Skip to main content
Find a Lawyer

Corporate-Attorney Client Privilege

At a recent seminar conducted by the Association of Corporate Counsel of America (ACCA), the lesson being taught was how in-house (or out-house) counsel could investigate shortcomings of their company and maintain the protection of the attorney-client privilege for the factual information received. The seminar leaders were questioned whether it was necessary for the interrogated employee to be seeking legal advice or services in order for the privilege to attach. The response was: "No, read Upjohn." Because a seminar leader should always say "re-read", I decided to actually read Upjohn. What I found was a United States Supreme Court decision demonstrating a profound misconception of the attorney-client privilege.

Upjohn related to factual information provided in writing to corporate counsel by field employees at the direction of senior management. The field employees were instructed to keep the transmitted information confidential. The Supreme Court rejected the "control group"(signifying corporate operatives who could actually use the advice being sought) test to determine whether the field employee's communication could qualify as privileged. Instead, a broad, unfocused, corporate-attorney-client privilege was recognized.

One may assume that this unfocused "zone of silence" was welcomed by the corporate community, but imagine the staggering impact of the Court's next relevant decision, Weintraub, which permitted a trustee in bankruptcy to waive all prior corporate-attorney-client privileges.

Both decisions are wrong for the same reason: Both erroneously assume that a robot has (needs, wants, requires, deserves) an attorney-client privilege. With this clearly unbiased view in mind, it is apparent that an attorney-client relationship can and should exist only between real people, and it is the fictionalization of the concept that has led to the may patently absurd results.

Texas Law

In Texas the evidentiary rules of privilege are codified in rules of evidence. In both a civil and criminal context, Texas recognizes a privilege for:

  1. statutory reporting;
  2. lawyer-clients;
  3. husband-wives;
  4. clergy;
  5. voters;
  6. trade secrets; and
  7. informers.

Under the civil, but not the criminal rules, a physician-patient privilege is recognized. Under the criminal, but not the civil rules, a drug counsel privilege is recognized. The underlying rationale seems grounded on the personal and private relationship that exists between these persons. Indeed, the right of personal privacy sits at the core of our constitutional freedoms.

But what is it about these relationships that causes us to protect the involved interaction? With the clergy, the confession of wrongdoing can be protected because religious counsel may cause a change or otherwise purify one's soul; a physician may cause a cure; a spouse may do all three (we'll ignore the rib-partnership theory); trade secrets protect ideas from unworthy exploitation; informers would be uncooperative if unfettered disclosure was permitted; a political vote is one's own business. To justify the lawyer-client relationship, it is commonly said that full and frank disclosure is necessary for the adversary system to work; the lawyer must known the good and the bad; a lay prson may not know what the lawyer needs to know or what is and what is not relevant. Therefore, what the client tells the lawyer in search of legal advice, as well as what the lawyer tells the client as legal advice, is private, confidential, and privileged.

Most striking about these relationships is the material prejudice that would result from adverse testimony given by the clergy, wife, physician, or lawyer. It is with these persons that honest communication is expected, and revealed admission against interest could well have devastating effects. Of course, what is more relevant than an admission against interest that is probably true? But these are not the only circumstances in which relevant evidence is deemed inadmissible. For example, statements made in settlement discussions are inadmissible on issues of liability or damages. Payment of medical expenses, remedial measures, pleas of nolo contendere, and discussions with the prosecutor are likewise inadmissible.

It is beyond the scope of this article to question the wisdom of these grants of confidentiality or irrelevancy. But it seems important to recognize that available proof of the "truth" is being limited.

Texas law specifically incorporates the control test in determining the existence of a corporate-attorney-client privilege. A "representative of a client is one authorized to obtain professional legal services, or to act on advice . . . on behalf of a client." An "agent" of the "representative of a client" (a field agent, for example) can provide privileged information as a conduit but not as a source, and substantive statements confidentially given by field agents are not considered attorney-client communications. Consistent with Hickman v Taylor, Texas requires production of factual information in civil cases even when gathered by the lawyer in a litigation context if there is "good cause". This does not suggest that factual information imparted by the actual client or client representative can be compelled from the lawyer if the information is personal to the client (or client representative). The client may be required to reveal the facts but not the facts "told to the lawyer."

There are four categories of protected information:

  1. attorney-client;
  2. factual "work product" of the lawyer associated with litigation;
  3. litigation "work product" not included above; and
  4. all mental impressions, theories, and strategies of the lawyer whether reflected in the lawyer's file or that of the client.

Number (1) is never revealed, numbers (2) and (3) are revealed on a showing of "good cause," and number (4) is revealed on a showing of "real good cause" (really never).

The Texas Supreme Court reaffirmed these guidelines in Nat'l Tank Co. v Brotherton, Judge. All the Justices seem to agree on the issue. A clear majority of the court did, however, modify the rules of the road to determine the existence of the investigative privilege. (Only Justices Doggett, Spector, and Gammage voted to maintain the objective litigation standards applied in Stringer v Eleventh Court of Appeals.) This decision should have little impact in actual practice. While it purports to expand the protection of the attorney and non-attorney, investigative communications under Texas Rules of Civil Procedure, Rule 166b(3) [now Rule 192 et.seq.], revelation is still required upon a showing of "good cause." Recently, the Houston (14th) court in Pittsburgh Corning Corp v. Caldwell misinterpreted and misapplied Nat'l Tank. There the communications were between "Relator's representatives and its lawyers." 861 S.W.2d 424. These "representatives," were not shown to be "representatives of the client" to qualify the communications for attorney-client status. Clearly, the communications were "work product associated with litigation," and the "factual information" was disclosable upon of showing of "good cause."

Confidential factual information from field agents provided the attorney at the direction of a client or a "representative of a client" reflects the underlying facts of Upjohn. Pittsburgh Corning notwithstanding, Upjohn communications in a nonlitigation civil context are confidential but not privileged. In a criminal proceeding, however, the evidentiary privilege extends to all confidential information obtained in the attorney-client relationship. Thus, the rationale of Upjohn, if any, is inapplicable to a civil proceeding in a Texas state court. So what is Upjohn, and what does it want?

Federal Law

Confronted with the possibility that its field operatives paid bribes to foreign dignitaries, Upjohn launched an internal investigation headed by its counsel. Field employees were contacted by senior management to fill out and return to counsel a questionnaire providing the who, what, when, and how much of all questionable payments. The field employees were instructed to keep the transmitted information confidential. In a later IRS investigation, the responses were sought by administrative subpoena. A magistrate ordered compliance, and the court of appeals determined that the information was not provided within the control group sphere and that it was therefore not a protected attorney-client communication and affirmed the order of production. The Supreme Court disagreed, stating:

The control group test adopted by the court below thus frustrates the very purpose of the privilege by discouraging the communication of relevant information by employees of the client to attorneys seeking to render legal advice to the client corporation. The attorney's advice will also frequently be more significant to noncontrol group members than to those who officially sanction the advice, and the control group test makes it more difficult to convey full and frank legal advice to the employees who will put into effect the client corporation's policy. See e.g., Duplan Corp. v Derring Milliken, Inc., 397 F.Supp. 1146, 1164 (DSC 1974) . . .

While it would probably be more convenient for the Government to secure the results of petitioner's internal investigation by simply subpoenaing the questionnaires and notes taken by the petitioner's attorneys, such considerations of convenience do not overcome the policies served by the attorney-client privilege. As Justice Jackson noted in his concurring opinion in Hickman v Taylor, 329 U.S. at 516, 67 S.Ct, at 396: "Discovery was hardly intended to enable a learned profession to perform its function . . . on wits borrowed from the adversary." . . .

Needless to say, we decide only the case before us, and do not undertake to draft a set of rules which should govern challenges to investigatory subpoenas. Any such approach would violate the spirit of Federal Rule of Evidence 501. See S. Rep. No. 93-1277, p.13 (197)("the recognition of a privilege based on a confidential relationship . . . should be determined on a case-by-case basis"); Trammel v U.S., 445 U.S. 40, 47; United States v Gillock, 445 U.S. 360, 367, 100 S.Ct. 1185, 1190, 63 L.Ed.2d 454 (1980). While such a "case-by-case" basis may to some slight extent undermine desirable certainty in the boundaries of the attorney-client privilege, it obeys the spirit of the Rules. At the same time we conclude that the narrow "control group test" sanctioned by the Court of Appeals, in this case cannot, consistent with "the principles of the common law as . . . interpreted . . in the light of reason and experience." Fed.Rule Evid. 501, govern the developmentof the law in this area.

Further, the Court denied revelation of the lawyers notes or memoranda of personal interviews of field people as litigation "work product materials" protected by Rule 26(b)(3). Now, frankly, this latter holding, at least, makes some sense. What a party, person or corporation may do and what information it obtains to protect itself in a litigation arena ought to be protected in an adversary system. But the compete absence of litigation at the time of Upjohn's investigation renders the Court's conclusion logically mysterious. Rule 26(b)(3) protects the work of lawyer and nonlawyer alike, and revelation of factual information may be ordered only upon a showing of "good cause." A privilege, however, insulates the information from disclosure even when "good cause" exists. Thus this inviolate treatment ought to be viewed with some circumspection. For this reason, perhaps, mitigating exceptions have been devised in federal court. The privilege is easily waived; contention interrogatories or Rule 30(b)(6) depositions may be used to get the factual information; the lawyer must be acting in the capacity of a lawyer and not a regulatory decision maker, business advisor, or record custodian. In this same vein, the Supreme Court held that the privilege belongs to and can be waived by a corporate successor and even a trustee in bankruptcy. This last bothers me the most by the extent to which it impersonalizes the attorney-client relationship.

My solution is simple. In truth, a lawyer represents senior management of a company, and they ought to be the client. In this context, internal investigations beyond the "client" would be "work product" or at least confidential if litigation was absent. A successor could never require revelation of confidential information from a predecessor's lawyer, except when judicially ordered, and could never obtain attorney-client information, except in limited circumstances. The only down side to that scenario would be the death of the perpetual, institutional law firm. There would be an inherent conflict in interest for the lawyer to assume a representational relationship with the successor without the willing consent of the former client, and even that consent might be insufficient if the conflict in interest was real.

Who's to say that ain't good?


I believe the seminar leaders gave bum advice to a group of Texas lawyers. The shield provided by Upjohn is quite porous and totally ineffective in a civil proceeding in state court.

Was this helpful?

Copied to clipboard