The Texas Tornado, Mark Bennett, in the finest tradition of the criminal defense bar, has sunk his teeth in the  Boucher discussion and won’t let go until it’s been ripped to shreds.  Mark raises some very worthy points, both in his post and  subsequent comments (a lousy place to have a discussion). 

Since the issues raised by this case involve a path less traveled, but one that may well prove to come up with far greater frequency in the future as computer crimes as the interplay between old law and new technology requires us to confront an ever-changing landscape, it’s worth some additional discussion and debate.

The traditional application of the  5th Amendment is to preclude compelling a person to make incriminatory statements.  As this is how most people, and most lawyers, think of the 5th, it tends to box us in when it applies in a less traditional fashion.  What the 5th provides (in relevant part) is:

No person . . . shall be compelled in any criminal case to be a witness against himself, nor be deprived of life, liberty, or property, without due process of law . . ..

On its face, the 5th has no such limitation to statements, but its interpretation by courts makes it both broader and narrower than some might think.  It speaks to being made “a witness against himself.”  While one might well believe that taking a defendant’s blood (or fingerprints, DNA, voice recordings, handwriting exemplars, etc.) compels a person to be a witness against himself, courts have held that these do not violate the privilege. 

The distinction under the 5th Amendment has been defined as “testimonial”:  Does it require a defendant to provide knowledge he possesses or speak his guilt.  While the latter is what people normally understand to be within the privilege, the former presents a concept that is often misunderstood by lawyers and laypeople alike. 

The Boucher decision rests upon the “testimonial” nature of the “act of production” of a password.  By compelling Boucher to produce a password, it provides  evidence that Boucher (1) knows the password, (2) has access to the kiddie porn on the computer, and (3) demonstrates his control over the computer.  The Boucher decision rejects the government’s argument that these are not testimonial, and hence unprivileged.

Mark was getting bogged down by the distinction between compelled incriminatory statements and compelled incriminatory “testimonial” acts.  This leads to the effort to try to plug all 5th Amendment privilege issues into the Miranda paradigm, which just doesn’t work.  These are just not statements.

The failure to fit non-statement testimonial evidence can be seen by the government’s various offers in Boucher to “overcome” the testimonial nature of the password.  They offered everything from letting the defendant (or someone on his behalf) input the password in private to granting use immunity.  Mark argued that this would solve the situation:

If Mr. Boucher had not moved to quash the subpoena, or if the magistrate judge had denied the motion to quash, he could have taken the Fifth before the Grand Jury. Then the government could have given him immunity (technically or practically [by getting Judge Niedermeier to order him to answer the question]) and forced him to answer the question or go to jail for contempt (which might be less onerous than going to prison for possession of child pornography).

Because this didn’t require Boucher to give statements, this shows why it is critical to understand the difference between testimonial acts and testimony.  Not only was Boucher not subpoenaed to appear and give testimony before a grand jury (the typical grand jury subpoena situation), but he could not have invoked the 5th a second time after having a ruling from the court to the contrary.  True, Boucher could have refused to comply with the subpoena, but the court could have jailed him until such time as he complied, and then he would have been prosecuted for the porn.  This wouldn’t have provided much of a solution.

But Mark astutely notes why the invocation of privilege matters so much:

If Mr. Boucher had been compelled to provide his password and Agent Curtis became unavailable for trial, the only evidence the government would then have of the contents of the hard drive would have been derived from Mr. Boucher’s compelled testimony, and would therefore have been inadmissible.

This is the reason why Judge Niedermeier rejected the government’s proposal to give use immunity, but be permitted to enjoy the fruits of derivative use, the computer files accessed after the password was provided.  Additionally, the strength of evidence against Boucher, Agent Curtis’ testimony about a file that he claims to have seen on Boucher’s computer versus presenting the actual file to the jury, is substantially different.  Clearly, the latter is far more damning than the former.

Mark is also correct in noting that the implicit assumption that Boucher input the password, unseen by Agent Curtis at the border, may not in fact have happened. 

[D]rive Z may already have been decrypted and mounted when Mr. Boucher woke his computer from sleep at the crossing. After seeing the suspect images, according to Magistrate Judge Niedermeier’s opinion, “Agent Curtis arrested Boucher. He then seized the laptop, after shutting it down.” This would have triggered the encryption software’s unmount of drive Z.

This is a plausible explanation, though somewhat problematic since it is up to the defendant to raise his factual disputes, and we indulge in fiction when we engage in post hoc factual recharacterizations to support our positions.  In other words, since this isn’t what Judge Niedermeier found to be the facts, it isn’t the fact for the purpose of discussing his decision. 

But similarly, Orin’s (and my) acceptance of the implicit idea that Boucher had affirmatively input the password is not a hard finding by Judge Niedermeier either, meaning that I’ve done the same thing as Mark, overlayed my understanding of the facts which are not expressly stated.  I stand corrected.

The problem, for purpose of the Boucher decision being upheld or reversed, is that the computer somehow went from being observable by Agent Curtis to locked by a password.  One way of the other, it implicates Boucher’s involvement with a password, whether by inputting it originally or having a password protected computer that locked up upon shutdown. 

Either way, Boucher finds himself responsible for the existence of a password, unless he contends that this was not his computer at all.  Since that’s not the case, we’re still stuck with the password problem, that Boucher’s computer is password protected, but was accessed by Boucher at the border.   Somehow, Boucher ends up having the password that the government needs, and this fact remains proven by the earlier conduct.

This is where the final, and most important part, of the Boucher decision comes into play.  Even though it is clear (at least to me, though Mark would disagree) that Boucher has already shown that he had the power to access the contents of his password protected computer, the contents of which are clearly not privileged, does Boucher have the constitutional right to invoke his privilege against disclosing the password and allowing the government to access the contents the second time around?

Orin argues that a waiver happened when Boucher allowed Agent Curtis to see the content of his computer the first time around.  Mark argues that Boucher can invoke the 5th regardless of what happened the first time around.  The court held that because the password is not a physical thing but rather something that exists in Boucher’s mind thus rendering it testimonial, it cannot be compelled.

This disagreement revolves around the question of whether a privilege once waived can be restored or recaptured.  Unless we indulge in the fiction that Agent Curtis is lying about having seen the content of the computer (a place where Mark is willing to go but I am not since it was never raised or argued in Boucher), Boucher waived his right once by allowing access to the contents of his computer.  Having let the horse out of the barn, the genie out of the bottle, whatever trite phrase you prefer, can he now put (horse, genie, whatever) back in?

Judge Niedermeier never expressly deals with this.  Mark says he doesn’t have to.  Orin says this is a huge gap in the decision.

In his first post on Boucher, Mark analogizes the scenario to sex (agreeing to sex once doesn’t mean that it’s agreed to forever).  This has superficial appeal, but simply isn’t the law.  There are waivers that put an end to the issue, and waivers that don’t.  Mark then analogizes it to having initially agreed to answer police questions, but subsequently invoking the 5 to refuse to the same questions again (say with a video, just to explain why they would want to do this).

This is where Mark’s analogy has legs.  While it deals with statements rather than non-statement testimonial acts, the waiver aspect is essentially identical.  Both are testimonial, yet the defendant cannot be compelled to “do it again” just because he did it once. 

Orin, I suspect, would argue that because there is no harm to Boucher by doing it again, having already created the proof of production via his initial waiver, and since it is not the password but the computer content that is the object of this search for evidence, the testimonial nature of the act of production should be subject to waiver analysis and, indeed, deemed waived for this purpose.

But Mark has sufficiently, and correctly, shown why there is harm to Boucher, whether because Agent Curtis is later unable to testify or because the Agent’s testimony is of lesser weight than the actual computer files themselves.  It is important, however, to note that this is hardly as cut and dry as it might appear based on Mark’s application of the privilege relating to statements as opposed to the act of production. 

While I agree with Mark’s conclusion (and I hope the reviewing court in Boucher does as well), the path taken to reach it was more than a little dubious and misdirected not recognizing the difference between statements and non-statement testimonial evidence under the 5th Amendment. 

The point of all this (again) is that criminal defense lawyers, dealing with application of old law to new technologies, need to understand the nuances if we are going to mount successful opposition to the government’s efforts to intrude on a digital world.  Getting stuck in a simplistic and erroneous analysis, even if we accidentally arrive at the right conclusion, opens the door to the creation of bad law as well as the conviction of a defendant.  We need to appreciate the subtleties so that we can successfully defend the Constitution in the face of novel applications.