data base, and ordered every agency involved to submit declarations confirming what they had done to ensure thatBrady documents had been produced. United States v. W.R. Grace, 402 F. Sup. 2d 1178, 1180 (D. Mont. 2005), r'vsd 493 F.3d 1719 (on the witness list issue), aff'd 526 F.3d 499 (9th Cir. 2008 (en banc) (on the witness list issue).
The government cases cited from outside the Ninth Circuit, primarily from the Fifth Circuit, do stand more for an outright rejection of any governmental need to identify Brady in a large documents case, but those cases are unpersuasive for the reasons set forth herein. The Fifth Circuit cases do no more analysis than to say there is no authority which requires such a disclosure. See United States v. Mmahat, 106 F.3d 89, 94 (5th Cir. 1997), rvs'd. on other grounds United States v. Parsons, 367 F.3d 409 (5th Cir. 2004). Subsequent cases simply cite to the earlier cases. Even in Skilling, supra however, the court found that there may be some circumstances which might warrant theBrady/Giglio identification of documents in a mass disclosure, such as the bad faith of the government in purposefully hiding exculpatory or impeaching documents in a massive dump of documents. Skilling, 554 F.3d at 577. But as is well established by the Supreme Court, a Brady violation does not depend on the good or bad faith of the prosecution. District Atty's Office, etc. v. Osborne, ___ U.S. ___, 129 S.Ct. 2308, 2335 (2009) citingBrady. Thus, if there is a non-disclosure occasioned by the massiveness of the document production to which the defense is given access, it should make no difference whether such was accompanied by good or bad faith — a non-disclosure is a non-disclosure no matter what the motivation.
In light of the above, the undersigned affirms his original order. However, in view of the government's burden argument, the logistics of implementation are modified. No Giglio identification need be made until the time that the district judge orders the pre-trial identification of the government's witnesses or otherwise sets a time for disclosure of Giglio material. Brady identification of already disclosed materials must be made no later than 90 days from the date of this reconsideration order, or as otherwise ordered by the district judge. New disclosures shall be identified as Brady and/or Giglio, if appropriate, at the time the disclosures are made. The government shall keep appropriate records of the review methodology occasioned by this order and its results.
The undersigned, of course, does not require just one attorney to perform the review, nor is appropriate delegation to non-attorneys precluded. Where appropriate, electronic searches of data bases may be utilized to conduct the review in a meaningful fashion, they may be used. Also, it may well be the case that a person can affirm his or her previous recollection of a review of a category of documents, e.g., data compilations, that no Brady/Giglio material exists without undertaking a re-review of the same material. A notation in the records for this review could simply be made that no disclosable material exists, if that is the case, or a specific identification could be made.
Due to the requirement not to narrowly construe Brady/Giglio as defined by the undersigned herein, i.e., there will be some identifications made in an abundance of caution, the undersigned orders that the act of identification by the government is not an admission which can be utilized before the trier of fact for any substantive purpose unless otherwise ordered by the district judge. Furthermore, nothing in this order excuses the defense from exercising reasonable diligence to discover for itself information disclosed, but not identified by the government, as Brady/Giglio , which it might later contend to be Brady/Giglio information.
The undersigned emphasizes that the initial order and this reconsideration order is limited to the circumstances of this case. The undersigned does not find, nor would he, that the identification requirements of this case would apply to other cases not similarly situated in factual circumstances.
Conclusion
The undersigned grants reconsideration of its order filed July 9, 2010 (Docket 132) at the request of the government, but upon reconsideration, it is ordered that the previous order is affirmed as explained herein with appropriate logistical modifications.
IT IS SO ORDERED.
Of course, the AUSA did not use those words, but his response indicated that grand jury subpoenas are purposefully, broadly worded, and that in a business case, the documents to be seized via search warrant are very voluminous as well.