UNITED STATES OF AMERICA v. KIELAN BRETT FRANKLIN
No. 20-30136
United States Court of Appeals for the Ninth Circuit
November 23, 2021
D.C. No. 6:19-cr-00006-SEH-1
Appeal from the United States District Court for the District of Montana
Sam E. Haddon, District Judge, Presiding
Argued and Submitted May 4, 2021
Seattle, Washington
Filed November 23, 2021
Before: Danny J. Boggs,* Marsha S. Berzon, and Mary H. Murguia, Circuit Judges.
Opinion by Judge Boggs;
Concurrence by Judge Berzon
Criminal
The panel affirmed a criminal judgment in a case in which Kielan Frankin pleaded guilty to aiding and abetting the possession of a firearm in furtherance of a crime of violence, in violation of
The panel wrote that, as Franklin conceded, binding precedent forecloses his contention that Hobbs Act robbery is not categorically a crime of violence under
Franklin contended that the district court violated his due-process rights at sentencing by relying on his codefendants’ unsworn hearsay statements, which accused him of trying to influence their testimony, in imposing an obstruction-of-justice enhancement. As this court had not clearly enunciated the standard for reviewing a district court‘s determination of whether coconspirator hearsay is unreliable, the panel took the opportunity to clarify it. After examining the development of the minimal-indicia-of-reliability doctrine over the last half-century, the panel concluded that there are two distinct questions that this court answers in examining a hearsay statement at sentencing: (1) whether the statement is “procedurally reliable” and (2) whether the statement is “substantively reliable.” If the court answers either question in the affirmative, then the
The panel applied that framework to this case. Reviewing procedural reliability de novo, the panel concluded that the government provided enough specifics so that Franklin was not put to the burden of proving that the enhancement did not apply, and that there were adequate procedural opportunities for Franklin to challenge the extrinsic, nonhearsay evidence corroborating codefendant Gerald Hiler‘s hearsay statements. Perceiving no error in the district court‘s conclusion that this evidence sufficiently corroborated Hiler‘s statements, the panel concluded that the admission of those statements at sentencing did not deprive Franklin of due process. As to substantive reliability, the panel considered the government‘s argument that Hiler‘s and codefendant Morgan Pitsch‘s hearsay statements corroborate each other enough to be admissible at sentencing. The panel concluded that the district court did not clearly err in implicitly finding the two coconspirators’ statements to corroborate each other enough to be substantively reliable, and that their admission at sentencing thus did not violate due process.
COUNSEL
Dwight J. Schulte (argued), Schulte Law Firm P.C., Missoula, Montana, for Defendant-Appellant.
Timothy J. Racicot (argued) and Julie R. Patten, Assistant United States Attorneys; Leif M. Johnson, Acting United States Attorney; United States Attorney‘s Office, Billings, Montana; for Plaintiff-Appellee.
OPINION
BOGGS, Circuit Judge:
Kielan Franklin pleaded guilty to one count of aiding and abetting the possession of a firearm in furtherance of a crime of violence, in violation of
I. Background
A. Events Leading Up to the Robbery
Kielan Franklin and Arielle Cowser were an unmarried couple who had one child together. They both used heroin. While living in Helena, Montana, they developed a relationship with B.G. and S.G. (the “victims“), a married couple who also used heroin.
In March 2019, Mr. Franklin gave the victims $1,200 to go to Spokane, Washington, and buy an “ounce,” meaning twenty-four grams, of heroin.1 But the price of heroin had increased, and the money would buy only twenty grams. During the trip, Mr. Franklin and Ms. Cowser frequently texted and called the victims using Ms. Cowser‘s phone to get status updates, but—losing patience—the victims eventually stopped responding. It took longer than Mr. Franklin had expected for the victims to return from Spokane, partially because of heavy snow and partially because B.G. had a habit of doing tasks slowly. The victims also testified that they had sampled some of the heroin before their return. When the victims finally returned, B.G., believing he was eight grams shy of an ounce, tried to “cut” the heroin—that is, add sugar to it—to make twenty-eight
Already upset by the victims’ lack of communication, Mr. Franklin went to the victims’ house; Ms. Cowser went with him. Mr. Franklin became even more upset after he found that the quality of the heroin was much worse than he had expected—heavily diluted and poorly blended with the sugar. Mr. Franklin told the victims that they owed him double his money back. After leaving, Mr. Franklin continued to text the victims, his messages becoming increasingly threatening.
B. The Robbery2
On the morning of March 8, 2019, Mr. Franklin, Ms. Cowser, and Gerald Hiler and Morgan Pitsch (the two other codefendants in this case) were at the house of Corissa Soltis. At some point, Mr. Franklin and Ms. Cowser had texted Mr. Hiler “about going out and making a collection on these people.” Mr. Franklin, still upset about the missing heroin and money, asked Mr. Pitsch to “go over to a house with him” to get either the drugs or the money. The four left Ms. Soltis‘s house in Mr. Franklin‘s Jeep.
While they drove, Mr. Franklin told the other three his plan to rob the victims. Ms. Cowser‘s job was to enter the victims’ house first and leave the door unlocked for the other three. Mr. Hiler was armed with a silver handgun, and he was to act as the “muscle,” the most aggressive of the three.
The group dropped Ms. Cowser off outside the victims’ house. Ms. Cowser lied to the victims, saying that she and Mr. Franklin had been fighting and that he had left her “in the middle of nowhere” on the frigid, snowy night. Over B.G.‘s protest, S.G. let Ms. Cowser inside. A few minutes later, on their security cameras, the victims noticed someone outside. Ms. Cowser approached the front door, and B.G. told her: “Don‘t open the door.” Before the victims could stop her, Ms. Cowser had let in Mr. Franklin, Mr. Hiler, and Mr. Pitsch.
Ms. Cowser told the three men that the drugs were upstairs and advised them to make sure that the victims did not have their phones. Even so, S.G. discreetly called 911 with a cell phone that she hid under a blanket. During the robbery, she made statements such as “I‘m really, really scared” to alert the operator about what was happening. Eventually, Mr. Hiler found the phone and hung up the call. The four defendants decided to make a quick getaway with a small amount of the victims’ property—S.G.‘s wedding rings, her phone, a purse, about $100, and about a gram of heroin.
C. Proceedings Below
Mr. Franklin and Ms. Cowser were each charged with one count of conspiracy to commit Hobbs Act robbery, in violation of
Mr. Franklin moved to dismiss the firearm count against him. He argued that neither the conspiracy charge nor the substantive robbery charge, either on its own or under an aiding-and-abetting theory or Pinkerton-liability theory,3 was a crime of violence under
Mr. Franklin and Ms. Cowser both initially intended to try their cases to a jury, while Mr. Hiler and Mr. Pitsch entered into plea agreements. Shortly after the district court accepted those guilty pleas, Mr. Franklin also entered into a plea agreement, admitting guilt to the substantive robbery count and the firearm-possession count.
But during his change-of-plea hearing, when asked to admit to the factual basis the government had proposed, Mr. Franklin balked. He instead maintained, under oath, that he had only wanted to talk to the victims to get his money back from them. Mr. Franklin asserted that he had first sent Ms. Cowser into the house to try to talk with the victims. Then, after telling Mr. Hiler and Mr. Pitsch to stay in the car, Mr. Franklin went to the house himself. Mr. Franklin claimed that he did not know that the other two codefendants had been armed and that they had disobeyed him by running into the house about a minute later, waving their guns. Mr. Franklin also claimed not to have known that anyone had stolen anything until after the four had all returned to Mr. Franklin‘s Jeep. In light of these contentions, the court
In November 2019, shortly after the abortive plea hearing, Mr. Hiler met with FBI Special Agent Jason Bowen for an in-person interview. At that interview, he told Special Agent Bowen that Mr. Franklin had sent him a note (which the jail staff would provide to Special Agent Bowen). The note said, in part:
Now that I know you have plead [sic] out I want to respectfully request your help.... All I need from you is to state the truth[.]
- I never asked you guys to rob anyone
- I didn‘t know there were guns on anyone
- I said to stay in the car so I could talk to homeboy
- Curly had his own peice [sic] and I had nothing to do with it. He got it before the alleged incident.
The note‘s author was also aware that “Curly” (a nickname for Mr. Pitsch, according to Mr. Hiler) had been “squawking“—talking to the government about the robbery in this case.
Mr. Hiler later told Special Agent Bowen that he had been assaulted by two other inmates after telling the government about the note. The attackers told Mr. Hiler that Mr. Franklin had told them to “slap [him] around.” Mr. Hiler sustained “[v]ery minor injuries” to the inside of his mouth as a result.
During an in-person interview, Mr. Pitsch told Special Agent Bowen that Mr. Franklin had also contacted him. Mr. Franklin had warned him that “he would be on paper, and that he should be careful.” Interpreting Mr. Franklin‘s statement as a threat that he would be branded as a snitch, Mr. Pitsch asked for a transfer to a different detention facility.
Based on these reports from Mr. Hiler and Mr. Pitsch, the government obtained a superseding indictment charging Mr. Franklin with two counts of witness tampering, in violation of
At a second change-of-plea hearing, contrary to his statements at the first hearing, Mr. Franklin admitted that he had intended to “relieve” the victims of their money, drugs,
At Mr. Franklin‘s sentencing hearing, the government presented hearsay statements by Mr. Hiler and Mr. Pitsch, given secondhand through Special Agent Bowen‘s testimony, about Mr. Franklin‘s attempts to influence those two men‘s testimony. The district court relied on those statements to find that an obstruction-of-justice enhancement applied, overruling Mr. Franklin‘s objection to the presentence report (PSR).
The court computed Mr. Franklin‘s offense level for the robbery count to be twenty-two, yielding an advisory Guidelines range of forty-six to fifty-seven months of imprisonment for that count (as his criminal history category was II). The firearms count carried a mandatory seven-year (eighty-four-month) consecutive sentence. The court then sentenced Mr. Franklin to fifty-five months of imprisonment on the robbery count; combined with the mandatory consecutive sentence, his total prison sentence was 139 months. Upon the government‘s motion, the district court dismissed the remaining counts against Mr. Franklin.
Mr. Franklin‘s timely appeal followed.
II. Hobbs Act Robbery as a Crime of Violence
Mr. Franklin contends that Hobbs Act robbery, as defined in
Possession of a firearm “in furtherance of” a “crime of violence” carries additional penalties beyond the sentence imposed for the underlying crime.
We have held that robbery under
Mr. Franklin concedes that Mendez and Dominguez are binding on us; he says he presents the issue solely to preserve it for potential review on certiorari. Thus, we hold that the district court did not err.
III. Due-Process Violation Through Hearsay Evidence at Sentencing
Mr. Franklin contends that the district court violated his due-process rights at sentencing by relying on his codefendants’ unsworn hearsay statements, which accused him of trying to influence their testimony, in imposing the obstruction-of-justice enhancement. Finding the appropriate standard of review unclear from our precedents, we take this opportunity to clarify it. Under the correct standard, Mr. Franklin‘s challenge fails.
A. Appeal Waiver
Mr. Franklin begins by arguing that his appeal waiver does not bar our review of this issue. But the government does not address waiver at all, instead arguing the merits of his claim. We likewise proceed to the merits, the government having forfeited any claim of waiver it might have had. United States v. Garcia-Lopez, 309 F.3d 1121, 1123 (9th Cir. 2002); United States v. Lewis, 798 F.2d 1250 (9th Cir. 1986).
B. Legal Background
Hearsay is generally admissible in sentencing hearings, as neither the Confrontation Clause nor the Federal Rules of Evidence apply to such hearings. United States v. Petty,
But a statement by a coconspirator that inculpates the defendant “is inherently unreliable.” United States v. Huckins, 53 F.3d 276, 279 (9th Cir. 1995) (quoting Lee v. Illinois, 476 U.S. 530, 546 (1986)); see also United States v. Vera, 893 F.3d 689, 693–94 (9th Cir. 2018) (discussing “widespread” reluctance among the courts of appeals to rely on admissions of coconspirators). We presume such statements unreliable because the coconspirator “may very well have been hoping to curry favor with law enforcement officials by implicating his accomplice.” Huckins, 53 F.3d at 279; see also Petty, 982 F.2d at 1369–70 (approving of applying a rebuttable presumption of unreliability to a coconspirator‘s proffer to the government). Although the “fact that a statement is self-inculpatory does make it more reliable,” any increased reliability is limited to the self-inculpatory aspects of the statement, not “collateral” statements about others’ guilt. Huckins, 53 F.3d at 279
Whether the government rebuts that presumption turns on whether the coconspirator‘s statements have independent corroboration. See United States v. Berry, 258 F.3d 971, 976 (9th Cir. 2001) (“One factor evidencing the reliability of hearsay statements by co-defendants is external consistency.“). That corroboration may come from trial testimony, United States v. Egge, 223 F.3d 1128, 1132–35 (9th Cir. 2000), the defendant‘s own testimony, United States v. Littlesun, 444 F.3d 1196, 1198–99, 1201 (9th Cir. 2006), or even from other codefendants’ hearsay accounts, Berry, 258 F.3d at 976–77.
C. Standard of Review
Reviewing our cases, we find that we have not yet clearly enunciated the standard by which we review a district court‘s determination of whether coconspirator hearsay is unreliable. The government cites United States v. Ayers, 924 F.2d 1468, 1481 (9th Cir. 1991), for the proposition that we review for abuse of discretion. Indeed, we have said “[c]onsideration of evidence outside the record of conviction for sentencing purposes is reviewed for an abuse of discretion,” and “[r]eliance on materially false or unreliable information is an abuse of discretion.” Ibid.
We have also said, in a general way, that “[a] district court abuses its discretion when it makes an error of law, when it rests its decision on clearly erroneous findings of fact, or when we are left with ‘a definite and firm conviction that the district court committed a clear error of judgment.‘” United States v. 4.85 Acres of Land, More or Less, 546 F.3d 613, 617 (9th Cir. 2008) (quoting United States v. Hinkson, 526 F.3d 1262, 1277 (9th Cir. 2008), vacated, 547 F.3d 993
[I]f the district court‘s application of fact to law “requires an inquiry that is essentially factual,” we review it as if it were a factual finding; if the district court‘s application of fact to law requires reference to “the values that animate legal principles,” we review it as if it were a legal finding.
United States v. Hinkson, 585 F.3d 1247, 1259 (9th Cir. 2009) (en banc) (quoting United States v. McConney, 728 F.2d 1195, 1202 (9th Cir. 1984) (en banc), abrogated in part by Pierce v. Underwood, 487 U.S. 552, 557–63 (1988)). That is to say, we review an essentially factual finding for clear error and an essentially legal finding de novo. Id. at 1259–60.
An essentially factual finding is one “requir[ing] an inquiry ... that is founded ‘on the application of the fact-finding tribunal‘s experience with the mainsprings of human conduct.‘” Id. at 1259 (quoting McConney, 728 F.2d at 1202). Examples in Hinkson of essentially factual findings include determinations of “motive, intent, and negligence.” Id. at 1260.
By way of contrast, “questions such as whether defendants’ conduct constituted a conspiracy in violation of the Sherman Act,” “questions that implicate constitutional rights,” and “the meaning of due diligence or the conceptual basis for granting new trials” are questions that “require[] us to consider legal concepts in the mix of fact and law and to exercise judgment about the values that animate legal principles.” Id. at 1259–60 (quoting McConney, 728 F.2d
To decide whether we have a factual or legal question before us in Mr. Franklin‘s case, we consider the history and application of the minimal-indicia-of-reliability doctrine in our case law.
1. Origin and Early Development
The origin of the rule that evidence against a defendant in a sentencing hearing must bear sufficient indicia of reliability is Townsend v. Burke, 334 U.S. 736 (1948). There, a pro se criminal defendant had been “sentenced on the basis of assumptions concerning his criminal record which were materially untrue.” Id. at 741. One of the charges relied upon by the sentencing court had been dismissed, and the defendant had been acquitted of two others. Id. at 740. Although the sentence had been “within the limits set by statute” and “its severity would not be grounds for relief,” it was “the careless or designed pronouncement of sentence on a foundation so extensively and materially false, which the prisoner had no opportunity to correct by the services which counsel would provide, that render[ed] the proceedings lacking in due process.” Id. at 741.
We applied and extended Townsend in United States v. Weston, 448 F.2d 626 (9th Cir. 1971), holding that the district court‘s reliance on not just materially false but also unreliable information violated due process. (See infra pp. 22-24 for more discussion of Weston.)
In United States v. Ibarra, 737 F.2d 825 (9th Cir. 1984), we attempted to clarify the meaning of “false or unreliable” by defining it to mean lacking in “some minimal indicium of reliability beyond mere allegation.” Id. at 827 (quoting
Ibarra also stated that “[w]e review a sentence for abuse of discretion if [the] defendant can show that the district court relied on information that should not have been considered during the sentencing phase.” 737 F.2d at 826–27 (citations omitted). That statement expressed the prevailing standard of review before the introduction of the Sentencing Guidelines, which came about shortly after we decided Ibarra. See, e.g., United States v. Wilson, 900 F.2d 1350, 1354 (9th Cir. 1990) (“Prior to the Guidelines’ promulgation, ... district courts employed a gestalt approach to sentencing where a single factual determination rarely had a sufficiently significant impact on the overall sentence to warrant an appeal.“); United States v. Sanchez-Murillo, 608 F.2d 1314, 1319 (9th Cir. 1979) (noting that the “only exception to this rule is where the defendant can establish that information presented to the court prior to sentencing should not have been considered,” which would supply the basis for a claim of an abuse of discretion); United States v. Kearney, 560 F.2d 1358, 1369 (9th Cir. 1977) (“A federal trial judge has wide discretion in imposing sentence, and where ... the sentence pronounced is within statutorily-prescribed limits, it is generally not subject to review.“).
2. Effect of the Sentencing Guidelines and Booker
With the advent of the Guidelines, which were binding until United States v. Booker, 543 U.S. 220 (2005), made them advisory, district courts no longer exercised traditional sentencing discretion, adjusting the weights of sentencing factors “to reflect the differences in the relative certainty of information on which the sentences [we]re based.” Wilson, 900 F.2d at 1352 (quoting United States v. Davis, 715 F. Supp. 1473, 1477 (C.D. Cal. 1989), aff‘d in part and vacated
[A] defendant‘s due process right to ensure the reliability of information used at sentencing includes the requirement that facts underlying sentencing factors be proved according to a specified standard of proof....
We hold ... that district courts are constitutionally required to make factual determinations underlying application of the Guidelines by at least a preponderance of the evidence. As so interpreted, the Guidelines do not violate due process.
Id. at 1354. That was perhaps our clearest statement tying reliability of sentencing evidence to the sentencing court‘s factfinding function.
After Booker made the Guidelines advisory rather than binding, much of district courts’ traditional sentencing discretion was restored. But Booker did not much affect our review of reliability determinations. District courts still must correctly compute the Guidelines range as a part of the sentencing process. United States v. Carty, 520 F.3d 984, 991–93 (9th Cir. 2008) (en banc). For that computation, the government must still prove by at least a preponderance any
...
We now review all sentences for an abuse of discretion, whether they are inside or outside the correctly computed Guidelines range. Carty, 520 F.3d at 993. That means “ensur[ing] that the district court committed no significant procedural error, such as . . . selecting a sentence based on clearly erroneous facts,” Gall v. United States, 552 U.S. 38, 51 (2007), and then considering the substantive reasonableness of a procedurally sound sentence, duly deferring to a district court‘s finding that the
3. Historical Application of the Minimal-Indicia-of-Reliability Doctrine
A survey of our application of the minimal-indicia-of-reliability doctrine over time demonstrates two overarching concerns that sometimes come into tension. The sentencing court should have as much information as possible so that it can effectively discharge its duty—to sentence each person as an individual. But there must also be sufficient procedural protections to allow a defendant the opportunity to question and refute evidence against him at sentencing. Our survey also reveals two separate inquiries that we have developed to accommodate these two concerns. The first was primarily an examination of the process afforded a defendant during sentencing. Later, we relaxed the bar on hearsay evidence at sentencing to allow statements that appeared substantively reliable, even if the defendant did not receive the same
a. At First, We Were Concerned Primarily with Procedure
i. United States v. Weston, 448 F.2d 626 (9th Cir. 1971). In one of our first cases discussing due-process rights at sentencing, the defendant, convicted of trafficking heroin, had been sentenced to the statutory maximum of twenty years of imprisonment. Id. at 630. The district court based its sentence on assertions by federal narcotics agents in the PSR that she had been “the chief supplier [of heroin] to the Western Washington area.” Id. at 628. The district court, opining that the probation officers who had compiled the report were “extremely objective,” and noting that the defendant had not provided any “contrary factual information, rather than simply a vehement denial,” accepted the PSR as true. Id. at 629. It advised the defendant that she could bring a motion later if she obtained any information to refute the report, to which her counsel responded, “I can‘t conceive of what type of investigation I can do to come back and say that she isn‘t” the “biggest dealer in the Western states.” Ibid. Nevertheless, the district court imposed the maximum sentence.
We vacated that sentence on appeal. Upon reviewing the sealed record, we observed that the PSR‘s allegations came from a narcotics agent‘s “unsworn memorandum” that merely “quote[d] a named informant, described only as ‘previously identified as a reliable cooperating individual,’ who indicate[d]” only that the defendant was about to make—not that she had made—a trip to Mexico to obtain heroin. Id. at 630. We found that the report “contain[ed] nothing to show, rather than to assert, that the information was reliable, or otherwise to verify the very serious charge
to saying that once a defendant has been convicted of offense A, narcotics agents can say to the probation officer, and the probation officer can say to the judge, “We think that she is guilty of much more serious offense B, although all we have to go on is an informer‘s report,” and the judge can then say to the defendant, “You say it isn‘t so; prove that to me!” In addition to the difficulty of “proving a negative,” we think it a great miscarriage of justice to expect Weston or her attorney to assume the burden and expense of proving to the court that she is not the large scale dealer that the anonymous informant says that she is.
Id. at 634. Thus, we held, “a sentence cannot be predicated on information of so little value as that here involved. A rational penal system must have some concern for the probable accuracy of the informational inputs in the sentencing process.” Ibid. We remanded for a new sentencing hearing, directing the district court not to consider the information from the PSR “unless it is amplified by information such as to be persuasive of the validity of the charge there made.” Ibid.
Weston stands for the proposition that if the sentencing process effectively puts the burden of proof on the defendant to refute a damaging hearsay allegation, particularly when the factual basis for believing such a charge is practically nonexistent, that process is legally flawed. See id. at 633. The district court treated as controlling the inherent credibility of
ii. United States v. Petty, 982 F.2d 1365 (9th Cir. 1993). This was a post-Guidelines case in which the kingpin of a cocaine-trafficking scheme, Mr. Kessack, gave a statement during plea negotiations with the government. Id. at 1366. Negotiations broke down, and Mr. Kessack and his coconspirators were convicted at trial. Id. at 1367. After Mr. Kessack was sentenced, he refused to testify at the other defendants’ sentencing hearing, despite an order compelling his testimony. Ibid. So the district court, upon the government‘s motion, unsealed and reviewed Mr. Kessack‘s statement. Ibid. Mr. Kessack then made a second, sworn statement that “cast doubt on the accuracy and reliability of his first [s]tatement.” Ibid. But the district court, after determining that his first statement “was corroborated by other evidence,” found that the conspiracy “involved more than 50 kilograms of cocaine” based on information in the first statement. It sentenced the other defendants based on that amount. Ibid.
We affirmed, approving the district court‘s procedure of treating the hearsay statement as “presumptively unreliable” and finding that presumption rebutted upon consideration of other, corroborating evidence. Id. at 1369. Indeed, we expressly endorsed the use of “extrinsic corroborating evidence to establish the reliability of hearsay . . . at sentencing,” including in-court witness testimony and admissions by members of the conspiracy. Ibid.
...
Notably, the external corroborating evidence considered by the district court in Petty was evidence that the defendants could have attempted to impeach via cross-examination in the normal course of the hearing. Had that evidence been false, the defendants had the usual tools of criminal procedure to challenge and refute it. And, had the defendants successfully refuted such evidence, there would have been no external corroboration supporting the hearsay statement, and it would have been inadmissible. Thus, the district court‘s procedure did not offend due process because, unlike in Weston, the burden was not on the defendants to disprove the government‘s allegations, and the government had proffered sufficient evidence tending to corroborate the coconspirator‘s hearsay statement.
iii. United States v. Huckins, 53 F.3d 276 (9th Cir. 1995). In Huckins, we found insufficient indicia of reliability to support hearsay statements by the defendant‘s accomplice that the defendant had been armed with a gun during two bank robberies. Id. at 278-79. The accomplice‘s statements “were not made under oath, nor at trial where he could be cross-examined” but rather “in the context of plea negotiations with the government, in which [he] may very well have been hoping to curry favor with law enforcement officials.” Id. at 279. Moreover, the part of the accomplice‘s statements attesting that the defendant had been armed was collateral to the part inculpating the accomplice. Ibid. We therefore removed those statements from consideration, leaving a single bank teller‘s statement that the defendant put his hands in his pocket during a third robbery, which we found was not enough to “prove by a preponderance of the evidence that Huckins was armed” at any of the robberies. Ibid.
It is also noteworthy that we rejected the government‘s argument on appeal that the bank teller‘s statement about one robbery was partially corroborative of the coconspirator‘s statement about the other two robberies. See Brief for the United States at 19, Huckins, 53 F.3d 276 (No. 94-30052), 1994 WL 16059689, at *19. Instead, we agreed with the government‘s statement in the district court that the bank teller‘s “perception that Mr. Huckins had a gun did not constitute evidence that he in fact had one.” Defendant-Appellant‘s Opening Brief at 9, Huckins, 53 F.3d 276 (No. 94-30052), 1994 WL 16059688, at *9; Huckins, 53 F.3d at 279. That foreshadowed the way in which our substantive
iv. United States v. Garcia-Sanchez, 189 F.3d 1143 (9th Cir. 1999). Although we noted in Garcia-Sanchez that “[w]e show great deference to trial court factual determinations, reviewing [them] only for clear error,” id. at 1148, we nevertheless vacated the district court‘s sentence on the grounds that the evidence did not have sufficient indicia of reliability. Indeed, Garcia-Sanchez reads much like a replay of Weston. We held that the district court erred by relying solely on “conclusory testimony” from a case agent at the sentencing hearing, who stated:
[a]s a result of interviewing Lawrence Bertolino, who is the principal middle person here in Spokane for Cipriano, [i.e.] Zavala[,] and Rutilio Garcia[-Sanchez]‘s distribution of cocaine and heroin, we determined that cocaine was being sold anywhere from two to three ounces a week from the Bertolino residence. In addition to one to two ounces of heroin.
Id. at 1149 (alterations in original). In particular, the agent “had no first-hand knowledge of the conspiracy‘s sales,” “did not explain how he arrived at his estimates,” “did not reveal the hearsay upon which he relied,” “did not produce the contemporaneous . . . reports of his interviews,” “was not cross-examined,” and “was not tested or challenged” on his opinion. Ibid.
The procedural defect in Garcia-Sanchez was the same as in Weston. Although the case agent had identified the source interviewed for the report, the report was nevertheless conclusory, not breaking down the steps of the agent‘s
Thus, up to the turn of the millennium, our cases on the minimal-indicia-of-reliability doctrine generally treated it as a procedural issue, even if we also implicitly considered the substantive reliability of the hearsay statements at issue. The main concern was whether the district court unfairly put the burden on the defendant to produce evidence to disprove government allegations. Although in some cases we allowed corroboration by external evidence, the defendants in such cases had an opportunity to challenge that external evidence through cross-examination. The “indicia of reliability” flowed from the procedural mechanisms allowing the defendant to challenge government allegations.
b. Later, We Began Considering Substantive Indicia of Reliability
Starting in the late 1990s, we expanded our understanding of “minimal indicia of reliability” to include whether the hearsay statement itself was likely substantively reliable. If so, its admission at sentencing satisfied due process, even if the defendant did not have the opportunity to challenge the source or other evidence corroborating the source in open court.
i. United States v. Chee, 110 F.3d 1489 (9th Cir. 1997). Chee gives an early example of our substantive-reliability review. There, the defendant pleaded guilty to assaulting his girlfriend on an Indian reservation. Id. at 1491-92. An FBI agent interviewed the girlfriend at a hospital, where she said
We held that the district court permissibly relied on the statements. The district court had found that the statements were “credible and trustworthy” at least partially “because they were made ‘immediately upon contact with her mother, with the contact by the police, and to hospital staff.‘” Ibid. We also noted that the girlfriend‘s statements were “corroborat[ed] by other statements, including [the defendant]‘s statement.” Id. at 1493. Moreover, her statements were consistent with the state police officer‘s description of her injuries. Thus, in Chee, we began to primarily consider the intrinsic reliability of the hearsay statement itself, albeit still in combination with external corroborating evidence that the defendant could challenge at the sentencing hearing.
ii. United States v. Berry, 258 F.3d 971 (9th Cir. 2001). This case established that even presumptively unreliable statements can permissibly corroborate one another. The
We affirmed. Although the district court had not made express findings about the reliability of the coconspirators’ statements, we held that reversal was not warranted because the “reliability of the hearsay statements [wa]s apparent from the record.” Id. at 976. Citing cases from the First and Eleventh Circuits, we adopted the rule that “hearsay statements by co-defendants that are consistent with each other may be deemed sufficiently reliable even if such statements are self-serving and contrary to the testimony of the defendant.” Id. at 976-77. And three coconspirators’ statements that the defendant had directed the enterprise in the same manner with respect to each defendant was enough consistency to satisfy us that the statements had “some minimal indicia of reliability.” Ibid. (quoting Petty, 982 F.2d at 1369).
c. Balancing Procedural and Substantive Protections
Following Berry, our cases continued to show some concern with affording the defendant procedural protections against coconspirator hearsay, generally in cases with no substantive indicia of reliability.
i. United States v. McGowan, 668 F.3d 601 (9th Cir. 2012). In McGowan, we vacated the sentence of a prison guard convicted of assaulting inmates. One inmate had testified in a different case that he had given the defendant drugs to smuggle into prison, and the district court admitted that testimony during sentencing in transcript form. Id. at 607. That same inmate had also given similar statements to FBI agents. Ibid. The government tried to corroborate the inmate‘s statements by noting that the inmate had known where the defendant‘s house was located. Ibid. Disagreeing with the district court, we held such knowledge to be insufficient corroboration, noting that the defendant had
In rejecting the inmate‘s hearsay, we emphasized the procedural defects. Even though the inmate had testified—“fleetingly“—under oath in a different case, the sentencing judge had only the bare transcript and therefore no opportunity to perceive the inmate‘s demeanor during that testimony. Id. at 607. Moreover, the defendant had no opportunity to cross-examine the inmate‘s testimony because it was in a different case, and neither did the defendant (nor the government) in that case have any incentive to cross-examine the inmate. Id. at 607-08. Thus, we held, the “allegations were made under oath but absent any other procedural mechanism that would ensure that a witness with the incentive to lie was telling the truth.” Id. at 608 (emphasis added).
ii. United States v. Pimentel-Lopez, 859 F.3d 1134 (9th Cir. 2016). In Pimentel-Lopez, we applied the rule from Berry in considering the statements of a codefendant, Mr. Elizondo, as related through a government agent at sentencing. The agent testified that Mr. Elizondo had said that Mr. Pimentel-Lopez had directed Mr. Elizondo‘s fiancée and her sister “to rent a house ‘to be used . . . to distribute drugs‘” and that he had “directed two individuals to deposit the proceeds of the drug sales into a bank account.” Id. at 1144. The fiancée had corroborated those allegations in a police interview before trial. Ibid. But, at trial, neither the fiancée nor her sister testified that Mr. Pimentel-Lopez had directed them to rent any house or directed anyone to deposit drug proceeds. Ibid. In fact, the
Because Mr. Elizondo‘s “statements were not made under oath, nor at trial where he could be cross-examined,” id. at 1144 (quoting Huckins, 53 F.3d at 279), we applied the presumption that “a codefendant‘s confession inculpating the accused is inherently unreliable,” ibid. (quoting Lee, 476 U.S. at 546). And, following Berry, we asked whether external evidence corroborated Mr. Elizondo‘s hearsay statements. Ibid. The only external corroboration of his statements was another out-of-court statement by his fiancée. Ibid. Recognizing that even “self-serving” statements can constitute minimal indicia of reliability, ibid. (quoting Berry, 258 F.3d at 976-77), we nevertheless had sufficient “doubt” of Mr. Elizondo‘s purported statements to deem them uncorroborated, ibid. For one thing, even though the fiancée had corroborated those hearsay statements before trial, she did not testify to them at trial. Ibid. For another, the sister‘s testimony that she could not speak Spanish made the proposition that Mr. Pimentel-Lopez had “directed” her to do something unlikely. Ibid.
Our inquiry in Pimentel-Lopez was essentially factual. Although we noted that the statement at issue and the corroborating statement were both out-of-court hearsay, the defendant‘s inability to challenge them did not drive our decision. Nor could it have after Berry. Instead, contrary to the district court, we assigned greater weight to the statements given under oath in open court, even considering silence on some issue in court to be enough to discount an earlier hearsay statement. And we considered the logical
iii. United States v. Vera, 893 F.3d 689 (9th Cir. 2018). In Vera, the government sought to prove the amount of drugs that the defendant brothers had sold by using the factual statements from twelve codefendants’ plea agreements. Id. at 691, 694. Again applying Berry, we examined those plea agreements to determine whether they sufficiently corroborated one another. We extensively analyzed the facts alleged in the plea agreements. Although those allegations adequately established that the Veras had some part in the drug scheme, only four specific drug transactions were referenced in more than one plea agreement (out of the forty total transactions represented in the agreements). Id. at 694-95. The plea agreements attributed none of those four transactions specifically to either brother. Id. at 695. Although the drug quantities listed in one plea agreement were verified independently by DEA laboratory reports, neither the plea agreement nor the laboratory reports
4. Synthesizing the Case Law
Having examined the development of the minimal-indicia-of-reliability doctrine over the last half-century, we conclude that there are two distinct questions that we answer in examining a hearsay statement at sentencing: (1) whether the statement is “procedurally reliable” and (2) whether the statement is “substantively reliable.” This is a disjunctive test: If we answer either question in the affirmative, then the statement may be considered at sentencing.
First, procedural reliability. We ask whether there are sufficient procedural protections so that the defendant does not have to “prove a negative” in the face of government allegations. This is an essentially legal question because whether the defendant is in that position “implicate[s] constitutional rights” and requires us “to exercise judgment about the values that animate legal principles.” Hinkson, 585 F.3d at 1260 (citation omitted).
Generally, if the government supports the hearsay statements with extrinsic evidence that the defendant can challenge on cross-examination, then we have found the process to be adequate to ensure that the defendant is not sentenced on the basis of unreliable or false information. See Petty, 982 F.2d at 1366-69. The district court may then consider the hearsay statement under the rubric of procedural
Second, if the government offers no corroboration of a hearsay statement that the defendant can challenge at sentencing through the normal adversarial process, we proceed to the substantive inquiry. As our cases show, substantive indicia of reliability can be enough to safeguard the defendant‘s right not to be sentenced on the basis of unreliable or false information. Thus, hearsay from a source that is self-demonstrably reliable is permissible on its own. See Chee, 110 F.3d at 1492. And even if the hearsay is from a presumptively unreliable source, such as a coconspirator, the government can prove its reliability by exhibiting other, independently obtained, consistent hearsay statements—even other presumptively unreliable statements, as in Berry. But, unlike procedural reliability, substantive reliability is an essentially factual issue. It requires judging whether a statement is probably truthful in light of all the circumstances—that is, “the application of the fact-finding tribunal‘s experience with the mainsprings of human conduct.” Hinkson, 585 F.3d at 1259 (quoting McConney, 728 F.2d at 1202).
The upshot is this. A determination of procedural reliability—that the hearsay in question does not put the burden on the defendant to prove a negative and that the defendant has adequate opportunity to confront corroborative evidence of the hearsay—is an essentially legal question that we review de novo. A determination of
D. Application
Here, the government gives two reasons that the use of Mr. Hiler‘s and Mr. Pitsch‘s statements8 at Mr. Franklin‘s sentencing did not violate his due-process rights. One invokes procedural reliability; the other invokes substantive reliability.
1. Procedural Reliability
We first consider the government‘s argument that external, nonhearsay evidence corroborates Mr. Hiler‘s hearsay statements. The statements at issue here are (1) that
Mr. Franklin did not bear the burden of disproving conclusory government allegations. Unlike in Weston and Garcia-Sanchez, the government did not use a government agent as a mere mouthpiece for unsourced obstruction-of-justice allegations. Instead, Special Agent Bowen related detailed, specific statements and identified the sources of those statements—Mr. Hiler and Mr. Pitsch. The government also offered extrinsic, nonhearsay evidence to corroborate those hearsay statements: (1) Special Agent Bowen‘s personal observation of Mr. Hiler‘s mouth injuries, (2) the copy of the note (purportedly from Mr. Franklin) that Special Agent Bowen received, and (3) the transcript of Mr. Franklin‘s first change-of-plea hearing.9 Combined with the rebuttable presumption of unreliability that we impose on coconspirators’ inculpatory statements, the government was indeed saddled with the burden to prove the obstruction enhancement‘s applicability.
As for opportunities to challenge the government‘s external, nonhearsay evidence: First, Special Agent Bowen testified at the sentencing hearing, and Mr. Franklin cross-examined him. Cross-examination is the “gold standard” of
Second: The copy of the note was present at the sentencing hearing and entered into evidence. If the note really had been forged, Mr. Franklin could have tried to challenge the note‘s provenance by, for example, introducing an exemplar of his own handwriting or having a handwriting analysis expert testify to show that he had not written the note. Or he could have attempted to find some detail stated in the note that he would not have known but someone else (presumably whoever forged it) would have known. And so on. In other words, Mr. Franklin had a full and fair opportunity to challenge the premise that he was the source of the note. So he could adequately meet that evidence, too.
And third: The transcript of the first plea hearing. Although Mr. Franklin could not reasonably dispute the reliability of the transcription, that is not a procedural problem—it is substantive. His remedy would have been to argue in the district court, as he does now on appeal, that the corroborative inference is weak—perhaps because Mr. Hiler had easy access to a transcript of the hearing and could have used it to forge the note. Cf. United States v. Matta-Ballesteros, 71 F.3d 754, 766-67 (9th Cir. 1995) (defendant‘s objection—that forensic expert‘s finding that hairs at crime scene were consistent with defendant‘s did not prove defendant was the only person who could have left the hairs—went to weight, not admissibility); id. at 768-69 (objection to defect in chain of custody went to weight, not admissibility).
2. Substantive Reliability
We also consider the government‘s argument that Mr. Hiler‘s and Mr. Pitsch‘s hearsay statements corroborate each other enough to be admissible at sentencing. Although the district court did not make an express finding of reliability, we read the court‘s ruling on the obstruction-of-justice enhancement as making such a finding implicitly.10 We review it for clear error.
And we find none. Each coconspirator reported that Mr. Franklin pressured him to testify a certain way—or not to testify—following the first change-of-plea hearing. Although the two accounts were not uniformly consistent, as was the case in Berry, they both flowed from the same triggering incident: Mr. Franklin‘s first plea hearing. They shared the common premise that Mr. Franklin had a consistent underlying motivation: that he wanted no one to testify contrary to his statements at that plea hearing, whether to avoid a perjury charge or to potentially win an acquittal at trial. And Mr. Hiler‘s statement that “Curly” is a
IV. Conclusion
Binding precedent forecloses Mr. Franklin‘s claim that Hobbs Act robbery is not a crime of violence under
AFFIRMED.
BERZON, Circuit Judge, concurring in the judgment:
I agree with the majority‘s conclusion that the district court did not err in relying on hearsay statements from Hiler and Pitsch as the basis for an obstruction-of-justice enhancement to Franklin‘s sentence. I write separately because my interpretation of our case law on the minimal-indicia-of-reliability doctrine is somewhat different from the
I
We have regularly stated that we “review for abuse of discretion the district court‘s evaluation of the reliability of evidence at sentencing.” United States v. Vera, 893 F.3d 689, 692 (9th Cir. 2018); see, e.g., United States v. McGowan, 668 F.3d 601, 606-08 (9th Cir. 2012); United States v. Berry, 258 F.3d 971, 976 (9th Cir. 2001); United States v. Chee, 110 F.3d 1489, 1492-93 (9th Cir. 1997). As the majority observes, however, we have not clearly explained how we determine whether a district court abused its discretion by basing a sentence on hearsay. Majority op. 15.
The majority carefully traces the development of our case law in this area and demonstrates that our earlier cases often focused on the procedural injury to a defendant when the district court based a sentence on unreliable hearsay, while our later cases were more likely to engage in a close factual examination of a hearsay statement to evaluate its reliability. Id. at 21-35. Based on this inquiry, the majority concludes that due process allows a district court to base a sentence on a hearsay statement if either the statement is “procedurally reliable,” that is, “there are sufficient procedural protections so that the defendant does not have to ‘prove a negative’ in the face of government allegations,” id. at 35, or the statement is “substantively reliable,” meaning, for example, that the statement is “from a source that is self-demonstrably reliable” or there is independent evidence corroborating it, id. at 36.
For example, the majority points to United States v. Weston, 448 F.2d 626 (9th Cir. 1971), as “stand[ing] for the proposition that if the sentencing process effectively puts the burden of proof on the defendant to refute a damaging hearsay allegation, particularly when the factual basis for believing such a charge is practically nonexistent, that process is legally flawed.” Majority op. 23. But, as the majority acknowledges, the reason we were concerned about putting the burden on the defendant in Weston to refute the hearsay statements in the presentence report is that we doubted the statements’ substantive reliability: “the factual basis for believing the charge was almost nil.” 448 F.2d at 633. In other words, the government had failed to produce substantive indicia of reliability to support the hearsay statements, instead leaving it to the defendant to refute them. We vacated the sentence and forbid the district court from relying on the presentence report on resentencing “unless” the government “amplified [it] by information such as to be persuasive of the validity of the charge there made“—i.e., unless the government produced substantive indicia of reliability. Id. at 634. We did not suggest that reviewing the sentencing hearing but applying the correct burden of proof would suffice.
Petty concluded that the district court “did not abuse its discretion in concluding that the [hearsay] Statement, when viewed in light of the corroborating evidence [introduced by the government], was sufficiently reliable.” 982 F.2d at 1369. The extrinsic evidence included trial testimony, post-arrest admissions by codefendants, and “tape recorded admissions of . . . one of the defendants.” Id.
The majority suggests that “the external corroborating evidence considered by the district court in Petty was evidence that the defendants could have attempted to impeach via cross-examination in the normal course of the hearing.” Majority op. at 25. Thus, the majority concludes, the hearsay statements were “procedurally reliable.” Id. at 35-36.
Although it does appear that in Petty the defendants had an opportunity to test some of the extrinsic evidence by cross-examination (such as the trial testimony), we did not rely on that circumstance to decide the case. Instead, we concluded, albeit with scant reasoning, that the extrinsic evidence “corroborat[ed]” the hearsay statements “sufficiently” to establish their substantive reliability, or at least that the district court did not err in so finding. 982 F.2d at 1369. If we had concluded that the extrinsic evidence did not corroborate the hearsay statements (as Judge Noonan, in dissent, suggested it did not, 982 F.2d at 1372), I do not see how we could have deemed the hearsay statements reliable,
It may well be true that we are more likely to accept a hearsay statement as substantively reliable if the government introduces corroborating evidence that is subject to cross-examination. But I read our cases as consistently requiring at least some indicium of substantive reliability before a hearsay statement may form the basis of a sentence. Procedural protections—such as the ability to cross-examine evidence the government introduces to corroborate hearsay—are a useful adjunct, but cannot alone establish the reliability of a hearsay statement.
II
I agree with the majority that substantive reliability is an essentially factual issue that we review for clear error. Majority op. 36-37. Here, although there may be a procedural dimension to our inquiry, see id. at 37-40, the questions whether Hiler‘s and Pitsch‘s statements corroborate each other, and whether the note purportedly from Franklin corroborates any of the statements, are essentially factual questions. Although I view the questions as close, I cannot say that the district court clearly erred in finding the statements sufficiently corroborated to be reliable.
Hiler‘s and Pitsch‘s statements were different from each other: Hiler said that Franklin sent him a note asking him to testify to specific facts, while Pitsch said that Franklin warned him that what he said “would be on paper, and that he should be careful.” Hiler also said that after he shared Franklin‘s note with the government, he was assaulted by two other inmates who told him that Franklin had told them to “slap [him] around.”
The note purportedly from Franklin is corroborative of Hiler‘s first statement and of Pitsch‘s statement. According to Hiler, the note referred to Pitsch as “Curly.” For example, the note asked Hiler to testify to the fact that “Curly had his own peice [sic] and I had nothing to do with it.” The note also says, “I know you said nothing against my best interest and it was Son and Curly that was squaking [sic] like birds.” The note‘s suggestion that its author believed Pitsch had implicated Franklin but Hiler had not provides an explanation for why Franklin would ask Hiler to testify to specific facts, but would simply warn Pitsch to watch what he said.
As the majority explains, the fact that Franklin had an opportunity to challenge the note at his sentencing hearing increases its value as corroborating evidence. Majority op. 39. Franklin could have attempted to prove that he was not the author of the note in various ways—for example, through a handwriting analysis—but he did not.
Given that Hiler‘s and Pitsch‘s statements were at least somewhat corroborative of each other, that the note was corroborative of both men‘s statements, and that “Franklin had a full and fair opportunity to challenge the premise that he was the source of the note,” Majority op. 39, I cannot say that the district court clearly erred in finding that Hiler‘s and Pitsch‘s statements bore adequate indicia of reliability to
Notes
Separately, Mr. Pitsch told Special Agent Bowen that Mr. Franklin had warned him that he was going to be branded as a snitch and advised him to be cautious. Taking the warning as a threat, Mr. Pitsch asked for a transfer to a different detention facility.