Cornell Law Professor Michael Heise, Venderbilt Law Professor Nancy King, and University of Chicago Law student Nicole Heise recently published an article entitled Criminal State Appeals Revealed, 70 Vand. L. Rev. 1939 (2017). Their work offers a detailed accounting of factors which tend to increase [the] chance of successful criminal appeals.

The authors used state criminal appeals data collected in 2010 by the United States Department of Justice’s Bureau of Justice Statistics. The data was gathered from 51 courts of last resort and 92 intermediate appellate courts. The authors excluded data from proceedings involving juveniles, post-conviction relief, interlocutory matters, probation or parole revocation, and civil commitment. Capital appeals, government appeals, and misdemeanor appeals were also excluded from the data set. This left only appeals from felony-level criminal judgments. Then, the authors separately evaluated appeals of right before the intermediate courts and appeals which occurred before discretionary courts of last resort. Id. at 1944-1945. The data set, and the manner in which the authors broke that data apart, invites the drawing of parallels, and potentially the application of the study’s findings, to practice in the military justice system.

When determining whether an appeal was successful, “[the authors] defined a decision as favoring the defendant if it involved anything other than an affirmance, a dismissal, a denial of review, or a withdrawal.” Id. at 1948. They also captured a wide range of variables for each case, including such factors as whether the defendant retained civilian counsel for the appeal, the nature of the conviction (i.e., homicide, sex offenses, robbery, etc.), the nature of the appellate issues raised (i.e., challenging the charge, statute, venue, raising pretrial issues, challenging evidence suppression rulings, etc.), whether the defendant’s pleadings sought sentence review, whether oral argument was held, and whether reply briefs were filed. Id. at 1949-1952.

The overall rate of success on appeal was 15 percent before intermediate appellate courts where review was a matter of right. The rate of success before discretionary courts of last review was less than 3 percent. Id. at 1967.

Before the intermediate courts, challenging the sentence was favorably correlated with the defendant obtaining relief. However, before courts of last resort, it was better for the defendant to just challenge the underlying conviction. Id. at 1965-1966. The authors also found that hiring private counsel did not significantly increase the likelihood of success on appeal. In fact, before courts of last resort, the opposite was true – public defenders were far more likely to obtain relief than private counsel. Id. at 1966.

Before the intermediate courts, oral argument and the filing of reply briefs were positively correlated with defense success. Id. at 1968. Further, appeals from drug trafficking and sexual offense convictions were more likely to receive appellate relief than other crimes. Id. at 1968. The particular types of claims most likely to yield relief for defendants were: challenging the sufficiency of the evidence, challenging the appropriateness of the sentence, and raising “other trial errors” (a broad category of post-arraignment issues including challenges to competency, mistrial, cumulative error, motions for findings of not guilty, and motions for new trials). Id. at 1968.

In cases which made it to the discretionary court of last appeal, and where the defendant’s request for review was granted, the factors most correlated with defense success were: appealing a sex offense conviction, filing “other trial errors” claims, and being represented by public counsel. Id. at 1964-1965.

This blog has traditionally compiled end of term data regarding the Court of Appeals for the Armed Forces. It is tempting to try to use that compilation to draw parallels to the findings reached by Heise, King and Heise. Or, at least their findings as to courts of last resort. But, that would be folly. Completing a comparable analysis would require the use of a data set as detailed as those authors used. Their data was collected, in excruciating detail, and no doubt at great cost, by the Department of Justice.

No such data exists for the military appellate jurisdiction. Or (perhaps more accurately) if it does, it has not been compiled, packaged, and made available for public use. Indeed, even the public’s access to such basic information as the service courts of criminal appeals’ decisions can be, at times, spotty. Assessing deeper factors, such as the sorts of claims made by military appellants in order to ascertain the rate of relief for those claims, would require access to pleadings and data from a military justice system that has long been criticized as opaque.

Nonetheless, it is likely that Heise, King and Heise’s findings, which are generally and statistically applicable throughout the civilian jurisdiction, are at least relevant to military appellate practice. Certainly their work warrants the attention of military justice practitioners.


2 Responses to “Scholarship Saturday: Successful appeals by the numbers”

  1. Shawn says:

    Your post is interesting, Isaac; however the article itself was not.  By no means did it offer “a detailed accounting of factors which tend to increase the the [sic] chance of successful criminal appeals.”  Nor, by any stretch, are its results “generally and statistically applicable throughout the civilian jurisdiction.”
    Using regression — or any other statistical technique — in this way is basically torturing the data until it confesses.  It is like using the roster of TV programs across all channels to predict the weather.  You might find correlation but cannot possibly find meaningful causation; not because it isn’t there, but because it’s the wrong way to look.
    The article makes no mention of confidence levels, for example, or whether the docket of appeals afforded a large enough sample size to achieve sufficient power.  More broadly, it took a classical (frequentist) approach to what is clearly a Baysian situation where an appeal passes through various stages of a process.
    You know the proverb about a man who is his own lawyer?  It’s also true of a lawyer who is his own statistician.  I would love to play poker with these guys.

  2. John Marshall says:

    I have always thought it a bit misleading when CAAFLog compares the success of the military appellate defense divisions to “civilian counsel,” but does not appear to account for the fact that military counsel assist civilian counsel in almost every case.