While the commentary to S 3C1.1 does not explicitlyinclude false statements made to state administrative agencies within its examples of obstructive conduct, the examples set forth are not exhaustive. See id n. 3. "Obstructive conduct can vary widely in nature, degree of planning, and seriousness.. . . [C]omparison of the examples set forth in Application Notes 3 and 4 should assist the court in determining whether application of this adjustment is warranted in a particular case." Id. n. 2.

The district court determined that Luca had intentionally submitted false prospectuses in response to an investigative subpoena issued by the Arizona Securities Division of the Arizona Corporations Commission ("ACC"). The court concluded that "based on the documents given to the Arizona Corporations Commission, which were in fact false documents . . . it was [Luca's] intent to divert the Arizona Corporation Commission from beginning its preliminary investigation, and it allowed for additional investors to be procured during this time."

For the obstruction of justice enhancement to apply, the district court must find that the defendant willfully provided a materially false statement to law enforcement officers that actually obstructed or impeded the official investigation or prosecution of the instant federal offense. See U.S.S.G. S 3C1.1 & n. 3(g). See also United States v. Solano-Godines, 120 F.3d 957, 963-64 (9th Cir. 1997) (interpreting "statement" and "significant hindrance" requirements), cert. denied, 118 S. Ct. 722 (1998); United States v. Zagari, 111 F.3d 307, 329 (2d Cir.) (perjury in state civil environmental enforcement action could justify obstruction of justice enhancement where material to federal crime), cert. denied sub nom., Herzog v. United States, 118 S. Ct. 445 (1997) & cert. denied sub nom., Shay v. United States, -U.S.-,119 S. Ct. 1585, 143 L.Ed.2d 679(1999). It is important to stress that the state agency was investigating "the offense," namely the Ponzi scheme that Luca generated and which resulted in his conviction and sentence in this case. For purposes of applying the obstruction enhancement, we can find no convincing reason to distinguish between state and federal law enforcement officers who are investigating the same offense.

Luca's argument that an implicit distinction exists between obstruction of state and federal law enforcement officials misapprehends the focus of the obstruction enhancement, which addresses the effect of the obstructive conduct rather than the level of law enforcement that was obstructed. For example, the guideline's commentary indicates that some obstructive conduct, such as providing a false name or identification at arrest or making false unsworn statements to law enforcement officials, should not ordinarily result in an enhancement. Where "such conduct actually resulted in significant hindrance to the investigation or prosecution," however, the severity of the obstruction justifies a sentence enhancement. Compare U.S.S.G. S 3C1.1 n. 4(a) and (b), with id. n. 3(g). Under the obstruction guideline's commentary, it is immaterial which law enforcement agency was obstructed. So long as the defendant provided information that was materially false, with a willful intent to obstruct or attempt to obstruct the discovery of his or her scheme, and which had the effect of significantly impeding the investigation, the obstruction enhancement is appropriate.

Willfully providing material false statements to federal law enforcement officials undisputedly justifies the obstruction enhancement where actual obstruction occurs. See, e.g., United States v. McNally, 159 F.3d 1215, 1217 (9th Cir. 1998) (false statements to FBI denying whereabouts of kidnapped child constituted obstruction); United States v. Ancheta, 38 F.3d 1114 (9th Cir. 1994) (false statements to DEA agents denying involvement in drug transaction constituted obstruction); United States v. Benitez, 34 F.3d 1489, 1497 (9th Cir. 1994) (enhancement appropriate where defendant provided false date and place of birth to federal pretrial services and failed to disclose outstanding warrant). In Zagari, the Second Circuit held that perjury in a state regulatory suit could support the obstruction enhancement in a related federal criminal conviction. See 111 F.3d at 328. We agree with the Second Circuit: so long as the district court found that the defendant "willfully and materially impeded the search for justice in the instant offense ," the enhancement should apply, even if the obstruction occurred before state rather than federal law enforcement officials. See id. at 328-29 (emphasis added).

Certain types of crimes, for example securities crimes and violations of environmental laws, are commonly investigated jointly by state and federal agencies. Others, such as some drug offenses, are investigated initially by local or state law enforcement officials, and only later do federal officials join the investigation. In either circumstance, efforts made to obstruct a local or state investigation may also obstruct the federal investigation. To limit the application of the obstruction of justice enhancement to circumstances where the defendant directed his or her obstructive conduct to federal law enforcement officials would interfere with the guideline's goal of punishing more severely those whose actions "significantly obstruct or impede the official investigation" of the federal crime.

The Supreme Court has interpretedS 3C1.1 as requiring sentencing courts to "review the evidence and make independent findings necessary to establish a willful impediment to, or obstruction of, justice, or an attempt to do the same." See United States v. Dunnigan, 507 U.S. 87, 95 (1993). The enhancement is mandatory where the court finds that the defendant provided false testimony on a material matter with willful intent. See Ancheta, 38 F.3d at 1118. Here, the trial court's conclusion, which finds ample support in the record, justifies imposing the enhancement to Luca's sentence. First, he trial court found that Luca's submissions to the ACC were false. The court also found that Luca intentionally submitted the documents in order to forestall the investigation and that the obstruction was material since the documents diverted the ACC and gained Luca enough additional time to secure other victims. Finally, the court found that Luca's obstruction was related to the instant offense because the State was investigating the same scheme to which Luca eventually pled guilty in federal court. We affirm the two-level enhancement for obstruction of justice.

The district court enhanced Luca's offense level by four levels after determining that U.S.S.G. S 3B1.1(a) applied because Luca's fraud was "otherwise extensive. " The guideline provides:Based on the defendant's role in the offense, increase the offense level as follows:

(a) If the defendant was an organizer or leader of a criminal activity that involved five or more participants or was otherwise extensive, increase by four levels.

See U.S.S.G. S 3B1.1(a). The commentary to this enhancement provides that "[t]o qualify for an enhancement under this section, the defendant must have been an organizer, [or] leader . . . of one or more participants." See id n. 2. Accord United States v. Anderson, 942 F.2d 606, 615 (9th Cir. 1991) (en banc); United States v. Avila, 95 F.3d 887, 889 (9th Cir. 1996) ("there must be evidence that the defendant exercised some control over others involved in the commission of the offense [or was] responsible for organizing others for the purpose of carrying out the crime"). "A `participant' is a person who is criminally responsible for the commission of the offense, but need not have been convicted." U.S.S.G. S 3B1.1 n. 1. Unknown facilitators of crimes will not be considered criminally responsible participants. See United States v. Cyphers, 130 F.3d 1361, 1363 (9th Cir. 1997).

The court explicitly concluded that it did not need to identify any "participants" whom Luca oversaw or managed. Luca argues that the opinion in Anderson, 942 F.2d 606, requires us to invalidate the S 3B1.1(a) enhancement because the district court did not explicitly find that there was another participant in the crimes. Luca also contends that alternate grounds which might exist in the record are insufficient to support an enhancement as a matter of law under U.S.S.G. S 3B1.1(a). We agree.

When the district court sentenced Luca, it based the S 3B1.1(a) enhancement on Luca's leadership role in an"otherwise extensive" criminal activity:

[B]y the overwhelming evidence -- not by preponderance, but by the overwhelming nature of the evidence -- before me, including factual basis at the change of plea, this fraud was extensive, and it involved a whole host of entities and people, wittingly and unwittingly, to keep this thing rolling for the period of time that it did.

While noting that "there is probable cause to believe that others helped Luca perpetrate some of these frauds that would qualify them as participants," the district court elected not to identify any other participant. The court suggested that it would need to hold an evidentiary hearing in order to identify any additional participants. "[T]hose people haven't been charged and I don't know who they are." The district court's decision to impose the enhancement without identifying another participant constituted legal error.

U.S.S.G. S 3B1.1(a) cannot apply unless the district court identifies a participant over whom the defendant exercised managerial or organizational control. See Anderson, 942 F.2d at 615. Here, the district court chose not to identify any other participant in Luca's scheme. It would be inappropriate for this court to make necessary findings of fact that the district court expressly declined to make. We vacate this enhancement and remand for further findings.

Luca argues that the district court erred in enhancing his sentence by two points pursuant to the vulnerable victim provision of U.S.S.G. S 3A1.1(b). The vulnerable victim enhancement provides: "[i]f the defendant knew or should have known that a victim of the offense was unusually vulnerable due to age, physical or mental condition, or that a victim was otherwise particularly susceptible to the criminal conduct, increase by 2 levels." U.S.S.G. S 3A1.1(b). The district court, principally relying on the P.S.R., concluded that "many of the victims" were vulnerable based on their ages and their membership in the church in which Luca held a leadership position. Luca argues that the court's conclusion regarding his victims' church affiliation could not give rise to a finding of vulnerability and that there was insufficient evidence in the record to support the enhancement based on his victims' ages.4

A trial judge considering an enhancement on this basis must make essentially two determinations. First, the judge must determine whether one or more of the victims belong to a class that is particularly vulnerable to the criminal activity in question, and second, is there one or more specific individual victims whom the defendant knew or should have known were unusually vulnerable by virtue of membership in the class identified. Thus the trial judge must first validate a class of individuals as unusually vulnerable and then identify one or more individuals whom the defendant knew or should have known were members of the class and victims of the offense. Here, the two guideline bases for enhancement -- age and church membership -- require slightly different analyses.

The commission has already determined that age on either end of a person's life qualifies as a protected class; thus in order to impose the vulnerable victim enhancement where a known victim was very young or very old, the trial judge need only identify the victim and find, based upon evidence in the record, that the individual was in fact unusually vulnerable because of youth or advanced age.5 In contrast, where the government relies upon a classification not spelled out in the guidelines -- such as membership in a particular religious group -- to show that a particular victim was "otherwise particularly susceptible," the class must first be validated and only then can the trial court determine whether in the particular case an identified victim's membership in the class rendered him or her unusually vulnerable. Here, it is necessary to remand the case to permit the trial court to reconsider the vulnerable victim enhancement in light of this court's decisions in Castellanos, 81 F.3d 108, and United States v. Peters, 962 F.2d 1410 (9th Cir. 1992).

This court has explained its determination of vulnerability as follows:

Section 3A1.1 will apply to increase the offense level where (1) a victim was either (a) unusually vulnerable due to age, physical or mental condition, or (b) otherwise particularly susceptible to the criminal conduct, and (2) the defendant knew or should have known of such vulnerability or susceptibility.

See United States v. Randall, 162 F.3d 557, 560 (9th Cir. 1998) (internal citations omitted). The record here does not demonstrate that the district court made factual findings sufficient to support the enhancement.

The vulnerable victim enhancement is not appropriate based on Luca's victims' ages because the trial court did not identify in his findings of fact specific individuals whose ages rendered them particularly vulnerable. The government is correct that a finding by the district court that a specific victim was unusually vulnerable due to age is enough to support a S 3A1.1(b) enhancement. As the Castellanos court noted, age "may per se render a victim worthy of the special protection" of U.S.S.G. S 3A1.1. See 81 F.3d at 110. Age standing alone, however, does not establish someone as a specific victim. Here, the district court made insufficient findings to support an age-based enhancement. While in some circumstances it is appropriate for a sentencing court to base sentence enhancements on the factual findings in presentence reports, see, e.g., United States v. Morgan, 164 F.2d 1235, 1238 (9th Cir. 1998), the P.S.R. here did not mention Luca's victims' ages. It was not until Luca's sentencing hearing that the court found Luca's victims unusually vulnerable due to their age; the district court made no other age-based findings. See United States v. Chavez-Gutierrez, 961 F.2d 1476, 1480-81 (9th Cir. 1992); United States v. Rigby, 896 F.2d 392, 394 (9th Cir. 1990).

We also find insufficient evidence to support the district court's finding that Luca's victims' church membership made them "otherwise particularly susceptible to the criminal conduct." See Randall, 162 F.3d at 560. The P.S.R. noted that

members of the Eagle's Nest Christian Assembly were convinced of the propriety of investing with Luca, in part, by and through Luca's prominent position and leadership role in the church. The direct and implied representations Luca was a `man of God' effectively lowered the guard of church members to the fraud

In response to Luca's objection to the P.S.R., the probation officer noted that "in some Pentecostal Christian Protestant churches, the importance and role of spiritual gifts, which would include prophetic utterances directly from God, are magnified and they are often relied on to assess situations and make decisions." We conclude that these findings were insufficient to support the enhancement.

We have enunciated two means of determining whether a defendant's victim is "otherwise particularly susceptible" for purposes of the vulnerable victim enhancement. In Peters, we directed the district courts to analyze "the characteristics of the defendant's victim, the victim's reaction to the criminal conduct, . . . the circumstances surrounding the criminal act [and] whether the defendant could reasonably have anticipated the victim's reaction." See 962 F.2d at 1416. In Castellanos, however, we cautioned against employing the vulnerable victim enhancement where there was no finding that the defendant's victims "are those who are in need of greater societal protections" or where some condition of the defendant's chosen victims "render the defendant's conduct more criminally depraved," see 81 F.3d at 111 (internal citations omitted). The district court's findings are insufficient to support the enhancement under either analysis.6

There is evidence that Luca's leadership position in the church caused some victims to trust him. A number of Luca's victims commented during allocution that Luca manipulated their faith to gain access to their money. For example, one victim testified that "the church out there where we went, where Mr. Luca went, endorsed him highly, the pastor did. I trusted the pastor, and thus, we trusted Mr. Luca." Another victim described how Luca prayed with her just before he showed her fraudulent layouts for his purported developments. The P.S.R. quoted a letter written by a former member of the church:

Luca was constantly being praised from the pulpit as an "anointed Christian businessman," with visiting prophets prophesying about his future successes and blessings from God. His later legal problems were called demonic attacks by these same people. . . Normally I could spot someone like Luca a mile away, but believing the EN [Eagle's Nest] active promotion of him, I turned off my internal alarms. . . Luca skillfully manipulated my faith in God to his advantage, looking me in the eye while praying to God to bless the investment, all the while stealing my life savings. . . To summarize, Luca is an expert at using people's faith in God as a means of getting to their savings, reaching through their souls to pick their pockets, taking not only their savings but also their faith.

Moreover, Luca clearly used his victims' faith to target investors. A brochure printed by the church invited parishioners to invest with Luca, announcing that "[i]n almost every case, our plan will be able to at least match or out perform your current yields, and at the same time earn dividends for our church and its future. These funds will become the backbone of our plan to build the church campus and retire all debt within five years." In soliciting parishioners' investments, Luca announced:

We can take . . . individuals who have whether its [sic] $2 hundred dollars [sic] in a savings account or $2 hundred thousand dollars [sic] in mutual funds, and we can allow you to retain the principal, but you use that interest . . . to help build God's kingdom, and also receiving the same rate that you're receiving currently from the bank. . . . I consider it a real honor and a privilege to be able to be an elder of this church and to be able to take part . . . in a vision that . . . will allow us . . . to quadruple in size . . . and when we finally get this facility, were [sic] gonna be able to minister to so much [sic] more people.. . .

However, the district court did not point to any facts that made the victims less able to defend themselves than a typical victim. Regardless of whether we employ the Peters or Castellanos analysis, we cannot conclude that the district court addressed thus court's concern that Luca's victims be "particularly susceptible" to the fraud.

In Castellanos, the defendant formed a company that targeted Spanish-speaking customers for investments and billed itself as a "proudly Hispanic company." The company turned out to be fraudulent. The district court made no specific findings as to the susceptibility of any of the individual victims, but enhanced the sentence because the defendant "targeted Spanish-speakers, many of whom may have placed trust in the defendant because he was one of their own." See 81 F.3d at 110. We vacated the enhancement, noting that while district courts do not need to make specific finding of individual victims' susceptibility,

[n]othing in the record supports a finding that the Spanish-speaking population of Southern California as a whole shares some unique susceptibility to fraud that warrants the law's protection, or that makes Castellanos' crime especially reprehensible. Nor is there any showing that the entire Hispanic population of Southern California shares any such susceptibility.

See id. at 112.

We recognized that criminals typically direct their activities toward those who are more likely to succumb to the scheme and indicated that the district court must consider more than the fact that a defendant would likely succeed in defrauding his or her targeted victim. See id. at 111. We noted that an affinity scam might support the enhancement where a finding was made that victims possessed specific vulnerabilities, but refused to find an entire ethnicity vulnerable because of their shared language or cultural heritage. See id. By looking to whether the victims are in greater need of social protection, the analysis set forth in Castellanos directs the trial court to focus on which of the victims' traits makes the defendant's criminal activity more depraved. See id. The analysis distinguishes the average victim, who may be "more likely to fall victim to the scheme," see id., from the victim who is "otherwise particularly susceptible," see U.S.S.G. S 3A1.1(b). While the Peters factors allow the sentencing court to consider a broader range of factors than does the analysis described in Castellanos, the focus must still remain on which characteristics made the victims especially liable to fall for the scheme. See Peters, 962 F.2d at 1416.

The district court here did not make any findings regarding Luca's victims susceptibility to his scam. At the sentencing hearing the district court concluded that Luca's victims were vulnerable as "a result of their association with Mr. Luca through church affiliated activities, where he was .. . a church official who was elected to be in charge of finances." The court later indicated that it applied the vulnerable victim enhancement because Luca "use[d] religious connections to aid in the perpetuation of the scheme. . . ." On remand the district court should evaluate the record in the light of Castellanos and Peters and make specific fact findings regarding the vulnerability of Luca's victims.7

Luca's sentence is VACATED and REMANDED for resentencing consistent with this decision.