Filed On: April 25, 2003 UNITED STATES OF AMERICA v. Robert Gale

Notice: This opinion is subject to formal revision before publication in the
Federal Reporter or U.S.App.D.C. Reports. Users are requested to notify
the Clerk of any formal errors in order that corrections may be made
before the bound volumes go to press.
United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
–————
No. 01–3011 September Term, 2002
95cr00297–02
Filed On: April 25, 2003
UNITED STATES OF AMERICA,
A PPELLEE
v.
R OBERT G ALE,
A PPELLANT
–————
BEFORE: R ANDOLPH and R OGERS , Circuit Judges, and
W ILLIAMS , Senior Circuit Judge.
O R D E R
Upon consideration of appellant’s petition for rehearing
filed February 21, 2003, and the opposition thereto, it is
ORDERED that the petition be denied.
Per Curiam
FOR THE COURT:
Mark J. Langer, Clerk
BY:
Michael C. McGrail
Deputy Clerk

-- 1 of 5 --

2
A separate statement of Senior Circuit Judge WILLIAMS,
explaining the denial of rehearing, in which Circuit Judge
R ANDOLPH joins, is attached.
Circuit Judge R OGERS would grant the petition for rehear-
ing.
A separate statement of Circuit Judge R OGERS dissenting
from the denial of rehearing is also attached.

-- 2 of 5 --

1
WILLIAMS , Senior Circuit Judge, with whom R ANDOLPH,
Circuit Judge, joins: Petitioning for rehearing, Gale calls our
attention to a sentence in United States v. Smith, 77 F.3d 510
(D.C. Cir. 1996). The government there had failed to disclose
some powerful impeachment evidence against a key govern-
ment witness, including the government’s agreement to dis-
miss certain D.C. Superior Court felony charges. The wit-
ness, in turn, testified falsely, omitting this agreement from
his testimony about potential benefits from cooperation.
Finding a Brady violation that required a new trial, we
said, ‘‘We must assume that, had the government disclosed
the information regarding the dismissed Superior Court
charges, the witness would have testified exactly as he did.’’
77 F.3d at 516. Gale argues that this line in Smith precludes
us from inferring that if Brown’s prior perjuries had been
disclosed to the defense here, the government would have
simply used another expert witness. See 314 F.3d at 4–5.
But in Smith the sentence now pinpointed by Gale ap-
peared in the context of a government claim that the wit-
ness’s false statement was unintentional. Responding to the
claim of mistake, the court stated that a witness’s good faith
is irrelevant to materiality under Brady, and that the court
would assume that his testimony would have remained un-
changed. But there is nothing in Smith to suggest that the
government was arguing that disclosure would have been
accompanied by a change in government strategy. So far as
appears, we spoke of the necessity of assuming unchanging
testimony simply in response to the thought that the witness,
having acted unintentionally (rather than being a case-
hardened perjurer), would likely have come clean.
Gale’s broader construction of Smith would render it incon-
sistent with our approach in United States v. Bowie, 198 F.3d
905, 909–12 (D.C. Cir. 1999), where we evaluated the effects
of disclosure dynamically, taking into account readily foresee-
able changes in prosecution strategy and witness testimony.
There it seemed obvious to us that if the defense used the
impeachment evidence that had been withheld in violation of
Brady, the prosecution would have responded with rehabili-
tating testimony, rendering the Brady violation immaterial.
198 F.3d at 911–12. Here the obviously foreseeable change in

-- 3 of 5 --

2
strategy is replacement of an easily replaced witness rather
than addition of new testimony, but we see no distinction.
Gale’s proposed reading of Smith would preclude consider-
ation of such changes and is therefore incorrect in light of
Bowie.
Also, Gale argues that we were required to perform the
materiality inquiry ‘‘in light of the trial that occurred, not
some hypothetical trial that might have taken place.’’ Peti-
tion at 10. But the materiality inquiry plainly requires us to
hypothesize about what ‘‘would have’’ occurred in the event of
disclosure. Smith, 77 F.3d at 514 (quoting United States v.
Bagley, 473 U.S. 667 (1985)). Gale does not really deny that
some hypothesizing is required, but rather insists that we
consider only changes the defense would have made, not
those of the government. Petition at 9–10.
We reject this ‘‘defense-only’’ view of the materiality test.
The test requires us to consider whether, with disclosure by
the government, there would be ‘‘a ‘reasonable probability’ of
a different result.’’ Kyles v. Whitley, 514 U.S. 419, 434
(1995) (citations omitted). That standard plainly requires
analysis of ‘‘reasonable’’ alternatives. Gale asked us to con-
sider one alternative: the government discloses and then
plunges doggedly ahead. The government proposed another:
the government discloses and makes a rather obvious strate-
gic adjustment, picking a different, untainted expert. His al-
ternative is unreasonable, the government’s reasonable.
Kyles requires that we pick the latter. Contrary to the claim
of his petition, there is no issue whether to consider an
alternative scenario; the only issue is whether we should
focus on reasonable alternatives, and Kyles says we should.

-- 4 of 5 --

1
ROGERS, Circuit Judge: I would grant rehearing to elimi-
nate any suggestion in United States v. Gale, 314 F.3d 1 (D.C.
Cir. 2003), that this circuit has adopted a rule of ‘‘readily
replaceable’’ witnesses as part of its materiality analysis of
Brady violations under Kyles v. Whitley, 514 U.S. 419, 434
(1995). Petition for Rehearing and Rehearing En Banc at 3.
Under Kyles, ‘‘[a] ‘reasonable probability’ of different result
is TTT shown when the government’s evidentiary suppression
‘undermines confidence in the outcome of the trial.’ ’’ Id.
(quoting United States v. Bagley, 473 U.S. 667, 678 (1985)).
Suffice it to say, the court in United States v. Smith, 77 F.3d
511 (D.C. Cir. 1996), explained, ‘‘we must decide whether the
undisclosed information could have substantially affected the
efforts of defense counsel to impeach the witness, thereby
calling into question the fairness of the ultimate verdict.’’ 77
F.3d at 515. Thus, in United States v. Bowie, 198 F.3d 905,
909–12 (D.C. Cir. 1999), the court considered how the trial
might have proceeded had there been no Brady violation and
concluded that because Federal Rule of Evidence 801 afford-
ed the government a means to rehabilitate the fact-witness’s
credibility, the defendant had failed to show that he was
denied a fair trial. At no point, however, did the court in
Bowie rely on a revision of either the trial witnesses who
testified or the testimony that those witnesses provided.
That is consistent with the admonition in Smith that ‘‘[w]e
must assume that, had the Government disclosed the informa-
tion TTT, the witness would still have testified exactly as he
did.’’ 77 F.3d at 516. It is also consistent with Kyles, which
at no point suggests that the reviewing court should focus
other than on the trial that actually happened. 514 U.S. at
433–34. Accordingly, discussion of how the government
might have proceeded at a new trial with a different expert,
much less the good faith of the prosecutor and testimony in
another case, is besides the point.

-- 5 of 5 --