Thursday, December 14, 2017

Making a Murderer Postscript: The Perversion of Henry Friendly's Innocence Concern

by Michael Dorf

In 1970, the University of Chicago Law Review published an article titled Is Innocence Irrelevant? Collateral Attack on Criminal Judgments by federal appeals court judge Henry Friendly. Judge Friendly was a judicial conservative in the small-c sense, non-ideological, committed to deciding cases narrowly, and an expert legal craftsman. As a young lawyer, Chief Justice John Roberts clerked for Friendly during Friendly's later years, and Roberts is fond of quoting (though not always abiding by) Friendly's aphorism that if it is not necessary to decide an issue to decide a case it is necessary not to decide the issue.

Is Innocence Irrelevant? was somewhat uncharacteristic of Friendly in that it offered a controversial policy proposal on a politically contentious issue. Writing in a period of transition from the Warren Court to the Burger Court, Friendly lamented that federal habeas corpus petitions by prisoners sentenced under state law were too often succeeding based on procedural irregularities that had no connection to innocence. To use the more nakedly political argot, federal courts were letting guilty state prisoners off on technicalities. Quoting Justice Hugo Black's dissent in a then-recently-decided case, Judge Friendly offered what he regarded--and what many still regard--as a self-evidently sensible proposition: "the defendant's guilt or innocence is at least one of the vital considerations in determining whether collateral relief should be available to a convicted defendant."

The ensuing nearly five decades have proven Judge Friendly prophetic--but probably not in a way that he would have approved. In 1976, the Supreme Court held that habeas corpus would not be available at all for petitioners claiming that otherwise-reliable evidence obtained in violation of the Fourth Amendment was used to convict them. The next year, the Court would make it considerably harder for petitioners who had failed to raise their objections in compliance with state court rules to obtain relief in federal court on otherwise meritorious claims. Other judicial narrowings followed and then, in 1996, Congress passed and President Bill Clinton signed the Antiterrorism and Effective Death Penalty Act (AEDPA), which further limited habeas corpus.

The Court decisions and AEDPA have made it much more difficult for prisoners without plausible constitutional claims that bear on innocence to obtain relief via habeas corpus. To my mind, that is understandable if not ideal: understandable because innocents serving prison sentences or awaiting execution suffer a much graver injustice than guilty parties whose proceedings were tainted by constitutional error; not ideal because habeas review once served, but no longer serves, as a means of ensuring that state court judges under political pressure to be tough on crime give full effect to the constitutional rights of criminal defendants.

But even if one thinks that Congress and the courts were right to cut back on habeas in cases where prisoners raise claims that speak only to the fairness of the proceedings, not to guilt or innocence, there is cause for alarm. Modern habeas law honors only half of Judge Friendly's agenda. It makes the bringing of habeas petitions by guilty defendants considerably harder than in the Warren Court era. But it also makes it extremely difficult for the innocent to obtain habeas relief. That proposition was on full display late last week in an en banc ruling by the US Court of Appeals for the Seventh Circuit.

Spoiler Alert: I will now discuss a case that figures in the Netflix documentary series Making a Murderer. If you intend to watch it but have not yet done so, you might want to bookmark the column and come back here after viewing.

For those readers who did not (and do not intend to) watch or have forgotten the basic story of Making a Murderer, it goes like this: (1) From 1985 to 2003, Steven Avery of Manitowoc County, Wisconsin, served a prison sentence for a sexual assault he did not commit; (2) after he was exonerated, Avery sued the county and various police officials who were, the documentary indicates, at least grossly negligent in the handling of his case; (3) while the civil suit was pending, Avery was arrested for the murder of photographer Teresa Halbach; (4) the case against Avery was largely circumstantial, including evidence that appeared to result from police tampering and other improper procedure; (5) that said, the police sometimes try to frame guilty people, and I came away from Making a Murderer thinking that Avery might be guilty of murdering Halbach, even if the evidence was thin; (6) the thinnest evidence was the confession of Avery's nephew Brendan Dassey, who is close to intellectually disabled, who was interviewed by police in an extremely suggestive manner, and who told a story that was internally inconsistent and did not match the physical evidence on key points; and (7) if the conviction of Avery appears dubious--I think the jury ought to have found reasonable doubt but at least he might be guilty--the conviction of Dassey as an accomplice in the murder seems like a grave injustice, because Dassey is very likely innocent.

Much of the Dassey-focused portion of Making a Murderer illustrates how Dassey's first lawyer did a terrible job. The lawyer seems to have concluded that Dassey was guilty without even meeting with him and without ever considering the possibility that he might be innocent, gave damning statements to the press, and hired a private investigator who was more or less working for the police. However, the key evidence against Dassey was the video of an interview before his ineffective lawyer began representing him. On the basis of Dassey's recorded confession, the jury convicted him of murder, and his state court appeals failed. Despite the very high burden, Dassey obtained habeas relief from a federal district judge, who concluded that the confession was involuntary. A panel of the Seventh Circuit affirmed, but last week the en banc Seventh Circuit reversed by a vote of 4-3.

There is a difference between a documentary film and a full trial record, so I read the majority opinion with an open mind, expecting to learn that Making a Murderer had perhaps left out key details of the case against Dassey. To my surprise, I found none. Indeed, to the contrary, Judge Hamilton's opinion for the court dutifully recites the inconsistencies in Dassey's story, his seeming not to realize the nature of the interrogation (indicating he was eager most of all to get back to class even after he had confessed to a murder), the language used by the officers to indicate that if only Dassey told them what they wanted to hear he would be free, and . . . nothing else. Although the written account in the en banc opinion lacks the full drama of Making a Murderer, like the documentary, the opinion paints a picture of the events that strongly suggests that Dassey is innocent.

So why does the majority allow to stand a conviction based on a confession that is at best of questionable reliability when there is no other real evidence tying Dassey to the crime? Because Dassey's likely innocence is, in a word, irrelevant.

Unlike the Fourth Amendment, which protects values like privacy and property, which are not linked to a defendant's guilt or innocence of the crime for which he is accused, the Fifth Amendment right against a coerced confession is linked to guilt or innocence. That's one reason why, in a 1993 case, the Supreme Court refused to extend the no-Fourth-Amendment-exclusionary-rule-claims-on-habeas rule to habeas petitions based on a claimed Miranda violation. The Miranda warnings serve to mitigate the inherent coerciveness of custodial interrogation with the aim of preventing coerced confessions, and coerced confessions are unreliable evidence. Strong evidence of guilt might be obtained in violation of the Fourth Amendment, but evidence obtained via a coerced confession is not strong evidence of guilt because the coercion, rather than the suspect's conscience, will have been the basis for the confession.

Yet even though petitioners can bring Fifth Amendment claims on habeas, the restrictive rules that were adopted by  the Court beginning in the early 1970s and then tightened further by Congress in AEDPA make it difficult to prevail on Fifth Amendment claims, just as they make it difficult to prevail on more "technical" claims that do not correlate with guilt or innocence. The Court and Congress heeded Judge Friendly's call to make habeas relief much more difficult for guilty prisoners to obtain, but in so doing they threw the baby out with the bathwater by also making habeas relief much more difficult for innocent prisoners to obtain.

Thus, Judge Hamilton's en banc opinion notes that AEDPA sets a high standard for relief and that Dassey hasn't met that standard. End of story.

Is that right? Not necessarily. Here's how Judge Wood begins her dissent:
Psychological coercion, questions to which the police furnished the answers, and ghoulish games of ”20 Questions,” in which Brendan Dassey guessed over and over again before he landed on the “correct” story (i.e., the one the police wanted), led to the “confession” that furnished the only serious evidence supporting his murder conviction in the Wisconsin courts. Turning a blind eye to these glaring faults, the en banc majority has decided to deny Dassey’s petition for a writ of habeas corpus. They justify this travesty of justice as something compelled by AEDPA. If the writ, as limited by AEDPA, were nothing more than a dead letter, perhaps they would be correct. But it is not. Instead, as the Supreme Court wrote in Harrington v. Richter, “[t]he writ of habeas corpus stands as a safeguard against imprisonment of those held in violation of the law.” It is, the Court went on to say, “a guard against extreme malfunctions in the state criminal justice systems.” 
As the district court and the panel majority recognized, we have before us just such an extreme malfunction. Dassey at the relevant time was 16 years old and had an IQ in the low 80s. His confession was coerced, and thus it should not have been admitted into evidence. And even if we were to overlook the coercion, the confession is so riddled with input from the police that its use violates due process. Dassey will spend the rest of his life in prison because of the injustice this court has decided to leave unredressed. I respectfully dissent.
Kudos to Judge Wood for trying to make lemonade from a lemon, in particular Harrington v. Richter, in which the rhetoric she quotes begins an opinion that goes on to deny relief to the habeas petitioner. Kudos as well for strongly suggesting--even without exactly saying--that Dassey should be granted relief because he is probably innocent. Judge Wood only just barely failed. The case was 4-3, after all. But Dassey would have had a better chance of obtaining habeas relief if the full Friendly program had been implemented and innocence were made an explicit basis for placing a thumb, or better yet, an entire arm, on the scale in favor of relief.

Is this the end of the habeas road for Dassey? Technically not. He could file a cert petition with the Supreme Court, but his prospects there are doubtful. Ex ante, one would have predicted a better chance of success before the Seventh Circuit. Moreover, Dassey's case is mostly about the application of law to facts and evidence, and thus not obviously cert-worthy on any issue of larger importance. Sure, it presents the larger question whether, as Friendly asked, innocence is irrelevant, but the Supreme Court doesn't seem much interested in that question.

There remains the desperate possibility that Dassey could obtain federal habeas relief by filing an "actual innocence" claim. The Supreme Court suggested the possibility of such a claim in Herrera v. Collins in 1993, but it is not clear that actual innocence is a basis for relief from a sentence of life imprisonment rather than only from a death sentence. And even if so, the Court has set an almost impossibly high standard for relief based on actual innocence. In the one case in which a habeas petitioner obtained any relief from the SCOTUS on a Herrera claim, the Court said his lawyers needed to go back to the district court and produce "evidence that could not have been obtained at the time of trial [that] clearly establishes petitioner’s innocence." That petitioner, Troy Davis, was subsequently found not to have produced such evidence and was executed.

Dassey could not, in any event, take advantage of the Herrera/Davis opening, even if it were broader than a pinhole, because he is not adducing new evidence. Dassey's lawyers claim that the evidence that was adduced at trial should not have been deemed adequate to convict him because his confession was coerced and thus unreliable.

All that is left for Dassey, it seems, is the hope that lightning strikes twice--that once again someone wholly unconnected to Steven Avery is shown to have committed the crime for which Avery (and this time Dassey as well) was convicted.

2 comments:

Asher Steinberg said...

You should write more fed courts stuff here. Two things. One, I don't think this post ever explains exactly (or even inexactly) what AEDPA does or why it precluded habeas relief in this case. People who don't know would probably find it more instructive if you addressed this.

Two, I imagine the purpose of 2254(d) had little to do with Friendly, but rather went something like this: circuit courts that didn't like the death penalty or other aspects of the state criminal justice system were making up new criminal-procedural rights in order to set aside death sentences or long prison terms, and it was very difficult for the Supreme Court to correct all the errors, given how many there were, the lack of cert-worthiness in most such cases, and the inadequacy of summary reversal, especially where the errors are novel enough as to not be obvious enough for summary reversal. Mandatory Supreme Court jurisdiction in habeas or even in death habeas wasn't a realistic option, so Congress arrived at the alternative of freezing the law, for purposes of habeas review of state-court convictions, at whatever the Supreme Court had said in the past; that way, lower courts can't make up new bad law, and when they do, error correction will be much easier and more fit for summary reversal because the question is whether they relied on clear Supreme Court precedent, not the substantive correctness of the lower court rule. I wasn't paying attention to what the Sixth or Ninth Circuits were doing in habeas in the early '90s or '80s, but if it's anything like what they've done since AEDPA, I can understand what Congress was thinking. Of course, there's a lot of good law that remains to be made that the Court hasn't made yet, but I think Congress could (a) reasonably doubt how much new criminal-procedure law remained to be made, and (b) think that, given state appellate review, direct Supreme Court review, state habeas, comity, and all the bad law that lower courts were making in pre-AEDPA habeas, they'd accept a certain amount of wrongly denied relief in return for eliminating a lot of wrongly granted relief.

Michael C. Dorf said...

On point 1, thanks and I'll try, but it's hard to explain even in Fed Courts! I'm working on an article on the Holy Grail of Fed Courts: Jurisdiction stripping. I'll say more about it when I have a draft ready to share, in about a month, when the footnotes are cleaned up. Exciting, I know!

On point 2, I think two things:
a) For people who were acting in good faith and supported the changes, I think that gets about half of it, but there was also a concern that lawyers were gaming the system. The charge of "sandbagging" that Rehnquist made in Wainwright v. Sykes is an example of this. So was the addition of the one-year Statute of Limitations in AEDPA. I think the latter was arguably a legit concern w/r/t some death-sentenced prisoners, who benefited from delay. But the approach was misguided with respect to the people serving prison sentences, who therefore had every incentive to file their best claims asap.

b) Even as to the new rule point (both in Teague v. Lane and in AEDPA), the Court and Congress defined new rules more broadly than needed to achieve the goal.

All that said, I do think a fair compromise would be to distinguish between innocence-related claims (such as coerced confession claims) and non-innocence-related claims.