Ernesto Cirilo-muñoz, Petitioner, Appellant, v. United States of America, Respondent, Appellee, 404 F.3d 527 (1st Cir. 2005)

In 1995, Ernesto Cirilo-Muñoz ("Cirilo") was convicted of aiding and abetting, during the commission of a drug crime, the murder of an on-duty policeman. This court affirmed the conviction. United States v. Mangual-Corchado, 139 F.3d 34 (1st Cir. 1998). Cirilo thereafter sought relief under 28 U.S.C. § 2255 (2000), which was denied by the district court, and now seeks review in this court, attacking his sentence based on ineffective assistance of counsel and under Blakely v. Washington, ___ U.S. ___, 124 S. Ct. 2531, 159 L. Ed. 2d 403 (2004), and United States v. Booker, ___ U.S. ___, 125 S. Ct. 738, 160 L. Ed. 2d 621 (2005).

Our earlier opinion sets forth the facts in detail but, in substance, the following is what occurred. Jose Lugo-Sanchez ("Lugo"), Saul Mangual-Corchado ("Mangual") and David Silva worked regular shifts selling drugs outside Cafetin El Ideal — a retail shop where customers could also drink and play pool — in Trujillo Alto, Puerto Rico. Cirilo also frequented El Ideal and was said (by El Ideal's owner) to be "almost always" with Mangual and Lugo, although there was no evidence that he sold drugs.

Lugo correctly suspected that one drug customer — Agent Ivan Mejias-Hernandez ("Mejias") — was in fact an undercover police officer. In October 1994, one of Lugo's suppliers ordered Lugo to kill Mejias, and Lugo in turn told Luis Antonio Ramirez-Ynoa ("Ramirez") of his plan. On November 1, 1994, Mejias arrived at El Ideal between 10:00 a.m. and 11:00 a.m. driving a white Suzuki. At around 11:15 a.m., Lugo called Ramirez on the phone from El Ideal, telling him to drive to El Ideal and to bring "the revolver."¹  Ramirez arrived by 12:30 p.m. with a black Oldsmobile and a revolver.

Ten to fifteen minutes later, Cirilo arrived at El Ideal and was greeted by Lugo. Shortly thereafter, Lugo and Mejias walked to the stoop of a nearby building, where Lugo accused Mejias of being an informant. Silva and Ramirez arrived at the stoop, and Lugo walked back toward El Ideal, retrieving the revolver from the Oldsmobile and concealing it. Walking back toward El Ideal again, Lugo encountered Cirilo, Mangual and one Yito Morales, and tried unsuccessfully to incite them to "beat up" Mejias.

Then Lugo, with several others (possibly including Cirilo), returned to the stoop. There Lugo ordered Mejias at gunpoint to turn over the keys to the Suzuki. Ramirez and Mangual then searched the Suzuki and retrieved Mejias' gun. Cirilo was standing "real close to the car but not searching," just "looking"; his fingerprints were found on the Suzuki, although no evidence was presented as to when he touched the car.

Mejias, escorted by Lugo, then returned to the Suzuki from the stoop and retrieved his keys. Lugo told him to get into the car and "not to come around there anymore." However, as Mejias was about to leave, another man called "Papilin" told Lugo, "You have to take him or kill him because he might come back." There was no direct evidence that Cirilo heard this exchange, and the evidence was unclear as to where Cirilo was standing (or where Papilin and Lugo were standing) when this command was given.

Lugo then ordered Mejias into the Suzuki, and Mangual drove the Suzuki onto the highway with Mejias and Lugo in the back seat. Shortly thereafter, Lugo shot Mejias in the abdomen and in the head. Cirilo and Ramirez followed in the Oldsmobile, with Cirilo driving, although there was no evidence as to why. Lugo later testified that he had not told Cirilo about the planned murder and that he (Lugo) had not asked that Cirilo follow.² 

The cars stopped at a cemetery. Whether Mejias was still alive is unclear but in any event Ramirez shot Mejias twice more in the head. The men then drove in the two cars to a quarry (during the drive Cirilo ingested cocaine provided by Lugo); the Suzuki (with Mejias' body in it) was pushed into the quarry. The men then left the quarry in the Oldsmobile driven by Cirilo. Lugo split the $240 he had taken from Mejias' wallet with the others. Cirilo finally drove Lugo home.

Mangual, Ramirez, Lugo and Cirilo were apprehended and, in September 1995, Cirilo was convicted after a jury trial in federal court of aiding and abetting the murder of an on-duty law-enforcement officer during the commission of a drug offense, 21 U.S.C. § 848(e) (1) (B) (2000); 18 U.S.C. § 2 (2000). On appeal, Cirilo argued that the evidence was insufficient to establish the aiding and abetting offense but, with one judge dissenting on this issue, the conviction was affirmed. Mangual, 139 F.3d at 44-49; id. at 49-56 (McAuliffe, J., dissenting in part).

At sentencing in January 1996, the district judge found that Mejias' killing was motivated by his status as a police officer, resulting in a three-level enhancement under the sentencing guidelines, U.S.S.G. § 3A1.2(a). This raised Cirilo's sentencing range from about 27 to 34 years, to required life imprisonment. At sentencing Cirilo's trial counsel objected to the enhancement, arguing that the government had not shown that Cirilo knew that Mejias was a police officer. The issue was not raised on appeal.

On September 29, 1999, Cirilo filed a petition for post-conviction relief under 28 U.S.C. § 2255 (2000),³  attacking his sentence in various respects. The district court denied the petition, concluding inter alia that the court at sentencing had correctly applied the enhancement and that the lawyer's failure to raise the issue on appeal did not violate Cirilo's constitutional rights. We granted a certificate of appealability, id. § 2253(c), directed to this claim and later broadened review to encompass a Sixth Amendment claim as well.

An ineffective assistance claim requires the defendant — who bears the burden of proof, Scarpa v. DuBois, 38 F.3d 1, 8-9 (1st Cir. 1994) — to show (1) that counsel's performance fell below an objective standard of reasonableness, and (2) that but for counsel's failures, the outcome would likely have been different. Strickland v. Washington, 466 U.S. 668, 687, 104 S. Ct. 2052, 80 L. Ed. 2d 674 (1984); Cofske v. United States, 290 F.3d 437, 441 (1st Cir. 2002). We review the district court's underlying factual findings for clear error, but review its ultimate legal conclusions de novo. Reyes-Vejerano v. United States, 276 F.3d 94, 97 (1st Cir. 2002); Cody v. United States, 249 F.3d 47, 52 (1st Cir. 2001).

Because counsel is entitled to exercise professional judgment, Cirilo must show that an attack on his sentencing enhancement on direct appeal "was so obvious and promising that no competent lawyer could have failed to pursue it." Arroyo v. United States, 195 F.3d 54, 55 (1st Cir. 1999). We believe that test is met in this instance: the enhancement, which had a dramatic effect on the sentence, rested on very thin evidence and a possible misinterpretation of the jury verdict by the district court.

During the sentencing hearing, the judge asked (and was answered in the affirmative), "Didn't the jury find that this victim was killed ... because he was a police officer?" Later (during a discussion of a minor-participant adjustment), Cirilo's lawyer suggested that even though Cirilo followed in the Oldsmobile, "that doesn't mean that [Cirilo] knew that [Lugo] was going to kill and that [Mejias] was a police officer. He may have been surprised when — ". The court responded "that's what the jury decided."

If the district judge was referring to knowledge of Mejias' status, this is not what the jury had decided; the court's instructions to the jury stated that "[k]nowledge of the victim's status as a law enforcement officer is not necessarily an element of the offense." Although the jury was instructed that Cirilo needed to know that murder was intended, the jury was not required by the instruction to find that Cirilo knew that the victim was a police officer.⁴  The distinction between levels of knowledge is a fine one but it mattered to the particular enhancement.

Whether this possible misperception affected the district judge's ruling is unclear. In the sentencing hearing, the district judge did make a formal finding that Cirilo knew that the victim was a police officer when he (Cirilo) assisted in the venture. But there was no detailed discussion by the district judge of the evidence on which such a finding might rest. Ordinarily it is enough that sufficient evidence exist, but in this instance the evidence is thin and the basis for the inference drawn by the district judge is not apparent to us.

The conviction itself rested on fairly limited evidence of scienter, but an inference that Cirilo was involved in the plot could be drawn from Cirilo's presence at the scene of incitement and threat, his prints on the car, his otherwise unexplained pursuit of the Suzuki with a party to the plot, his presence when the victim was shot again and his sharing of the proceeds. Other contextual clues were the other perpetrators' willing acceptance of Cirilo's presence during the events leading up to Mejias' death, and evidence of Cirilo's indebtedness to Lugo's supplier which may have given Cirilo a motive to assist.

It is much harder, on what we can find in this record and without more explanation, to see why Cirilo should be taken to have known that the intended victim was a police officer.⁵  Cirilo's conduct and continued presence may be difficult to explain if he were not aware that Mejias would be killed; the panel that affirmed the conviction so reasoned. But even with this awareness, Cirilo might have thought that Mejias was an informant or a member of a rival drug supply organization; or Cirilo might have assisted his associates' efforts to kill a man without knowing just why they were doing it. Had any of these been Cirilo's motivation, the enhancement would not apply.

Although Lugo at one point testified that he suspected Mejias of being "a police officer," he also testified that he had confronted Mejias and told him that "as far as I was concerned he was a police informant" and "a snitch," and that he told Silva to find out if Mejias was really "a police informant or what." Particularly if Lugo's expressed concern was that Mejias was an informant (and not that he was an officer), it is unclear why the district judge thought that Cirilo knew or believed Mejias was a police officer and thus that his assistance in the murder was motivated by Mejias' official status.

The government says that the omission of the enhancement issue on direct appeal was simply a "tactical judgment." Murchu v. United States, 926 F.2d 50, 58 (1st Cir. 1991). Yet the argument would not have detracted from the evidentiary challenge to the conviction but would have built upon it. It had the further advantage of focusing on the district court's arguable conflation of the two different scienter issues. And it represented the difference between a long jail sentence and a life sentence. One would need a potent reason for omitting the enhancement argument from the direct appeal.

Assuming that the omission of the argument was deliberate, the best one can say for counsel is this: that in some situations lawyers think — usually in error — that by omitting a good argument, they can thereby increase the chance of prevailing on a more doubtful argument directed to a more far-reaching result. However, in this instance, such a calculation would have been manifestly unreasonable under an objective standard, given the comparative strengths of the two different attacks, the opportunity to make both, and the stakes for the defendant.

The second stage of the Strickland inquiry requires a "reasonable probability that, but for counsel's unprofessional errors, the result ... would have been different." Epsom v. Hall, 330 F.3d 49, 53 (1st Cir. 2003) (quoting Strickland, 466 U.S. at 694, 104 S. Ct. 2052) (omission in original). We think that had the enhancement issue been pressed on direct appeal, it would have altered the outcome of the appeal. This is so because of the thinness of the evidence to support the enhancement, the lack of detailed explanation for the finding of knowledge, and the court's apparent error in relying at least in part upon the jury verdict for a finding that the jury did not visibly make.

Some errors that result in a defendant losing the benefit of his appeal may be remedied by reinstating the appeal, see United States v. Torres-Otero, 232 F.3d 24, 30-32 (1st Cir. 2000), but in this instance our second-stage Strickland ruling renders this interim step pointless. The government has already had its opportunity on this appeal to defend the enhancement, and we already know that our own disposition of the direct appeal would have been to remand for re-sentencing. We therefore do so now.

We have had to make our assessment of the constitutional claim based on our own unaided review of the record and without any clear understanding of the district judge's reasoning in finding that Cirilo knew in advance of the crime that the victim was a police officer. Conceivably, on re-sentencing, the government may again urge that such knowledge existed. Without foreclosing such an argument by the government, we are highly skeptical that such a premise can be established or that a life sentence can be justified on the known facts.

Cirilo's brief may be taken to raise a separate claim under section 2255 that his sentence should be vacated because the enhancement was supported by inadequate evidence. This issue — unlike the ineffective assistance claim covered by the certificate of appealability and the Sixth Amendment claim that we later expressly permitted Cirilo to pursue — is not properly before us as an independent claim. Further, the use of section 2255 to attack sentencing findings is, if permissible, at best limited to very unusual cases. It also is likely that on remand the issue will disappear from the case.

As part of his section 2255 petition, Cirilo filed a supplemental brief seeking additional review of his sentencing in light of the Supreme Court's decision in Blakely v. Washington, ___ U.S. ___, 124 S. Ct. 2531, 159 L. Ed. 2d 403 (2004), which we allowed. Cirilo urges resentencing because the jury did not find beyond a reasonable doubt that he believed that Mejias was a police officer. Blakely claims are now viewed through the lens of United States v. Booker, ___ U.S. ___, 125 S. Ct. 738, 160 L. Ed. 2d 621 (2005).

Only in limited circumstances do new rules apply to convictions that have already become final. Schriro v. Summerlin, ___ U.S. ___, 124 S. Ct. 2519, 2522, 159 L. Ed. 2d 442 (2004).⁶  These exceptions include rules that "prohibit criminal punishment for certain types of primary conduct," and those that "forbid the imposition of certain categories of punishment for particular classes of defendants." Sepulveda v. United States, 330 F.3d 55, 59 (1st Cir. 2003); see Schriro, 124 S. Ct. at 2522-23. Neither rubric describes this case.

Otherwise, new rules are applied retroactively to cases already final only if they are "watershed rules of criminal procedure implicating the fundamental fairness and accuracy of the criminal proceeding." Schriro, 124 S. Ct. at 2523 (internal quotation marks omitted). Cirilo's version of the error (under Blakely) was that the enhancement finding was made by the judge based on a preponderance of the evidence; Booker has preserved the use of judge-made findings by directing that the guidelines hereafter be treated as advisory rather than mandatory guidelines. See United States v. Antonakopoulos, 399 F.3d 68, 75 (1st Cir. 2005).

In our view, the use of judge-made findings at sentencing does not undermine "accuracy" (in terms of substantially different outcomes) or undermine fundamental fairness. Such judge-made findings have been the conventional practice throughout our nation's history. They will, post-Booker, continue to be the rule where the sentence is within statutory limits. We have already decided that Apprendi, 530 U.S. at 490, 120 S. Ct. 2348, which provides jury trials for increasing statutory maximums, would not apply retroactively. See Sepulveda, 330 F.3d at 61-63. This resolves any comparable Blakely-like claim in this circuit.

As for the application of mandatory rather than advisory guidelines, it is unclear that advisory guidelines will alter a great number of sentences; mandatory minimums imposed by statute are another question altogether. The guidelines remain a central consideration in sentencing; and sentencing courts must still consider the same statutory factors that the Sentencing Commission was required to use in promulgating the guidelines. See 18 U.S.C. § 3553(a) (2000); 28 U.S.C. § 994(m) (2000). To describe the use of mandatory guidelines as generating serious inaccuracy or fundamental unfairness would not be easy.

Realistically, it is unlikely that the Supreme Court will adopt a retroactivity analysis that opens up to required reexamination practically all of the federal sentences imposed since the guidelines went into effect in 1987. This would comprise tens of thousands of sentences imposed under a regime whose lawfulness was assumed during most of this period. If such a vast reopening of final judgments is required, it must await a decision of the Supreme Court. Certainly Booker itself does not give any clear hint that retroactive effect is intended.

Every other circuit that has considered this issue has agreed that Booker does not apply retroactively. See Varela v. United States, 400 F.3d 864, 866-68 (11th Cir. 2005); Humphress v. United States, 398 F.3d 855, 860-63 (6th Cir. 2005); McReynolds v. United States, 397 F.3d 479, 480-81 (7th Cir. 2005); United States v. Mitchell, 122 Fed.Appx. 539, 540 (2d Cir. 2005) (unpublished); United States v. Leonard, 120 Fed.Appx. 759, 761 (10th Cir. 2005) (unpublished).

Cirilo's sentence is vacated and the matter remanded to Judge Laffitte for resentencing. We leave it to the parties and to the district court to resolve in the first instance whether this new sentencing, which will occur post-Booker, should be governed by the advisory guideline regime. Although failure to use advisory guidelines is not the basis for the remand, the issue of their use once the remand is ordered on other grounds remains open for resolution.⁷ 

It is so ordered.

TORRUELLA, Circuit Judge (Concurring).

Lugo's cold-blooded murder of Officer Mejías was as horrible a crime as can be committed. On that point we can agree. If the horrible nature of the murder added some weight to the record evidence supporting the aiding and abetting charge against Cirilo-[Muñoz], then I would gladly join my colleagues in affirming his conviction. But it does not.

United States v. Mangual-Corchado, 139 F.3d 34, 49 (McAuliffe, J., dissenting in part). This is the dissenting judge's view on direct appeal regarding the sufficiency of evidence on the aiding and abetting charge, a matter which we are presently foreclosed from revisiting on the merits. Nevertheless, the facts surrounding this conclusion are not irrelevant to the issues before us. This is particularly so in light of the majority's recognition, especially in its discussion of appellant's claim of ineffective assistance of counsel, of the "thinness" of the evidence supporting Cirilo-Muñoz's knowledge that the victim was a police officer. See maj. op. at 530, 531.

I write separately in part because I have considerable disagreement with the majority's glossing of the events as they unfolded prior to the police officer's murder by Lugo. Although my view of the relevant facts might, at first glance, appear to bog down on minor details, those facts are significant precisely because of the "thinness" of the case against Cirilo-Muñoz, and because that "thinness" is on a broader scale than my colleagues in the majority acknowledge.

First, there is no evidence that Cirilo-Muñoz was present at "El Ideal" on the day in question as part of Lugo's drug-selling gang. Cirilo-Muñoz lived only five minutes' walking distance from this neighborhood hangout, which he regularly frequented to listen to music, play pool and talk to other youths. There is no evidence that he sold drugs, although the record does show that he was a drug user. Second, there is no evidence that Cirilo-Muñoz's presence at "El Ideal" on November 1, 1994, was at the behest of Lugo or of anyone in his gang. When he arrived, Lugo was already there, and the evidence is uncontradicted that Cirilo-Muñoz was unaware of the reason for Lugo's presence and his intent to harm the undercover police officer, Mejías. In fact, the record is clear that later, when Lugo tried to incite several of those present to "beat up" Mejías, appellant flat out refused to do so.⁸  Furthermore, there is no evidence, and no valid inference can be made from the record, that Cirilo-Muñoz was on the stoop at any time on November 1st when Lugo and his cohorts were accosting Agent Mejías. The majority's statement to the effect that Cirilo-Muñoz "possibly" was on the stoop, maj. op. at 528, is pure speculation. In fact, the evidence is to the effect that Cirilo-Muñoz did not exit "El Ideal" until after Ramirez and Mangual came out of the stoop to search Mejías's car for the agent's gun. Cirilo-Muñoz did not participate in the search and "was just looking." The fingerprints found on Mejías's car matching Cirilo-Muñoz supports this account, as they were all lifted from the exterior portions of the car. No evidence exists as to when they were imprinted there.

It is with regard to the evidence surrounding the murder of Agent Mejías that I find the majority's statement of the facts most problematic because I believe it unfairly places an aura of knowledge of the events to follow on appellant. This implication of knowledge is unsupported by the "thin" record, so "thin" that the government, who is the one properly called upon to meet this burden under our system of justice, has verily skated through the ice.

In this regard, the majority states that "[t]here was no direct evidence that Cirilo-[Muñoz]" heard the exchange between Lugo and "Papilín" in which "Papilín" told Lugo to "take or kill" Mejías, and that "the evidence was unclear as to where Cirilo-[Muñoz] was standing (or where Papilin and Lugo were standing) when this command was given." Maj. op. at 529. If that is so — i.e., if the evidence is not direct and is unclear — what should follow, considering that this is a criminal case in which the unstated premise of these statements carries serious consequences for the appellant, is a statement by the majority to the effect that no inference of knowledge regarding this conversation may be imputed upon Cirilo-Muñoz. But no such disclaimer is made. Instead, we are left with the worst kind of imputation, a half-inference, a lingering doubt as to what appellant has to answer for, and a vague question mark as to what weight is actually being given by the court to this non-evidence.

This rustling of soft "evidence" is followed by the majority's downgrading of the only direct evidence presented regarding Cirilo-Muñoz's lack of knowledge of Mejías's fate. See Maj. op. at 529, n. 2. Lugo, the person who actually murdered Mejías, and who took the stand as a government witness, testified that after he shot Mejías upon leaving "El Ideal" in a car with Mangual, he observed that Cirilo-Muñoz and Ramirez were following them in another car. Lugo testified that he had not asked Cirilo-Muñoz and Ramírez to follow them. In fact, Lugo testified that Cirilo-Muñoz only knew about beating Mejías, but that at that time "didn't know that he was going to be killed."⁹  Although the majority points to Lugo's earlier statement to the F.B.I. that incriminated appellant, this statement was made when Lugo was looking to make a deal for himself with the government. By the time of trial, however, Lugo no longer had anything to gain by wrongly implicating Cirilo-Muñoz.

In light of the full factual background of this appeal, I am concerned about the majority's decision to remand for resentencing to the same judge who sentenced appellant originally and who also decided his § 2255 petition. This judge had two bites at the apple to do so, and in each instance insisted that the "whole picture" supports Cirilo-Muñoz's enhancement.¹⁰  Yet our own examination of the entire trial transcript failed to produce any evidence establishing Cirilo-Muñoz's knowledge of the victim's status as an officer.¹¹  The majority's failure to address the district judge's reasoning and obstinate conclusion has provided an implicit roadmap for the same judge to reimpose a similarly harsh sentence, thus possibly making this remand an exercise in futility.

In fact, there is little said in the majority opinion, or for that matter, in this concurrence, that the trial judge, who heard the flimsy evidence against appellant, was not aware of when he sentenced appellant. Furthermore, from the transcript of the sentencing hearing, a small flavor of which is reproduced in the majority opinion, maj. op. at 530, it is apparent that this judge was not greatly interested in hearing the argument of counsel.¹²  This tone is reproduced again when we see the short shrift given by the same judge to appellant's § 2255 petition, in which he found without much ado not only that he had properly applied the enhancement based on appellant's knowledge of the victim's status as an officer — of which there is not a shred of evidence anywhere on the record — but also that defense counsel's failure to raise this issue on appeal did not violate Cirilo-Muñoz's constitutional rights.

In my opinion, these recurrent conclusions by the trial judge are not only legally erroneous, which the majority recognizes, but also demonstrate an obstinate predisposition to reach a set conclusion in this case. Resentencing should therefore take place before a different judge. See, e.g., United States v. Muniz, 49 F.3d 36, 41 (1st Cir. 1995) (remanding for resentencing by a different judge where, inter alia, the original judge based his sentence on unsupported factfindings); Mawson v. United States, 463 F.2d 29, 31 (1st Cir. 1972) ("It is difficult for a judge, having once made up his mind, to resentence a defendant, and both for the judge's sake, and the appearance of justice, we remand this case to be redrawn."). I simply see no reason why this "reassignment would entail waste and duplication out of proportion to any gain in preserving the appearance of fairness." See Maldonado Santiago v. Velázquez Garcia, 821 F.2d 822, 832-33 (1st Cir. 1987). See also United States v. Hanono-Surujun, 914 F.2d 15 (1st Cir. 1990) (remanding to a different judge where a rule has not been fully complied with and there has been a sharp upwards departure from the Guidelines).

Even stretching it to the breaking point, Cirilo-Muñoz's "participation" and his knowledge of events was minimal and paled when compared to the actions of Lugo — who not only was the leader and principal culprit, even if the government gave him a sweetheart deal — but he was the cold-hearted killer of Officer Mejías. Yet Cirilo-Muñoz received a sentence of life imprisonment as compared with the twenty years imposed upon Lugo. This, after the district judge's "possible misperception[s]" led him to make "a formal finding that Cirilo knew that the victim was a police officer when he (Cirilo) assisted in the venture. But there was no detailed discussion by the district judge of the evidence on which such a finding might rest." Maj. op. at 530. Of course, there is no evidence in the record that Cirilo-Muñoz possessed the requisite knowledge that the district judge attributed to him, unless he acquired it by extra sensory perception, or that he "assisted in the venture,"¹³  unless this term has acquired a new meaning.

This is a case which started out on the wrong foot. Unfortunately it appears destined to continue to suffer permanently from this handicap. "Nothing is more damaging to a new truth than an old error." Goethe, Spruche in Prosa.