The Great Writ

This last historical post is more consistent with the types of habeas issues that have predominated in this blog over time – a use of habeas to challenge a criminal conviction. It was first published on July 28.

The Second Circuit Court of Appeals issued an interesting habeas corpus opinion on Monday in Hernandez v. McIntosh. The posture of the case is bizarre insofar as the—in my opinion—obvious investigatory misconduct of the police was ignored, except to the extent that a judicial response to a jury question rendered this conduct important. In many ways, this illustrates the cascading effect of poor judicial decisions coupled with “impossible to meet” standards of review. In the end, a defendant with multiple mental disorders and an IQ between 67 and 76 was convicted of a murder that took place 18 years earlier with no physical evidence or contemporaneous witness testimony—only a confession obtained after seven hours of interrogation at the police station without the benefit of Miranda warnings. Amazingly, none of this mattered. What did matter was the answer provided by the trial judge to a question posed by the jury. 

The jury asked: “If we find that the confession … before the Miranda rights was not voluntary, [must we] disregard the two later videotaped confessions …?”The trial court simply said: no. The District Court, in habeas, concluded that this answer was correct because “the jury was not required to disregard the subsequent confessions.” They could disregard the later confessions or not—it was up to them.  Relying upon Missouri v. Seibert, the Second Circuit pointed out that the tactic utilized by the police in this case was clearly a constitutional violation. Seibert prohibits the police from eliciting an un-Mirandized confession, then administering Miranda warnings, and subsequently obtaining a “rehash” of the first confession. The Second Circuit emphasized that the Supreme Court’s conclusion in Seibert applies not only to suppression motions contesting the admissibility of the subsequent confession, but also to a jury’s assessment of the voluntariness of a confession relied upon at trial. From that perspective, the answer to the jury’s question was not “no.” Indeed, the answer “no” was manifestly inaccurate—dramatically so with respect to the confession immediately after Hernandez was Mirandized, and also with respect to Hernandez’s later statement to the district attorney.”

Yet, the petitioner could only prevail if he could overcome the District Court’s conclusion that the error—if it was an error—was harmless. An error can only lead to relief if it has a “substantial and injurious effect or influence in determining the jury’s verdict,” and if “every fairminded jurist would harbor at least a reasonable doubt that the error was harmless.” The Second Circuit rather easily concluded that the error meaningfully affected the verdict—after all, the question to which the erroneous answer was given was posed by the jury, and there was no other contemporaneous evidence of guilt in the case.  The more interesting discussion surrounds the second part of the test. As the Second Circuit observed, the AEDPA requires such deference to state court rulings that even rulings on harmless error at the state court level must be deferred to. As the Court noted: “When a state court makes a harmless error determination on direct appeal, we owe the ‘harmlessness determination itself’ deference under AEDPA.” This requires the reviewing court to decide if the state court harmlessness determination was contrary to, or an unreasonable application of, the applicable standard regarding harmlessness found in Chapman v. California. Chapman requires the state to show beyond a reasonable doubt that the constitutional error was harmless. When the state court concludes that the state has met this burden, it can only be overturned by a habeas reviewing court when “every fairminded jurist” would agree that the state had actually failed to meet its burden.  In other words, as expressed by the Second Circuit: “we must ask whether any ‘fairminded jurist’ applying the Chapman standard would find the error harmless beyond a reasonable doubt”—if the answer is yes, then the lower court’s determination should stand.

This may be the textbook example of begging the question. Clearly, some jurists did find the error was harmless beyond a reasonable doubt (the state court judges), and in a case such as this, in which a Federal Court of Appeals is reviewing the decision of the District Court, another jurist—the District Court judge—reached the same conclusion. When this occurs, the reviewing court has to conclude that prior jurists are not “fairminded.” Unfortunately, courts don’t have the option of concluding that the standard itself is nonsensical.  The “fairminded jurist” standard was intended to assist in defining an “unreasonable” application of law or finding of fact. One could argue whether unreasonable needs to be amplified with an additional test, given that “reasonableness” is a standard courts deal with all the time. In fact, it is hard to see how (if it really means what it says) the “fairminded jurist” test is any better. Do reviewing courts really want to implicitly state that a lower court was not fairminded?  If the argument is that the test doesn’t actually mean what it says, and is rather just a stilted way of suggesting that a lower court decision was unreasonable—what is the point of the “fairminded jurist” test at all? Courts have expressed concern with this standard since its adoption by the Supreme Court in Harrington v. Richter. As critics of the standard have questioned: when state court judges disagree on a particular issue, does this mean that the “fairminded jurist” standard can never be met, no matter what decision was reached by the state court majority? This conundrum is amplified when there is a split among Federal Circuit Court judges on an issue—or a split among the Circuits themselves.

A cynic could argue that the real purpose of this test is to stack the deck against any petitioner challenging a lower court harmless error determination. If the test is so awkward as to require a reviewing court to challenge the fairmindedness of other judges, perhaps it is intended to make such a determination so nonsensical as to render a lower court’s finding of harmless error unreviewable.  This same impossible bar to review is found in another area where it could be argued that courts don’t wish to get involved in lower court determinations—namely, the first part of the ineffective assistance of counsel test found in Strickland v. Washington. As the Supreme Court has observed in this context, in order to overturn a lower court determination that defense counsel was not ineffective, a reviewing court “may grant relief only if every ‘fairminded juris[t]’ would agree that every reasonable lawyer would have made a different decision.” Such a test is overwhelmingly difficult to satisfy, thus suggesting the outcome: defense attorneys are usually not found to be ineffective. Similarly, if courts are reluctant to overturn state court decisions when using the “fairminded jurist” test to assess harmless error, then meaningful review of lower court decisions on this issue is also hampered. Deference may have its place, but this test is not only logically meaningless, but also unduly suggestive of the outcome when utilized. The Court would be better served by simply returning to the “unreasonable” test.

Fortunately for Hernandez, the Second Circuit was undeterred by the nature of the “fairminded jurist” test and concluded that: “[n]o fair-minded jurist could reach the conclusion that the error was harmless beyond a reasonable doubt.”