Friday, April 26, 2013

You’ve Been Warned

By David Rangaviz

State v. Brooks, 2013 VT 27

With the tragic events in Boston, the eyes of the country have turned toward Miranda.

Miranda, as I summarized in a previous post, requires police to read a series of rights (“you have the right to remain silent, you have the right to an attorney … etc.”) to a suspect in custody, and then obtain a knowing and voluntary waiver of those rights, before any interrogation can proceed.  Of course, police can still question a suspect without providing the warning, but any incriminating statements made during that interrogation will later be inadmissible in court.  To obtain a useable confession, Miranda warnings must be given and the rights described therein, waived.

The surviving suspect in the Boston Marathon bombing was not initially read this warning.  It has led to a significant amount of press coverage of something called the “public safety exception,” under which police can engage in a limited unwarned interrogation when confronted with an imminent threat to public safety and still have the resulting statements be admitted.  (More detail on the origin and doctrine underlying the public safety exception can be found here and here.)

The refusal to provide the warning has sparked considerable debate in the legal community about the proper limits of the public safety exception.  Some see it as a trampling of rights that could distort the law of Miranda for everyday cases, while others look solely at the confines of the Boston case and view the decision as correct under the circumstances.

Although I’ll remain agnostic in this debate, I will say this: there is nothing stopping the FBI from gathering intelligence from the suspect without providing Miranda warnings.  They can ask him whatever they want; they just shouldn’t necessarily expect his statements to be admissible in court.  And, given the apparent strength of their case against him, they probably don’t need his confession anyway.  (He is alleged to have confessed to the carjacking victim anyway, which would be admissible because it was not the result of police interrogation.)  So the FBI doesn’t seem to need to invoke the public safety exception in this case—just question him, gather intelligence, and live with the suppression of his statements by convicting him with the mountains of other evidence.  (More coverage of these issues can be found here.)

Today’s case presents a related issue (obviously in a very different context): If police provide a mid-interrogation Miranda warning, can statements made after the warning be admitted in court?  Or does an initial, unwarned statement so taint the entirety of the interaction that even subsequent post-warning statements cannot be admitted?

The facts are straightforward.  In 2009, Defendant was called to the police station for an interview, during which he denied any improper, sexual contact and intercourse with his 12-year old daughter.  He was then arrested and placed in a holding cell.  Six hours later, a detective approached Defendant to plan his dinner.  When the detective told Defendant about the ongoing police investigation, Defendant volunteered, “Well, if everyone said I did this I must have.” 

Up to this point, Defendant had not yet been Mirandized.

Ten minutes later, the detective set up an interview, in which he provided a full warning, secured a waiver, and then (after a seventy-five minute interrogation) obtained a confession.  During the interrogation, Defendant initially denied all of the allegations, but eventually wrote out a full confession to the sexual intercourse.  The trial court did not permit the introduction of the unwarned statement, but allowed the later confession into evidence.

Defendant was convicted following a jury trial on two counts of aggravated sexual assault.

The SCOV now has to decide whether the initial, unwarned statement (“I must have”) tainted the subsequent confession.  If so, the post-Miranda confession would be inadmissible and the conviction would have to be reversed.

In a unanimous opinion, the SCOV concludes that the mid-stream Miranda warning was effective in safeguarding Defendant’s rights, and therefore affirms the conviction.

To determine whether the statements should have been admitted, the SCOV employs what it describes as a “totality-of-the-circumstances analysis.”  The basic test: Was the cat already out of the bag?  For example, if the suspect has already made a full unwarned confession, doing so again, after a warning, would still not be admissible.  If a suspect thinks the cat is out of the bag (or the barn door open, or the camel in the tent … pick your metaphor), he’ll probably confess again because (1) he’s already given the information and (2) he doesn’t understand the legal significance of the Miranda warnings (i.e., that warned statements can be admitted in court).  There is also a recognition of the possibility of police misconduct in this line of cases—if officers can extract an unwarned, admissible confession, they will try to do so out of a fear that proper warnings will stop a suspect from speaking.  Permitting such a practice would sanction manipulation of the dictates of Miranda.

The SCOV highlights five factors that weigh in this analysis:
(1)   “the completeness and detail of the questions and answers in the first found of interrogation”
(2)   “the overlapping content of the two statements”
(3)   “the timing and setting of the first and second”
(4)   “the continuity of police personnel”
(5)   “the degree to which the interrogator’s questions treated the second round as continuous with the first”

Employing these five factors, the SCOV concludes that the two interrogations were sufficiently distinct to allow admission of the post-Miranda statements.  The initial statement—“Well, if everyone said I did this I must have”—was simply too vague, particularly when compared with the full, detailed confession provided later.  Also, in the period between the two incriminating statements, Defendant denied the abuse, which, according to SCOV, “eliminat[ed] any residue of guilt from the initial unwarned statement.”  This period of denial effectively bifurcated the two interrogations.

The exchanges also took place in different settings, and the detective never referred back to the initial statement during the second interrogation.  In other words, he didn’t try to leverage the earlier statement to get Defendant to incriminate himself a second time.  Instead, he treated the second interrogation as a separate discussion, and so . . . it was a separate discussion.

Because the Miranda warnings were effective, and the subsequent waiver voluntary, there is no reason to suppress Defendant’s statement.  The confession was properly admitted.

Moving to a secondary issue, the SCOV briefly addresses the admission of Defendant’s browsing history into evidence.  The trial court had allowed the jury to hear evidence of Defendant’s internet history of browsing pornographic and incest websites.

The SCOV opts to sidestep this legal question and simply assume the admission was error.  According to the SCOV, any such error must have been harmless (i.e., did not alter the jury’s verdict) because the browsing history was so tangentially relevant compared with the other evidence admitted, particularly the full, detailed confession.  Because the evidence of guilt was overwhelming, any error here was harmless.

As is the case here, sometimes you can tell what a court is thinking more by what it assumes than what it explicitly decides.  Why would the court avoid the issue and jump right to the harmlessness of the error?

The admission of this browsing history is clearly troubling; just because someone visits a website doesn’t mean they like, agree with, or wish to emulate everything on that page.  (Except for readers of SCOV Law Blog . . . your love of this page says a great deal about your intellectual prowess and innate good looks!)

Not only does this raise the evidentiary problems that Defendant identified, but it may also provoke more difficult First Amendment questions.  The First Amendment protects both the right of people to engage in speech and the right of an audience to receive it.  Free speech advocates would argue that the audience should not have their receipt of information—even heinous information—potentially used against them in court.  First Amendment protection extends beyond majoritarian or conventional thinking.  For the marketplace of ideas to remain a truly free market, its shoppers should not be penalized for their choices.

The admission of such history could also have a chilling effect on browsing itself; if you know your browsing history can be used against you, you might change your browsing habits.  This is particularly likely given the ease with which such browsing history can be catalogued and tracked.  If we can analogize browsing history to, say, library borrowing history, these potential problems become even more apparent.

Particularly here, the salacious nature of the websites would seem to prejudice the jury against Defendant from the outset of trial, distorting the lens through which the jury viewed the remainder of the otherwise “overwhelming” evidence.  (The browsing history was admitted as “prior bad act” evidence, the possible dangers of which I have discussed in a past post.)

In short, the admissibility of browsing history is a potentially thorny legal question.

The SCOV, in seeming recognition of how tough this issue could get, simply assumes the problem away.  Courts often do this; if one way of deciding a case is easy and another way is hard, the easy way usually wins out.  The hard way often provokes dissent, dividing a court over an issue it needn’t have even reached, and requiring unnecessary effort in the drafting and re-drafting of multiple opinions by overworked judges and law clerks.

But, particularly given the internet’s proliferation and ubiquity in everyday life, this issue is not going away. 

Sooner or later, the SCOV will have to decide the extent to which a person’s browsing history may be used against them in court.  The importance of this question cannot be overstated because of the clear slippery slope it portends. 

Where do we draw the line in the admission of browsing history?  Can a defendant’s movie preferences be used against him?  What about taste in music?  To what extent can we assume that a person’s criminality may affect that individual’s choice of entertainment?  Or does the entertainment itself cause the criminality?  If that’s the case, why shouldn’t victims be able to sue the purveyors of such entertainment for the crimes of their customers?  Such civil liability would undoubtedly chill protected expression.

Of course, I don’t blame SCOV for refusing to step into the breach.  Today’s case is easy and SCOV’s time is limited: whether it was error or not, it would not have changed the jury’s verdict.  Hard cases may make bad law, but there is no easy case corollary.  For now, the “assumption” by SCOV does serve one purpose—flagging the issue for enterprising defense attorneys as an indication of a possible point of contention at the court.

These two main issues thus resolved, SCOV affirms the conviction.  Despite the affirmance, the message to prosecutors is fairly clear: don’t manipulate Miranda, and tread lightly when using website browsing history to obtain a conviction.

Also: stop reading Playboy . . . even for the articles.


  1. So if you confess>deny>confess, then a midstream warning doesn't render a subsequent confession a product of the initial confession, but if you just confess>confess then the statements are excluded? That is absurd.

    Seem to me like most of the Elstad/Seibert factors go in favor of the defendant here anyway, but math or logic have never really been totems of the "totality-of-the-circumstances" test.

    Why don't they just require a rule for midstream warnings to include a warning that if the suspect has already made unwarned statements, then those statements can't be used later during trial or interrogation? The Courts don't seem to have a problem requiring Miranda rights be read in the first place, so it's not any additional burden on the cops. There may be exceptions of course, but at least by that point the suspect knows they have the right to speak with an attorney to find out whether he/she's already dug his/her own hole.

  2. Yeah, I guess the logic of the court is that if you confess>deny>confess, the intervening denial makes it less likely that you think that the ship has already sailed. If the goal of the test is to get at what is in the defendant's mind, I think that makes a measure of sense, though I agree that the facts of the case are tough.

    I actually think that's a really thoughtful idea: if police are going to make a mid-interrogation Miranda warning, they must include the statement that "all things you've already said cannot be used against you" (or something to that effect). It's definitely no greater burden on the cops, because, as you say, they're already required to read Miranda anyway. If the point of Miranda is as a prophylactic requirement to protect the underlying rights, it definitely seems like adding an element to the mid-interrogation would serve this purpose even better. Now that I think about it, that's probably something that SCOV, as well as other courts, should consider.