You get Bagley expounding on Agurs opining that a “Napue” violation – which really should be termed a “Mooney” or “Pyle” violation – and this is what you wind up with in a District Court in 2020, in this case the District of Maryland:

However, a Napue violation is evaluated under the harmless-error standard, Bagley, 473 U.S. at 679 n.9, while a Brady violation is not; instead, an alleged Brady violation only requires reversal if a reasonable probability exists that, had undisclosed evidence been disclosed to the defense, the result of the proceeding would have been different, id. at 680-81, 682.

A Mooney violation, which would of course include the erroneously termed “Napue” violation, has never been explicitly subject to any limitation at all, including a “harmless error” analysis, and neither Bagley nor Agurs can change that because they were both Brady cases and everything they said about the Mooney line of cases was dicta, and the portion of the Bagley opinion cited in the District of Maryland wasn’t even part of the opinion of the SCOTUS!

In fact, as we pointed out in an amicus brief to the SCOTUS last year (Actually, that link is to a motion for “divided argument” but let’s not get too technical, just look at paragraphs 6 and following after you click the link, which we notice few of you ever do!), the better argument is that the SCOTUS already implicitly held that deliberate lying and cheating by police and prosecutors would result in automatic dismissal and/or reversal.  The harmless constitutional error doctrine was established for the first time in Chapman v. California.  Chapman was decided virtually simultaneously with Miller v. Pate, which was the last case strictly in the Mooney line and which contained this otherwise kind of inexplicable language:

More than 30 years ago this Court held that the Fourteenth Amendment cannot tolerate a state criminal conviction obtained by the knowing use of false evidence. Mooney v. Holohan, 294 U. S. 103. There has been no deviation from that established principle. Napue v. Illinois, 360 U. S. 264Pyle v. Kansas, 317 U. S. 213; cf. Alcorta v. Texas, 355 U. S. 28. There can be no retreat from that principle here.

Why was the question of a “retreat from that principle” even on the table?  Because Chapman was the very next case in the US reports (386 US 1 v. 386 US 18), and held that some constitutional errors can be subject to harmless error analysis.  Meaning that they might not result in a reversal on appeal.

Mooney violations always, always require reversal.  And dismissal.

But the Rehnquist SCOTUS, ignoring Miller v. Pate, beat a retreat from that principle for years, although never quite succeeding.

Obviously, the SCOTUS is going to have to be clarify the law here at some point.  We hope soon, but we have already been waiting so long.

Taking up the second question on this petition might help.