Brown v. Davenport

LII note: the oral arguments in Brown v. Davenport are now available from Oyez. The U.S. Supreme Court has now decided Brown v. Davenport .


May a federal court grant a defendant’s petition for a writ of habeas corpus upon finding a trial error had a “substantial and injurious effect” on the defendant; or, must the court also determine that the state court’s interpretation of Chapman v. California was unreasonable under the Antiterrorism and Effective Death Penalty Act of 1996?

Oral argument: 
October 5, 2021

This case asks the Supreme Court to resolve a difference in judicial opinion among several federal courts of appeal regarding which standard is appropriate for granting federal habeas relief. Petitioner Ervine Lee Davenport (“Davenport”) contends that the approach taken by the U.S. Court of Appeals for the 6th Circuit in Brecht v. Abrahamson, which requires that a defendant experience a “substantial and injurious effect” due to a trial error, is satisfactory. Respondent Mike Brown (“Brown”), Acting Warden, argues that the standard invoked by the U.S. Court of Appeals for the 2nd, 3rd, 7th, 9th, and 10th Circuits in Chapman v. California should instead apply. For a federal court to grant relief under Chapman v. California, a trial error must not be “harmless,” and the state court’s interpretation of Chapman v. California must be “unreasonable” under the Antiterrorism and Effective Death Penalty Act of 1996 (“AEDPA”). The outcome of this case will affect how much deference federal courts give to state courts’ interpretations of AEDPA, as well as the ability of defendants to successfully obtain relief in federal habeas proceedings.

Questions as Framed for the Court by the Parties 

Whether a federal habeas court may grant relief based solely on its conclusion that the test from Brecht v. Abrahamson is satisfied, or whether the court must also find that the state court’s application of Chapman v. California was unreasonable under 28 U.S.C. § 2254(d)(1).


On January 13, 2007, Ervine Lee Davenport and Annette White were drinking alcohol and using cocaine at a friend’s house when White began acting belligerently. Davenport v. MacLaren at 2. Several of White’s friends asked her to leave, and Davenport offered to drive White home. Id. Davenport testified that during the drive, White grabbed the steering wheel and sliced his arm with a box cutter. Id. Still driving, Davenport pushed firmly against White’s neck to subdue her. Id. Davenport soon observed that White had stopped moving and breathing. Id. at 3. Realizing that he had killed White, Davenport was frightened and left White’s body in a field. Id.

Parts of Davenport’s statement, such as White’s aggressive behavior, were supported by witness testimony presented at trial. Id. However, some medical evidence contradicted Davenport’s account. Id. A forensic pathologist testified that it would take at least four to five minutes to suffocate a person to death, but it would take only thirty seconds to cut off someone’s air supply so that they pass out. Id. Further, the pathologist identified that White’s injuries were more consistent with a victim who had been choked than with one who had been pressed against a car door. Id.

At trial, Davenport wore shackles on his left hand, ankles, and waist. Id. at 4. The trial court did not provide any explanation on the record for Davenport’s restraints. Id. The jury found Davenport guilty of first-degree murder. Id.

Davenport appealed his conviction, arguing that the requirement that he wear shackles at trial violated his due process rights. Id. The Michigan Court of Appeals reviewed the trial court decision for plain error, finding that it was erroneous for the trial court to restrain Davenport. Id. However, the Court of Appeals held that Davenport had not proven that the jury could see his restraints and, consequently, had not shown evidence of prejudice. Id. The Michigan Supreme Court reversed the Court of Appeals’ decision and remanded Davenport’s case to the trial court. Id. The trial court conducted a hearing to determine whether the twelve jurors saw Davenport’s shackles and, if so, whether the restraints impacted their verdict. Id. at 5. Although several of the jurors noticed the shackles, all swore that Davenport’s restraints did not alter their decision. Id. The trial court held that the prosecution proved beyond a reasonable doubt that the jury’s deliberations were not impacted by Davenport’s shackling. Id. The Michigan Court of Appeals affirmed. Id. The Michigan Supreme Court denied leave to appeal. Id. at 6.

Davenport filed a petition for a writ of habeas corpus in federal district court pursuant to 28 U.S.C. § 2254 to address his shackling. Id. The district court denied Davenport’s petition, holding that the state court’s decision was “neither contrary to, nor involve[d] an unreasonable application of, clearly established federal law.” Id. Davenport appealed to the United States Court of Appeals for the Sixth Circuit. Id.

The Sixth Circuit reviewed the district court’s decision to deny Davenport’s writ of habeas corpus de novo. Id. Duncan MacLaren, Warden at Michigan Department of Corrections, contended that the Sixth Circuit should utilize the tests articulated by Chapman v. California and the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) to evaluate Davenport’s petition. Id. at 7. Alternatively, Davenport asserted that the Court should use the test under Brecht v. Abrahamson. Id. The Sixth Circuit ruled in Davenport’s favor, finding that the Brecht standard applies. Id. MacLaren filed a petition for rehearing en banc, which was denied. Davenport v. MacLaren, On Petition for Rehearing En Banc at 2.

The United States Supreme Court granted certiorari to Mike Brown, the Acting Warden where Davenport is currently incarcerated, on April 25, 2021.



Petitioner Brown argues that the U.S. Court of Appeals for the Sixth Circuit should have applied Chapman v. California to find that Ervine Davenport’s unconstitutional shackling was a harmless error. See Brief for Petitioner, Brown at 34. Brown explains that a court must find evidence beyond a reasonable doubt to declare a “federal constitutional error” under Chapman. Id. at 18. Further, Brown contends that a federal court should apply Chapman when reviewing a state court’s finding of an error, and disregard Deck v. Missouri, which “provided a framework for identifying a shackling error.” Id. at 34. Brown argues that the Deck Court’s holding is not applicable here because Deck did not explain how to determine when “visible shackling” was harmless error, and only examined whether it violated due process. Id. at 42. Brown explains that the Sixth Circuit should have applied Chapman exclusively because the State concedes that Davenport was shackled, and the due process analysis from Deck was irrelevant in such a context. Id.

Moreover, Brown asserts that the Sixth Circuit erroneously extended Holbrook v. Flynn to the present case. Id. at 36. Brown argues that the Sixth Circuit should not have applied Flynn because the State concedes that an error occurred. Id. at 38. Brown explains that the court in Flynn considered whether the presence of law enforcement in the courtroom was “inherently prejudicial.” Id. Brown notes that jurors in Flynn were asked to speculate as to whether law enforcement presence during trial would affect a jury verdict. Id. at 37. Brown explains that in Flynn, the Court noted that the jurors might not consciously understand the effect of seeing uniformed officers during trial. Id. at 38. Brown distinguishes this case from Flynn by explaining that the twelve jurors here “unequivocally” said that the shackling error did not affect their deliberations. Id.

Respondent Davenport counters that the Sixth Circuit reasonably relied on Deck and Chapman. See Brief for Respondent, Davenport at 31–32. Davenport argues that the Sixth Circuit used Deck as the basis for its decision in finding an error when the State visibly shackled him. Id. at 32. Davenport explains that the Deck Court held that shackling was “inherently prejudicial,” and the State has the burden to prove beyond a reasonable doubt that the shackling error did not impact the jury’s verdict. Id. Davenport adds that the Sixth Circuit relied on Chapman to review the trial record. Id. Davenport concludes that the Sixth Circuit correctly relied on these two cases to find that his due process rights were violated and that he suffered “actual prejudice.” Id.

Although Davenport agrees with the Sixth Circuit’s holding, he argues that the Michigan Supreme Court should have applied Chapman to find whether the unconstitutional shackling was an error. Id. at 37-38. Davenport explains that the Michigan Supreme Court erroneously applied Flynn to hold that his visible shackling was harmless because the Court could not find “impermissible factors” that were involved. Id. at 38. Davenport contends that the Flynn Court established an “unacceptable risk test” that determines whether an error occurred based on the number of factors that affect a jury’s verdict. Id. Davenport argues that the Flynn test is not applicable in the present case because the State had conceded under Deck that his visible shackles were a due process violation. Id. at 39. Davenport further explains that the Michigan Supreme Court’s decision to rely on Flynn contravened federal law that was established in Deck v. Missouri and Flynn regarding the use of juror testimony. Id. at 43. Davenport explains that in Deck, the Court found that visible shackles are a due process rights violation and could affect a jury’s verdict. Id. As such, Davenport explains that unconstitutional shackling affects jurors, and the Michigan Supreme Court reliance on juror’s testimony was “improper.” Id. at 43–44. Davenport concludes that the Michigan Supreme Court’s holding should not be upheld because the Court erroneously relied on juror testimony, which was not permitted in Deck and Flynn. Id. at 44.


Brown argues that Brecht v. Abrahamson, Chapman, and the Antiterrorism and Effective Death Penalty Act of 1996 (“AEDPA”) compel the Court to deny Davenport habeas relief. See Brief for Petitioner, Brown at 34. Brown explains that a court must use the Brecht and AEDPA/Chapman tests to determine if habeas relief is available. Id. Brown notes that before AEDPA was passed, Brecht was used to determine if an error demands habeas relief, and Brown contends that habeas relief was available under Brecht if the error had a “substantial or injurious” effect on a jury's verdict. Id. at 19–20. Brown explains that the Brecht standard is met when a defendant shows that a trial error influenced the jury so substantially as to result in “actual prejudice” Id. at 28, 19.

Brown argues further that AEDPA contains a separate test that courts must apply in addition to Brecht before habeas relief is granted. Id. at 34. Brown explains that this test determines that habeas relief is available under the statute if (1) the state court’s decision is contrary to federal law, (2) the state court’s decision is an “unreasonable application” of federal law, or (3) the state court’s decision was made on the basis of an “unreasonable determination of the facts.” Id. at 22. Brown counters that Brecht alone “could not” encompass AEDPA because AEDPA requires that federal courts defer to state court’s “adjudication of the same claim.” Id. at 33. Brown notes that after Congress enacted AEDPA, it amended 28 U.S.C. § 2254, a federal statute that governs when a federal court can grant a “writ of habeas corpus” to review state court decisions. Id. According to Brown, 28 U.S.C. § 2254(d) bars federal courts from granting habeas relief based on an independent review unless the federal court first finds that the state court was unreasonable in rejecting the habeas claim. Id. at 23.

Davenport counters that a federal habeas court may choose to formally apply only the Brecht standard because Brecht has a higher standard to meet that “subsumes” AEDPA. See Brief for Respondent, Davenport at 16. Davenport explains that when a federal court finds actual prejudice under Brecht, the AEDPA/Chapman standard is also met because a reasonable jurist could agree with the state court’s finding of harmlessness. Id. Davenport argues that the relationship between Brecht and AEDPA/Chapman is defined by Fry v. Pliler and Davis v. Ayala. Id. at 22. Davenport concedes that the Fry Court recognized the Brecht and AEDPA/Chapman standards but explains that the Court also found that a formal application of both tests is unnecessary. Id. at 22. Similarly, in Ayala, Davenport explains that the Court found that a federal habeas court may apply only Brecht because Brecht requires a petitioner to show actual prejudice that would encompass AEDPA/Chapman standards. Id. at 16.

Moreover, while Davenport concedes that the Supreme Court has not yet explicitly laid out how a federal habeas court can comply with a habeas grant under 28 U.S.C § 2254(d)(1) without applying AEDPA/Chapman, he argues that a federal court may grant habeas relief through a Brecht analysis which relied on “legal and factual material” stemming from state court proceedings. Id. at 16-17. Davenport explains that a federal court that conducts a 28 U.S.C § 2254(d)(1) analysis would look at whether a state court “unreasonably applied” federal law that was established by the Supreme Court. Id. Thus, according to Davenport, if a federal court relies on evidence provided during the federal habeas trial, then a separate AEDPA/Chapman analysis is appropriate. Id. at 17. However, Davenport concludes that a separate application of Brecht and AEDPA/Chapman is not always necessary if a court relies only “on the record that was before the state court”. Id. at 18.



The Criminal Justice Legal Foundation (“CJLF”), in support of Brown, contends that Brown should prevail because federal courts must respect state court decisions regarding federal constitutional claims. See Brief of Amicus Curiae Criminal Justice Legal Foundation (“CJLF”), in Support of Petitioner at 5–6. CJLF asserts that balance between state and federal courts is important because it is the framework that the Founding Fathers intended to create through the Constitution. Id. at 6. CJLF expresses concern that a ruling in favor of Davenport could result in an unequal balance of power between the state and federal governments. Id. Consequently, CJLF posits that individuals bringing lawsuits in state courts could risk having future decisions overturned by the Supreme Court due to a preference for federal court rulings. Id. Two law professors, Jonathan F. Mitchell and Adam K. Mortara, (“Mitchell and Mortara”) agree with CJLF’s concerns, explaining that state and federal court decisions should co-exist, and federal court rulings should not simply replace those of state courts. See Brief of Amici Curiae Mitchell and Mortara, in Support of Petitioner at 20.

Davenport argues that Brown misconstrues his approach as undermining respect for state court decisions. See Brief of Respondent at 26. In agreement, the United States Court of Appeals for the Sixth Circuit notes in its opinion that Brown’s assertion that Davenport’s approach does not respect state court judgments is misguided. See Davenport v. MacLaren at 15. Davenport asserts that respect for state court decisions is already included in the Brecht standard because a federal court must satisfy specific criteria before it can overturn a state court’s decision. Brief of Respondent at 26. Thus, Davenport contends that a ruling in his favor would not create an imbalance in power between state and federal courts because a federal court could only overturn a state court’s decision in very limited circumstances. Id.


The States of Arkansas, et al. (“The states”), on behalf of Brown, assert that Davenport’s approach is harmful because it permits a court to consider decisions from lower courts in addition to decisions from the United States Supreme Court. See Brief of Amici Curiae The States of Arkansas, et al., in Support of Petitioner at 5. The states explain that courts utilizing Brown’s approach may only consider rulings from United States Supreme Court cases. Id. The states contend that this variation between the two standards is problematic because it may produce inconsistent outcomes among lower courts by allowing lower courts using Davenport’s approach to create new precedent instead of deferring to United States Supreme Court decisions. Id. at 5–7. Accordingly, the states conclude that it could become easier for habeas defendants in states using Davenport’s approach to obtain relief than defendants in states using Brown’s approach. Id.

Davenport counters that it is harder for a habeas defendant to obtain relief using his approach. See Brief for Respondent at 18–19. Davenport argues that Brown’s framework is more favorable toward defendants because a court using Brown’s approach must reverse a conviction when there is a trial error unless the prosecution can show that the error was harmless (meaning that the error did not impact the jury’s verdict). Id. at 19. Alternatively, Davenport’s approach only grants relief to defendants if the defendant’s attorney can show that the trial error actually prejudiced the jury’s deliberations. Id. at 21. Therefore, Davenport asserts that using his approach will not lead to conflicting outcomes between defendants in lower courts because his standard is more difficult to prove. Id.



The authors would like to thank Professor Keir Weyble for his guidance and insights into this case.

Additional Resources