Peter Ventura v. Attorney General, State of Florid ( 2005 )

[PUBLISH]
IN THE UNITED STATES COURT OF APPEALS
FILED
FOR THE ELEVENTH CIRCUIT U.S. COURT OF APPEALS
________________________ ELEVENTH CIRCUIT
August 9, 2005
THOMAS K. KAHN
No. 04-14564 CLERK
________________________
D. C. Docket No. 02-01159-CV-ORL-19KRS
PETER VENTURA,
Petitioner-Appellant,
versus
ATTORNEY GENERAL, STATE OF FLORIDA,
SECRETARY, DEPARTMENT OF CORRECTIONS,
Respondents-Appellees.
________________________
Appeal from the United States District Court
for the Middle District of Florida
_________________________
(August 9, 2005)
Before EDMONDSON, Chief Judge, MARCUS and PRYOR, Circuit Judges.
MARCUS, Circuit Judge:
In this capital case, Peter Ventura has petitioned for federal habeas corpus
relief on the ground that the state prosecutor’s knowing failure to correct a key
government witness’s false testimony that he received no consideration in
exchange for his testimony violated Ventura’s due process rights, as established in
Giglio v. United States, 405 U.S. 150, 92 S. Ct. 763, 31 L. Ed. 2d 104 (1972). The
Supreme Court of Florida denied post-conviction relief, reasoning that Giglio’s
materiality element was unsatisfied, since there was no reasonable likelihood that
the false testimony could have affected the judgment of the jury.
After careful review of the record, we are convinced that the Florida court’s
disposition of Ventura’s Giglio claim was neither contrary to nor an unreasonable
application of clearly established federal law, and therefore we affirm the district
court’s denial of Ventura’s petition.
I.
Peter Ventura was convicted by a Florida jury of the 1981 murder of Robert
Clemente and sentenced to death. Clemente’s body was discovered on April 15,
1981, in a rural area of Volusia County off state route 44. The body was found in a
truck bearing the logo of Clemente’s employer, Crow’s Bluff Marina. Clemente
had been shot four times, beaten, and possibly stabbed.
Clemente’s former employer, Jerry Wright, had taken out a “key man”
insurance policy on Clemente’s life, without Clemente’s knowledge, when
2
Clemente worked at Wright’s tire store. Wright then hired Jack McDonald to
murder Clemente in exchange for half of the insurance proceeds. McDonald, in
turn, hired the petitioner, Ventura, to commit the murder.
On June 25, 1981, Ventura was arrested in Chicago and McDonald was
arrested in Florida. Both were indicted for first-degree murder on June 30, 1981.
Ventura was released on bond in Chicago while awaiting extradition, and failed to
appear at his extradition hearing on August 18, 1981. He remained a fugitive until
June 11, 1986, when he was apprehended in Austin, Texas. In the meantime, the
state determined that it could not successfully prosecute McDonald without
Ventura’s cooperation. McDonald had remained incarcerated without trial until
the expiration of Florida’s speedy-trial deadline, which requires the state to
commence trial within six months of indictment. Thereafter, by operation of
Florida’s law, McDonald became immune from prosecution for Clemente’s
murder.1
After being released unconditionally by Florida, however, McDonald was
indicted by a federal grand jury sitting in the Northern District of Illinois in 1983
on charges arising out of his involvement in a Chicago bank fraud scheme. He
pled guilty and was sentenced to three consecutive five-year terms of
1
Jerry Wright was tried and convicted of the first-degree murder of Robert Clemente in
February, 1990, and sentenced to life imprisonment.
3
imprisonment. After sentencing, McDonald was released on bond, and
subsequently failed to report for his incarceration. He remained a fugitive until he
was apprehended in Georgia some four years later on September 23, 1987.
On January 11, 1988, Ventura was brought to trial for Robert Clemente’s
murder in the Circuit Court for the Seventh Judicial Circuit in Volusia County.
McDonald, by then back in federal custody, testified against Ventura. Among
other things, the state prosecutor, Stark, asked McDonald on direct examination,
“Any promises been made to you concerning your testimony here?” to which
McDonald replied, “None whatsoever.”
Then, on redirect, Stark asked McDonald, “And what is your motivation for
testifying here today?” McDonald explained: “Well, I’m nearing sixty years of
age. This is probably, undoubtedly, the most horrendous thing I have ever been
involved in, and I think it is about time we cleared the air and it might give Mr. and
Mrs. Clemente a little peace of mind knowing exactly what happened.”
Ventura’s counsel, Cass, followed up on this on recross-examination, in this
exchange with McDonald:
Q. You have just said that your motive for testifying is simply to
clear the air, and bring the truth out; is that correct?
A. That’s correct.
Q. And for no other reason?
4
A. None.
Q. And because you’re in a position where it doesn’t really cost
you anything to say it?
A. That’s correct.
A jury convicted Ventura of Clemente’s murder and then recommended the
death penalty. The trial judge adopted the jury’s recommendation and sentenced
Ventura to die on January 21, 1988. The Supreme Court of Florida affirmed both
Ventura’s conviction and the death sentence on direct review. Ventura v. State,
560 So. 2d 217 (Fla. 1990).
Thereafter, it came to light that McDonald had, in fact, been testifying
pursuant to a deal with the prosecution. A series of letters document this
arrangement. First, on December 19, 1986 -- while McDonald was still a fugitive -
- Stark wrote the U.S. Attorney’s Office in Chicago (the office that prosecuted the
bank-fraud case against McDonald) explaining the Clemente case and asking for
assistance in securing McDonald’s testimony. Stark’s letter stated, in part:
While I understand the obvious reluctance of the sentencing court to
show lienience [sic] upon Mr. McDonald especially in light of the fact
that he took advantage of the court’s prior lieniency [sic] by failing to
report to prison after having been given time to get his affairs in order,
I feel that the interests of justice could be better served by having Mr.
McDonald on lengthy probation with a short jail term if necessary,
available to testify at the trial of Peter Ventura and possibly Jerry
Wright (in the event he is indicted).
5
I would appreciate any consideration your office could give in the
effort to locate Jack McDonald, or coax him out of hiding.
On March 6, 1987, the U.S. Attorney replied to Stark, explaining that it was
too late to modify McDonald’s sentence in the bank fraud case, but that if
McDonald testified against Ventura, “his cooperation and truthful testimony in that
case can be made known to the Federal Parole Board at his first parole hearing.”
The letter further stated: “Should Mr. McDonald surrender to federal authorities
and also appear as a witness at Mr. Ventura’s trial, this office will consider the
nature of Mr. McDonald’s cooperation and truthful testimony in evaluating
whether to pursue further prosecution of Mr. McDonald on bond jumping
charges.” No promise was made, however, at that time concerning the prosecution
of McDonald on federal bond-jumping charges.
Several months later, on September 3, 1987, McDonald wrote a letter to one
of the investigative agents, Detective Hudson of the Volusia County Sheriff’s
Office, stating:
If anyone is interested in my testimony at this point in time it will
have to be a two for one trade. In other words I will cooperate fully
provided I am released by court order from all federal charges
including the IRS. . . . If you can get a court order to this effect out in
front I will promptly turn myself in and cooperate fully. . . . Should it
be decided that the U.S. Gov’t and the State of Florida can handle this
I will give you whatever time is necessary to clear this up. But again,
I can only do it as a free man. If by some fluke I am apprehended
without any deal being made I will rot in hell before I would give any
6
testimony on anything. This is a promise.
Soon thereafter, on September 23, McDonald was taken into federal custody.
Two days later, Stark again wrote the U.S. Attorney’s Office, in an effort to secure
a deal for McDonald on possible federal bond-jumping charges in exchange for his
testimony at the Ventura trial. This letter stated, in pertinent part:
Pursuant to our telephone conversation of today’s date, I would like to
formally request that you consider dismissing the bond jumping
charges against Jack McDonald. . . . On September 24th, David
Hudson, the lead investigator on the Ventura case[,] interviewed Mr.
McDonald and was assured of McDonald’s cooperation with us in the
prosecution of Peter Ventura and others involved in the murder of one
Robert Clemente. Needless to say, Mr. McDonald is a crucial witness
in both cases. The case of the State of Florida vs. Peter Ventura is
presently scheduled for trial in Circuit Court, Volusia County on
October 12, 1987. Mr. McDonald’s cooperation is essential.
The U.S. Attorney then agreed not to prosecute McDonald for bond
jumping.2 An October 5, 1987 letter from the U.S. Attorney to Stark stated:
2
In 1983, when McDonald jumped bond, bond jumping carried a penalty of a fine of up
to $5000, and/or up to five years’ incarceration. The statute in effect at that time, 18 U.S.C. §
3150 (which was enacted in 1966 and repealed in 1984), stated:
Whoever, having been released pursuant to this chapter, willfully fails to appear
before any court or judicial officer as required, shall, subject to the provisions of
the Federal Rules of Criminal Procedure, incur a forfeiture of any security which
was given or pledged for his release, and, in addition, shall (1) if he was released
in connection with a charge of felony, or while awaiting sentence of pending
appeal or certiorari after conviction of any offense, be fined not more than $5,000
or imprisoned not more than five years, or both, or (2) if he was released in
connection with a charge of misdemeanor, be fined not more than the maximum
provided for such misdemeanor or imprisoned for not more than one year, or
both, or (3) if he was released for appearance as a material witness, shall be fined
not more than $1,000 or imprisoned for not more than one year, or both.
7
Pursuant to your request, my office will not pursue bond-
jumping charges against Jack McDonald as long as he cooperates
fully with your office in the upcoming murder case referred to in your
letter of September 25, 1987. Should Mr. McDonald fail to testify
truthfully in that case or in some other way fail to cooperate with your
office, we will then be free to pursue bond-jumping charges.
Moreover, this agreement does not affect Mr. McDonald’s
obligation to serve the federal sentence which has been imposed for
his prior criminal conduct in this district.
Immediately after Ventura’s trial, on January 20, 1988, Stark wrote another
letter to the U.S. Attorney’s Office explaining that “[w]hile there were no promises
made to Mr. McDonald in return for his testimony,” he “fe[lt] a compelling
obligation to advise [the U.S. Attorney’s Office] and the Court in Chicago of the
assistance provided by Mr. McDonald to the State of Florida in the prosecution of
those persons involved in the homicide of Robert G. Clemente.” Among other
things, Stark said:
While I realize that Jack McDonald is equally responsible for
the death of Robert Clemente, along with his two co-defendants, I also
recognize the fact that he did not have to provide information to the
State of Florida detailing each player’s participation in the overall
scheme to commit murder. He has testified in Mr. Ventura’s trial and
I anticipate calling him in Mr. Wright’s trial. He has previously stated
in deposition and at trial that his motive for giving his testimony has
been to clear the air and set the record straight so that the family of the
victim can have some peace of mind. He has also stated that “this is
Section 3150 did not require the trial judge to run the bond-jumping sentence consecutively to
the sentence for the underlying charges. However, under the current version of the statute, 18
U.S.C. § 3146, the sentences must run consecutively. Id. § 3146(b)(2).
8
the most horrendous crime he has ever been involved with.”
I believe I understand Mr. McDonald’s motivation to testify and
I further believe that his testimony has been extremely valuable in the
prosecution of his co-defendants. Whatever consideration can be
given him at any future hearings in his two Federal cases in return for
this assistance would, in my opinion, be in the interest of justice.
Corporal David Hudson of the Volusia County Sheriff’s Office
and I would appreciate the courtesy of a telephone call regarding the
scheduling of any future hearings to be held for Mr. McDonald so that
we can make arrangements to be heard by the Court considering Mr.
McDonald’s cooperation in Florida.
Finally, Stark wrote the U.S. Attorney’s Office again on October 31, 1988,
explaining that McDonald was cooperating in the case against Jerry Wright, and
that “[a]ny considerations [sic] that the Federal Courts could show Mr. McDonald
for his efforts in this regard for his cooperation to date and in the future would be
appreciated.” No record evidence indicates that the prosecutor at any time
communicated McDonald’s cooperation either to the federal district court in
Chicago or to the United States Parole Commission. McDonald was not indicted
for federal bond jumping.
On March 2, 1992, Ventura filed his first motion for postconviction relief in
the state trial court pursuant to Florida Rule of Criminal Procedure 3.850. Ventura
claimed that he was unable to file a proper postconviction motion because several
agencies had not satisfied his public records requests, and he listed the claims he
9
intended to raise once the requests were fulfilled. The trial court dismissed all of
Ventura’s claims, but the Supreme Court of Florida reversed the dismissal as
premature, directing the trial court to permit Ventura to amend his postconviction
motion after the public records issues were resolved. Ventura v. State, 673 So. 2d
479 (Fla. 1996).
On August 19, 1996, Ventura filed an amended postconviction motion in the
state trial court, this time raising fifteen claims, including a Giglio claim alleging,
based on the series of letters described above, that prosecutor Stark knowingly
offered false testimony that McDonald had received no consideration in exchange
for his trial testimony.
The trial court summarily dismissed ten of Ventura’s claims and, on June 1,
1998, held an evidentiary hearing on the remaining claims, including the Giglio
issue. At this hearing, Ventura’s counsel introduced the above-described letters,
and attorneys Stark and Cass both testified. Stark offered no explanation for
allowing the false testimony, claiming to have little recollection of the relevant
events. He could not definitively state whether the letters were disclosed to
Ventura. He said that his office maintained an “open-file policy,” which made
their entire file available to the defense, but could not recall whether the letters in
question were part of the file.
10
Defense counsel Cass testified that he did not recall ever being told about a
deal for McDonald in exchange for his testimony, that he was unaware of any
communications between the state attorney and the U.S. Attorney’s Office, and
that he remembered the frustration of not being able to impeach McDonald based
on his motive for testifying.
Following the hearing, the trial court denied Ventura’s remaining claims,
including his Giglio claim, with little explanation. Order, July 28, 1998. The trial
court apparently interpreted the Giglio claim as an ineffective assistance of counsel
claim, rejecting it on the ground that Ventura failed to demonstrate a reasonable
probability of a different result had counsel performed adequately. Id. at 1-2. In
reaching this conclusion, the trial court lumped the Giglio claim together with
several others, finding that “the other issues are not worthy of any specific
discussion.” Id. at 5.
The Supreme Court of Florida, by vote of 4-3, affirmed the denial of
postconviction relief. Ventura v. State, 794 So. 2d 553 (Fla. 2001). As to
Ventura’s Giglio claim, the supreme court found that the state had indeed secured a
deal for McDonald in exchange for his testimony, and that the prosecutor
knowingly presented false testimony to the contrary, but that the testimony was not
material. Because McDonald had been significantly impeached at trial, and
11
because McDonald’s testimony was extensively corroborated by other pieces of
compelling evidence, the Florida Supreme Court reasoned, evidence of the deal
was not “material” under Giglio. Id. at 562-65. The United States Supreme Court
subsequently denied cert. Ventura v. Florida, 535 U.S. 1098 (2002).
Ventura then petitioned for a writ of habeas corpus in the United States
District Court for the Middle District of Florida. After careful review, the district
court denied the petition. Order, Ventura v. Moore, No. 6:02-cv-1159-Orl-19KRS
(M.D. Fla. July 30, 2004). The district court expressed “grave doubts” about
whether the state supreme court had applied the correct materiality standard in
evaluating Ventura’s Giglio claim, since the state court had asked whether there
was “a reasonable probability that the false evidence may have affected the
judgment of the jury,” id. at 18 (quoting Ventura, 794 So. 2d at 563), rather than
whether “the false testimony could . . . in any reasonable likelihood have affected
the judgment of the jury,” id. (quoting Giglio, 405 U.S. at 154). Nevertheless,
although the state court had “misstated the Giglio rule,” the district court found
that “the decision itself” was neither contrary to nor an unreasonable application of
clearly established federal law. Id. at 20. Applying Giglio’s “any reasonable
likelihood” standard, the district court agreed with the state supreme court’s
conclusion that the substantial impeachment of McDonald at trial, as well as the
12
extensive corroboration of his testimony, established that “presentation of the
correct facts could not have altered the outcome of the trial,” and thus that the false
testimony was immaterial. Id. at 21.
On October 28, 2004, the district court granted a certificate of appealability
only as to Ventura’s Giglio claim. That claim is the only one now before us.
II.
Giglio error is a species of Brady error that occurs when “the undisclosed
evidence demonstrates that the prosecution’s case included perjured testimony and
that the prosecution knew, or should have known, of the perjury.” United States v.
Agurs, 427 U.S. 97, 103, 96 S. Ct. 2392, 49 L. Ed. 2d 342 (1976). “If false
testimony surfaces during a trial and the government has knowledge of it, . . . the
government has a duty to step forward and disclose.” Brown v. Wainwright, 785
F.2d 1457, 1464 (11th Cir. 1986). “In order to prevail on a Giglio claim, a
petitioner must establish that the prosecutor knowingly used perjured testimony, or
failed to correct what he subsequently learned was false testimony, and that the
falsehood was material.” Tompkins v. Moore, 193 F.3d 1327, 1339 (11th Cir.
1999).
The origins of the Giglio doctrine lie in the Supreme Court’s decision in
Napue v. Illinois, 360 U.S. 264, 79 S. Ct. 1173, 3 L. Ed. 2d 1217 (1959), which
13
held that a prosecutor’s failure to correct false testimony by the principal state
witness that he had received no promise of consideration in return for his testimony
violated the defendant’s Fourteenth Amendment due process rights and required a
reversal of the judgment of conviction. The Court explained that “it is established
that a conviction obtained through use of false evidence, known to be such by
representatives of the State, must fall under the Fourteenth Amendment.” Id. at
269 (citing Mooney v. Holohan, 294 U.S. 103, 55 S. Ct. 340, 79 L. Ed. 791(1935)). “The same result obtains when the State, although not soliciting false
evidence, allows it to go uncorrected when it appears.” Id. This principle, the
Court observed, “does not cease to apply merely because the false testimony goes
only to the credibility of the witness,” since “[t]he jury’s estimate of the
truthfulness and reliability of a given witness may well be determinative of guilt or
innocence, and it is upon such subtle factors as the possible interest of the witness
in testifying falsely that a defendant’s life or liberty may depend.” Id. Reversal
was required because “the false testimony used by the State in securing the
conviction may have had an effect on the outcome of the trial.” Id. at 272.
Subsequently, in Giglio v. United States, 405 U.S. 150, 92 S. Ct. 763, 31 L.
Ed. 2d 104 (1972), the Supreme Court held that the government’s failure to correct
false testimony that its key witness (the defendant’s coconspirator) had received no
14
promise of nonprosecution in exchange for his testimony, as well as the
prosecutor’s false statement to this effect in closing argument, required that the
defendant be granted a new trial. The Court explained that “deliberate deception of
a court and jurors by the presentation of known false evidence is incompatible with
rudimentary demands of justice.” Id. at 153 (citation and internal quotation marks
omitted).
The Giglio Court made clear, however, that such errors do not require
automatic reversal, and articulated a “materiality” standard to guide the
determination of whether a new trial is warranted:
We do not . . . automatically require a new trial whenever a combing
of the prosecutors’ files after the trial has disclosed evidence possibly
useful to the defense but not likely to have changed the verdict. A
finding of materiality of the evidence is required under Brady. A new
trial is required if “the false testimony could . . . in any reasonable
likelihood have affected the judgment of the jury.”
Id. (citations and internal quotation marks omitted) (quoting Napue, 360 U.S. at
271). Because “the Government’s case depended almost entirely on [the falsely
testifying witness’s] testimony,” the Court reasoned, his “credibility as a witness
was therefore an important issue in the case, and evidence of any understanding or
agreement as to a future prosecution would be relevant to his credibility and the
jury was entitled to know of it.” Id. at 154-55. Accordingly, the Court reversed
the judgment of conviction.
15
Since its decisions in Napue and Giglio, the Supreme Court “has
consistently held that a conviction obtained by the knowing use of perjured
testimony is fundamentally unfair, and must be set aside if there is any reasonable
likelihood that the false testimony could have affected the judgment of the jury.”
United States v. Agurs, 427 U.S. 97, 103, 96 S. Ct. 2392, 2397, 49 L. Ed. 2d 342(1976) (footnote omitted) (emphasis added); see also Kyles v. Whitley, 514 U.S.
419, 433 & n.7, 115 S. Ct. 1555, 131 L. Ed. 2d 490 (1995); United States v.
Bagley, 473 U.S. 667, 677, 105 S. Ct. 3375, 87 L. Ed. 2d 481 (1985); accord
Brown, 785 F.2d at 1465-66.3
The “any reasonable likelihood” standard differs from the materiality
standard applicable to other types of Brady violations because of the nature of the
error. As the Supreme Court has explained, “the Court has applied a strict standard
of materiality [to Giglio violations], not just because they involve prosecutorial
misconduct, but more importantly because they involve a corruption of the truth-
seeking function of the trial process.” Agurs, 427 U.S. at 104; accord United
3
Giglio error does not fall within the category of constitutional error that the Supreme
Court has characterized as “structural,” requiring automatic reversal of a tainted conviction. A
structural error is a “defect affecting the framework within which the trial proceeds, rather than
simply an error in the trial process itself.” Arizona v. Fulminante, 499 U.S. 279, 310, 111 S. Ct.
1246, 113 L. Ed. 2d 302 (1991). The Supreme Court has found structural error “only in a very
limited class of cases,” including those involving complete denial of counsel, a biased trial
judge, racial discrimination in selection of a grand jury, denial of self-representation, denial of a
public trial, and a defective reasonable doubt instruction. Neder v. United States, 527 U.S. 1, 8,
119 S. Ct. 1827, 144 L. Ed. 2d 35 (1999) (citation and internal quotation marks omitted).
16
States v. Alzate, 47 F.3d 1103, 1110 (11th Cir. 1995).
As we have explained, Ventura’s Giglio claim arises out of his post-trial
discovery that the state prosecutor had worked out a deal with the U.S. Attorney’s
Office to immunize McDonald from prosecution on federal bond-jumping charges
in exchange for his testimony at Ventura’s trial, and that McDonald’s testimony
that he had received no promises in exchange for his testimony was therefore false.
The State does not contest the Florida Supreme Court’s factual determination that
this deal indeed existed, and that McDonald’s testimony to the contrary was false.
See Ventura, 794 So. 2d at 562. The only issue the parties have briefed and the
only one now before us is whether the error was “material.”
III.
We review the district court’s denial of Ventura’s habeas petition de novo.
Harrell v. Butterworth, 251 F.3d 926, 930 (11th Cir. 2001). However, under the
Antiterrorism and Effective Death Penalty Act (“AEDPA”), which governs
Ventura’s petition, we may grant the writ only if we find that the Florida Supreme
Court’s resolution of Ventura’s Giglio claim “resulted in a decision that was
contrary to, or involved an unreasonable application of, clearly established Federal
law, as determined by the Supreme Court of the United States.” 28 U.S.C. §
2254(d)(1).
17
A.
The phrase “clearly established Federal law,” as used in § 2254(d)(1),
encompasses only the holdings of the United States Supreme Court. Williams v.
Taylor, 529 U.S. 362, 412, 120 S. Ct. 1495, 146 L. Ed. 2d 389 (2000). Section
2254(d)(1) expressly “restricts the source of clearly established law to [the
Supreme] Court’s jurisprudence.” Id. As we have previously explained, “§
2254(d)(1) provides a measuring stick for federal habeas courts reviewing state
court decisions. That measuring stick is ‘clearly established Federal law.’ Clearly
established federal law is not the case law of the lower federal courts, including
this Court. Instead, in the habeas context, clearly established federal law ‘refers to
the holdings, as opposed to the dicta, of [the Supreme Court’s] decisions as of the
time of the relevant state court decision.’” Putman v. Head, 268 F.3d 1223, 1241
(11th Cir. 2001) (alteration in original) (quoting Williams, 529 U.S. at 362).
As far as Giglio materiality is concerned, the clearly established law of the
Supreme Court is simply that reversal of a conviction is required when “there is
any reasonable likelihood that the false testimony could have affected the judgment
of the jury.” The Supreme Court has repeated this standard numerous times, but
has not elaborated further on its meaning; indeed, no Supreme Court case since
Giglio itself has squarely addressed a Giglio claim. Accordingly, the proper
18
inquiry for our Court is whether the Florida Supreme Court’s treatment of
Ventura’s Giglio claim was contrary to or an unreasonable application of the “any
reasonable likelihood” standard.4
4
Our Court has held, on several occasions, that Giglio’s “any reasonable likelihood”
standard is equivalent to the harmless error inquiry of Chapman v. California, 386 U.S. 18, 87 S.
Ct. 824, 828, 17 L. Ed. 2d 705 (1967), which provides that a conviction must be reversed if
constitutional trial error occurs, unless the beneficiary of the error establishes that the error was
“harmless beyond a reasonable doubt.” Id. at 24. We have stated at least twice in cases before us
on direct review, see Alzate, 47 F.3d at 1110; United States v. Rivera Pedin, 861 F.2d 1522,
1529 n.13 (11th Cir. 1988), and at least once in a case before us on habeas review, see Carr v.
Schofield, 364 F.3d 1246, 1255 (11th Cir. 2004), that the harmless-error inquiries required by
Giglio and by Chapman are indistinguishable.
Although this is the law of our Circuit, it is not clearly established federal law as
determined by the Supreme Court, since no majority has ever held that the Giglio standard is
equivalent to the Chapman standard, and since the Supreme Court’s decision in Brecht v.
Abrahamson, 507 U.S. 619, 113 S. Ct. 1710, 123 L. Ed. 2d 353 (1993), casts substantial doubt
on whether Chapman is ever the appropriate standard for harmless-error analysis conducted on
collateral review.
In concluding that the two standards are equivalent, we have drawn on Justice
Blackmun’s opinion in United States v. Bagley, 473 U.S. 667, 105 S. Ct. 3375, 87 L. Ed. 2d 481(1985), which observed that Giglio’s “any reasonable likelihood” standard “may as easily be
stated as a materiality standard under which the fact that testimony is perjured is considered
material unless failure to disclose it would be harmless beyond a reasonable doubt.” Id. at 678-
80. Thus, Justice Blackmun concluded, “the standard of review applicable to the knowing use of
perjured testimony is equivalent to the Chapman harmless-error standard.” Id. at 680 n.9.
Although Justice Blackmun clearly equated the Giglio and Chapman standards, this portion of
his opinion was joined only by Justice O’Connor and, consequently, does not constitute a
holding of the Court.
The Chapman Court itself suggested some similarity between the standards, stating:
“There is little, if any, difference between our statement in Fahy v. State of Connecticut about
‘whether there is a reasonable possibility that the evidence complained of might have contributed
to the conviction’ and requiring the beneficiary of a constitutional error to prove beyond a
reasonable doubt that the error complained of did not contribute to the verdict obtained.”
Chapman, 386 U.S. at 24 (quoting Fahy v. Connecticut, 375 U.S. 85, 84 S. Ct. 229, 11 L. Ed. 2d
171 (1963)).
This statement does not, however, suffice to render it “clearly established” that the
19
Chapman and Giglio standards are indistinguishable. For one thing, this statement in Chapman
was dictum. Moreover, Fahy involved a different kind of constitutional error -- admission of
unconstitutionally obtained evidence -- and the Fahy standard, while very similar to the Giglio
standard, is not identical. (Fahy asks “whether there is a reasonable possibility that the evidence
complained of might have contributed to the conviction”; Giglio asks whether “there is any
reasonable likelihood that the false testimony could have affected the judgment of the jury.”)
Indeed, when Giglio was decided nine years after Fahy, the Court made no mention of Fahy.
Because the Supreme Court has never definitively stated that Giglio’s “any reasonable
likelihood” inquiry is the same as the Chapman “harmless beyond a reasonable doubt” inquiry,
the only clearly established law governing Giglio materiality analysis is that Giglio error requires
reversal when there is any reasonable likelihood that the false testimony could have affected the
judgment of the jury. As the Supreme Court recently emphasized, “[a] federal court may not
overrule a state court . . . when the precedent from this Court is, at best, ambiguous.” Mitchell v.
Esparza, 540 U.S. 12, 17, 124 S. Ct. 7, 157 L. Ed. 2d 263 (2003). Accordingly, the “any
reasonable likelihood” formulation is the only one the Florida court was obligated to apply in
resolving Ventura’s Giglio claim.
Moreover, even if Chapman is seen as establishing the proper materiality standard for a
Giglio claim on direct review, Brecht casts substantial doubt on whether this standard has any
application on habeas review. Brecht held that Chapman’s “harmless beyond a reasonable
doubt” standard is not the proper harmless-error inquiry on collateral review; rather, habeas
courts must apply the test laid out in Kotteakos v. United States, 328 U.S. 750, 66 S. Ct. 1239,
90 L. Ed. 1557 (1946), which asks whether the error “had substantial and injurious effect or
influence in determining the jury’s verdict.” Brecht, 507 U.S. at 637 (quoting Kotteakos, 328
U.S. at 776). The Court explained that “an error that may justify reversal on direct appeal will
not necessarily support a collateral attack on a final judgment,” id. at 634 (citations and internal
quotation marks omitted), citing “the State’s interest in the finality of convictions that have
survived direct review within the state court system,” “comity and federalism” concerns, and the
integrity of the trial process as reasons for distinguishing between direct and collateral review.
See id. at 633-36. In addition, “[s]tate courts are fully qualified to identify constitutional error
and evaluate its prejudicial effect on the trial process under Chapman, and state courts often
occupy a superior vantage point from which to evaluate the effect of trial error.” Id. at 636.
Thus, “it scarcely seems logical to require federal habeas courts to engage in the identical
approach to harmless-error review that Chapman requires state courts to engage in on direct
review.” Id.
As several of our sister Circuits have observed, Brecht continues to apply on habeas
review even after AEDPA when a state court either fails to conduct any harmless error review
because it unreasonably concludes that no constitutional error has occurred, or conducts
harmless error review applying the wrong standard (i.e. when the state court’s decision is
contrary to or an unreasonable application of clearly established federal law). See, e.g., Aleman
v. Sternes, 320 F.3d 687 (7th Cir. 2003); Herrera v. Lemaster, 301 F.3d 1192 (10th Cir. 2002)
20
B.
The “contrary to” and “unreasonable application” clauses of § 2254(d)(1)
are separate bases for reviewing a state court decision. Williams, 529 U.S. at
412-13; see also Putman, 268 F.3d at 1241. “Under the ‘contrary to’ clause, a
federal habeas court may grant the writ if the state court arrives at a conclusion
opposite to that reached by [the Supreme] Court on a question of law or if the state
court decides a case differently than this Court has on a set of materially
indistinguishable facts.” Williams, 529 U.S. at 412-13; see also Bell v. Cone, 125
S. Ct. 847, 851, 160 L. Ed. 2d 881 (2005) (“A state court’s decision is ‘contrary to
. . . clearly established Federal law’ if the state court applies a rule that contradicts
the governing law set forth in our cases, or if the state court confronts facts that are
materially indistinguishable from a relevant Supreme Court precedent and arrives
at a result opposite to ours.” (citation and internal quotation marks omitted)).
Thus, there are two possible bases for concluding that the Florida Supreme
Court’s resolution of Ventura’s Giglio claim was contrary to clearly established
(en banc). Thus, even when a state court improperly fails to use the Chapman standard in
conducting harmless error review, a federal habeas court reviews the constitutional error only
under the more lenient Kotteakos standard. Cf. Penry v. Johnson, 532 U.S. 782, 795, 121 S. Ct.
1910, 150 L. Ed. 2d 9 (2001) (“Even if our precedent were to establish squarely that the
prosecution[] . . . violated Penry’s Fifth Amendment privilege against self-incrimination, that
error would justify overturning Penry’s sentence [on federal habeas review] only if Penry could
establish that the error “had substantial and injurious effect or influence in determining the jury’s
verdict.” (quoting Brecht, 507 U.S. at 637)). Chapman, therefore, has little application to a case
before us on collateral review.
21
federal law: either (1) it reached a conclusion of law that contradicts one reached
by the Supreme Court; or (2) it reached a result opposite one reached by the
Supreme Court on a set of materially indistinguishable facts.
The second possibility is more easily dismissed. Whether a Giglio violation
is material is a highly fact-dependent inquiry, and we can find no Supreme Court
case whose facts may fairly be characterized as “materially indistinguishable” from
those of Ventura’s case. Whether the false testimony offered in a particular case
could in any reasonable likelihood have affected the judgment of the jury must be
analyzed in light of a number of highly context-specific factual considerations,
including the importance of the testimony of the falsely testifying witness to the
government’s case, the nature and significance of the falsehood, and, notably, to
what extent the witness’s testimony is substantially corroborated by other
evidence.
The context-specific nature of this inquiry is evident from the Supreme
Court’s analysis in Giglio and Napue -- the only two cases in which the Supreme
Court has squarely addressed a Giglio claim -- as well as from our own analysis in
Giglio cases. In Giglio, the Supreme Court found reversal of the defendant’s
conviction required because “the Government’s case depended almost entirely on
[the falsely testifying witness’s] testimony; without it there could have been no
22
indictment and no evidence to carry the case to the jury.” Giglio, 405 U.S. at 154.
In Napue, the Court again based its conclusion that reversal was required on its
“own evaluation of the record” before it. Napue, 360 U.S. at 272. Likewise, all of
our cases reversing convictions on the basis of Giglio error -- even those involving
false testimony that a witness did not cut a deal in exchange for his testimony --
have conducted careful analyses of the trial record to render a case-specific
conclusion as to whether the violation might have made a difference. See, e.g.,
United States v. Rivera Pedin, 861 F.2d 1522 (11th Cir. 1988); Haber v.
Wainwright, 756 F.2d 1520 (11th Cir. 1985); DeMarco v. United States, 928 F.2d
1074 (11th Cir. 1991); United States v. Sanfilippo, 564 F.2d 176 (5th Cir. 1977).
Neither the Supreme Court nor our Court has ever adopted a per se rule that
the government’s failure to correct a witness’s false testimony that he did not
receive any consideration in exchange for his testimony requires reversal of the
conviction. We can find no case on all fours with Ventura’s, and therefore cannot
conclude that the Florida Supreme Court’s treatment of Ventura’s Giglio claim was
“contrary to” clearly established federal law in the sense of reaching an opposite
result on materially indistinguishable facts.
The second inquiry -- whether the state court reached a conclusion of law
that contradicts one reached by the Supreme Court -- requires a careful analysis of
23
the Florida Supreme Court’s decision. In analyzing Ventura’s Giglio claim, the
Florida court began its analysis by correctly noting: “To establish a violation of
Giglio Ventura must show: ‘(1) that the testimony was false; (2) that the prosecutor
knew the testimony was false; and (3) that the statement was material.’” Ventura,
794 So. 2d at 562 (quoting Robinson v. State, 707 So. 2d 688, 693 (Fla. 1998)).
The court also accurately explained “that ‘[t]he thrust of Giglio and its progeny has
been to ensure that the jury know the facts that might motivate a witness in giving
testimony, and the prosecutor not fraudulently conceal such facts from the jury.”
Id. (quoting Robinson, 707 So. 2d at 693).
However, the Florida Supreme Court’s phrasing of Giglio’s materiality
standard deviates from the formulation favored by the United States Supreme
Court. As the Florida court explained it: “Under Giglio, a statement is material if
‘there is a reasonable probability that the false evidence may have affected the
judgment of the jury.’” Id. (quoting Routly v. State, 590 So. 2d 397, 400 (Fla.
1991)). Routly cites Giglio and Napue, but nevertheless formulates the standard in
this altered manner. The Florida Supreme Court’s formulation differs from
Giglio’s in three respects: it uses “a” instead of “any”; “probability” instead of
“likelihood”; and “may” instead of “could.”
The last distinction is immaterial, since “may” and “could” are both words
24
used to express “possibility.” Oxford English Dictionary (2d ed. 1989). Indeed,
“can” -- the present tense of “could” -- is “sometimes used interchangeably with
may.” Merriam-Webster’s Collegiate Dictionary (10th ed. 1994); accord Oxford
English Dictionary (2d ed. 1989) (listing “may” as a synonym of “can”). Nor do
we see a significant difference between “a” and “any.” “Any” is simply an
“indeterminate derivative” of “a” or “an,” “in which the idea of unity . . . is
subordinated to that of indifference as to the particular one or ones that may be
selected.” Oxford English Dictionary (2d ed. 1989). For purposes of articulating
the Giglio standard, this distinction is of no consequence.
The key issue is the Florida court’s use of the term “probability” instead of
“likelihood.” After careful review, we conclude that because the court proceeded
to describe the Giglio standard accurately, and because AEDPA plainly requires us
to give state courts the benefit of the doubt on habeas review, the Florida court’s
single use of the phrase “a reasonable probability” does not render its decision
contrary to clearly established Supreme Court law. Insofar as the district court
suggested that the Florida Supreme Court erroneously conflated Giglio’s standard
with Brady’s, we disagree, since the Florida court understood the critical difference
between the two and correctly applied the more lenient Giglio standard.
The Florida Supreme Court explicitly acknowledged the essential difference
25
between the Giglio standard that applied to Ventura’s claim and the “reasonable
probability” standard that governs more general Brady claims -- namely, that
Giglio sets the materiality bar lower than Brady. The Florida Supreme Court
observed the distinction this way:
In denying Ventura’s claim, the trial court incorrectly relied on the
materiality standard appropriate to Brady claims. See United States v.
Alzate, 47 F.3d 1103, 1109-10 (11th Cir. 1995) (“Where there has
been a suppression of favorable evidence in violation of Brady v.
Maryland, the nondisclosed evidence is material: ‘if there is a
reasonable probability that, had the evidence been disclosed to the
defense, the result of the proceeding would have been different.’ . . . A
different and more defense-friendly standard of materiality applies
where the prosecutor knowingly used perjured testimony, or failed to
correct what he subsequently learned was false testimony.”) (citations
omitted).
Ventura, 794 So. 2d at 563.
Indeed, the Alzate case on which the Florida court relied recites the Giglio
standard in exactly the terms the United States Supreme Court has repeatedly used:
[T]he falsehood is deemed to be material “if there is any reasonable
likelihood that the false testimony could have affected the judgment of
the jury.” United States v. Agurs, 427 U.S. 97, 103, 96 S. Ct. 2392,
2397, 49 L. Ed. 2d 342 (1976) (emphasis added); accord Giglio v.
United States, 405 U.S. 150, 154, 92 S. Ct. 763, 766, 31 L. Ed. 2d 104(1972); Napue v. Illinois, 360 U.S. 264, 271, 79 S. Ct. 1173, 1178, 3
L. Ed. 2d 1217 (1959).
Alzate, 47 F.3d at 1110.5
5
The Florida Supreme Court also suggested that in determining whether the Giglio
standard is satisfied, “courts must focus on whether the favorable evidence could reasonably be
26
The Florida Supreme Court concluded that “based on this record of ample
impeachment and corroboration, we hold the evidence of the deal immaterial under
Giglio.” Ventura, 794 So. 2d at 565. In reaching this result, the court cited two
cases -- its own decisions in Routly v. State, 590 So. 2d 397 (Fla. 1991), and White
v. State, 729 So. 2d 909 (Fla. 1999) -- that recited the standard in terms of a
“reasonable probability,” but also another case -- United States v. Petrillo, 821 F.2d
85 (2d Cir. 1987) -- that used the well-established “any reasonable likelihood”
formulation. Petrillo explained: “Where the government uses perjured testimony
in obtaining a conviction, and knew or should have known of the perjury, the
defendant’s due process right to a fair trial is violated ‘if there is any reasonable
likelihood that the false testimony could have affected the judgment of the jury.’”
Id. at 87 (quoting Agurs, 427 U.S. at 103).
Against this backdrop, we cannot say that the state court’s imprecision in
explaining the Giglio standard for materiality means that it “arrive[d] at a
taken to put the whole case in such a different light as to undermine confidence in the verdict.”
Ventura, 794 So. 2d at 563 (quoting White v. State, 729 So. 2d 909, 913 (Fla. 1999)). Although
no Supreme Court case explains the Giglio standard in these terms, it bears noting that our Court
has, on at least one occasion, explained the Giglio standard using nearly identical terms. In
United States v. Dickerson, 248 F.3d 1036, 1041 (11th Cir. 2001), a panel of this Court stated:
“The materiality element [of Giglio] is satisfied if the false testimony ‘could reasonably be taken
to put the whole case in such a different light as to undermine confidence in the verdict.’” Id. at
1041 (quoting Strickler v. Greene, 527 U.S. 263, 290 (1999)) (internal quotation marks omitted);
see also Carr v. Schofield, 364 F.3d 1246, 1255 (11th Cir. 2004) (explaining the standard as
whether “the defendant failed to receive ‘a trial resulting in a verdict worthy of confidence’”
(quoting Kyles, 514 U.S. at 434).
27
conclusion opposite to that reached by [the Supreme] Court on a question of law,”
Williams, 529 U.S. at 412, or -- as this part of the “contrary to” standard has
alternatively been stated -- that it “applie[d] a rule that contradicts the governing
law set forth in [Supreme Court] cases,” Bell, 125 S. Ct. at 851.
To be sure, a “reasonable probability” standard differs from a “reasonable
likelihood” standard. See, e.g., Alzate, 47 F.3d at 1110 n.7 (noting that “the
district court applied the ‘reasonable probability of a different result’ standard,”
which “is substantially more difficult for a defendant to meet than the ‘could have
affected’ standard we apply”); Stephens v. Hall, 407 F.3d 1195, 1206 (11th Cir.
2005) (“This reasonable likelihood standard imposes a ‘considerably less onerous’
burden on [the petitioner] than the Brady standard.”).
However, the fact that the state court failed to track precisely the language
used by the Supreme Court does not mean that it applied the wrong standard here.
Indeed, the fact that the state court explicitly recognized the critical difference
between the Giglio and Brady standards strongly suggests that it did indeed apply
the more lenient Giglio standard. The Florida Supreme Court criticized the state
trial court for improperly applying Brady’s standard rather than Giglio’s, and cited
our Alzate case for the proposition that the Giglio standard is more defense
friendly than the Brady standard.
28
Moreover, neither the U.S. Supreme Court nor this Court has ever attempted
to articulate precisely what the difference is between the Brady standard and the
Giglio standard, and we doubt whether the distinction between a “reasonable
probability” and a “reasonable likelihood” is readily susceptible of quantification.
It has always sufficed for our purposes to observe that one standard is appreciably
more stringent than the other. Alzate, for example, called the Brady standard
“substantially more difficult for a defendant to meet than the ‘could have affected’
standard we apply.” Alzate, 47 F.3d at 1110 n.7. Our opinion in Stephens v. Hall
called Giglio’s standard “considerably less onerous” than Brady’s. Stephens, 407
F.3d at 1206 (quoting Kirkpatrick v. Whitley, 992 F.2d 491, 497 (5th Cir. 1993)).
And the Florida Supreme Court plainly captured the essence of this inexact
distinction when it observed that the trial court’s reliance “on the materiality
standard appropriate to Brady claims” was “incorrect[],” and cited Alzate for the
proposition that a “different and more defense-friendly standard of materiality
applies” to Giglio claims. Ventura, 794 So. 2d at 563.
We add that the phrase “contrary to,” as the Supreme Court explained in
Williams v. Taylor, means “diametrically different,” “opposite in character or
nature,” or “mutually opposed.” 529 U.S. at 405. The Court offered the following
example of what would constitute a decision “contrary to” governing law:
29
Take, for example, our decision in Strickland v. Washington, 466 U.S.
668, 104 S. Ct. 2052, 80 L. Ed. 2d 674 (1984). If a state court were to
reject a prisoner’s claim of ineffective assistance of counsel on the
grounds that the prisoner had not established by a preponderance of
the evidence that the result of his criminal proceeding would have
been different, that decision would be “diametrically different,”
“opposite in character or nature,” and “mutually opposed” to our
clearly established precedent because we held in Strickland that the
prisoner need only demonstrate a “reasonable probability that . . . the
result of the proceeding would have been different.” Id., at 694, 104
S. Ct. 2052.
Id. at 405-06. In contrast, a state court decision that is merely “contrary to the
federal court’s conception of how Strickland ought to be applied in that particular
case . . . is not ‘mutually opposed’ to Strickland itself.” Id. at 406.
A misstatement of a standard whose substance a court applies correctly falls
within the latter category. When a court recognizes the substantive difference
between two competing standards but conflates their language, the resulting
decision is not “diametrically different” from clearly established precedent. A
habeas court’s focus is properly on the substance rather than the form of the state
court’s decision. See, e.g., Mitchell v. Esparza, 540 U.S. 12, 16, 124 S. Ct. 7, 157
L. Ed. 2d 263 (2003) (“A state court’s decision is not ‘contrary to . . . clearly
established Federal law’ simply because the court did not cite our opinions. We
have held that a state court need not even be aware of our precedents, ‘so long as
neither the reasoning nor the result of the state-court decision contradicts them.’”
30
(citation omitted) (quoting Early v. Packer, 537 U.S. 3, 7-8, 123 S. Ct. 362, 154 L.
Ed. 2d 263 (2002))); Bell, 125 S. Ct. at 853.
Finally, the Supreme Court has explained that Ҥ 2254(d) requires that
state-court decisions be given the benefit of the doubt. Readiness to attribute error
is inconsistent with the presumption that state courts know and follow the law.”
Holland v. Jackson, 124 S. Ct. 2736, 2739, 159 L. Ed. 2d 683 (2004) (alteration,
citations, and internal quotation marks omitted). Thus, the Court has several times
reversed grants of federal habeas relief on the ground that the state court’s
incorrect articulation of the Strickland standard -- which requires a defendant
claiming ineffective assistance of counsel to demonstrate “a reasonable probability
that, but for counsel’s unprofessional errors, the result of the proceeding would
have been different,” Strickland, 466 U.S. at 694 (emphasis added) -- was
“contrary to” clearly established Supreme Court precedent. See, e.g., Woodford v.
Visciotti, 537 U.S. 19, 123 S. Ct. 357, 154 L. Ed. 2d 279 (2002); Holland v.
Jackson, 124 S.Ct. 2736, 2738-39, 159 L. Ed. 2d 683 (2004).6
6
In Woodford, the Court observed that the state court’s “occasional shorthand reference
to that standard by use of the term ‘probable’ without the modifier may perhaps be imprecise,
but if so it can no more be considered a repudiation of the standard than can this Court’s own
occasional indulgence in the same imprecision.” Id. at 23-24 (citing Mickens v. Taylor, 535
U.S. 126, 166, 122 S. Ct. 1237, 152 L. Ed. 2d 291 (2002) (using the phrase “probable effect upon
the outcome”), and Williams, 529 U.S. at 393 (using the phrase “probably affected the
outcome”)).
This imprecision, the Court held, did not warrant the conclusion that the state court’s
31
These decisions stand in contrast to Williams v. Taylor, where the Court
found that the state supreme court’s application of an incorrect Strickland standard
was “contrary to” clearly established law, since the state court’s decision “turned
on” its erroneous conception of the relevant standard. The Supreme Court
explained: “While [the state court] may also have conducted an ‘outcome
determinative’ analysis of its own, it is evident to us that the court’s decision
turned on its erroneous view that a ‘mere’ difference in outcome is not sufficient to
establish constitutionally ineffective assistance of counsel.” Williams, 529 U.S. at
397 (emphasis added) (citation omitted).
These cases suggest that a misstatement of the law may be insufficient to
find a state court decision contrary to clearly established law; rather, the state
decision was contrary to clearly established law, and indeed the Court criticized the federal
appeals court for not giving the benefit of the doubt to the state court:
The Court of Appeals made no effort to reconcile the state court’s use of the term
“probable” with its use, elsewhere, of Strickland’s term “reasonably probable,”
nor did it even acknowledge, much less discuss, the California Supreme Court’s
proper framing of the question as whether the evidence “undermines confidence”
in the outcome of the sentencing proceeding. This readiness to attribute error is
inconsistent with the presumption that state courts know and follow the law. It is
also incompatible with § 2254(d)’s “highly deferential standard for evaluating
state-court rulings,” Lindh v. Murphy, 521 U.S. 320, 333, n. 7, 117 S. Ct. 2059,
138 L. Ed. 2d 481 (1997), which demands that state-court decisions be given the
benefit of the doubt.
Id. at 24.
32
court’s decision must have resulted from its application of an incorrect standard.7
In this case, the Florida Supreme Court understood the law and applied it correctly.
The Florida court’s decision is undeniably entitled to deference and is not contrary
to clearly established Supreme Court precedent. The only remaining question is
whether the state court decision was an “unreasonable application” of such
precedent.
C.
Under § 2254(d)’s “unreasonable application” clause, “a federal habeas
court may grant the writ if the state court identifies the correct governing legal
principle from [the Supreme] Court’s decisions but unreasonably applies that
principle to the facts of the prisoner’s case.” Williams, 529 U.S. at 412-13.
However, “a federal habeas court may not issue the writ simply because that court
7
This conclusion is consistent with the numerous cases in our Court concluding that
federal courts are required to defer to state court decisions even when they offer no explanation
or reasoning as to how they reached their result. As we have explained: “The plain language of
§ 2254(d)(1) requires only that the federal claim have been ‘adjudicated on the merits in State
court proceedings’ and have ‘resulted in a decision’ that is neither contrary to nor involves an
unreasonable application of Supreme Court precedent. That is all the text of the provision
requires.” Wright v. Moore, 278 F.3d 1245, 1254 (11th Cir. 2002). “The statutory language
focuses on the result, not on the reasoning that led to the result, and nothing in that language
requires the state court adjudication that has resulted in a decision to be accompanied by an
opinion that explains the state court’s rationale. Accordingly, all that is required is a rejection of
the claim on the merits, not an explanation.” Id. at 1255 (citations omitted). Accord Herring v.
Sec’y, Dep’t of Corrs., 397 F.3d 1338, 1347 (11th Cir. 2005) (“To be entitled to deferential
review under 28 U.S.C. § 2254, it is not necessary for the state court to explain its decision.
Even a summary, unexplicated rejection of a federal claim qualifies as an adjudication entitled to
deference under § 2254(d).”); Parker v. Sec’y, Dep’t of Corrs., 331 F.3d 764, 775-76 (11th Cir.
2003); Isaacs v. Head, 300 F.3d 1232, 1259-60 (11th Cir. 2002).
33
concludes in its independent judgment that the state-court decision applied [the
governing legal principle] incorrectly. Rather, it is the habeas applicant’s burden
to show that the state court applied [that principle] to the facts of his case in an
objectively unreasonable manner. An unreasonable application of federal law is
different from an incorrect application of federal law.” Woodford, 537 U.S. at 24-
25 (citations and internal quotation marks omitted).
Put differently, the reviewing court may not “substitute[] its own judgment
for that of the state court.” Id. at 25; see also, e.g., Mitchell, 540 U.S. at 18 (“We
may not grant respondent’s habeas petition, however, if the state court simply erred
in concluding that the State’s errors were harmless; rather, habeas relief is
appropriate only if the [state court] applied harmless-error review in an
‘objectively unreasonable’ manner.”); Lockyer v. Andrade, 538 U.S. 63, 75, 123 S.
Ct. 1166, 155 L. Ed. 2d 144 (2003) (“The gloss of clear error fails to give proper
deference to state courts by conflating error (even clear error) with
unreasonableness.”).
In this case, the Florida Supreme Court identified the correct governing legal
principle and did not apply it unreasonably. Its conclusion that there was no
reasonable likelihood that the false testimony presented at Ventura’s trial could
have affected the judgment of the jury was not “objectively unreasonable” in light
34
of the extensive and powerful corroborating evidence introduced and the
substantial impeachment of McDonald at trial.8
As for corroboration, the state’s evidence linking Ventura to Clemente’s
murder consisted of McDonald’s testimony; the testimony of three other witnesses:
Joseph Pike, Reginald Barrett, and Timothy Arview; and documentary evidence
including motel receipts and a record of a wire transfer placing Ventura -- then a
resident of Chicago -- in the vicinity of the murder at the time Clemente was killed.
McDonald testified at length about the details of Clemente’s murder. He
explained that Wright was having financial problems and recruited him to kill
Clemente in exchange for half of the proceeds of a key man life insurance policy
Wright held on Clemente. McDonald then brought in Ventura to commit the
murder in exchange for half of McDonald’s share of the proceeds. McDonald said
that he met with Ventura in Atlanta in April 1981, a few days before the murder, to
work out the details. They planned that Ventura would travel to Daytona Beach,
where McDonald would meet him. McDonald supplied Ventura with some money
8
Ventura makes a great deal of the fact that prosecutor Stark, in the state-court hearing on
Ventura’s 3.850 motion, referred to McDonald numerous times as the state’s “star witness.”
Indeed, the State concededly felt that it could not proceed with the prosecution of either
McDonald or Ventura without the other. Nevertheless, the State’s case against Ventura was
compelling even in light of McDonald’s false testimony, since the numerous sources of powerful
corroboration make it implausible that the jury would have wholly discarded McDonald’s
testimony had they known that he was testifying pursuant to a deal. We cannot substitute one
participant’s conclusory characterization of a witness’s testimony for a thorough review of all of
the evidence presented at trial.
35
for expenses, and later wired him an additional $100 in DeLand, Florida, when
Ventura ran out of funds.
McDonald testified that he met Ventura outside the Days Inn on April 15,
1981, and they drove to Barnett Bank in DeLand. Ventura had called Clemente,
who was a boat salesman at Crow’s Bluff Marina, and, posing as a potential
customer, arranged to meet him at the bank. Clemente arrived in a black pick-up
truck and left with Ventura. McDonald followed along at a distance in his own
car. McDonald and Ventura had planned that Ventura would ask Clemente to pull
off the road at a prearranged spot along state route 44 so that Ventura could relieve
himself, and Ventura would then shoot Clemente. Clemente’s truck pulled off the
road at the designated spot, and McDonald waited nearby. About ten minutes later,
Ventura returned to McDonald’s car. McDonald dropped Ventura off at a
restaurant and proceeded to Daytona Beach to see Wright about getting $2000 for
Ventura, who wanted money to travel. According to McDonald, Ventura planned
to go to California to see his daughter.
McDonald collected $2000 and met Ventura at the Atlanta airport to give
him part of it. At that meeting, McDonald said, he informed Ventura that it would
take thirty to sixty days to collect the insurance proceeds. McDonald returned to
Daytona Beach around June 20 when he learned that Wright was about to be paid
36
by the insurance company, and was arrested for the murder upon his arrival.
Interlocking physical and testimonial evidence introduced at trial
corroborated most of the critical aspects of McDonald’s account of the murder.
First, McDonald’s testimony that Jerry Wright had taken out a key man insurance
policy on Clemente and that the purpose of this murder-for-hire was to collect the
proceeds is corroborated on numerous levels. A representative of Midwestern
National Life in Cleveland, Ohio confirmed that Jerry Wright held a key man life
insurance policy on Robert Clemente issued by Midwestern National Life on July
2, 1980, and that he filed a claim on the policy shortly after the murder, on April
25, 1981. Police searching Ventura’s residence in Maywood, Illinois after his
arrest found a piece of paper inside Ventura’s “Weekly Minder” with the name
“Midwestern Life” written on it. The record evidence affords no innocent
explanation for this physical evidence linking Ventura to the insurance company
that issued the policy for which Clemente was killed.
Moreover, Reginald Barrett, a neighbor and acquaintance of Ventura’s,
testified that in February 1981 -- approximately two months prior to the murder --
the petitioner, Ventura, asked him to contact Midwestern Life Insurance in Ohio
“to inquire about a certain insurance called ‘keyman insurance.’” Specifically,
Ventura wanted to know how long it would take to pay a claim, and, notably, “if an
37
employee . . . was insured under a keyman policy, but was to leave the place of
employment, would this insurance policy still cover him after he had left.”
(Clemente had left Wright’s employ and worked for Crow’s Bluff Marina at the
time of his death.) Barrett never placed the call, but a representative of
Midwestern National Life confirmed at trial that the fact Clemente no longer
worked for Wright at the time of his death did not affect Wright’s ability to make a
claim under the policy. Again, the record evidence offers no benign explanation
for why Ventura would ask his friend Barrett, some two months before the murder,
to inquire of the insurance company whether it would honor a key man insurance
policy after the employee had left his place of employment.
Joseph Pike, still another acquaintance and one-time business partner of
Ventura, provided further independent corroboration of McDonald’s account of the
murder. Pike testified that on May 6, 1981 -- shortly after the murder -- Ventura
told Pike about a plan that Ventura, McDonald, and another acquaintance of
McDonald’s had devised to murder a man and collect a key man insurance policy
on his life. When Pike directly asked what Ventura’s role was, Ventura explained
that he “handled the extermination.”
McDonald’s testimony that he met with Ventura in Atlanta in April 1981,
and arranged to meet him shortly thereafter in Daytona Beach, was also
38
corroborated by documentary and testimonial evidence. Among other things,
Barrett testified that in late February or early March of 1981, Ventura asked him
for a gun, which he said he needed for “Jack in Atlanta.” (Barrett in fact provided
Ventura with a gun, but not the one used to kill Clemente.) Then, during the first
week of April, just before the murder, Ventura told Barrett that he needed some
money “to go to Atlanta to meet with Jack.” Barrett helped Ventura to get a job
that paid him $250, and the next day Ventura told Barrett “that he was on his way
to Atlanta” because, “Jack wanted him to come down and burn someone,” which
Barrett took to mean “to murder someone.”
Subsequently, around April 10 or 11, Ventura left a message asking Barrett
to call him back. When Barrett returned the call, he learned that the phone number
Ventura had left was for a motel in Daytona or DeLand, Florida. At that time,
Ventura asked Barrett to take some money to Ventura’s son, explaining that he was
still doing business and would call when he was finished.
Motel records introduced at trial further corroborated McDonald’s testimony
that Ventura traveled to the Atlanta area in early April and then continued on to the
Daytona area, where he was staying when Clemente was murdered on April 15.
Indeed, motel records place Ventura in the Atlanta area from April 5-7 and 9-10,
and then in Volusia County from April 10-18. Records from Motel 1 in Chamblee,
39
Georgia -- just outside Atlanta -- showed that Ventura had checked in April 5,
departing April 7, and that he checked in again April 9, departing April 10.
Another motel registration card established that Ventura had registered at the
Boulevard Motel in DeLand the nights of April 10, 11, and 12, 1981. Records
from a Days Inn in the Daytona area established that Ventura had stayed there the
nights of April 13 and 14, checking out April 15, 1981. More records from Motel
1 in Chamblee showed that Ventura checked back into that motel on April 16,
departing April 18. On this record, we can find no benign explanation for why
Ventura, a resident of the Chicago area, had been in Atlanta or Volusia County
during this period.
The records from the Days Inn, where Ventura stayed from April 13-15, also
corroborate McDonald’s testimony that he picked up Ventura in front of the Days
Inn on April 15, the day of Clemente’s murder. In addition, Corporal David
Hudson of the Volusia County Sheriff’s Department testified that he checked out a
phone number from which Reginald Barrett had received a call around 9:00 a.m.
on April 15, and traced the number to a pay phone outside the same Days Inn.
The records indicating that Ventura stayed at the Boulevard Motel the nights
of April 10-12 interlock with Barrett’s testimony that he reached a motel in
Daytona or DeLand when he dialed the return number Ventura had left when he
40
called Barrett around April 10 or 11. Corporal Hudson further verified that his
investigation revealed that Barrett had placed a return call to Ventura at the
Boulevard Motel, although he could not remember the precise date of the call.
Additional documentary evidence corroborated McDonald’s testimony that
he wired $100 to Ventura in DeLand, Florida, when Ventura ran out of expense
money. Western Union records introduced at trial showed that on April 13, Peter
Ventura picked up a $100 money order, which had been sent on April 12 by Jack
McDonald, at an office located in DeLand’s Greyhound bus station.
In addition, McDonald’s account of the details of the Clemente murder was
entirely consistent with the forensic evidence presented at trial and with the
testimony of several of Clemente’s co-workers. Several law enforcement officers
dispatched to the crime scene reported that Clemente’s body was recovered in a
secluded area off state route 44; that it was found in a black pick-up truck bearing
the Crow’s Bluff Marina insignia; and that Clemente had been shot to death.
Denise Jorgenson, the owner of Crow’s Bluff Marina, confirmed that Clemente
had told her on the day of his murder that he was meeting a client at Barnett Bank,
an occurrence Jorgenson considered unusual since clients typically came to the
marina. Charles W. Coccia, a mechanic at the marina, testified that he lent
Clemente his company truck to go meet a client on the day of Clemente’s murder.
41
Coccia identified the truck in which Clemente’s body was found as the one he had
given Clemente.
The delay McDonald recounted in collecting the insurance proceeds after the
murder was also corroborated by the testimony of both Pike and Barrett. Pike
testified that when Ventura told him about the murder in early May, “there were
indications that there were delays in collecting,” and Ventura “was expecting to
collect it within a few weeks.” Barrett testified that in early May, upon Ventura’s
return from a trip to California, Ventura mentioned “that he would be getting his
payment in about thirty days,” when “Mr. McDonald was to receive some funds
from an insurance proceed in Florida.” Then, in late May, Ventura told Barrett that
“Jack should be coming to Chicago within just a few days to make a payment” for
the job Ventura had done. Ventura explained that “there was some delay” in
receiving the money that “Jack . . . was going to collect . . . from someone who had
an insurance policy.” A representative of the insurance company, Midwestern
National Life, confirmed that the key man insurance claim had been filed on April
25, 1981, and that payment had been issued initially on June 19, 1981, but that
payment on the check had been stopped and that the claim had not been paid until
September 14, 1981.
Barrett also confirmed McDonald’s testimony that Ventura intended to
42
travel to California after the murder. Ventura called Barrett around April 19 or 20,
1981, and said that he would be going to California as soon as his job was finished.
Ventura called Barrett again in early May “indicating that he had gotten back from
California,” and Barrett picked him up at the bus station.
A letter Ventura sent to Barrett further corroborates Ventura’s involvement
in the murder with McDonald. Around May 4 or 5, when Barrett returned from a
trip, he had received a letter addressed to him at his office in Maywood, Illinois,
postmarked April 17. The return address read, “P. Ventura, Atlanta, Georgia.”
The letter, which was introduced at trial, stated, “In the event something should
happen to me! Jack McDonald is responsible.” The letter then supplied
McDonald’s phone number. Included in the mailing envelope along with the letter
was an envelope bearing the return address of Motel 1 -- the motel in Chamblee,
Georgia where Ventura was staying on April 17. On that envelope was written,
“Jack lives near here.”
Ventura’s subsequent flight after his indictment for the Clemente murder is
further evidence of his role in the crime. Ventura was released on $50,000 bond
after his indictment in late July, 1981, and failed to appear at a hearing scheduled
for August 18, 1981. In spite of the efforts of state and federal law enforcement to
track him down, Ventura remained a fugitive for five years. He lived under several
43
assumed names, including Juan Gadaya and Juan Contras, until he was
apprehended in Austin, Texas in June 1986. As we have noted repeatedly,
“evidence of resistance to arrest and flight is admissible to demonstrate
consciousness of guilt and thereby guilt.” United States v. Wright, 392 F.3d 1269,
1277 (11th Cir. 2004) (quoting United States v. DeParias, 805 F.2d 1447, 1454
(11th Cir. 1986)); see also, e.g., United States v. Frazier, 387 F.3d 1244, 1266 n.20
(11th Cir. 2004) (en banc) (finding that defendant’s flight from police “was strong
evidence of consciousness of guilt”); United States v. Beard, 775 F.2d 1577, 1581
(11th Cir. 1985) (“There is no question that evidence of flight can raise a
permissive inference of consciousness of guilt of the crime charged.”).
Finally, Ventura’s role in the murder was independently corroborated by the
trial testimony of Timothy Arview, a young man who became acquainted with
Ventura while Ventura was living in Austin, Texas as a fugitive. Arview, who did
not know McDonald, Barrett, or Pike, testified at trial that he first met Ventura,
who was using the name Juan Gadaya, in December 1985 or January 1986, and
worked for him for six or seven months doing construction. Around April or May
of 1986, Arview and Ventura were wrestling and “playing around” until Ventura
“got upset and started playing roughly.” Ventura then specifically told Arview that
he had killed a man in Florida five years earlier, and that it was a contract killing.
44
Lieutenant Juan Gonzales of the Austin Police Department confirmed that Arview
reported this information -- which matched the date, location, and nature of the
Clemente killing exactly -- to Austin police, which in turn led to Ventura’s rearrest.
In addition to this extensive chain of corroborating evidence, the Florida
Supreme Court took into account the fact that McDonald was substantially
impeached in a variety of ways at trial. First, McDonald was impeached by his
admitted involvement in the Clemente murder. McDonald testified at trial that he
had been indicted for the murder along with Ventura, but that he had escaped
prosecution because of Florida’s speedy trial rule.
Second, McDonald was impeached by evidence of his prior convictions.
McDonald conceded on direct examination that after his release on speedy trial
grounds, he had been indicted by the federal government and had pled guilty, in
1983, to four counts of postal theft. He was sentenced in May or June of 1983 to a
term of imprisonment of fifteen years, then fled before he started serving his
sentence, and was not taken back into custody until September 23, 1987. On cross-
examination, McDonald also acknowledged another prior conviction, for
embezzlement involving bank fraud, in 1964. Finally, McDonald admitted that he
had been involved in the illegal sale of a load of stolen truck tires to Mr. Wright,
although he was apparently never prosecuted for this offense.
45
Third -- and perhaps most significantly -- McDonald was extensively
impeached on cross-examination with his own prior inconsistent deposition
testimony. McDonald was deposed at the Sheraton Inn at the Atlanta Airport on
May 4, 1983, by the prosecutor, Stark. At that deposition, McDonald said that he
had not been with Ventura on the day of the murder, in direct contradiction of his
trial testimony. At trial, McDonald admitted that in spite of being sworn to testify
truthfully, he “did not tell the truth” in that deposition. McDonald explained that
his “intention was just to leave some clouds in the issue,” and that perhaps he “did
not want to face the reality of being virtually, you might say, at the scene.”
Defense counsel chipped further at McDonald’s credibility by insinuating that he
had lied in the deposition “because [he was], in fact, at the scene and did commit
the murder,” a suggestion that McDonald emphatically denied.
Finally, the defense attacked McDonald’s motive for testifying by
suggesting that McDonald harbored “some feeling of rancor towards Mr. Ventura
as a result of the Federal bank scam that resulted in [McDonald’s] conviction.”
Again, McDonald flatly denied this. Defense counsel then asked whether
McDonald had agreed to testify because he was angry that Wright never paid him
for his participation in the murder, another suggestion that McDonald strongly
denied.
46
We are bound by the state court’s decision unless it was not only erroneous
but “objectively unreasonable.” Woodford, 537 U.S. at 24-25. As the Supreme
Court has admonished, a federal habeas court is not free to “substitute[] its own
judgment for that of the state court.” Id. at 25. We cannot, in this case,
characterize as unreasonable the state court’s determination that the corroboration
and impeachment were so substantial that there was no reasonable likelihood
McDonald’s false testimony could have affected the judgment of the jury.
Virtually every step of McDonald’s testimony as he took the jury through
the planning, execution, and aftermath of Clemente’s murder was corroborated by
documentary and testimonial evidence. The multiple admissions Ventura made
over many years to Barrett, Pike, and Arview, as well as the testimony of
Clemente’s co-workers and of police who responded to the crime scene, mirrored
McDonald’s account of the crime. Motel receipts, Western Union records, and
phone calls placed Ventura -- a Chicago-area resident -- in Atlanta and in Volusia
County at precisely the times McDonald said he was there, including the time of
Clemente’s murder. A conversation between Ventura and Barrett and the note
found in Ventura’s Weekly Minder directly linked Ventura to the insurance
company that issued the key man policy on Clemente’s life. A subsequent note in
which Ventura suggested McDonald might have it in for him, as well as Ventura’s
47
flight after his indictment, further indicated Ventura’s involvement in the murder.
In addition, McDonald was extensively impeached by evidence of his own
involvement in the Clemente murder, his prior felony fraud convictions and bond
jumping, and his prior false deposition testimony in the Clemente case, as well as
by the implication that he was motivated to testify by ill will toward Ventura and
Wright.9
The substantial volume of interlocking corroborating and impeachment
evidence soundly supports the Florida Supreme Court’s conclusion that a jury
would not have discounted McDonald’s testimony even if it knew he had been
promised a deal on potential bond-jumping charges in exchange for his testimony,
and thus that there was no reasonable likelihood that McDonald’s false testimony
could have affected the judgment of the jury. We therefore agree with the district
court that the Florida Supreme Court’s disposition of Ventura’s Giglio claim was
not an unreasonable application of clearly established federal law.
9
The effect of these other means of impeachment was not, however, to render evidence of
McDonald’s deal cumulative. See, e.g., United States v. Sanfilippo, 564 F.2d 176, 178 (5th Cir.
1977) (“The fact that the history of a witness shows that he might be dishonest does not render
cumulative evidence that the prosecution promised immunity for testimony. A jury may very
well give great weight to a precise reason to doubt credibility when the witness has been shown
to be the kind of person who might perjure himself.”); accord Brown, 785 F.2d at 1466. Rather,
the other sources of impeachment are simply relevant to the determination of whether there is
any reasonable likelihood that evidence of McDonald’s deal could have affected the judgment of
the jury.
48
IV.
The conduct of the prosecutor in this case was inexcusable, and wholly
discordant with his paramount obligation to seek the truth. Nevertheless, the
Florida Supreme Court’s determination that McDonald’s false testimony was not
“material,” as required by Giglio, was neither contrary to nor an unreasonable
application of clearly established federal law. We, therefore, affirm the district
court’s denial of Ventura’s petition for a writ of habeas corpus.
AFFIRMED.
49