If the court has reached the merits of a petitioner’s claims, in a habeas or direct appeal proceeding, and had found constitutional error, then the court subjects that error to a “harmless error” analysis. The state can argue that a finding of constitutional error does not warrant relief because the error was harmless. Although there are some errors that the Court considers per se prejudicial, or per se reversible, apart from prejudice, the court subjects most constitutional errors to a harmless error analysis. This harmless error analysis has gone through subtle, but important changes, beginning with Chapman v. California in 1967 (see below), through the decision of the Brecht court in 1993, to the enactment of the Antiterrorism and Effective Death Penalty Act in 1996.
The standard under the rule is that constitutional violations required relief unless the government proved that the error was harmless beyond a reasonable doubt. The Supreme Court adopted this standard, formally, in Chapman v. California, 386 U.S. 18 (1967), to fulfill the federal courts’ responsibility to “protect people from infractions by the States of federally guaranteed rights,” in habeas law.
However, this standard changed in 1993, when the Court decided Brecht v. Abrahamson. A majority of the Court ruled that there was a new and different measure of harmless error that should apply in federal habeas proceedings. The applicable standard now is the one the Court fashioned in 1946 in Kotteakos v. United States for assessing the harmlessness of non-constitutional errors. At the present time, an error may be deemed harmless if the reviewing court fins that “the error did not influence the jury, or had but very slight effect” and that “the judgment was not substantially swayed by the error.” The phrase that the Brecht court borrowed form the Kotteakos opinion was, “the standard for determining whether habeas relief must be granted is whether… the error had substantial and injurious effect or influence in determining the jury’s verdict.” When “the matter is so evenly balanced that [the federal judge] feels himself in virtual equipoise as to the harmlessness of the error,” the court should find that the error is not harmless and rule in favor of the petitioner.
In 1996, the Antiterrorism and Effective Death Penalty Act (hereinafter “AEDPA”), Pub. L. 104-32, 110 Stat. 1214 (1996), which otherwise made substantial changes in habeas corpus practice procedure, is silent on harmless error questions. This is apparently leaves intact the rule of Brecht and its progeny. However, although AEDPA’s enactment did not overrule the Brecht court standard of harmless error, it did complicate the standard. The courts have since had to grapple with whether to employ the Chapman or Brecht standards of harmless error when assessing under 28 U.S.C. § 2254(d) whether a state court’s analysis of harmless error was “contrary to” or an “unreasonable application” of Chapman.  This means that the state court will apply Chapman, and the federal court will apply Brecht, in its analysis of harmless error. This leaves a distinct disparity between the standards of review of the state and federal courts, which are supposed to review the collateral attack in a manner that assures fairness and justice. Split standards do little to assure that the goals of fairness and justice are reached.
However, the Brecht rule has survived all challenges, to be at the forefront of habeas litigation, if the petitioner can get the court to the merits of the claims. This rule is extremely important now, because the Brecht court has drawn a fine distinction between the rule of error which applies on direct appeal, and the rule of harmless error which applies on habeas review. To justify this distinction, the Brecht majority pointed to “the State’s interest in the finality of convictions that have survived direct review within the state court system.” And concerns of “comity and federalism.” The Brecht court majority apparently premised these justification on an assumption that a finding of harmlessness by the state court under the more stringent Chapman rule will precede habeas corpus review of the harmlessness question under the less stringent Brecht rule. However, these rationales do not seem to justify Brecht’s departure from over 200 years of direct appeal and habeas corpus parity in the scope and standard of review of constitutional issues.
With these standards, what does it mean for the practitioner today? It means that the federal practitioner benefits from the less stringent standard of Brecht. It means that the federal courts are reviewing a state court decision at a lower standard than the state court applied. However, state courts rarely, if ever, grant a habeas petition. It is more common for the federal courts to grant the habeas petition. This might be because the state court has a harder time finding that its own courts have committed error, while the federal courts review these decisions independent of the concerns of “their courts” committing error, but rather concerned with whether, “a court” committed an error. This means that the standard in State court is higher now, using the Chapman standard, and the federal court uses the less stringent, Brecht standard. It might behoove the petitioner to fail in state court habeas proceedings and wait until federal court, if the constitutional error is clear enough that the court will reach the merits of the petition, rather than dismiss on procedural grounds.
 See, e.g., Brecht v. Abrahamson, 507 U.S. 619, 648 (1993) (White, J. dissenting) (trial errors are … subject to harmless error analysis… and… most constitutional errors are of this variety (citations omitted)); id at 653 (O’Conner, J., dissenting) (By now it goes without saying that harmless error review is of almost universal application: there are few errors that may not be forgiven as harmless) (discussing the standard adopted in Arizona v. Fulimante, 499 U.S. 279 (1991), for determining whether constitutional errors are or are not subject to harmless error rule).
 See e.g. Yates v. Evatt, 500 U.S. 381, 402-03 (1991); Rose v. Clark, 478 U.S. 570 (1986)
 Chapman at 21 (Chapman standard protects those rights that are “rooted in the Bill of Rights, offered and championed in Congress by James Madison, who told the Congress that the ‘independent’ federal courts would be the ‘guardians of those rights’”).
 507 U.S. 619 (1993)
 328 U.S. 750 (1946)
 Kotteakos, 328 U.S. at 764
 Id. at 765
 Brecht, supra, 507 U.S. at 623; quoting Kotteakos, supra, 328 U.S. at 776)
 O’Neal v. McAninch, 513 U.S. 432, 435 (1995). O’Neal, which resolved a question left open in Brecht by making it clear that the state bears the burden of proving harmlessness.
 See, e.g, Penry v. Johnson, 121 S. Ct. 1910, 1919-20 (2001) (Court first applies s 2254(d)(1) to state court’s ruling on the merits and, after concluding that state court opinion was not objectively unreasonable, and that petitioner therefore is not entitled to relief, observes that relief would not be available in any event because claimed error was harmless under Brecht.)
 See, Nevers v. Killinger, 169 F.3d 352, 370-72 (6th Cir. 1999) and Mangus v. Edwards, 40 F. Supp 2d 908, 920 (N.D. Ohio 1999) (determining whether to apply the Chapman standard of the Brecht standard under § 2254(d) analysis).
 Brecht, supra, 507 U.S. at 635.
 See Brecht, supra, 507 U.S. at 649 (White, J., dissenting) (Our habeas jurisprudence is taking on the appearance of a confused patchwork in which different constitutional rights are treated according to their status, an in which the same constitutional right is treated differently depending on whether its vindication is sought on direct or collateral review).