Last week, the Wisconsin Supreme Court threw up its hands in a case it had agreed to adjudicate. At issue were what consequences attend a criminal complainant’s refusal to waive the privilege of keeping personal mental health treatment records private. See State v. Lynch, 2016 WI 66. This echoed the last time the Court considered the same issue. See State v. Johnson, 2014 WI 16 (per curiam), reconsidering 2013 WI 59 (per curiam). Both Lynch and Johnson presented thorny issues of competing constitutional and statutory rights. And it is little surprise that those issues would divide the Court. But the substantive issues aside, the Court’s method of handling those divisions, in Johnson and now differently in Lynch, raises far more questions than it answers.

As a threshold matter, it is striking that the Court has twice been unable to reach a decision on this issue when both times an odd number of justices participated. Courts usually comprise odd numbers by design, to avoid tie outcomes. And yet, though five justices participated in Johnson and all seven justices participated in Lynch, both times the Court split in paralyzing ways and left the court of appeals decision in effect. Typically, that undesirable outcome results only when recusals or vacancies yield an even number of justices participating, with a tie vote among them. But neither Johnson nor Lynch involved a tie vote. A separate opinion in Johnson addressed this distinction, noting that the Court “fail[ed] to explain how an odd number of justices can be deadlocked” and that the outcome in that case was “not like the three-three split cases cited” as authority. 2014 WI 16, ¶16 (Bradley, J., concurring in part and dissenting in part). None of the five separate opinions in Lynch addresses the impossibility—in mathematical or legal terms—of an even split among seven justices.

Substantively at issue in Lynch, and in Johnson before, was the validity of State v. Shiffra, 175 Wis. 2d 600, 499 N.W.2d 719 (Ct. App. 1993), modified by State v. Green, 2002 WI 68, and cases that follow it. Shiffra created a process under which a criminal defendant in Wisconsin can, by demonstrating a reasonable likelihood that a complainant’s mental health treatment records contain relevant information to the determination of the defendant’s guilt or innocence, review those records. Because Wisconsin law shields such records from disclosure without the patient’s consent, see Wis. Stat. §905.04(2), Shiffra also contemplates what happens if the defendant makes the required showing and the complainant refuses to consent to review of the records. In such circumstances, Wisconsin courts have precluded the complainant from testifying at the criminal trial. That is precisely what occurred in Lynch, leading the State to appeal.

In Lynch, the State asked the Supreme Court to decide three questions of law:

  1. Whether Shiffra is based on an improper extrapolation from the constitutional analysis in Pennsylvania v. Ritchie, 480 U.S. 39 (1987), such that Shiffra should be overruled.
  2. Whether, assuming Shiffra is legally sound, preclusion of the complainant’s testimony is the only remedy available to a court seeking to ensure the defendant receives a fair trial when the complainant denies access to mental health treatment records.
  3. Whether a trial court may rely on Wis. Stat. §146.832(2)(a)4. to order production of privileged mental health treatment records to the defendant notwithstanding the complainant’s refusal to consent.

In response to those questions, the Court generated five separate opinions totaling 135 pages.[1] This table shows how the justices line up on each of the questions presented.

As the table shows, four justices espoused the belief that Shiffra should not be overruled. And a majority of the justices who considered the question concluded that the Shiffra process should allow a remedy other than excluding the complainant’s trial testimony. The only issue that evenly divided the justices to consider it was the third question.

This raises a handful of procedural questions about the Court’s operations.

First, what does the term “lead opinion” mean when the Court uses it, and how does an opinion so denominated differ from other opinions issued by the Court?

Justices Abrahamson and A.W. Bradley addressed this issue in their opinion, see Lynch, 2016 WI 66, ¶¶123-46. They also addressed this issue in another opinion released last week. See Office of Lawyer Regulation v. Riley, 2016 WI 70, ¶¶92-95 (Abrahamson, J., concurring). In addition to Lynch and Riley, they identified six other opinions from this term with “lead opinions” joined by fewer than a majority of the participating justices. See Lynch, 2016 WI 66, ¶141 (citing cases). In Lynch, if four justices agreed that Shiffra and its progeny are valid and the Court’s mandate affirmed the application of Shiffra below, by what criteria did the Court decide to affix the term “lead opinion” to the contrary view?

Second, how did the lead opinion’s conclusion that Shiffra should be overruled justify not considering the other two questions presented for review?

Courts usually do not—and should not—reach legal questions unnecessary to resolving the case under consideration. But continued application of Shiffra, either through affirmance or reliance on the decision below, leads directly to the next question—the scope of remedies available to a circuit court when the defendant makes the requisite showing and the complainant refuses to consent to review of his/her mental health treatment records. The Court offers no clear answer to that question because the lead opinion did not “address the second and third issues presented for review.” 2016 WI 66, ¶8. To support this approach, the lead opinion noted that, when the Court’s “resolution of one issue disposes of a case, [the Court] will not address additional issues.” Id., n.10. But, given that the lead opinion’s conclusion on question 1 had only minority support and is in tension with the decision below, how could it dispose of the case or obviate the need to examine question 2?

Third, given the complexity of the issues often presented to the Court and the apparent fractiousness of the current Court, is it time to consider reversing the prohibition on minority vote pooling?

The Court has defined “minority vote pooling” as “the approach that even when there is no majority as to a particular ground for reversal, reversal is still required if the minorities unite and constitute a majority of the court.” State v. Gustafson, 121 Wis. 2d 459, 461, 359 N.W.2d 920 (1985) (per curiam), reconsidering 119 Wis. 2d 676, 350 N.W.2d 653 (1984). Gustafson rejected minority vote pooling “to avoid [] arbitrary and illogical results.” Id. at 462. But this desire to avoid hypothetically absurd results has created concrete oddities. In Gustafson, though four justices deemed Gustafson’s criminal conviction legally improper, it was affirmed “in conformity with the wishes of a minority of three.” Id. at 464 (Abrahamson, J., dissenting). And in Lynch the Gustafson rule led to the Court—again—not deciding a complex and important question of criminal procedure, despite four votes to reaffirm Shiffra and likely a majority to expand the circuit court’s remedial options when a complainant denies access to mental health treatment records. In light of the facts that only one justice who participated in Gustafson remains on the Court and that fewer than a handful of courts nationwide follow the Gustafson approach, see id. at 461 n.2 (citing cases), should the Court reexamine its approach to counting votes?

Fourth, if the Court continues to adhere to the prohibition on minority vote pooling, what will that mean going forward?

As discussed above, the Court has been unable to reach majority agreement in a significant percentage of cases decided this term. Does a rejection of vote pooling cast all such cases into doubt? Why is pooling votes no more problematic where the Court splinters 3-1-3—with the 3 votes in the lead opinion and the single vote in the concurrence forming a majority bloc for the mandate—than where the Court splinters 3-2-2—such that the 2 votes in each of the dissents would combine into a majority bloc to overcome the view of the lead opinion? And if vote pooling is problematic across the board, how will the Court proceed, given the frequency with which the Court seems unable or unwilling to build consensus among a majority?

Finally, did the Court need to—and was it right to—publish 135 pages of opinions in Lynch?

Set aside the previous questions about how the Court counts votes and accept at face value the Court’s assertion just above the lead opinion in Lynch that it deadlocked such that “the law remains as the court of appeals has articulated it.” 2016 WI 66. That assertion suggests that all five opinions in Lynch are unnecessary to the resolution of the case. Indeed, Justice Ziegler derided them as “a collection of several law review articles by members of this court.” Id., ¶231. Why does it not follow that all five writings are advisory opinions? In similar circumstances, the U.S. Supreme Court, for example, issues one-sentence orders noting only that an equal division among the Justices precludes a decision on the merits. See, e.g., United States v. Texas, 579 U.S. ___ (June 23, 2016) (per curiam). In light of established Wisconsin law that the Supreme Court “declines to render advisory opinions,” Smith v. Town of Pershing, 10 Wis. 2d 352, 357, 102 N.W.2d 765 (1960), why isn’t anything beyond such a short, per curiam order improper in a case like Lynch?


[1] Here is the breakdown of the opinions:

  • Justice Gableman wrote an opinion (joined by Chief Justice Roggensack and Justice R.G. Bradley) concluding that Shiffra should be overruled. Though this opinion represented a minority view on question 1 and did not address questions 2 and 3, the Court labeled this the lead opinion.
  • Chief Justice Roggensack wrote separately, criticizing the tone of the separate opinion written by Justices Abrahamson and A.W. Bradley. The Court labeled this a concurrence.
  • Justices Abrahamson and A.W. Bradley wrote jointly, concluding that Shiffra is valid, that exclusion is not the sole remedy, and that circuit courts can order disclosure of a complainant’s mental health treatment records without consent. This opinion also criticizes the Court’s decision to label a “lead opinion.” The Court labeled this a concurrence in part and a dissent in part.
  • Justice Prosser wrote separately, concluding that Shiffra is valid, that exclusion should not be the sole remedy, and that circuit courts lack authority to order disclosure of a complainant’s mental health treatment records without consent. The Court labeled this a dissent.
  • Justice Ziegler wrote separately, concluding that Shiffra is valid, that exclusion is the sole remedy, and that circuit courts lack authority to order disclosure of a complainant’s mental health treatment records without consent. Though only Justice Ziegler’s opinion agrees in full with the court of appeals decision—which the Court’s mandate affirmed—the Court labeled this a dissent.

Find a Professional