Split Tenth Circuit panel finds mandatory five-year prison term for violation of supervised release itself violates Fifth and Sixth Amendments

I just saw that an interesting and  important constitutional procedure opinion was handed down by the Tenth Circuit last week in US v. Haymond, No. 16-5156 (10th Cir. Aug 31, 2017) (available here).  Here is how the panel's majority opinion gets started and some of the opinion's substantive analysis:

The district court revoked Andre Ralph Haymond’s supervised release based in part on a finding that Haymond knowingly possessed thirteen images of child pornography. The district court imposed the mandatory minimum sentence required by 18 U.S.C. § 3583(k). Haymond appeals and argues that the evidence was insufficient to support a finding by a preponderance of the evidence that he possessed child pornography, and that 18 U.S.C. § 3583(k) is unconstitutional because it violates his right to due process.

We conclude that the evidence was sufficient to support the district court’s finding that Haymond violated the conditions of his supervised release, but we agree that 18 U.S.C. § 3583(k) is unconstitutional because it strips the sentencing judge of discretion to impose punishment within the statutorily prescribed range, and it imposes heightened punishment on sex offenders based, not on their original crimes of conviction, but on new conduct for which they have not been convicted by a jury beyond a reasonable doubt. Thus, we affirm the district court’s revocation of Haymond’s supervised release, but we vacate Haymond’s sentence and remand for resentencing....

We conclude that 18 U.S.C. § 3583(k) violates the Fifth and Sixth Amendments because (1) it strips the sentencing judge of discretion to impose punishment within the statutorily prescribed range, and (2) it imposes heightened punishment on sex offenders expressly based, not on their original crimes of conviction, but on new conduct for which they have not been convicted by a jury beyond a reasonable doubt and for which they may be separately charged, convicted, and punished....

By requiring a mandatory term of reimprisonment, 18 U.S.C. § 3583(k) increases the minimum sentence to which a defendant may be subjected. For example, when Haymond was originally convicted by a jury, the sentencing judge was authorized to impose a term of imprisonment between zero and ten years.  See 18 U.S.C. § 2252(b)(2). After the judge found, by a preponderance of the evidence, however, that Haymond had violated a particular condition of his supervised release, the mandatory provision in § 3583(k) required that Haymond be sentenced to a term of reincarceration of at least five years, up to a maximum term of life. This unquestionably increased the mandatory minimum sentence of incarceration to which he was exposed from no years to five years, yet the jury did not make the factual finding required to change his statutorily prescribed sentencing range. Instead, that finding was made by a judge by only a preponderance of the evidence. This violates the Sixth Amendment....

In Johnson v. United States, 529 U.S. 694 (2000)

, the Supreme Court made clear that, in order to avoid serious constitutional concerns, revocation of supervised release must be viewed as punishment for the original crime of conviction, not as punishment for the violation of the conditions of supervised release....

Regardless of the nature or severity of the defendant’s original crime of conviction, § 3583(k) imposes a mandatory minimum five-year term of imprisonment for only those specific offenses enumerated, while all other violations are subject to the maximum terms set in § 3583(e)(3). By separating these crimes from other violations, § 3583(k) imposes a heightened penalty that must be viewed, at least in part, as punishment for the subsequent conduct — conduct for which the defendant has not been tried by a jury or found guilty beyond a reasonable doubt.  This, the Court has said, is not permitted. 

To be sure, the sentencing judge can and, according to the Sentencing Guidelines, should consider the severity of the conduct by which a defendant violated the conditions of his or her supervised release.  A more serious violation might well recommend a longer term of reimprisonment.  But, if we wish to maintain the premise that revocation of supervised release is a punishment for the original crime of conviction, Congress must set the authorized term of reimprisonment based on the severity of that original crime.

Notably, Judge Kelly dissents in part because he is (reasonably) concerned that the majority's reasoning might impact any and all judicial fact-finding supporting the revocation of supervised release:

Were the court correct [in its constitutional analysis], the problem it identifies seems like it would be true of all revocation proceedings: if a defendant is sentenced to any term of supervised release, the fact that the release can then be revoked and the defendant be sent back to prison for an additional term means that “the penalty to which a defendant may be subjected” has been increased based on facts not found by a jury. Id. (emphasis added).

In other words, unless either (a) all revocation proceedings must empanel juries for fact-finding (which the Supreme Court, with good reason, has told us is not the case) or (b) the revocation proceeding is treated as a new criminal prosecution (which the Supreme Court also has told us is not the case), it is hard to understand why under current precedent Booker would apply but Apprendi and Alleyne would not. While postrevocation penalties might be considered attributable to the original conviction, the revocation proceeding is neither part of that criminal prosecution nor is it a new criminal prosecution. See Johnson, 529 U.S. at 700....

[According to the majority], the distinction, apparently, is that the terms of revocation differ based on what kind of new crime the defendant committed. But I see no reason why Congress cannot make that distinction. As the Sentencing Guidelines explain, under the “breach of trust” theory applicable to the revocation of supervised release, “the nature of the conduct leading to the revocation [can] be considered in measuring the extent of the breach of trust.” U.S. Sentencing Guidelines Manual § 7A3(b) (2016). In my view, Congress can determine that the commission of certain crimes constitutes a more serious breach of trust warranting a longer term of revocation. Doing so does not thereby make the revocation proceeding a new criminal prosecution....

Ultimately, we should not jump ahead of the Supreme Court when it has already spoken on this issue. Any tension between the supervised release scheme approved in Johnson and the rationale of the Apprendi / Booker line of cases is for the Supreme Court itself to resolve.

Ever since the Supreme Court got serious about applying its Apprendi doctrine to various sentencing determinations in cases like Blakely and Booker, I have thought the judicial fact-finding that takes place in federal supervised release proceedings were on constitutionally shaky grounds.  Or, to parrot Judge Kelly's final statment, I have long believed that there is significant tension between the supervised release scheme approved in Johnson and the rationale of the Apprendi / Booker line of cases.  But, in various settings, various lower federal courts have found various ways to uphold the judicial fact-finding involved in supervised release revocations — revocations that result in a significant number of federal defendants getting sent back to prison.  (This 2010 USSC report found that roughly 1/3 of all released federal prisoners get revoked and sent back to prison, and that 6% of the federal prison population are serving revocation terms.)

It will be very interest to watch if the Justice Department seeks en banc or SCOTUS review of this Tenth Circuit ruling.  I hope they will, in part because this case seems like it might just get SCOTUS to finally take a look at what its modern Fifth and Sixth Amendment doctrines should mean for supervised release revocation proceedings.