The recurring and pervasive problem of flawed confidential witness (“CW”) allegations tops my list of the key issues in securities class action litigation.* I don’t mean just notorious situations such as those recently at issue in the Lockheed, SunTrust, and Boeing securities class actions – which I discussed in an earlier post and discuss further below. I also mean the garden-variety inaccuracies that are present in a great many cases.
After catching readers up on what has happened in Lockheed, SunTrust, and Boeing since my prior CW post, I’ll discuss why fixing this problem is so crucial, and then propose a solution. For a useful survey of CW decisions, see Bryan House, “The Fact Pattern Behind the Boeing Class Action Grounding,” Law360 (April 2, 2013).
Update on Lockheed, SunTrust, and Boeing
City of Pontiac General Employees’ Retirement System v. Lockheed Martin (S.D.N.Y. Case No. 11 CV 5026 (JSR)). In Lockheed, Judge Jed Rakoff denied defendants’ motion to dismiss. 875 F. Supp. 2d 359 (S.D.N.Y. 2012). During discovery, several CWs disputed telling the investigator for plaintiffs’ counsel (Robbins Geller) the facts the complaint attributed to them, and discovery revealed that certain of the CW allegations were not based on the CWs’ personal knowledge. Defendants moved for summary judgment, pointing out the flaws in the CW allegations on which Judge Rakoff relied in denying defendants’ motion to dismiss.
On October 1, 2012, Judge Rakoff held a day-long evidentiary hearing to determine “who the heck tried to pull a fraud on this court.” At the hearing’s conclusion, Judge Rakoff offered some tentative thoughts about the witnesses’ credibility. He remarked that some CWs were credible and others were not, and that plaintiffs’ investigator was credible “on the whole.” He asked for briefing by the parties on the issues raised at the hearing.
Following the parties’ post-hearing submissions, Judge Rakoff issued a summary order denying defendants’ summary judgment motion, and promised a longer order. The fact of the denial indicated that, to some extent, Judge Rakoff rejected defendants’ CW challenges, though we’ll never know his findings, because the case settled before he issued his longer order.
Belmont Holdings v. SunTrust Banks (N.D. Ga., Case No. 1:09-cv-01185-WSD). In SunTrust, Judge William Duffey denied defendants’ motions to dismiss, primarily because of allegations based on information provided by a CW, Scott Trapani, indicating that defendants knew SunTrust’s reserves were understated throughout 2007. During the motion-to-dismiss process, defendants pointed out that Mr. Trapani left SunTrust in August 2007, and therefore was not in a position to comment about the reserves throughout the year, but the court ruled it would leave that issue for discovery.
Defendants moved for reconsideration based on declarations from Mr. Trapani that he left SunTrust in August 2007, knew nothing about the challenged financial reporting thereafter, and never told the investigator for plaintiffs’ counsel (again Robbins Geller) that he discussed the individual defendants’ knowledge of SunTrust’s financial reporting thereafter. Based on Mr. Trapani’s declarations, the court reconsidered its motion-to-dismiss order and dismissed the action. The court “reluctantly” decided against sanctions because it appeared that notes from plaintiffs’ investigator, Desiree Torres, supported the FAC’s allegations related to Mr. Trapani.
Ms. Torres later contacted the court and indicated that she was concerned with the accuracy of information plaintiffs’ counsel submitted in their argument against sanctions. The court held a hearing to hear from Ms. Torres, her firm, and the parties, including Robbins Geller. Ms. Torres indicated that she had quit her job over the situation. She said her primary concerns were that, in her view, the submissions incorrectly suggested that plaintiffs’ counsel was not part of interviews of Mr. Trapani in which he indicated that he left SunTrust in August 2007, and that plaintiffs’ allegations inaccurately portrayed Mr. Trapani’s knowledge of matters after August 2007.
At the hearing, Robbins Geller examined Ms. Torres in detail about the interviews of Mr. Trapani and plaintiffs’ pleadings. If you are a securities litigation geek, you’ll find the transcript fascinating. In a nutshell, the problem seems to have been caused by a combination of vagueness in what Mr. Trapani said he knew after August 2007, which apparently was occasioned in part by his unwillingness to provide certain details to plaintiffs’ counsel and their investigators, and plaintiffs’ counsel’s interpretation, inferences, and extrapolation of the information he did provide – and then plaintiffs’ counsel’s failure to correct their prior allegations and argument once the scope of Mr. Trapani’s knowledge became clearer.
The court took the matter under advisement, and in a post-hearing order did not impose Rule 11 sanctions:
After hearing testimony and argument at the hearing, the Court concludes that, while not in keeping with the conduct expected of attorneys practicing before this Court, Plaintiff’s counsel’s actions in this matter did not constitute an actionable violation of the Federal Rules of Civil Procedure. The Court remains troubled by the conduct of Plaintiff’s counsel in failing to correct representations made in their pleadings or to notify the Court of them immediately after it became apparent that Trapani did not have knowledge after August 2007 of Defendants’ conduct or beliefs regarding the [reserves]. The decision not to correct the record after counsel became aware of the Court’s reliance on Plaintiff’s representations is perplexing and disappointing. Had Plaintiff’s counsel done so, Ms. Torres likely would not have felt compelled to contact the Court after reading the August 28th Order based on her understanding of the manner in which the Court interpreted the information that was provided to it by Plaintiff’s counsel.
City of Livonia v. Boeing (N.D. Ill., Case NO. 09 C 7143; 7th Cir. Case Nos. 12-1899, 12-2009). On the basis of allegations based on information allegedly obtained from a CW, Bishnujee Singh, the court denied defendants’ motion to dismiss. Boeing’s investigation revealed that plaintiffs’ allegations based on Mr. Singh were incorrect, including the allegations that he was employed by Boeing (he was not; he was employed by a contractor of Boeing), or highly improbable, including the allegation that he communicated with senior management. Boeing took his deposition. He denied almost everything the investigator for plaintiffs’ counsel (again Robbins Geller) had attributed to him.
Defendants filed a motion for reconsideration of the court’s order denying their motion to dismiss. The district court granted the motion and dismissed the complaint. Defendants didn’t file a motion for Rule 11 sanctions, and the court didn’t impose them on its own.
Plaintiffs moved for relief from that order on various grounds, including assertions that the documents Boeing produced confirmed the information Mr. Singh provided, and that Mr. Singh’s recantation was caused by his desire to work directly for Boeing. In support of their recantation theory, plaintiffs cited an email from Mr. Singh to senior Boeing employees, which plaintiffs characterized as follows:
Singh’s communications with Boeing employees also demonstrate his motive to change his story. He actively sought work at Boeing. Pl. Br. at 9, 13. On the very day of his deposition, Singh wrote directly to Michael Denton, who he had identified in his meeting with the investigator as the Vice President of Engineering for the 787 Program, and Jim Albaugh, defendant Carson’s replacement at Boeing, noting that he was “following up” with them, and stating that he deserved “[a]t least” a “THANK YOU!” for “trying my best to help in all possible ways to Boeing group in this disposition [sic] case by denying knowledge of the facts.” Pl. Br. at 4; Dkt. No. 173, Ex. 1. Eight days later, Singh filed another application for work at Boeing.
The Seventh Circuit, in an opinion by Judge Richard Posner, affirmed the dismissal, and remanded the case to the district court to address plaintiffs’ counsel’s compliance with Rule 11, which Judge Posner noted the Reform Act requires even without a motion by defendants. 2013 WL 1197791 (7th Cir. March 26, 2013). Judge Posner noted that the evidence suggested that plaintiffs’ counsel’s investigator had “qualms” about the information Mr. Singh provided, and that the failure of some of the evidence to check out “should have been a red flag.” Judge Posner’s criticism of plaintiffs’ counsel was blistering:
Their failure to inquire further puts one in mind of ostrich tactics—of failing to inquire for fear that the inquiry might reveal stronger evidence of their scienter regarding the authenticity of the confidential source than the flimsy evidence of scienter they were able to marshal against Boeing. Representations in a filing in a federal district court that are not grounded in an “inquiry reasonable under the circumstances” or that are unlikely to “have evidentiary support after a reasonable opportunity for further investigation or discovery” violate Rules 11(b) and 11(b)(3).
The plaintiffs’ law firm–Robbins Geller Rudman & Dowd LLP–was criticized for misleading allegations, concerning confidential sources, made to stave off dismissal of a securities-fraud case much like this one, in Belmont Holdings Corp. v. SunTrust Banks, Inc., No. 1:09-cv1185-WSD, 2012 WL 4096146 at *16-18 (N.D.Ga. Aug. 28, 2012). The firm is described in two other reported cases as having engaged in similar misconduct: Camp v. Sears Holdings Corp., 371 Fed. Appx. 212, 216-17 (2d Cir.2010); Applestein v. Medivation, Inc., 861 F.Supp.2d 1030, 1037-39 (N.D.Cal.2012). Recidivism is relevant in assessing sanctions. Reed v. Great Lakes Cos., 330 F.3d 931, 936 (7th Cir. 2003).
The Importance of Solving the Confidential-Witness Problem
Obviously, we can’t keep having problems like those at issue in these three cases. Although these cases have received significant attention because of the prominence of the companies and judges, as well as some extreme facts (e.g. Ms. Torres contacting the court and quitting her job), problems with CW allegations – from disagreement about information attributed to them, to vagueness and ambiguity in the complaint’s descriptions and allegations – exist in many cases. Finding a solution is important, for two main reasons.
First and foremost, CW allegations based on inaccurate information result in injustice; insufficient complaints aren’t dismissed. Yet defendants are procedurally limited in their ability to present evidence demonstrating inaccuracies unless and until their motion to dismiss is denied. This is so because the Reform Act’s stay of discovery during the motion to dismiss process applies to both plaintiffs’ and defendants’ discovery. So, even if the defendants know that the plaintiffs have completely misstated what a CW told plaintiffs’ counsel/investigator – indeed, even if a purported CW claims not to have spoken with the plaintiffs at all – defendants cannot take discovery to establish such facts without court permission. And a decision to seek court permission to conduct discovery can be tricky; there’s a risk that discovery will mushroom, and the defendants will lose the benefits of the stay.
Moreover, even if discovery demonstrates inaccuracies in the complaint, there is no completely satisfactory procedural mechanism for raising the issue before the court decides the motion to dismiss. A Rule 11 motion is the most procedurally suitable procedure, but it is a very serious one, and the defendants rightly approach it with caution, since it can backfire in the form of an angry judge and/or an inefficiently unworkable relationship between the lawyers. A Rule 12(f) motion to strike isn’t available in at least some courts, and is frowned upon by some others. And defendants can submit factual information on a motion to dismiss only in limited circumstances.
So, defendants often make the best motion-to-dismiss arguments they can and then address the CW problems after the motion to dismiss is denied. But, at that point, it’s too often late to effectively address the inaccurate CW allegations, because the case is in discovery, and the CW problems can fade into one of the myriad issues to be sorted out in discovery.
Second, I believe the greatest risk to the Reform Act’s protections has always been legislative backlash over a perception that the Reform Act is unfair to investors. The Reform Act’s heavy pleading burdens have caused plaintiffs’ counsel to seek out former employees and others to provide internal information. The investigative process is often difficult and is ethically tricky, and the information it generates can be lousy. This is so even if plaintiffs’ counsel and their investigators act in good faith – information can be misunderstood, misinterpreted, and/or misconstrued by the time it is conveyed from one person to the next to the next to the next. And, to further complicate matters, CWs sometimes recant, or even deny that they made their previous statements.
The result is an unseemly game of he-said/she-said between CWs and plaintiffs’ counsel, in which the referee is ultimately an Article III judge.
If Robbins Geller avoids sanctions in Boeing, it will be major, major news, especially given Judge Posner’s scathing criticism. I don’t predict that there would be an outpouring of sympathy for Robbins Geller. But I do believe that a decision not to impose sanctions, along with the outcomes of Lockheed and SunTrust, would prompt scrutiny by commentators and possibly legislators: how could there have been three cases in just the last year that turned from allegations of serious misconduct against plaintiffs’ counsel – accompanied by preliminary but harsh criticism from courts – to conduct the courts ultimately found did not violate Rule 11?
An examination of these CW hearings would necessarily involve a discussion of the Reform Act’s heightened pleading standards; they are the reason plaintiffs’ counsel turns to CWs. Indeed, in Lockheed, Judge Rakoff’s concluding remarks noted that extreme pleading standards involve “dangers” – for example, the “difficulties plaintiffs have in getting information that they know they’re going to have to get to meet the very high standard that the Supreme Court has now imposed on plaintiffs in these cases.”
It would be highly unfortunate, however, if there were serious discussion about reforming the Reform Act’s pleading standards or other protections. The CW problem can be solved through simpler means that do not undermine the Reform Act’s protections, which have created a system of securities litigation that is vastly superior to the one the Reform Act reformed.
Some Suggested Reforms to the CW Process
An effective solution to the CW problem could be achieved with three reforms:
First, plaintiffs’ counsel should be required to obtain from each confidential witness a declaration and/or a certification that he or she has read the complaint and agrees with the description of the information he or she provided. This simple requirement would prevent most CW problems, and make the ones that do arise much easier to resolve. Although some witnesses may balk at providing a declaration, few legitimate witnesses with accurate information to provide would hesitate to certify the accuracy of the relevant portions of the complaint – indeed, most would want to do so, to avoid the hassles that misunderstandings can cause.
Second, plaintiffs should be required to precisely describe, at a minimum, the following information for each CW: (1) employment dates – by day, month, and year; (2) employment responsibilities – including job title, job description, and a detailed list of job responsibilities, and the substance and exact date of any changes; and (3) how the CW knows the information the complaint alleges. There can be no reasonable objection to CWs providing these facts. With regard to employment information, CWs know it precisely and, if they don’t recall precisely, they have documents that reflect it. With regard to the basis of knowledge, if the CWs can’t recall the basis, there’s no good reason to credit the other information the complaint alleges. Indeed, under current law, vagueness in these three areas undermines the CW allegations.
Third, defendants should be allowed to seek limited discovery, without risk to their discovery-stay rights, and to offer evidence to address significant inaccuracies before the motion to dismiss, through a motion to strike the inaccurate allegations – or, alternatively, during the motion-to-dismiss process itself, without converting the motion to a summary judgment motion. Having a designated process for raising these concerns would help attorneys and the judge to navigate a moderate middle course between the two extremes that are pervasive today – either allowing inaccurate allegations to survive through a motion to dismiss, or taking the dramatic step of filing a Rule 11 motion against plaintiffs’ counsel. Defendants should not be required to resort to Rule 11 to raise these issues, though they should be permitted to use Rule 11 if the conduct of plaintiffs’ counsel makes it the most appropriate course of action.
These three measures would have prevented the problems at issue in Lockheed, SunTrust, and Boeing. For example:
- If the CWs in Lockheed had provided declarations or certifications, Judge Rakoff would not have had to hold a day-long hearing to determine whether the CWs, in fact, told plaintiffs’ counsel’s investigator the information alleged in the complaint.
- If plaintiffs’ counsel in SunTrust were required to more precisely allege Mr. Trapani’s employment dates, the court and parties could have avoided the hassle and spectacle of the hearing to settle this basic issue.
- In all three cases, it would have been more efficient and less costly if defendants had an effective means of raising these concerns before the motion to dismiss, or, at the very least, during the motion-to-dismiss process.
These reforms would not only prevent unseemly showdowns – between defense counsel and plaintiffs’ counsel, or among plaintiffs’ counsel, their CWs and their investigators. They would make all securities class action complaints more factually accurate and thus make the outcomes more just – and would help to avoid continued actions against plaintiffs’ counsel, which could eventually cause Congress to consider reforming the Reform Act.