Significant 2d Circuit Opinion on Lack of Waiver of Attorney-Client and Work Product Privileges in Common Interest Situation (11/14/16)

Note, this presentation was substantially revised on 11/15/16.

In Schaeffler v. United States, ___ F.3d ___, 2015 U.S. App. LEXIS 19617 (2d Cir. 2015), here, the Second Circuit rendered a major decision on the issue of waiver of the attorney client privilege, through § 7525, here, in a common legal interest context.  The Court's opening paragraph is:

Georg F.W. Schaeffler ("Mr. Schaeffler" or "Schaeffler") and associated entities ("Schaeffler Group") (collectively "appellants") appeal from Magistrate Judge Gorenstein's order denying a petition to quash an IRS summons. n1 We conclude that: (i) the attorney-client privilege was not waived by appellants' provision of documents to a consortium of banks ("Consortium") sharing a common legal interest in the tax treatment of a refinancing and corporate restructuring resulting from an ill-fated acquisition originally financed by the Consortium; and (ii) the work-product doctrine protects documents analyzing the tax treatment of the refinancing and restructuring prepared in anticipation of litigation with the IRS. We therefore vacate and remand.

So, how did the 2d Circuit justify that holding?  The opinion is relatively short, so that is the best source for its reasoning.  At the risk of oversimplification, I offer my short analysis of the points I think appropriate.

Schaeffler, a U.S. resident (perhaps not citizen), was 80% owner of a German corporation which attempted to acquire a minority interest in another German corporation by a tender offer.  German law requires such tender offers to at least offer to acquire all shares.  The offer, unfortunately, was made just before the 2008 financial crisis, hence "far more shareholders than expected or desired accepted [4]  the offer, leaving the Schaeffler Group the owner of nearly 89.9% of outstanding Continental AG shares."  The net result was that the Schaeffler Group had to re-group, so to speak, or refinance with its committed lenders.  That created a potential bust in the financing as a result of the decline in the market and German law.  That required coordination among all parties, including Schaeffler and the coordination in order to respond to the crisis.

From that re-grouping and refinancing and sharing among the parties of information, documents and legal analysis, this dispute arose.  Did that sharing among parties with a common legal interest waive the privileges -- attorney-client and work product?

The issue presented a subtlety in the application of the attorney-client privilege in a common-interest situation.  What exactly does it mean that sharing of otherwise attorney-client or work product privileged information among persons with a common interest preserves the privileges from waiver?  The opinion does not provide any black letter law on that issue, but does address the issue in the specific context before it.

The Second Circuit described the common interest rule as follows (omitting all citations and most quotation marks for easier readability):

While the privilege is generally waived by voluntary disclosure of the communication to another party, the privilege is not waived by disclosure of communications to a party that is engaged in a “common legal enterprise” with the holder of the privilege. Such disclosures remain privileged where a joint defense effort or strategy has been decided upon and undertaken by the parties and their respective counsel in the course of an ongoing common enterprise and multiple clients share a common interest about a legal matter. The need to protect the free flow of information from client to attorney logically exists whenever multiple clients share a common interest about a legal matter. 

Parties may share a “common legal interest” even if they are not parties in ongoing litigation. The common-interest-rule serves to protect the confidentiality of communications passing from one party to the attorney for another party where a  joint defense effort or strategy has been decided upon and undertaken by the parties and their respective counsel. It is therefore unnecessary that there be actual litigation in progress for the common interest rule of the attorney-client privilege to apply. However, only those communications made in the course of an ongoing common enterprise and intended to further the enterprise are protected. The dispositive issue is, therefore, whether the Consortium's common interest with appellants was of a sufficient legal character to prevent a waiver by the sharing of those communications.

The Saltzman Tax Procedure treatise, here, has a good discussion of the common interest rule.  Saltzman and Book, Tax Practice and Procedure, ¶ 13.04[3][a][viii][A] Express waiver [of the attorney-client privilege].

The common interest rule has a specific application in criminal investigations and prosecutions where parties who are subjects, targets or defendants with common interests may enter a "joint defense agreement" ("JDA") as a formal expression of their common interest and commitment to preserve the privilege with respect to privileged information shared among them.  I thought readers might like something on the JDA, so I add the following from the last iteration of my self-published Tax Federal Tax Crimes Book, here (which I suspended after preparing Chapter 12 on Tax Crimes from the Saltzman Tax Procedure publication, here).  Here is where I left it off (not covered in the Saltzman chapter 12):

We attorneys think that we understand the attorney-client privilege, and at a basic level in most situations, undoubtedly we do.  The classic statement of the privilege is (8 Wigmore on Evidence 2292 (McNaughton rev. 1961)): 

(1) Where legal advice of any kind is sought (2) from a professional legal adviser in his capacity as such, (3) the communications relating to that purpose, (4) made in confidence (5) by the client, (6) are at his instance permanently protected (7) from disclosure by himself or by the legal adviser, (8) except the protection be waived. 

Federal Courts apply a more generalized federal common law attorney-client privilege.  There is no definitive statement of this federal common law privilege, so Wigmore’s definition is often used as a starting point.  In addition, Proposed FRE 503(b), 56 F.R.D. 183, 326 (1972), although not adopted, is recognized as “a source of general guidance regarding federal common law principles.”   That proposed rule is: 

A client has a privilege to refuse to disclose and to prevent any other person from disclosing confidential communications made for the purpose of facilitating the rendition of professional legal services to the client, (1) between himself or his representative and his lawyer or his lawyer's representative, or (2) between his lawyer and the lawyer's representative, or (3) by him or his lawyer to a lawyer representing another in a matter of common interest, or (4) between representatives of the client or between the client and a representative of the client, or (5) between lawyers representing the client.

 I have bold-faced the key portion relevant to this blog entry.  The common interest privilege is the basis of Joint Defense Agreements ("JDA").  Here also is my discussion in the prior Federal Tax Crimes Book about Joint Defense Agreements (most footnotes omitted):

H. Joint Defense Agreements (“JDAs”). 

1. Theory of JDAs – Extension of Attorney-Client Privilege. 

Knowledge is power.  In a defense setting, obtaining information efficiently and effectively is power.  In multi-target investigations and multi-defendant prosecutions, each target or defendant may have information that would be important to the defense of the other targets or defendants.  Hence, the targets and defendants often feel a mutual need to share information for their mutual benefit.  Sharing the information, however, means the potential for damage if the information is misused.  No target or defendant wants to share information with other targets or defendants who might then turn it over to the Government.  And, no target or defendant wants the sharing of information to be treated as a waiver of the attorney-client privilege or work product privilege.  If these risks can be eliminated or at least mitigated sufficiently, the sharing of information among targets or defendants can be quite beneficial to them.  The JDA permits such sharing of information with at least acceptable tolerances for risk. 

 JDAs are based on the joint defense or common interest doctrine.  The joint defense doctrine permits parties investigated for or charged with a crime (hence a common interest) to share information pursuant to a joint defense agreement without waiving any of the parties’ attorney-client and work product privileges.  The doctrine is usually justified as an extension of the attorney-client privilege that makes, for some purposes (“some” is meant to be vague here), each attorney in the joint defense group (“JDG”) the attorney for each of the members of the JDG. fn1672  I  explore in this section some of the problems that this justification for the doctrine creates.
   fn 1672 United States v. Henke, 222 F.3d 633, 637 (9th Cir.2000); Wilson P. Abraham Construction Corp. v. Armco Steel Corp., 559 F.2d 250, 253 (5th Cir. 1977) (holding that an attorney  in a JDA is “in effect, counsel for all” and breaches his fiduciary duty if he uses information learned from one of the co-defendants pursuant to the JDA against that co-defendant); United States v. Melvin, 650 F.2d 641, 645-646 (5th Cir. 1981) (noting the “respectable body of law from other courts to the effect that the attorney-client privilege applies to confidential communications among attorneys and their clients for purposes of  a common defense” and string citing cases); United States v. BDO Seidman, 492 F.3d 806, 815-816 (7th Cir. 2007) (noting the doctrine “ extends the attorney-client privilege to otherwise non-confidential communications in limited circumstances.).”  Because, however, the JDA does not fit perfectly the paradigm of the attorney-client privilege (e.g,, can it be logically said that, by joining a JDA, an attorney for one participant becomes the attorney for all, thus raising conflicts questions and duty questions?), there have been various attempts at reformulating the justification for the doctrine in a way that does not implicate the attorney-client privilege.  See ABA Formal Opinion 95-395 (July 24, 1995);  Brown v. Doe, 2 F.3d 1236 (2d Cir. 1993) (describing the joint defense doctrine as creating a fiduciary relationship among the members of the JDG and their lawyers); Deborah Stavile Bartel, Reconceptualizing the Joint Defense Doctrine, 65 Fordham L. Rev. 871 (1996); and Amy Foote, Joint Defense Agreements in Criminal Prosecutions: Tactical and Ethical Implications, 12 Geo. J. Legal Ethics 377, 378 (1999).  Nevertheless, mainstream discussion by the courts and the commentators continue to emphasize the attorney-client privilege justification for the doctrine, even when they gerrymander the concept to avoid some of the problems from application of attorney-client privilege concepts.  

The purpose of the doctrine is to insure that members of the JDG can share otherwise privileged attorney-client communications or attorney work product without waiving either privilege.  The joint defense doctrine is thus more precisely is characterized as a derivative privilege to protect from waiver otherwise privileged information shared in the joint defense context.  This truism sets the limits of its application  – information that is not otherwise privileged does not become privileged simply because shared among parties who have entered a JDA.  Of course, there is also a truism that much of what will be shared will be otherwise privileged at least under the work product privilege. 

Most critically and immediately, of course, the members of the JDG do not want the Government to be able get to the information and use it against any member of the JDG in the criminal investigation and prosecution.  Beyond that, generally, if the JDA is to serve its intended purpose, there needs to be some assurance that the information will not be used in any context adverse to the members providing the information.  

 I will use a simple example to explore some of the issues presented by JDAs.  A and B are targets of a grand jury investigation.  A has engaged attorney X, and B has engaged you.  You and X are considering a JDA in which X will share with you otherwise privileged information he receives from A, and you likewise will share with X otherwise privileged information you receive from B.  A and B, and their respective attorneys, will commit under the JDA to maintain the confidentiality of the information so shared.  Is this really an attorney-client relationship between you and A?  If that is the case, can A object to your representing B if both are subsequently indicted or, worse, can the prosecutor urge that X and you are conflicted out in the criminal case because of that JDA?  Even if there is not strictly speaking a traditional full-bore attorney-client relationship between you and A, do you still have responsibilities to A with respect to using the information received from A or A’s attorney - specifically, can you use the information to benefit your client (B) even if it is adverse to A? n1675  On a more mundane level, do you have to do a conflicts check with respect to A and will you thereafter be conflicted in future representation based upon the relationship between you and A under the JDA?  Can you continue to represent B if A’s and B’s interests diverge?  Should your client decide to plea bargain, can you bring to the negotiating table the information you learned from A (either directly or through A’s lawyer, X)?  Do you have malpractice exposure to A?
   n1675 See ABA Comm. on Ethics and Prof'l Responsibility,  Formal Op. 95-395 (1995), titled “Obligations of a Lawyer who Formerly Represented a Client in Connection with a Joint Defense Consortium.”  This opinion reasons that, while the attorney has no ethical duties typical of the attorney-client relationship to the other parties to the JDA, the attorney does have fiduciary responsibilities limiting his use of information obtained pursuant to the JDA.  

I cannot provide here anything approaching a definitive discussion of these issues, but do address the more immediate ones that are raised by a JDA.  I do note at the outset, however, that, whatever the full ramifications of the JDA are, at a minimum, an attorney considering having his client enter a JDA should perform a conflict check for each client in the JDG and insist that each of the attorneys in the JDG do so likewise.  Furthermore, if possible, the issues raised above should be discussed and dealt with in the JDA in a way that all parties understand how the risks in the JDA are assigned among the parties.

In my now discontinued treatise, I discuss some of the subtleties of the JDA which are also present in the general common interest area but become accentuated in the criminal investigation or prosecution context.

In Schaeffler, the parties entered a common interest agreement (see p. 6, fn. 3):

3. When the Schaeffler Group and the Consortium agreed to share legal analyses, they signed an agreement, styled the “Attorney Client Privilege Agreement.” Of course, the title of that agreement was not binding on the district court and is not binding on us. The Agreement is relevant, however, to the issues of whether the Schaeffler Group and the Consortium maintained confidentiality with regard to third parties and were pursuing a common legal interest.

This is the same as a JDA which is the terminology used in criminal investigations and prosecutions. The following from the discussion of JDA's may be helpful in this respect (footnotes omitted):

2. Types of JDAs. 

There is no standard JDA.  The terms and scope of the JDA vary with the needs and risk-tolerances of the members of the JDG.  One author has noted: 

The cooperative arrangement can take a variety of shapes. Sometimes the lawyers exchange legal or factual memoranda without sharing client confidences. Sometimes the lawyers meet and disclose, either orally or via memoranda, their respective clients' confidential statements. Other times, the lawyers and co-defendants find it best to meet together to discuss joint defense strategy.  It also happens that in pursuit of a joint defense, the lawyer for one co-defendant, or one of the lawyer's agents - such as a criminal investigator or an accountant - may meet separately or communicate directly with a co-defendant who is not his client in the absence of that co-defendant's lawyer.  

My experience reflects other potential uses of the JDA.  In a document intensive investigation or prosecution, the parties may agree to keep a jointly accessible collection or database of documents and research which may be divided up among the lawyers in order to minimize costs and maximize efficiency.  The precise shape and terms of the JDA will be negotiated among counsel for the members of the JDG. 

Because there is no standard JDA, an informal oral JDA raises a real risk that the parties will be unable to prove the existence of a JDA with sufficient terms clearly agreed upon that a court would find the JDA existed.  That does not mean that a JDA must be written in all cases.  Oral JDAs the terms of which can be proved will perform the intended function (depending upon the terms).  Moreover, sometimes the ebb and flow of communication between counsel simply does not permit the time and effort required to hammer out a written JDA. Such informal JDAs should be entered only with an understanding that the benefits to be achieved by foregoing the effort to hammer out the written JDA outweigh the benefits of getting it in writing.

Work Product Privilege.

The Schaeffler also held that the work product privilege applied to an EY Tax Memo.  The Court held that its precedent in United States v. Adlman, 134 F.3d 1194 (2d Cir. 1998) controlled because there was sufficient nexus between the work performed and the prospect of litigation over the major transaction.

[The EY Tax Advice] was specifically aimed at addressing the urgent circumstances arising from the need for a refinancing and restructuring and was necessarily geared to an anticipated audit and subsequent litigation, which was on this record highly likely. See Adlman, 134 F.3d at 1195 (predicted litigation was virtually inevitable because of size of transaction and losses). 

We also disagree with the district court's characterization of the form of the advice EY would be ethically and legally required to give appellants even in the absence of anticipated litigation. Neither professional standards, tax laws, nor IRS regulations required that appellants' tax advisors provide the kind of highly detailed, litigation-focused analysis and advice included in the EY Tax Memo. Cf. id. at 1195 (noting extraordinary detail in 58-page memorandum). The standards relied upon by the district court all target concerns over the "audit lottery," in which aggressive tax advisers might recommend risky tax positions solely because the particular clients were statistically unlikely ever to be audited. See ABA Formal Op. 85-352 (1985) (establishing a governing standard requiring lawyers to advise clients whether a position is likely to withstand litigation). That policy concern is simply not implicated here where appellants would not have sought the same level of detail if merely preparing an annual routine tax return with no particular prospect of litigation.