The privilege as applied to in-house attorneys
Some in-house attorneys plainly have both legal and non-legal responsibilities. But many, particularly in larger legal departments, have only legal responsibilities. I should know, I once was one. These in-house attorneys have an exclusively legal function, whether it be supervising or managing litigation, reviewing company advertising or public statements, or revising contracts. In any real sense, their advice is every bit as exclusively legal advice as is the advice provided by most outside counsel, and there is no more reason to question whether their advice is legal advice than there is with outside counsel.
But in-house attorneys differ from outside counsel. They are on the payroll, so there is no need to retain them or write engagement letters setting forth the purpose and parameters of their engagement. They often advise the same clients on similar issues again and again over long time periods. Over time, the request for legal advice, although understood by all involved, becomes implicit, rather than explicit. That tendency has been exacerbated by the advent of e-mail, which encourages rapid, fragmented, and short communications so that more and more context never makes its way into the communication. E-mail also allows, indeed encourages, retransmission to large groups with a simple “Reply To All” or “Forward,” so an exchange involving legal advice from in-house counsel may be embedded with prior or subsequent e-mails that, while still confidential within the corporation, do not relate to legal advice. Further, and although e-mail seems transient, the reality is that it is more or less permanent.
Thus, when the opposing party’s discovery requests are sufficiently broad in scope, the number of responsive communications involving in-house counsel can be staggering. Many communications that legitimately fall within the scope of the attorney-client privilege will, on their face, provide precious little factual support for the privilege claim. That creates serious problems for attorneys charged with identifying, segregating, and logging the communications. Again, I should know; I’ve done this. And, while communication-by-communication investigation may well turn up the required factual support, undertaking that investigation is hugely time-consuming and expensive, particularly since outside litigation counsel, not the in-house attorneys involved, often are called on to assert and defend the privilege claims. The problem is magnified as the time period covered by the requests expands and, therefore, memories fade and persons with first-hand knowledge leave the company.
These practical problems, however real, are virtually never acknowledged by courts (or their designees) resolving privilege claims. Courts routinely find that the attorney-client privilege must not only be asserted, but supported, on a document-by-document basis and that, if the privilege-holder fails to support each of the required elements for maintaining privilege, privilege claims should not be sustained. Courts express surprise that affidavits supporting privilege claims as to thousands of e-mails are not immediately forthcoming when the opposing party questions the claims. And, when the attorney involved is in-house, some courts require an additional or “clear” showing that the attorney was providing legal, rather than business, advice.
Thus, corporations faced with broad discovery requests that focus on areas in which in-house attorneys have heavy involvement face an unenviable choice. On the one hand, they can risk waiving legitimate claims of privilege by failing to adequately investigate and support them (and risk subject matter waiver attacks). On the other hand, they can spend substantial time and money defending claims that, at best, will result in supportable claims relating to largely innocuous documents being sustained. Their litigation adversaries, particularly when they are individuals or groups of individuals, have no corresponding exposure. They may, and often do, challenge privilege claims as broadly as courts will let them – a low cost process that, at a minimum, imposes substantial burdens on the other side and, at best, produces additional substantive evidence.
The Vioxx privilege challenge
The recent order in the Vioxx litigation, In re Vioxx Litigation, MDL No. 1657, slip op. (E.D. La. Aug. 14, 2007), is a good example of this phenomenon. There, the defendant pharmaceutical manufacturer, Merck, produced over two million documents and withheld roughly 30,000 documents (or about 500,000 pages) based on attorney-client privilege or work product claims. The district court ordered Merck to submit the withheld documents for in camera review and then to categorize them and thereby enable the court to rule on “random samples of each category.” The district court reported that it reviewed the 30,000 documents in a two-week period, found less than 500 protected, and ordered the others produced. Merck sought review in the Fifth Circuit, which directed the district court to adopt Merck’s proposed procedure under which Merck would identify 2,000 documents for further examination and then provide additional support and input for the claims as necessary. In re Vioxx Liability Litig. Steering Comm., No. 06-30378, 2006 U.S. App. LEXIS 27587 (5th Cir. May 25, 2006).
On remand, the district court appointed Paul Rice, a law professor and the author of a leading attorney-client privilege treatise, to conduct the in camera review as a special master. Special Master Rice reviewed 2,000 documents Merck selected, along with 600 documents selected by plaintiffs’ counsel, issued tentative rulings, gave Merck an opportunity to provide additional support, and issued a final report that the district court reproduced in its slip opinion.
Perhaps due to the still-huge volume of communications at issue, Merck initially attempted to support its claims using “generic proof” applicable to multiple documents or classes of documents. For example, Merck asserted what Special Master Rice characterized as a “‘pervasive regulation’ theory” – that “because the drug industry is so extensively regulated by the FDA, virtually everything a member of the industry does carries potential legal problems vis-à-vis government regulators.” (Slip op. at 18-19). Although he acknowledged that “services that initially appear to be non-legal in nature … could, in fact, be legal advice within the context of the drug industry,” Special Master Rice found this analytic framework unhelpful. (Id. at 19).
Merck also advanced a theory the special master characterized as “‘reverse engineering’” – that communications with and relating to otherwise unprotected attachments were privileged “because … adversaries can discern the content of the legal advice that was subsequently offered.” (Id. at 24). Again, Special Master Rice did not adopt this rationale because, among other grounds, “the role of legal counsel … change[s] from legal advisor to corporate decision-maker” if lawyers have the ability to require revisions. (Id. at 24 (italics in original)).
Special Master Rice ultimately developed a number of his own rules or guidelines, many (but not all) of which are similar to those commonly used as guidelines for reviewing attorneys. The more significant of those include:
- When lawyers were “examining and commenting upon a legal instrument, like a patent application, contract for a study, or the retention of experts,” a wide range of lawyer comments, including those regarding “clarity, grammar, consistency, and organization” could be protected (id. at 21);
- Lawyers providing “grammatical, editorial, and word choice comments on non-legal type communications like scientific reports, articles, and study proposals” generally were not protected (id.);
- Following an FDA warning letter, communications about responding to the warning letter generally were privileged (id. at 21-22);
- “When e-mail messages were addressed to both lawyers and non-lawyers for review, comment, and approval, [the special master] concluded that the primary purpose of such communications was not to obtain legal assistance since the same was being sought from all” (id. at 31);
- “Relative to the circulation of legal advice within a corporate structure, it is not acceptable for a corporation to take a document and attachment that are privileged, because they were sent primarily to an attorney for legal advice, and then subsequently send the same document and attachment to other corporate personnel for non-legal purposes … and successfully claim that the document and attachment are privileged,” although the corporation is “permitted to redact the comments of the attorney as privileged legal advice (assuming the elements of the privilege would otherwise be established)” (id. at 32);
- Blind copies directed only to attorneys that seek legal advice are privileged even though the main copy of the communication may not be (id. at 26);
- “When lawyers make the same comments about technology, science, public relations, or marketing” that employees in those departments make, the documents are not privileged unless the corporation “demonstrates that those comments are primarily related to legal assistance” (id. at 28 (italics in original));
- Communications addressed solely to an attorney with limited circulation and an identifiable legal question were privileged (id. at 31);
- Communications that did not reveal the substance of either a confidential communication to an attorney or an attorney’s advice were not privileged (id. at 35);
- Each communication in an e-mail string or thread should be listed individually on a privilege log (id. at 35-36).
The district court, overruling a number of Merck’s objections, adopted Special Master Rice’s final report and the bulk of his recommended privilege rulings. In addition, the district court ordered Merck to produce not only the “sample” documents that Special Master Rice reviewed and found not privileged, but to “extrapolat[e]” those rulings to all “similar documents in the remaining census” within a month. (Id. at 41, 42 n.36).
Where does this leave us?
Before going further, I should note that I have reviewed neither the materials Merck submitted to support its privilege claims nor the underlying documents that were withheld. Thus, I speak with no small degree of ignorance as to specifics. But that will not stop me from advancing three general points that the In re Vioxx district court’s and special master’s rulings raise.
Point One: The Special Master’s general guidelines for assessing privilege claims are, in most (but not all) instances, reasonable general rules to apply in the first instance. In my view, however, the Special Master tilted the balance too far against the privilege-holder by starting from the premise that, for in-house attorneys, there must be “clear” evidence that a communication related to legal advice. See Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 592 n.10 (1993) (preliminary evidentiary questions “should be established by a preponderance of proof”) (citation omitted); Bourjaily v. United States, 483 U.S. 171, 176 (1987) (preliminary evidentiary issues should be resolved under Fed. R. Evid. 104(a) “by a preponderance of the evidence”) (footnote omitted). This perhaps was an artifact of the fact that Merck apparently claimed privilege as to, and withheld, entire e-mail chains based on either limited attorney involvement or attorney involvement in only small portion of the larger chain.
Point Two: The district court’s order directing Merck to apply the Special Master’s general guidelines developed in reviewing one set of documents to other sets of documents that were not reviewed places Merck and its attorneys in a difficult situation. If Merck’s attorneys believed there was a good faith basis in law and fact for asserting claims of privilege as to types of documents that, under the Special Master’s guidelines, were not privileged, does Merck now simply produce those documents? If so, is Merck producing those documents voluntarily, which should result in a waiver, or involuntarily, which may not result in a waiver? If not, has Merck violated the district court’s order? How similar must an unreviewed document be to one of the categories found not privileged before it must be produced? What if a document falls into multiple categories, including one found privileged and another found not privileged?
Point Three: The rules of engagement dictated the outcome. Merck produced an inspirational volume of non-privileged documents – over two million documents. It withheld based on claims of privilege or protection only a relatively small number of documents – 30,000 or 1.5% of the total. All of this occurred in time periods that, given the volumes of documents involved, were short.
Once the district court started down the road to resolving all of Merck’s privilege claims, whether by reviewing the documents itself or by having the special master review “sample” or “representative” documents that would serve as a basis for “extrapolating” to other documents, Merck had little or no chance. Given the volumes of privileged documents that are withheld in response to the sweeping document requests typically seen in major litigation – tens of thousands of documents in In re Vioxx – it is prohibitively expensive for a privilege-holder to explain and provide factual support for each privilege claim. Even properly logging that many documents is a major undertaking, particularly because, as the Special Master’s guideline required, each communication within in every multi-part e-mail string should be separately identified and logged, producing the non-privileged parts and redacting the privileged ones. While the In re Vioxx district court and special master repeatedly expressed surprise at Merck’s failure to provide more fulsome and timely factual support for its privilege claims, I don’t find it the least bit surprising. Indeed, I strongly suspect that Merck’s attorneys worked unbelievably long hours and, in their view, did all they could.
A more fair and reasonable procedure would have been to permit the plaintiffs to challenge the privilege claims for specified numbers of documents each month based on the information made available in the privilege logs about authors, addressees, dates, and subject matters. For example, the case management order entered in the federal government’s civil RICO action against the major cigarette manufacturers, an action I was involved in, permitted “[e]ach side … to challenge the assertion of privilege of up to five hundred (500) documents per thirty (30) day period” and provided that “no one Defendant shall be required to respond to more than 250 challenges to asserted privileges within a thirty (30) day period.” United States v. Philip Morris, Inc., No. 99-CV-2496, 2001 U.S. Dist. LEXIS 12603, *22 (D.D.C. Mar. 27, 2001). That process, although painful and expensive, allowed both sides to investigate and support privilege claims in a more reasonable fashion.
The sheer volumes of privileged documents created and retained in the age of e-mail have created an imbalance that requires limits on the challenging party’s ability to simply say “prove it” and thereby cause the privilege-holder, at best, to incur substantial attorneys’ fees and, at worst, to lose valid claims. For litigants with little or no electronic records of their own, merely insisting on proper logs and factual support for every one of their opponent’s withheld privileged documents not only imposes huge financial burdens on the adversary, but goes a long way toward overcoming the claims themselves. Thus, when the In re Vioxx district court allowed the challenge process to be directed at the entire universe of withheld documents, the plaintiffs had won. It may have taken (or may take) a while for that victory to become apparent, but a victory it was.