State v. Anthony James Jendusa, 2021 WI 24, affirming a court of appeals order denying interlocutory appeal; case activity (including briefs)

This litigation has been procedurally weird, as we’ve discussed before, but SCOW’s decision on the merits may turn out to be a momentous one for the future of ch. 980.

Commitment trials commonly turn on the application of tools called actuarial instruments. These are tests of a sort: an examiner gathers a certain set of facts about the respondent. The examiner then compares a “score” derived from those facts with the scores of various groups of other ex-offenders released in the past. The examiner relies on the rates of reoffense actually observed in those groups to inform predictions about whether the respondent is more likely than not to commit another sexually violent offense; this is one of the three criteria for indefinite commitment under the law.

So, for example, a very widely used actuarial instrument called the Static-99 (which appears with slightly different names reflecting revisions over the past two decades) will generate a raw score, say 6. The examiner will then look to a table showing observed recidivism rates for groups of people who also scored a 6 on the instrument. Crucially, though, there are different groups of people available for this comparison, and the recidivism rates observed in these groups vary widely. The original comparison groups used to develop the Static-99 consisted of about 1,000 people released between the late 1950s and the 1990s in UK and Canada; these groups had much higher rates of reoffense than have been observed in many more recent studies in various jurisdictions in this country.

But what about this jurisdiction? It turns out that the Wisconsin Department of Corrections has been maintaining a database of those inmates it had evaluated, pre-release, for possible commitment under ch. 980. As many of those inmates did not end up being committed, this database has the potential to establish a recent, Wisconsin-based sample of recidivism rates, both for released people as a whole and maybe for subsets of those people with various of the characteristics actuarial instruments consider.

We know about this database through the testimony of the state’s expert witness at the probable cause hearing in this case, a DOC evaluator named Christopher Tyre. On hearing of the database, Jendusa’s lawyer sought to get access to it–naturally, since Dr. Tyre also testified that his preliminary review of the data showed an overall reoffense rate of seven percent: just one-third of the reoffense rate he’d relied on to conclude that Jendusa was more likely than not to reoffend.

Jendusa was thwarted in his attempts–but not by the courts. He sought the data from the DOC itself under an administrative process. DOC initially agreed to provide it to him before backtracking and eventually ceasing communication. He subpoenaed Dr. Tyre to bring the database with him to a motion hearing; the DOC didn’t move to quash the subpoena but instead advised Dr. Tyre not to comply based on its (dubious) interpretation of federal privacy regulations. When the circuit court specifically ordered Dr. Tyre to bring the database, he again showed up without it, again saying DOC’s counsel had advised him to do this. (¶¶12-13).

So, finally, the circuit court ordered DOC itself to provide the database. The state petitioned the court of appeals for interlocutory review of that decision, which the court of appeals denied. As our prior post noted, the supreme court, against much prior precedent, granted review of that decision.

Four justices now hold that the court of appeals did not err in denying the interlocutory appeal. Preliminarily, the majority approves the lower court’s usual practice, when it denies such requests, of doing so without comment. The state (and the dissent here) argued that the court of appeals ought to give reasons for its decisions in order to demonstrate a valid exercise of discretion; the majority says that they (ordinarily) don’t review such denials anyway and, when they do, they do so for reasons unrelated to the lower court’s discretionary call, so an explanation is unnecessary. (¶¶21-22).

That would be enough to dispose of the appeal: if the court of appeals’ denial of interlocutory appeal was valid, then the circuit court’s ruling that the DOC must turn over the data must be effectuated. (Though perhaps we’re being naive; see the history above.) Nevertheless the majority proceeds to the merits, holding that the circuit court was correct to order the data turned over.

Jendusa points to various provisions, statutory and constitutional, requiring the discovery of the database, but he majority here is content to hang its hat on one: Wis. Stat. § 980.036(5), which provides that on a party’s motion, the court may order production of “any item of evidence or raw data that is intended to be introduced at the trial for testing or analysis under such terms and conditions as the court prescribes.” The database contains data, Jendusa wants to use it at trial, and it’s relevant, as it provides grounds to argue he’s less likely to recidivate than the state says. So, the majority says, the circuit court was right to say he gets it. (¶¶25-33).

The state put forth various theories about why it shouldn’t have to turn over the database. It claimed that it doesn’t “possess” the database: the database is actually possessed by the DOC; i.e. the state agency whose authorization is required to commence a ch. 980 commitment proceeding, the state agency that provided the state’s sole witness at the probable cause hearing, and the state agency that generated the report urging the circuit court to begin the proceeding and commit Jendusa to Sand Ridge. It also argued the data in the database would not be “introduced at trial” since it would not be meaningful without analysis. The majority points out that this reading would render the statute nonsensical, since it also contemplates a party’s “testing or analysis” of that data. (¶30). Three justices (Ziegler, R.W. Bradley, and Roggensack) dissent, saying this interpretation is, in fact, the correct one.

Facebooktwitterlinkedinmail