United States Court of Appeals For the First Circuit

No. 99-1811

MANUEL GONZÁLEZ-SOBERAL,

Petitioner, Appellant,

v.

UNITED STATES OF AMERICA,

Respondent, Appellee.

APPEAL FROM THE UNITED STATES DISTRICT COURT

FOR THE DISTRICT OF PUERTO RICO

[Hon. José Antonio Fusté, U.S. District Judge]

Before

Torruella, Chief Judge,

Lynch and Lipez, Circuit Judges.

Rafael F. Castro-Lang, was on brief, for appellant. Thomas F. Klumper, Assistant U.S. Attorney, with whom Guillermo Gil, United States Attorney, and Jorge E. Vega-Pacheco, Assistant U.S. Attorney, were on brief, for appellee.

April 5, 2001 TORRUELLA, Chief Judge. Appellant Manuel González-Soberal

was tried before a jury and convicted of conspiracy to possess with

intent to distribute multiple kilograms of cocaine as well as

distribution of eight, nineteen, and twenty-nine kilograms of cocaine

on three occasions. The government's case against González rested

entirely on the testimony of two cooperating witnesses, William Negrón

Zapata and Willy Maya Acosta.

González brings this appeal challenging the denial of a writ

of habeas corpus pursuant to

. His primary claim is

that his trial counsel provided ineffective assistance of counsel by

failing to use two pieces of documentary evidence with which to impeach

the government's chief witnesses, Negrón and Maya.1 We vacate the

denial of the writ and remand for further consideration of this issue

by the district court.

Background

In October 1994, pursuant to a cooperation agreement, Negrón

provided evidence against five individuals in a trial for conspiracy to

distribute and distribution of cocaine. All five defendants were

convicted. Maya, one of the co-defendants, began cooperating with the

government around the time of his sentencing. Maya provided the

1 Appellant also argued, in the alternative, that the government failed to meet its obligation under Brady v. Maryland,

(1963), in not providing these documents to the defense, but this claim was conceded by appellant at oral argument.

-2- government with González's name and identified him as the sixth

indicted co-conspirator, who up to that point had been known only as

John Doe, a.k.a. "Raúl". This identification ultimately led to

González's arrest in January 1995.

The government's case against González rested wholly on the

testimony of cooperating witnesses Maya and Negrón. Both Maya and

Negrón alleged that their introductions to and interactions with

González revolved around five cocaine transactions taking place in

October of 1991. From the record, we summarize the substance of their

respective testimonies.

Maya testified to having met González for the first time on

July 25, 1990. On that date, Maya claims to have been approached by a

man and a woman while he was in the process of tying his boat up to the

pier in the Boquerón bay area. At the trial, Maya identified this man

as González. Small talk was made, and the two men went for a ride in

Maya's boat, at which point the man asked Maya whether he knew of any

fisherman in the area who had found bales of cocaine or marijuana.

Maya replied that he "hadn't been that lucky." The man continued that

he would be interested in purchasing any bales that became available

and that he would pay Maya a $1,000 commission for any such sales that

Maya could arrange. After they returned to the pier, the man wrote

down his home and cellular phone numbers for Maya and asked him to call

if he was able to provide any cocaine or marijuana.

-3- Maya was introduced to Negrón at a family gathering in March

1991. The introduction ultimately resulted in arrangements being made

to facilitate cocaine sales from Negrón to Maya. The first two

transactions were an exchange of seven and eighteen kilograms of

cocaine between Negrón and Maya at a price of $10,000 per kilo. The

money for the cocaine was allegedly provided by González, with Maya

acting as the go-between in exchange for his $1,000 commission per

sale.

For the third transaction, Maya testified that González

arrived at Maya's house with his wife and daughter in a red Toyota

station wagon (or hatchback). Maya invited González to accompany him

to meet Negrón. Maya allegedly told González not to reveal his real

name, but to introduce himself as "Raúl". The two met Negrón,

exchanged the money for the cocaine, and returned to Maya's house,

where González remained for approximately one hour before leaving.

González went with Maya to meet Negrón for both the fourth

and fifth cocaine transactions. During the fifth exchange, Maya was

called away to drive a family member to the airport. Maya instructed

González to wait for Negrón's call (at Maya's house), which he did.

González then went to meet Negrón to pick up the cocaine without Maya.

Finally, Maya testified that he and his family had visited González at

González's home during 1992 and 1993.

-4- The trial transcript reveals that Negrón corroborated Maya's

story as to their introduction and as to the details of the five

October cocaine transactions. Negrón testified that, at the third

meeting, Maya was accompanied by an individual whom Maya introduced to

Negrón as "Raúl", one of his "friends from San Juan." At trial, Negrón

identified González as the person Maya had introduced to him. Negrón

stated that González (known to him as "Raúl") accompanied Maya for the

fourth and fifth exchanges as well. When he returned with the drugs on

the fifth date, only "Raúl" and an unidentified woman were there to

meet him. "Raúl" allegedly explained Maya's absence by saying that he

had been called away on a family matter. Negrón also testified that

"Raúl" had arrived at the transaction site in a burgundy Toyota station

wagon.

González's trial counsel cross-examined both Maya and Negrón

in several areas. In cross-examining Maya, González's attorney

questioned him about his agreement to testify and the fact that his

sentence might be reduced as a result of his cooperation. Maya also

admitted that he had heard Negrón's testimony prior to testifying

himself.2 During Negrón's cross-examination, González's trial counsel

asked him about his prior criminal activities and about the benefits he

was receiving from the government in exchange for his testimony.

2 It is unclear whether Maya heard Negrón's testimony at González's trial or at Maya's trial. Since Maya was in custody at the time of González's trial, it seems more likely that it was the latter.

-5- In addition, González's trial counsel elicited the

inconsistency between Negrón's testimony that "Raúl" had accompanied

Maya to the third cocaine transaction and his failure to mention

"Raúl"'s presence at this sale in his 92-page statement written in

connection with his cooperation agreement. Negrón admitted that no

encounters that he had with "Raúl" lasted longer than three minutes.

Finally, Negrón had suggested to agents at one point in the

investigation that the "Raúl" in question was a person named Raúl

Tiburón.3

As for González's case-in-chief, one of the government

agents, Héctor Ortiz-Rodríguez, was called to testify. A man by the

name of Raúl Tiburón was initially arrested, but released after Negrón

failed to identify him. Later, agents, seeking to arrest González,

mistakenly arrested his brother.4 Ortiz was likely called to testify

about these two prior arrests in order to suggest that the government

was not sure that they had the right person.

González also took the stand in his own defense. He

testified that he owned a brown Ford pickup, not a red or burgundy

3 Negrón explained that during his detention on the drug charges, an attorney visited him and stated that he was there "on behalf of Raúl Tiburón and of Boliche [Maya's alias]." Negrón supposedly concluded from this that the attorney was referring to the "Raúl" that had been with Maya for the last three cocaine transactions. 4 Apparently, part of the confusion resulted from the similarity between González's name (Manuel González-Soberal) and his brother's (Manuel Osvaldo González-Soberal).

-6- Toyota station wagon. He stated that he lived in Hato Rey, not San

Juan, and that he had never traveled to the Boquerón area of Puerto

Rico during 1990 or 1991. Under cross-examination, he stated that he

had never seen Negrón or Maya before and that he did not know them.

González was convicted based on Maya's and Negrón's testimony

on May 23, 1995 and subsequently sentenced to 188-month imprisonment.

His trial counsel then withdrew from the case, and a new attorney

represented him in the direct appeal of his conviction, United States

v. González-Soberal,

(1st Cir. 1997) (affirming), and

represents him in this § 2255 appeal.

González's appellate counsel purports to have discovered two

pieces of documentary evidence that could have been used to impeach

Negrón and Maya. The first document is a letter written by Negrón5 to

someone named Jorge. The main thrust of the letter is aimed at

assuring Jorge that, although Negrón is cooperating with the

government, he does not intend to implicate Jorge in any criminal

activity. The letter (translated from Spanish) states, in part:

Jorge I have not yet been sentenced because the federals interviewed me and I went and spoke to them and told them with all the persons I had dealt and they told me that if I cooperated with them they could get me less years and I am telling you but do not tell anyone what is important is that nothing will happen to you you are like a brother to me and I am not going to

5 It has not been confirmed that this letter was actually written by Negrón, but we assume so for the purposes of this appeal.

-7- tell anything that will harm you. Jorge please don't tell anyone for I may be screwed. What I am telling you do not tell anyone, keep it to yourself please. Of those I dealt with the ones that are going are: Perry, Papito, Elton, Pto. Real, Tito Morgan, Cucho, Alejandro, Boliche, Ram and Chamchi of the Parguera.

The letter, according to González, should have been used to show that

Negrón, despite his agreement to be honest and to reveal everything

that he knew, had not been entirely truthful with the government

agents. From this act of withholding information, the jury could infer

that Negrón might be willing to be dishonest in other regards. The

letter, therefore, could have had a detrimental effect on Negrón's

credibility.

González also contests his trial counsel's failure to impeach

Maya with a psychodiagnostic report that suggested that Maya had a

below-normal capacity to distinguish truth from fiction. In 1994, a

psychological evaluation was performed on Maya to determine whether a

diminished capacity defense was appropriate for his trial. The results

of the evaluation found that Maya suffered from Attention Deficit

Disorder with hyperactivity, the long-term effects of which included an

"[in]ability to perceive reality without distortion"6 and a tendency to

"bend reality to avoid dealing with perceived or anticipated harshness

6 When read in context, it appears that the statement should have said an "inability" or an "impaired ability" to perceive reality without distortion rather than an "ability".

-8- in the environment." This psychodiagnostic report might have caused

the jury to doubt the reliability of Maya's testimony.

The district court rejected González's claim of ineffective

assistance of counsel. In its Opinion and Order, the district court

properly set out the two-part test identified in Strickland v.

Washington,

(1983), used for determining ineffective

assistance of counsel claims. In a troubling statement, however, the

court went on to identify a standard for determining whether a

defendant has been prejudiced by his counsel's acts or omissions: "In

order to establish ineffective assistance of counsel, Petitioner must

demonstrate that but for the unprofessional error, he would not have

been found guilty." United States v. González-Soberal, Civ. No. 98-

1292 (JAF), at 4 (Feb. 8, 1999). This suggests a higher standard than

the one set out in Strickland, i.e., an outcome-determinative standard

that requires a defendant to show that it is more likely than not that

counsel's errors assured a guilty verdict. This higher standard was

considered, and explicitly rejected, by the Court in Strickland: "The

result of a proceeding can be rendered unreliable, and hence the

proceeding itself unfair, even if the errors of counsel cannot be shown

by a preponderance of the evidence to have determined the outcome."

.

In the next paragraph of the Opinion and Order, the district

court reverted to the appropriate Strickland test, that of reasonable

-9- probability, in stating that the failure to use the letter and

psychodiagnostic reports for impeachment purposes was not prejudicial

to González. This citation to the proper standard (and the other

correct references found in the opinion) would do much to assuage this

Court that the district court did not apply an incorrectly high

standard, had the district court opinion provided an explanation for

the finding of no prejudice. Unfortunately, no reasons were provided

to justify the district court's conclusion. The district court did not

point to the possible limited value of the impeachment testimony, to

the effectiveness of the cross-examination of the two government

witnesses otherwise, to the strength of the available evidence against

the defendant, or to any other factor or factors that may have been

observed and noted by the district court.

Analysis

Ineffectiveness "is a mixed question of law and fact."

Strickland,

. The district court findings of fact are

subject to the clearly erroneous standard.

In the past, we have

reserved the question as to the precise standard of review to be used

by this Court in evaluating ineffective assistance of counsel claims.7

7 In another case, alleging ineffective assistance of counsel based on a conflict of interest, we held that the de novo standard of review applied, while according the deference of the clearly erroneous standard to the district court's underlying findings of fact. Familia- Consoro v. United States,

(1st Cir. 1998) (actual, rather than constructive, conflict of interest).

-10- Bucuvalas v. United States,

, 657 n.1 (1st Cir. 1996);

United States v. McGill,

, 226 n.2 (1st Cir. 1993). We need

not decide the issue at this time, because, irrespective of whatever

standard we may employ, the actions taken by the district court compel

our conclusion.

In order for a defendant to succeed in an ineffective

assistance of counsel claim, he must show, by a preponderance of the

evidence, that defendant's trial counsel's conduct fell below the

standard of reasonably effective assistance and that counsel's errors

prejudiced the defense. Strickland,

. The district

court did not evaluate González's trial counsel's efforts, but instead

proceeded directly to the prejudice requirement, rejecting the claim on

that basis. Addressing the prejudice prong prior to evaluating

counsel's conduct is a permissible approach and even endorsed where

more efficient.

. We will follow the district court's order

in our review of González's ineffectiveness claim and turn now to

whether the alleged errors of González's counsel resulted in prejudice.

As stated previously, prejudice exists in a particular case

when there is "a reasonable probability that, but for counsel's

unprofessional errors, the result of the proceeding would have been

different." Strickland,

. A reasonable probability is one

"sufficient to undermine confidence in the outcome."

On one end,

it is not enough to show that the errors had "some conceivable effect

-11- on the outcome."

. Nor is it required, however, that the

defendant prove that the errors were more likely than not to have

affected the verdict.

It is important to maintain the focus of an

ineffectiveness inquiry on the "fundamental fairness of the

proceeding."

.

Three factors need to be considered in this case in order to

make the prejudice determination. The first is the strength of the

government's case against González. Second, we must evaluate the

effectiveness of the presentation of González's defense absent the

impeachment documents. Third, we must consider the potential

impeachment value of the two documents in undermining the credibility

of the government witnesses' testimony.

Examining the government's case, we note that a significant

factor weighing in favor of finding prejudice is the absence of any

corroborating evidence other than the testimony of Maya and Negrón.

See Phoenix v. Matesanz,

(1st Cir. 1999) (case resting

entirely on blood and fingerprint evidence). The government had no

surveillance, undercover agents, or hard evidence, such as phone

records, with which to prove that González participated in the five

reported drug transactions. "In making this determination [prejudice],

a court hearing an ineffectiveness claim must consider the totality of

the evidence before the judge or jury. . . . Moreover, a verdict or

conclusion only weakly supported by the record is more likely to have

-12- been affected by errors than one with overwhelming record support."

Strickland,

.

In addition, there were several weaknesses to Negrón and

Maya's stories that were highlighted by González's counsel's cross-

examination. The fact that Negrón identified the man accompanied by

Maya as "Raúl" or "Raúl Tiburón" until late in the investigation

weakens his later identification of González as Maya's companion for

the last three drug transactions. Negrón's contact with González

lasted only a few moments each time, and all the meetings occurred more

than three years before Negrón testified in this case. There is

certainly an argument that Negrón was mistaken in his identification,

or even implicated González in an effort to cover the tracks of others

or to ingratiate himself to the government agents.

Maya's testimony had flaws as well. His story of the chance

meeting of González, González's proposition that they engage in drug

transactions, and González's alleged entrustment of large sums of money

to Maya seems rather implausible. And under cross-examination, Maya

also admitted his self-interest in cooperating with the government in

González's case and to having heard Negrón's testimony prior to

providing his own version of events.

On the other hand, there was significant evidence presented

at trial indicating that González was involved in these drug deals.

From the outset of Maya's cooperation, he identified "Raúl" as

-13- González, correctly providing agents with González's full name and

address. Maya and Negrón both pointed to González during their

testimony when asked whether "Raúl" was present. And the two

witnesses, while telling different stories, corroborated one another's

accounts with respect to González's alleged participation.

Turning to the impeachment value of the two documents,

Negrón's letter suggests that he was not completely truthful in his

cooperation with the government, which was directly contrary to what he

had agreed. The terms of his agreement with the government required

that he "cooperate completely, candidly, and truthfully" and provide

"any and all information in his possession relating directly or

indirectly to any and all criminal activity concerning the importation,

possession, and distribution of controlled substances in Puerto Rico or

any other places and any related matters of which he has knowledge."8

Impeachment of Negrón with the letter could have created a suspicion in

the jury's mind that Negrón was not being honest in his testimony. The

jury might not have concluded from this, however, that Negrón would

falsely implicate González. There is a definite distinction between

Negrón withholding information to protect a friend and his lying under

oath as to González's involvement.

8 While it may be that Negrón did not sign his official statement (April 26, 1994) until after he wrote the letter to Jorge (March of 1993), it is clear that he had agreed to cooperate with the government as of August or October of 1992.

-14- It is even more difficult to know the proper weight to place

on the statements made in Maya's psychodiagnostic report. The

difficulties that Maya allegedly has in distorting reality are not

sufficiently explained in terms that a layperson can definitively

understand. They may completely undermine Maya's testimony and

identification of González, and they may not. A clinical explanation

of these comments would have been helpful to the jury at trial (if the

report had been used). This Court similarly lacks the background

necessary to predict what effect, if any, Maya's disabilities could

have on his testimony. Thus, we have little basis for estimating how

the jury would have perceived this report with appropriate medical

explanation.

Whether prejudice resulted from González's counsel's failure

to impeach government witnesses Negrón and Maya with these two

documents is a close call. That, coupled with the possibility that the

district court held González to an improperly stringent standard,

demands that we vacate the denial of González's § 2255 petition as to

his ineffective assistance of counsel claim.

We could make an independent determination of prejudice based

on our review of the record and of the impeaching documents, but

several factors caution us against this. First, the district court

presided over the trial and has a better perspective from which to

evaluate the possible impact of these two documents on the jury and its

-15- verdict. Especially in a case such as this, where the strength of the

government's case rests largely on the credibility of the cooperating

witnesses, the trial judge has had a unique opportunity to form a

judgment about the value of their testimony. Second, the impact of the

psychodiagnostic report would be more accurately assessed with the

benefit of clinical explanation. The district court, in holding a

prejudice hearing, could hear testimony explaining the possible impact

of Maya's diagnosis on his ability to testify truthfully.

As such, we remand the case to the district court for a

reevaluation of the possible prejudicial effect of González's counsel's

alleged errors. If the district court still finds that prejudice is

absent, we would urge the district court to explain its conclusion

given the closeness of the question as we have identified it. The

district court should consider whether to hold a hearing to determine

whether González's trial counsel's conduct met the standards

articulated in Strickland for reasonably effective assistance.

Vacated and remanded.

-16-