The History of Attorney-Client Privilege
History shows that the attorney-client privilege was not consciously designed—it evolved. According to legal scholarship, the privilege has its roots in the 16th century and at first belonged to the attorney. At that time in English courts, clients did not testify. Their lawyers told their stories. Evidence was given by barristers who had been briefed by solicitors who also sometimes gave evidence. Solicitors learned the facts, at least in part, from their clients. Requiring the lawyer to testify about specific communications with the client was a trespass on the attorney’s honor, as the attorney was presumed to be recounting the client’s story truthfully and would be dishonored if required to testify against the client.
In 1743, an English court in Annesley v. Angelsea articulated the first form of what we might today recognize as an attorney-client privilege. It was then no longer rooted in honor but in necessity, and it was narrow in scope. The dividing line between a privileged and non-privileged communication was whether the communication with the attorney was needed to carry on or defend the suit or prosecution in which disclosure was sought. If unnecessary for that purpose, it was not privileged.
By the middle 1800s, the privilege had broadened. The consensus view, at least in America, was that the privilege covered all matters that a client communicated to a lawyer in professional confidence, except when clients sought advice about committing a crime or fraud. By then, it was commonly understood that the privilege belonged to the client, not the lawyer, and only the client could waive it.
“Inadvertent waiver” seems to have been a 20th-century development. The term itself is somewhat of an oxymoron. In 1938, the Supreme Court told us in Johnson v. Zerbst that a waiver is “ordinarily an intentional relinquishment or abandonment of a known right or privilege.” The law on inadvertent privilege waivers, though, developed incrementally, and the relevant case law on the topic seems to have left the Johnson concept of intentional waiver by the roadside.
In the first half of the 20th century, most of the waiver cases involved clients who asserted an advice of counsel defense or who had offered portions of privileged communications in evidence. Those voluntary acts were held to have opened the door to interrogation on, or discovery of, the balance of the communications. By those voluntary acts, confidentiality had been broken.
That led to a broader view of inadvertent waivers in the second half of the century, developed in cases involving the accidental production of privileged documents in discovery. That was bound to happen after the 1937 adoption of the Federal Rules made document discovery available on motion. It came up even more often after the 1970 amendments made document discovery available upon a mere request to the opponent, which led to an explosion in the volume of documents sought and produced.
Like door-opening waivers, confidentiality concerns led to waiver findings when privileged documents were produced accidentally. Some courts reasoned that the client is strictly responsible for ensuring that confidential communications remain confidential. Thus, any accidental disclosure or failure to exercise due care to preserve confidentiality would void or waive the privilege. Other courts, though, declined to find a waiver from inadvertent disclosures, holding that no waiver could arise unless the client’s disclosure of privileged information was intentional.
Today, most courts fall into a middle ground: An inadvertent production does not automatically waive the privilege, but a waiver will be found if the privilege holder failed to take reasonable precautions to prevent the disclosure or failed to take reasonable, prompt action to recover the inadvertently produced material.
That philosophy—that privilege can be lost if the privilege holder fails to take reasonable precautions to preserve confidentiality—also finds oxygen beyond inadvertently produced documents. When attorneys and clients speak to one another where someone might overhear them, or when clients share their attorney-client communications with trusted confidants, conventional wisdom holds that the privilege is lost, and the communication becomes discoverable evidence. That is because, on matters of privilege, courts are hawkish when lawyers or clients purposely put those communications at risk.
But should courts be so hawkish? Why should a chink in the armor of confidentiality require that the armor be removed altogether? Why should it require that the sensitive information be turned over, wholesale, to the last person in the world with whom the client would want to share it?
Attorney Work Product
To answer those questions, it helps to consider how courts handle attorney work product. When work product is disclosed to a third party, the information is still immune from discovery—that is, there is no waiver—unless the third party is the litigation adversary or is a conduit to the litigation adversary. Courts rationalize the difference in treatment on the basis that the attorney-client privilege is meant to protect the confidentiality of the attorney-client relationship, while the attorney-work-product doctrine is meant to protect the fairness of the adversary process.
That distinction, though, is artificial, as both the privilege and the doctrine share a larger common denominator: the need to ensure a zone of privacy and protection from one’s litigation adversary so the attorney can provide effective representation to the client. Yet, the work-product doctrine enjoys a healthier and more realistic approach to waiver issues, consistent with this larger common denominator.
That approach recognizes a practical reality. Lawyers and clients sometimes need to share the protected information with third parties or to have confidential conversations in insecure places when the benefit of sharing the information outweighs the small risk that the information will fall into an opponent’s hands.
Why should that calculus protect attorney work product but not information covered by attorney-client privilege? The need for sharing the information can be just as important, and the risk of it finding its way to one’s litigation adversary can be just as small. Yet, waiver principles will protect the information in one case but not the other.
Consider this hypothetical: A lawyer recommends a litigation strategy that has some risks to the client and to the client’s family and business. But it also has an upside if the risks don’t materialize. To make an informed decision, the client discusses the protected information with, and seeks advice from, trusted non-attorneys—the client’s parents or grown children, a business partner whose interests might also be affected, or a close friend who has excellent judgment. If the information is work product, there is no waiver. Not so if it’s just privileged.
The attorney-client privilege seeks to enable lawyers to give sound legal advice. But the advice is not an end. It’s a means to an end. It allows clients to make sound decisions that serve their interests while also complying with the law, thus benefiting the client while serving the public interest.
Privilege in the Real World
In the real world, though, decision-making based on legal advice alone is not always easy or realistic. The law could be murky. The decision could affect people close to the client. The decision might be costly or could require the client to get financial or emotional support from others. For reasons like those, clients often need to consult others about the lawyer’s advice before acting on it. But clients can’t easily do that if they are forced to keep privileged information in the hermetically sealed jar that the law on waiver seems to require.
If we are genuinely interested in enabling clients to make good decisions based on good legal advice, then the law ought to make room for the client to share that advice with others in the client’s orbit. And it should give the client some discretion over whom to share it with. Clients should be able to judge for themselves when the risk of their confidential information falling into the wrong hands is so high as to warrant top secret treatment.
Nor should the law be so unforgiving when confidential lawyer-client communications take place within possible earshot of third parties. Lawyers and clients often have—and often can’t avoid having—important conversations in places like courthouse hallways, restaurants, street corners, airports, or airplanes, usually in hushed tones and generally out of earshot of others.
When that happens, the likelihood is remote that the content of the conversation will migrate to the client’s opponent. Although theoretically possible, it is doubtful that any would-be eavesdroppers are actually listening, paying attention, or able to orient the conversation to anything they could understand or place in context.
And what about when the conversation is actually overheard? Lawyers and clients talk to each other in taxis on the way to court or to an important meeting. Might the taxi driver overhear? Sure. Is the conversation likely to mean anything to the driver or motivate the driver to republish it? Almost never.
Do this reality check: Think of all the countless times you’ve been positioned to overhear someone else’s conversation—in a store or restaurant; on a bus or an airplane; in a line at a sporting event, a movie theater, or a concert hall. How many times have you actually listened in, known who the parties were, understood what they were talking about, been able to identify some adverse party, realized that what you overheard could be valuable to that adversary, been motivated to reveal to that adversary what you heard, tracked down the adversary, and spilled the beans? Probably never.
Why, then, do we have a rule that kills the attorney-client privilege, but not work-product protection, if the confidential information is exposed to third parties in circumstances unlikely to reach the party’s adversary? Probably because the two doctrines evolved independently. By the time courts realized they were treating similar and closely connected types of confidential information very differently, it was easier to develop a justification for the difference than to rethink it and treat them alike.
It’s now time to treat them alike, at least on waiver issues. The attorney-client privilege is owned by the client, not by the attorney and not by the court. It’s an important protection, and courts should not brush it aside lightly. If the client shares privileged information with a third party or has a confidential conversation with the lawyer near others but in settings unlikely to lead to the information falling into unfriendly hands, courts should not take the extra step of helping the adversary get that information.
With inadvertently produced privileged documents, the precedent has already been established that ordinarily there is no waiver. And that’s when the adversary has seen the documents. It is thus both illogical and undesirable to find waiver when a client shares protected information with friendly confidants or where non-adversaries might overhear it if the client’s adversaries would not learn of it.
If the privilege waiver rule were brought into line with the work-product waiver rule, then less ink—and less time and money—would be spent on privilege fights, the privilege would continue to provide its intended benefits, and clients would be the better for it. So would the legal system, for it too is best served when clients can talk to their lawyers free from the searching eyes and ears of adversaries. Courts often remark that the law abhors a forfeiture. That maxim should make no exception for the attorney-client privilege.