Comparing Wisconsin’s Supreme Court with the Courts of Other States: Felony Appeals, Part 2

Last month’s post made use of a forthcoming article (“State Criminal Appeals Revealed”) in the Vanderbilt Law Review, which presented findings from a national dataset on criminal appeals resolved in 2010 by state supreme courts and lower appellate courts throughout the nation.  Regarding state supreme courts, the authors (Michael Heise, Nancy J. King, and Nicole Heise) extracted a representative sample of 1,425 cases in which defendants sought review.  From these 1,425 cases, review was granted in 89, and it was this subset of 89 cases that provided a national “standard” or “average” against which we measured decisions filed by the Wisconsin Supreme Court in cases that met the criteria specified by the authors of the law review article: no capital cases, no appeals filed by the State rather than by defendants, and no misdemeanors.

After directing our attention last month to the frequency with which defendants received favorable decisions from the Wisconsin Supreme Court compared to state supreme courts nationwide, we now take a closer look at the data by dividing the cases into several categories of crimes.  The crimes in question here are the “original” or underlying crimes alleged to have been committed—as distinct from the issues before the supreme courts (which might center on any number of other things, such as the admissibility of evidence, ineffective assistance of counsel, or the jury-selection process).  From 2004-05 through 2016-17 the Wisconsin Supreme Court decided 163 cases fitting the criteria described above, and, in Table 1, I have divided these cases into the categories of crimes selected by the authors of the law review article.[1]  (The year 2010 did not yield enough decisions in Wisconsin to populate several categories in the table, so I drew Wisconsin’s decisions from the same 13-year interval covered in the previous post.)

Opinions may vary on this score, but I am not overwhelmed by the differences when comparing the column for the nationwide data and the column for Wisconsin.  The largest divergences appear to involve, on the one hand, drug crimes and robbery (a higher percentage in state supreme courts collectively) and, on the other hand, homicide and public order (a higher percentage in Wisconsin).  Perhaps one or another of these differences suggests something significant that has eluded me, but my caution stems not only from limited perception, but also from the fact that the national data pertain to a single year.  Although these data probably would not fluctuate from year to year as much as they do in a single state, I would have more confidence drawing conclusions from a sample that spanned a longer period of time.

Now that we have Wisconsin’s 163 cases divided into categories, we can calculate how frequently defendants received favorable decisions in cases associated with each of the underlying types of offenses.[2]  As displayed in Table 2, Wisconsin defendants had the best odds in cases where the underlying crimes were sex offenses, as well as the more diffuse “public order” and “other” categories.  It may be of interest to note here that 80% (28/35) of the “sex offense” cases involved child victims (as did eight of the nine decisions favorable to defendants), and that OWI charges figured in 45% (9/20) of the “public order” cases (and in four of the seven decisions favoring defendants).[3]

I have included in Table 2 a column for data from the nationwide sample, though it is of limited use at this level of detail.  The number of cases in the national sample is so small (89 decisions filed by state supreme courts in 2010) that the statistical power of the entries in any individual category must be rather low.  The authors of the law review article readily acknowledge this point, which is probably why no figures are provided in four of the categories.  It would be interesting to learn if a larger sample from state supreme courts around the country would continue to yield percentages of decisions favorable to defendants that are astonishingly high compared to outcomes in Wisconsin alone.

This difference between Wisconsin and the nationwide sample is all the more remarkable in light of another question raised by the law review article: Are decisions more likely to favor defendants in states where the justices are elected, or in states where they are appointed?  The data for Wisconsin suggest that favorable rates would be lower in states with elected justices, but the article reached the opposite conclusion.  For the nation as a whole, state supreme courts with elected justices were much more inclined to issue decisions favoring defendants—55% of the time, compared to 39% for justices in other states.  If we are prepared to assume that the article’s data are at all close to representative, Wisconsin emerges as an even more remote outlier.


[1] The table in the law review article includes an entry for “unknown felony” (10.1% of decisions) and an entry for “court order violation” (1.1% of decisions).  Among the 163 Wisconsin cases, none of the felonies were “unknown,” but there were a handful of cases in which the principal offense did not fit in any of the specified categories, and thus I created an “other” category for these cases.  I also combined the two categories mentioned above from the law review article—“unknown felony” and “court order violation”—and put them in Table 1’s “other” category, which explains the figure of 11.2% in that column (10.1 + 1.1).  The 14 Wisconsin decisions in the “other” category include various forms of “theft” (as opposed to “robbery” or “burglary”), such as “theft as a trustee/bailee” and “identity theft,” as well as “taking and driving a vehicle without the owner’s consent,” “felony misconduct in public office,” and a few instances of non-sexual child abuse or neglect.

Here are the most important guidelines that I followed for classifying cases.

(1) If a case included charges in more than one category, I identified the most serious of the alleged offenses—as determined by such things as the felony class of each charge and the length of sentence received for each offense.  One offense was nearly always more serious than the others, and it determined the category of classification.

(2) As one would expect, homicides (including attempted homicides) trumped all other categories for classification purposes.  OWI homicides, for instance, have been filed under “homicide” rather than “public order” (which encompasses “driving under the influence” cases).

(3) If a defendant was initially charged with an offense in one category, but later pleaded guilty to an offense in a difference category, I classified the case in the second category.

(4) The “assault” category includes battery crimes and a few other cases that involved an intentional, physical attack on another person (generally, cases with charges that included the words “recklessly endangering safety” or “reckless injury”).

(5) “Public order” crimes range across many offenses, including possession of weapons, drunk driving, commercialized vice, morals and decency transgressions, liquor-law violations, stalking, bail jumping, and fleeing/eluding a police officer.

(6) I omitted a couple of cases that were difficult to categorize (when, for example, no single charge was clearly more serious than the others).

[2] As in last month’s post, I have sought to preserve consistency with the statistics presented by the authors of the law review article by employing their method of determining whether a decision was “favorable” for the defendant: “We defined a decision as favoring the defendant if it involved anything other than an affirmance, a dismissal, a denial of review, or a withdrawal. It is worth noting that our coding convention captures many decisions that a defendant might not necessarily consider to be a ‘win,’ including remands and modest modifications of one of several sentences. The data do not offer a reliable method to distinguish significant modifications or remands from less meaningful ones, and the approach we take comports with prior empirical work examining appeals.”

[3] In three of the 28 “child” cases, the charge was possession of child pornography rather than the much more common “sexual assault of a child.”

About Alan Ball

Alan Ball is a Professor of History at Marquette University in Milwaukee, WI.

SCOWstats offers numerical analysis of the voting by Wisconsin Supreme Court justices on diverse issues over the past 91 years.

Speak Your Mind