PUBLISHED

UNITED STATES COURT OF APPEALS

FOR THE FOURTH CIRCUIT

UNITED STATES OF AMERICA,
Plaintiff-Appellee,

v.                                                                    No. 95-5359

CARLOS TREVINO,
Defendant-Appellant.

Appeal from the United States District Court
for the Eastern District of North Carolina, at Raleigh.
James C. Fox, Chief District Judge.
(CR-94-78-F)

Argued: February 2, 1996

Decided: July 12, 1996

Before HALL and HAMILTON, Circuit Judges, and PHILLIPS,
Senior Circuit Judge.

_________________________________________________________________

Affirmed by published opinion. Judge Hall wrote the majority opin-
ion, in which Judge Hamilton joined. Judge Phillips wrote a concur-
ring and dissenting opinion.

_________________________________________________________________

COUNSEL

ARGUED: George Alan DuBois, Assistant Federal Public Defender,
Raleigh, North Carolina, for Appellant. John Samuel Bowler, Assis-
tant United States Attorney, Raleigh, North Carolina, for Appellee.
ON BRIEF: Janice McKenzie Cole, United States Attorney, J. Doug-
las McCullough, Assistant United States Attorney, Raleigh, North
Carolina, for Appellee.
OPINION

HALL, Circuit Judge:

A jury found Carlos Trevino guilty of conspiring to traffic in mari-
juana and of traveling in interstate commerce in aid of racketeering.
Trevino appeals the convictions, contending that the district court
erred by denying his motion to disclose the Presentence Investigation
Reports (PSRs) of conspiracy members who testified against him.
Neither Trevino nor his counsel have seen the reports; the district
court examined the PSRs in camera prior to announcing its decision.
On appeal, we have reviewed the PSRs at issue, and we are satisfied
that the district court's ruling was not clearly erroneous; we therefore
affirm.

I.

According to the government, Trevino was involved in a marijuana
distribution conspiracy headed by Stephen Wilson. Wilson began sell-
ing marijuana in 1983 after his farming business became unprofitable;
he continued to traffic in large quantities of marijuana until sometime
in 1992 or 1993. On December 6, 1994, the grand jury indicted
Trevino, charging that he participated in the conspiracy, and that he
traveled in interstate commerce with the intent to facilitate an unlaw-
ful activity. See 21 U.S.C.A § 846 (West Supp. 1996); 18 U.S.C.A.
§ 1952 (West Supp. 1996).

The government's case against Trevino consisted of the testimony
of twelve witnesses; as many as nine of them were connected with the
conspiracy. Prior to trial, Trevino requested that the district court
release the presentence reports of seven eventual witnesses who had
previously entered into plea agreements with the government. The
court, after obtaining the PSRs and examining them in camera, denied
Trevino's request.

At trial, Wilson told how he had been introduced to marijuana
growers and dealers by a companion he had met in Florida, where he
had gone to purchase tomatoes to resell in North Carolina. Wilson
took his tomato profits and bought approximately twenty pounds of

                    2
marijuana from Joe Munyos in southern Texas. Wilson's new busi-
ness flourished, and he began to hire couriers to transport large quan-
tities of marijuana purchased from Munyos or another supplier,
Freddy Gonzales.

Wilson testified that he met Trevino in 1988 when Trevino accom-
panied Munyos on a trip to North Carolina. Wilson stated that he ini-
tially hired Trevino, who speaks Spanish, to help oversee the Mexican
migrants who worked on Wilson's farm. According to Wilson,
Trevino eventually became his "right-hand man" in the drug business,
coordinating the Texas-to-North Carolina runs. Wilson said that he
fired Trevino on January 1, 1991, upon learning that Trevino had
skimmed $16,000 from a drug payment that he was to deliver to Gon-
zales. Wilson estimated that, over the entire course of the conspiracy,
his operation had sold over fifteen tons of marijuana.

The other testifying co-conspirators detailed their sundry dealings
with Trevino. Some admitted having accompanied Trevino on drug-
buying trips; others mentioned having delivered drugs to him from
time to time.

Customs agent Michael Doherty concluded the government's case.
He testified that, during a telephone conversation, Trevino admitted
being involved with Wilson in the marijuana business, though simul-
taneously insisting that Wilson was the mastermind.

Trevino took the stand and denied any wrongdoing. The jury, how-
ever, decided otherwise; it found Trevino guilty of both charges, and
he was subsequently sentenced by the district court to 151 months'
imprisonment. Trevino appeals, maintaining that the court's denial of
his request for the PSRs warrants a new trial.

II.

A.

Due process requires that the government disclose to the accused
any favorable evidence in its possession that is material to guilt or
punishment. Brady v. Maryland, 

, 87 (1963).1 "Favorable"
_________________________________________________________________
1 Ordinarily, the accused must ask the government to produce Brady
materials; however, "there are situations in which evidence is obviously

                    3
evidence includes not only that evidence tending to exculpate the
accused, but also any evidence adversely affecting the credibility of
the government's witnesses. United States v. Bagley, 

,
676 (1985); Giglio v. United States, 

, 154-55 (1972),
citing Napue v. Illinois, 

(1959). 2 Evidence is "material"
if there is a reasonable probability that it will affect the result of the
proceeding. See Bagley at 682 (opinion of Blackmun, J.).

On occasion, the government may possess potential Brady material
that it deems privileged or that is otherwise confidential. If the
accused does not specifically request that it be produced, this material
is treated much like everything else in the government's file, i.e., "the
prosecutor's decision on disclosure is final." Pennsylvania v. Ritchie,

, 59 (1987). If, however, the accused is able to identify
the requested confidential material with some degree of specificity, he
may then attempt to convince the district court that it is subject to dis-
closure. See 

n.15 (requiring the accused to "at least make
some plausible showing" of how the evidence would be "both mate-
rial and favorable to his defense"), quoting United States v.
Valenzuela-Bernal, 

, 867 (1982); see also Love v.
Johnson, 

, 1313 (4th Cir. 1995) (district court's judg-
ment dismissing habeas petition vacated where state court declined to
undertake an in camera examination of subpoenaed materials, even
though the petitioner's showing "far exceeded that held sufficient by
the Ritchie Court.").
_________________________________________________________________

of such substantial value to the defense that elementary fairness requires
it to be disclosed even without a specific request." United States v.
Agurs, 

, 110 (1976).

2 Where the source of the potential impeachment material is the wit-
ness's own prior statement, the government is not required to disclose the
statement to the accused until concluding its direct examination of the
witness. 18 U.S.C.A. § 3500 (West 1985), codifying Jencks v. United
States, 

(1957); United States v. Lewis, 

, 151
(4th Cir. 1994).



                     4
Once the accused has made a plausible showing that the evidence
would be both material and favorable, the trial court must review the
information in camera to ascertain its true nature and determine
whether it must be disclosed. Ritchie at 58-60; Love at 1313.3 The
court conducts its examination in private because the Constitution
does not accord an accused the right of unrestricted access to the gov-
ernment's files. Ritchie at 59-60; Love at 1313; see also United States
v. Leung, 

, 583 (2d Cir. 1994) ("In the rare circumstances
where [an in camera] inspection is required, its purpose is not to pro-
vide a general discovery device for the defense[.]"). The trial court's
ultimate conclusion as to whether the information is subject to disclo-
sure -- whether the evidence is both material and favorable -- may
be disturbed on appeal only if it is clearly erroneous. United States v.
Mora, 

, 1139 (5th Cir.), cert. denied, 

(1993).

B.

Presentence reports represent a special subcategory of potentially
discoverable confidential information, and rules governing their dis-
closure have evolved apart from the relatively recent Ritchie decision
and the line of lower court cases that have followed. In United States
v. Figurski, 

(4th Cir. 1975), we held that presentence
reports prepared on behalf of government witnesses must be exam-
ined by the district court in camera before any of the information con-
tained therein is disclosed to the accused. 

We did not,
however, establish any guidelines to assist the courts in making the
necessary threshold determination, i.e., whether an accused has suffi-
ciently justified the court's involvement in the discovery process. In
view of the potential effect of Ritchie on the distinct body of law per-
taining to the disclosure of presentence reports, and in view of the
issue on appeal, the time is ripe to speak on the subject.
_________________________________________________________________
3 An in camera review is also warranted where the government, pursu-
ant to its duty under Agurs, see note 

, submits confidential mate-
rial to the court on its own initiative. See Mora, infra, at 1139 (agent's
field notes, not subject to disclosure under the Jencks Act, were produced
by the government and determined by the district court, following an in
camera review, to be not otherwise discoverable).

                    5
1.

The PSR may serve several purposes, but its principal function is
to assist the district court in imposing an appropriate sentence on the
criminal defendant who is the subject of the report. United States v.
Charmer Industries, Inc., 

, 1170 (2d Cir. 1983). Obvi-
ously, a PSR that is as accurate and complete as possible helps to
ensure a just sentence. PSRs must include, inter alia, pertinent infor-
mation about the defendant's life history and personal characteristics
(including any prior criminal record), the defendant's financial condi-
tion, a victim impact assessment, the probation officer's proposed
classification of the offense and the defendant's criminal history
under the Sentencing Guidelines, and "any other information required
by the court." Fed. R. Crim. P. 32(b)(4).

Indeed, concern over the accuracy of PSRs has resulted in several
amendments to Rule 32 over the past thirty years, most recently in
1994. Before 1975, most federal courts did not permit the defendant
to see his own PSR, much less comment on it. Fed. R. Crim. P. 32
advisory committee note (1974); see also Charmer at 1172. Rule
32(b) now provides that, except for the probation officer's final rec-
ommendation on the sentence (which the district court may direct be
excluded), the entire report must be provided to the defendant and his
counsel not less than 35 days prior to the sentencing hearing. Upon
receiving his PSR, the defendant is entitled to contest any material
information contained in or omitted from the report. Rule 32(b)(6)(B).

With the advent of the amendments requiring full disclosure to the
defendant, certain materials once commonly included in PSRs have
now been excluded. Rule 32(b)(5) requires the exclusion of "any
diagnostic opinions that, if disclosed, might seriously disrupt a pro-
gram of rehabilitation[,] . . . sources of information obtained upon a
promise of confidentiality[,] or . . . any other information that, if dis-
closed, might result in harm, physical or otherwise, to the defendant
or other persons." The district court may still receive and rely on such
material in determining sentence, but, if it does, the court must sum-
marize the information for the defendant and give him a reasonable
opportunity to comment. Fed. R. Crim. P. 32(c)(3)(A).

Notwithstanding that PSRs prepared in accordance with the 1994
amendments are somewhat watered-down versions of their earlier

                     6
incarnations, there are still valid reasons to prohibit the routine disclo-
sure of their contents to defendants in subsequent proceedings against
whom the subject of the report will testify. See U.S. Dep't of Justice
v. Julian, 

(1988):

          [I]n both civil and criminal cases the courts have been very
          reluctant to give third parties access to the presentence
          investigation report prepared for some other individual or
          individuals. . . . [O]ne reason for this is the fear that disclo-
          sure of the reports will have a chilling effect on the willing-
          ness of various individuals to contribute information that
          will be incorporated into the report. A second reason is the
          need to protect the confidentiality of the information con-
          tained in the report. Accordingly, the courts have typically
          required some showing of special need before they will
          allow a third party to obtain a copy of a presentence report.

(emphasis deleted) (citations omitted). Granted, the "chilling
effect" on contributors should be ameliorated somewhat now that
sources of confidential information can no longer be named in the
PSR. Nevertheless, we must point out that the rule only specifically
excludes mention of the source of the information; it does not neces-
sarily exclude the information itself.4 As the PSR is exposed to pair
upon pair of prying eyes, it becomes more likely that someone -- par-
ticularly a criminal confederate -- will be able to identify the source
of sensitive information contained in the report. 5
_________________________________________________________________
4 The rule does prohibit inclusion of any information that "might" result
in harm to anyone, but, unless all information is excluded, there are no
assurances that the desired effect will be achieved. We would indeed be
hesitant to embrace an approach that, through an overabundance of cau-
tion, is capable of producing only a plain vanilla PSR. Such a document
would be of little use to the district court, which would be constrained
to consult outside sources to obtain the needed information. The defen-
dant awaiting sentencing would be disserved as a result, because he
would only be permitted to view the court's written summary of the
information. We would then have come full circle to the pre-1975 days,
when the defendant was sentenced on the basis of information that he
could not effectively question.
5 Moreover, we would not overlook the common-sense observation that
perception often dictates reality. Even if the utmost care is taken to

                     7
In addition to the chilling effect on potential contributors, the
Julian Court noted the prevailing view that routine disclosure of con-
fidential PSR information may compromise the defendant's privacy
interest in that material. As we have already noted, the defendant's
criminal history and current financial condition are always contained
within the PSR. The reports also frequently detail the subject's child-
hood and family history, education and work resume, mental and
physical condition, and history of substance abuse. Needless to say,
these are all factors that the district court will often consider in fixing
a just sentence, and on which the Bureau of Prisons may properly rely
in classifying the defendant for incarceration. However, they are also
intensely personal matters.

Of course, the documents at issue in Ritchie-- state agency
records compiled on a juvenile rape victim -- no doubt also contained
highly sensitive and personal information. The difference between
those records and the PSRs under consideration here is that, in the
case of the former, the applicable state statute specifically authorized
unconditional disclosure to third parties by court order. Ritchie at 57-
58. The Supreme Court thus reasoned that, inasmuch as the state leg-
islature contemplated the use of the records in judicial proceedings,
there was no apparent state policy against disclosing them in criminal
prosecutions. 

There is simply no parallel provision of federal law that permits
such facile disclosure of presentence reports. To the contrary, the con-
fidentiality of PSRs has always been jealously guarded by the drafters
of the federal rules, and by the federal courts. See Julian at 12 (col-
lecting cases). Moreover, there is no indication in Ritchie that the
state agency's records, unlike PSRs, tended to contain material that,
if disclosed, could result in harm to the source. Consequently, we are
of the opinion that PSRs are entitled to a greater degree of protection
from examination and disclosure than the records before the Supreme
Court in Ritchie.
_________________________________________________________________
ensure their anonymity, contributors who are aware that what they say
may be subject to wide dissemination will naturally be reluctant to come
forward, and no amount of soothing assurances from the government are
likely to overcome their reticence.

                     8
2.

We also note that PSRs, by their very nature, do not fit particularly
well within the Brady/Giglio framework. The initiating procedure is
not unusual; if a report's submission to the district court is not com-
pelled by Agurs, see notes 1 and 

, then the accused must, as
with any other potential Brady material, request that it be produced.
However, unlike most potential Brady material, the existence of a
PSR is a foregone conclusion, and its general contents are predictable.

The result is that, although our adversary system of justice rightly
requires defense counsel to uncover potentially favorable evidence, an
experienced advocate may avail himself of a "free shot" by couching
what is actually a general request for a witness's PSR in seemingly
specific terms that might suffice as "some plausible showing" of
materiality and favorability under Ritchie. For instance, in the Eastern
District of North Carolina, every PSR contains a section entitled "Im-
pact of the Plea Agreement," which, in Trevino's words, "states in
concrete terms how a witness's guideline range was affected by the
plea agreement between the parties." Brief of Appellant at 19.

An accused armed with that knowledge would almost certainly be
able to satisfy Ritchie's threshold requirement with regard to any
cooperating witness by merely asserting that the witness's PSR con-
tains evidence of his or her potential bias in favor of the government.
If Ritchie controlled, such an assertion would trigger the district
court's duty to examine the entire PSR in camera , regardless of
whether the plea agreement (1) actually had an effect on the witness's
Guideline range that is (2) substantial enough to give rise to a fair
inference of bias.

This simply will not do. The adversary system does not permit
either party to "engage in groundless fishing expeditions, with the dis-
trict courts as their unwitting (and perhaps unwilling) agents." United
States v. Zolin, 

, 571 (1989) (holding that party asserting
crime-fraud exception to attorney-client privilege must make thresh-
old showing to trigger in camera review).

C.

We conclude that a district court is under no duty to conduct an in
camera examination of a requested PSR unless the accused has first

                    9
clearly specified the information contained in the report that he
expects will reveal exculpatory or impeachment evidence. Though the
imposition of such a requirement will not eliminate"fishing expedi-
tions" altogether, it may serve to curtail them.

Further, in light of the important policy concerns attendant to the
preparation and use of PSRs, we are not confident that the Ritchie
threshold standard of evaluating the accused's proffer of materiality
and favorability adequately protects the compelling public and private
interests in keeping confidential the data assembled therein; more
than "some plausible showing" is needed. We therefore hold that, as
a prerequisite to an in camera review, an accused must plainly articu-
late how the information contained in the PSR will be both material
and favorable to his defense.6

We recognize that an economy of words seldom conveys a depth
of meaning, and, thus, terms like "clearly specify" and "plainly articu-
late," in and of themselves, offer little guidance to a district court try-
ing to decide whether to grant an accused's request for in camera
review of a government witness's PSR. It is enough, however, if we
have adequately conveyed our wish that the court pause before grant-
ing the request. The decision to grant or deny review will necessarily
lie within the district court's sound discretion, and that is as it should
be.

III.

In the case at bar, the district court exercised its discretion in favor
of granting Trevino's request, and it decided to examine the PSRs of
the government's witnesses in camera. Upon evaluating the reports in
accordance with the standards set forth in Figurski, the court con-
cluded that it need not disclose any of the information contained
therein.
_________________________________________________________________
6 In evaluating the probable materiality and favorability of the
requested information, the district court may consider, among other
things, whether the material may be available from other sources and, if
impeachment material, whether the witness will effectively be
impeached otherwise.

                     10
We review only to see whether the district court's decision was
clearly erroneous.7 United States v. Strifler, 

, 1202 (9th
Cir. 1988), cert. denied, 

(1989). Our independent
examination of the PSRs has revealed no clear error. Trevino's con-
victions are, therefore, affirmed.

AFFIRMED

PHILLIPS, Senior Circuit Judge, concurring and dissenting:

I concur in the judgment and in all of the majority opinion except
Part II which, with all respect, I think is both unnecessary to our deci-
sion, and substantively wrong.

Appellant sought in the district court an in camera review, pursuant
to United States v. Figurski, 

(4th Cir. 1976), of the Pre-
sentence Reports (PSR's) of several previously convicted potential
government witnesses against him. The district court faithfully con-
ducted the review, found no material warranting disclosure under
applicable law, and so refused any disclosure of PSR contents to
appellant. Appellant's appeal properly sought review of that decision.
We have conducted that review of the properly sealed PSR's and
found no error. That is sufficient to decide the appeal, so I concur in
the court's judgment affirming the district court's ruling.

The majority goes beyond that, however and in Part II of its opin-
ion lays down a new rule (for prospective application only?) concern-
ing the showing that defendants must make in the district courts to
obtain in camera review of government-witness PSR's. That obvi-
ously is not necessary to our decision here where the district court
conducted in camera review on the showing made in conformity with
the procedure provided in Figurski.

More critically, the new rule is at odds with Figurski's in a way
that, if followed, effectively and improperly overrules that circuit pre-
_________________________________________________________________
7 In Trevino's case, we directed the district court to certify the PSRs as
supplements to the record on appeal. The better procedure would have
been for the court, upon concluding its in camera review, to immediately
seal the reports and designate them for inclusion in the record.

                     11
cedent in a critical respect. The majority assumes power to do this on
the basis (1) that Figurski left open the exact showing required, and
(2) (somewhat inconsistently) that, in any event, intervening Supreme
Court authority, Pennsylvania v. Ritchie, 

(1987), has sug-
gested that the majority's newly-adopted rule is the proper
constitutionally-based one for triggering in camera review of the
materials here at issue, and (3) that policy concerns support the new
rule. I respectfully disagree with each of these.

Figurski does not leave open the showing required. It simply
requires very little: no more than a request based upon an assertion
that there are PSR's in existence with respect to particular govern-
ment witnesses. But it states the requirement fully and clearly:
"[W]hen requested to exhibit [a PSR], the district court should exam-
ine it in camera and disclose only those portions, if less than all,
which [are exculpatory or, if only impeaching, reveal a reasonable
likelihood of affecting the trier of the fact]." 

.

Nor does Ritchie open the door for a re-examination of Figurski's
rule. Ritchie dealt only with the showing required to trigger in camera
review of a very specific and highly sensitive type of material--
government agency records concerning the alleged victim of sexual
abuse crimes. It did not purport to lay down a generally applicable
constitutional rule respecting all types of defense materials subject to
in camera review for possible disclosure. There are obvious reasons
for requiring a more stringent showing in that context than in this.
Furthermore, it is questionable whether Ritchie's requirement is, as a
practical matter, any more stringent than is Figurski's. See 

, 58 n.15. Ritchie cannot be read as implicitly overrul-
ing our Figurski precedent--certainly not by a panel of this court.

Finally, the policy concern drawn on by the majority simply does
not fit. The concern expressed is the protection of confidentiality of
PSR's with emphasis on the undoubtedly good reasons why they must
be protected from routine disclosure. But their confidentiality is
already fully protected in the only needed and relevant way--from
disclosure of their contents to the defendant or the general public--
precisely by Figurski's procedure for in camera review. The majori-
ty's new rule would only extend that already adequate protection to

                    12
an uncertain amount of PSR material that would not even be disclosed
to judicial eyes. That has never been, so far as I know, a proper con-
cern in judicial or legislative protections of confidentiality. Except
where confidentiality is absolute (if there be such) someone after all
has to have access to determine its limits. Confidentiality therefore is
not itself a justification for imposing the new rule.

Quite another policy concern than the expressed one of confidenti-
ality may be implicit in the majority's position. There are intimations
of concern with the undue burden imposed on the courts--particularly
the district courts--by Figurski's modest requirement for triggering
in camera review. That may be a legitimate concern--the district
court was at pains to disavow any precedential effect for its decision
to review in camera here--but Figurski has been on the books and
apparently in unquestioned operation now for twenty years. If its
procedure--grounded in constitutional concerns--is nevertheless on
balance too light on defendants and too heavy on the courts, the
answer for us can only properly lie in an en banc reexamination of its
rule, not by way of panel dictum in a case where its rule was applied
without objection.

                    13