HOWARD, Chief Judge.
These appeals arise out of a widespread corruption scandal at the Puerto Rico Board of Medical Examiners (the "Board"), the former licensing authority for doctors seeking to practice in Puerto Rico. Cesar Berroa, Julio Castro, Geraldo Castro, Raysa Pacheco-Medina, and Glenda Davila all sought medical licenses but failed to pass the required exams. Undeterred, they attempted to gain certification by obtaining falsified scores. A federal indictment and subsequent jury trial led to convictions on various charges against each defendant.
The appeals raise a litany of claims, and "[a]fter carefully considering each of the defendants' contentions and extensively reviewing the record," we address only those claims that are "worthy of discussion; the remainder lack arguable merit." United States v. Rose, 802 F.3d 114, 117 (1st Cir. 2015).1
We affirm the defendants' convictions for honest-services mail fraud conspiracy, but reverse the convictions for money or property mail fraud and aggravated identity theft, finding the government's theories of prosecution on those counts to be legally deficient.
All five defendants sought admission to practice medicine in Puerto Rico. The admissions process required applicants to pass a pair of gatekeeping tests: a basic exam and a clinical written exam. Applicants who achieved a minimum score of 700 on each of the two tests would then move on to a practical skills exam. Upon passage of the practical skills exam and completion of the remaining requirements, the Board would issue a regular medical license.
The government presented evidence that each of the defendants failed to achieve the required 700 score on at least one of the gatekeeping exams. As a result, they turned to Yolanda Rodríguez, an employee at the Board who had access to applicant files and the ability to create fraudulent score results. The process was decidedly low-tech: Rodríguez used a photocopier to superimpose passing scores of other applicants onto the failing students' exam sheets. She then placed the falsified exam sheets back into the applicants' files. The trial evidence supported a finding that each of the defendants' files contained a passing score sheet falsified by Rodríguez. Armed with passing scores, the previously unsuccessful applicants completed the remaining requirements and entered practice as medical doctors in Puerto Rico.
On April 20, 2010, a federal grand jury handed up an omnibus 138-count superseding indictment against the five defendants who have brought these appeals and a myriad of other applicants.2 All five defendants were indicted for conspiracy to commit honest-services mail fraud, money or property mail fraud, and aggravated identity theft. The government proceeded on consistent underlying theories for all of the defendants: (1) that they joined in a conspiracy to commit honest-services mail fraud in obtaining their medical licenses; (2) that they committed mail fraud by using the resulting licenses to practice medicine for financial gain; and (3) that they committed aggravated identity theft by issuing prescriptions to patients.
After trial, the jury convicted3 the defendants as follows:
Berroa: mail fraud, honest-services mail fraud conspiracy, and aggravated identity theft;
Julio Castro: mail fraud and honest-services mail fraud conspiracy;
Geraldo Castro: mail fraud and aggravated identity theft;
Pacheco: honest-services mail fraud conspiracy; and
Davila: honest-services mail fraud conspiracy.
II. Sufficiency of the Evidence
The defendants now attack the sufficiency of the evidence supporting their various convictions. These preserved challenges garner de novo review. United States v. Ridolfi, 768 F.3d 57, 61 (1st Cir. 2014). "Applying a familiar standard, we consider whether any rational factfinder could have found that the evidence presented at trial, together with all reasonable inferences, viewed in the light most favorable to the government, established each element of the particular offense beyond a reasonable doubt." Id. (citation omitted).
A. Money or Property Mail Fraud
Berroa, Julio Castro, and Geraldo Castro appeal their convictions for mail fraud in violation of 18 U.S.C. § 1341. This provision proscribes use of the mails in furtherance of "any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses." Because we find insufficient evidence to support the conclusion that the defendants obtained money or property "by means of" their alleged fraud, we reverse these convictions.
The Supreme Court has squarely held that the mail fraud statute is "limited in scope to the protection of property rights." McNally v. United States, 483 U.S. 350, 360, 107 S.Ct. 2875, 97 L.Ed.2d 292 (1987). Before this ruling, the statute had been used to prosecute "various forms of corruption that deprived victims of `intangible rights' unrelated to money or property." Cleveland v. United States, 531 U.S. 12, 18, 121 S.Ct. 365, 148 L.Ed.2d 221 (2000). McNally expressly curtailed this use of § 1341. Congress later passed a new statute, 18 U.S.C. § 1346, designed to cover one of the intangible rights recognized in the pre-McNally caselaw, namely, "the intangible right of honest services." Cleveland, 531 U.S. at 19-20, 121 S.Ct. 365 (quoting 18 U.S.C. § 1346). Here, the relevant counts of the indictment allege a scheme to deprive the victims of money or property. Accordingly, we restrict our inquiry to § 1341 for the time being.
The Supreme Court has broadly and unequivocally instructed that "[s]tate and municipal licenses" generally "do not rank as `property,'" sufficient to support a conviction under § 1341. Id. at 15, 121 S.Ct. 365. In Cleveland, the defendants were alleged to have made false statements in applications for state gaming licenses. The Court began its analysis by explaining that any interest the state had in the licenses was "regulatory," as opposed to proprietary, in nature. Id. at 20, 121 S.Ct. 365. It noted the government's concession that many other state licenses, including "medical licenses," are "purely regulatory." Id. at 22, 121 S.Ct. 365. But the Court did not rest solely on the fact that the government's theory of prosecution "stray[ed] from traditional concepts of property." Id. at 24, 121 S.Ct. 365. Rather, it went on to note that the government's preferred reading of the statute would result in "a sweeping expansion of federal criminal jurisdiction in the absence of a clear statement by Congress." Id. Indeed, "[e]quating issuance of licenses ... with deprivation of property would subject to federal mail fraud prosecution a wide range of conduct traditionally regulated by state and local authorities." Id. In short, Cleveland squarely precluded the government from seeking mail fraud convictions on the theory that the defendants defrauded the Board out of some property interest in the medical licenses.
Presumably cognizant of this restriction, the government charged a scheme to "depriv[e] unsuspecting consumers of health care services, health care benefit programs and health care providers, of property and money through the defendant[s'] knowing use of [their] fraudulently obtained medical license[s]." More specifically, the defendants allegedly used their fraudulent licenses to obtain payment for medical services and issue prescriptions. They continued to write prescriptions at least until about two to three years after receiving their licenses.
In its effort to circumvent Cleveland, the government runs headlong into another Supreme Court precedent. Loughrin v. United States, ___ U.S. ___, 134 S.Ct. 2384, 189 L.Ed.2d 411 (2014), involved a prosecution under the bank fraud statute, which prohibits schemes to obtain bank property "by means of false or fraudulent pretenses." 18 U.S.C. § 1344(2). The Court described the statute's "by means of" language, also present in § 1341, as a "textual limitation" on its scope. Loughrin, 134 S.Ct. at 2393. This limitation assuaged federalism concerns about infringing on state criminal jurisdiction. Id. at 2392-93. The Court explained that "by means of" "typically indicates that the given result (the `end') is achieved, at least in part, through the specified action, instrument, or method (the `means'), such that the connection between the two is something more than oblique, indirect, and incidental." Id. at 2393 (citing Webster's Third New International Dictionary 1399 (2002); 9 Oxford English Dictionary 516 (2d ed. 1989)). Accordingly, "not every but-for cause will do." Id. Rather, the "by means of" language requires that the defendant's fraud be "the mechanism naturally inducing a bank ... to part with money."4 Id. Here, the defendants' alleged fraud in obtaining their medical licenses cannot be said to have "naturally induc[ed]" healthcare consumers to part with their money years later.
The government correctly points out that Loughrin interpreted the bank fraud statute, while this case involves the separate prohibition on mail fraud. But, for aught that appears, this is a distinction without a difference. To be sure, these two provisions are not identical. See id. at 2391 (holding that the bank fraud statute, unlike the mail fraud statute, may be violated in two distinct ways). The government, however, offers no explanation at all for why the same "by means of" language should be read differently in these two contexts. See Smith v. City of Jackson, 544 U.S. 228, 233, 125 S.Ct. 1536, 161 L.Ed.2d 410 (2005) ("[W]hen Congress uses the same language in two statutes having similar purposes,... it is appropriate to presume that Congress intended that text to have the same meaning in both statutes."). In fact, to the contrary, the very same federalism concerns underlying this "textual limitation" in the bank fraud statute are equally applicable to mail fraud. See Cleveland, 531 U.S. at 24, 121 S.Ct. 365 (noting resistance to reading which would effect "a sweeping expansion of federal criminal jurisdiction in the absence of a clear statement by Congress"). Indeed, the issuance of licenses and permits is "traditionally regulated by state and local authorities." Id. And medical licenses, much like the gaming licenses at issue in Cleveland, almost invariably are sought and obtained in an effort to realize some monetary profit. Accordingly, under the government's theory, virtually any false statement in an application for a medical license could constitute a federal crime. Such a broad reading of the statute would impermissibly infringe on the states' "distinctively sovereign authority to impose criminal penalties for violations of" licensing schemes, "including making false statements in a license application." Id. at 23, 121 S.Ct. 365. Just as in Loughrin, the phrase "by means of" serves as a textual limitation preventing such a usurpation of state criminal jurisdiction.
Our dissenting colleague disagrees, suggesting that Loughrin's reading of "by means of" in the context of the bank fraud statute should not inform our interpretation of the identical language in § 1341. But, as the dissent readily concedes, the bank fraud statute was expressly "modeled on" the pre-existing prohibition on mail fraud. S. Rep. No. 98-225, at 378 (1983), reprinted in 1984 U.S.C.C.A.N. 3182, 3519. Both provisions "proscribe the conduct of executing or attempting to execute `a scheme or artifice to defraud' or to take the property of another `by means of false or fraudulent pretenses, representations, or promises.'" Id. (emphasis added). Perhaps unsurprisingly in light of this legislative history, other circuits have consistently applied precedents construing § 1341 to the bank fraud statute. See, e.g., United States v. Saks, 964 F.2d 1514, 1520 (5th Cir. 1992) ("It is well settled that Congress modelled § 1344 on the mail and wire fraud statutes, and that the usual practice is to look to precedents under those statutes to determine its scope and proper interpretation."); United States v. Young, 952 F.2d 1252, 1255 (10th Cir. 1991) (noting that the two statutes contain "virtually the same language"); United States v. Mason, 902 F.2d 1434, 1441 (9th Cir. 1990) ("[T]he bank fraud statute directly tracks or is parallel to the mail and wire fraud statutes."), abrogated on other grounds by Dixon v. United States, 548 U.S. 1, 126 S.Ct. 2437, 165 L.Ed.2d 299 (2006).
The dissent rejects this longstanding consensus, reasoning that, while the mail fraud and bank fraud statutes employ "equivalent language," the lack of "contemporaneous drafting" undermines any presumption that Congress intended the phrase "by means of" to have a similar meaning in both contexts. But we have never imposed any requirement of "contemporaneous drafting" to give rise to a presumption of similar meaning where two statutes employ identical language and one is expressly modeled on the other. We have, for example, held that the wire fraud statute should be construed according to our mail fraud precedents. See United States v. Fermin Castillo, 829 F.2d 1194, 1198 (1st Cir. 1987). We reached this result despite recognizing that the mail fraud statute was significantly "older" than its wire fraud counterpart. Id. Indeed, the substance of the federal prohibition on mail fraud has been in place since 1909. See Act of Mar. 4, 1909, ch. 321, § 215, 35 Stat. 1088, 1130-31; see also Jed S. Rakoff, The Federal Mail Fraud Statute (Part I), 18 Duq. L. Rev. 771, 821 n.225 (1980) (characterizing post-1909 amendments as "chiefly designed to remove `surplus' language from the statute"). The wire fraud statute was not enacted until more than four decades later. See Communications Act Amendments, 1952, ch. 879, § 18(a), 66 Stat. 711, 722. In Fermin Castillo, rather than treating chronology as dispositive, we noted that the wire fraud statute "tracks" the language of § 1341. 829 F.2d at 1198. We also cited legislative history indicating that the former provision was "patterned on" the latter. Id. Each of these considerations applies equally to the bank fraud statute.5
The dissent next takes the position that the federalism concerns underlying Loughrin are not transferrable to the mail fraud context. We disagree. Of course, the mail fraud and bank fraud statutes are predicated on different jurisdictional bases, but that does not mean that the scope of the former provision is unlimited. Indeed, the Supreme Court has expressly recognized that § 1341 "does not purport to reach all frauds." Schmuck v. United States, 489 U.S. 705, 710, 109 S.Ct. 1443, 103 L.Ed.2d 734 (1989) (citation omitted). Rather, it targets "only those limited instances in which the use of the mails is a part of the execution of the fraud." Id. (citation omitted). The mail fraud statute also requires that the fraudulent scheme seek to obtain money or property. See Cleveland, 531 U.S. at 18, 121 S.Ct. 365; McNally, 483 U.S. at 360, 107 S.Ct. 2875.
Adoption of the dissent's preferred construction of "by means of" would work "a sweeping expansion of federal criminal jurisdiction." Loughrin, 134 S.Ct. at 2392 (quoting Cleveland, 531 U.S. at 24, 121 S.Ct. 365). The present appeals provide a case in point. The government's mail fraud allegations are entirely predicated upon the falsification of test scores. With these falsified scores in hand, and after completing certain other requirements, the defendants received medical licenses. The Board mailed letters indicating that the licenses were ready for pick-up. Under Cleveland, the defendants had not yet committed mail fraud. The government, however, contends that the mail fraud charges are salvaged by evidence that, in the ensuing years after becoming licensed, the defendants practiced medicine for profit. The government points to no additional instances of fraudulent conduct, instead falling back on the defendants' "use of [their] fraudulently obtained medical license[s]." Endorsing such a prosecution theory would extend the scope of federal jurisdiction to cover cases where the underlying fraudulent scheme, and the mailing in furtherance thereof, is far removed from any money or property. The Loughrin Court's citation to Cleveland in discussing these federalism concerns makes clear that they remain relevant in the mail fraud context.
The dissent relies in large part on a string of cases refusing to read a so-called "convergence" requirement into the mail fraud statute. But this is a distinct issue from the causal nexus required under Loughrin. Our decision in United States v. Christopher, 142 F.3d 46 (1st Cir. 1998), illustrates the point. In that case, the defendant argued that, in order to constitute wire fraud, the alleged scheme had to "deceive the same person whom it deprive[d] of money or property." Id. at 52-53. We rejected this reading of the statute in Christopher, and we impose no such requirement in these appeals. Rather, our reversal of the money or property mail fraud convictions is based on the lack of a sufficiently direct causal nexus to satisfy Loughrin. In Christopher, after disposing of the convergence argument, we proceeded to address the separate issue of causation. See id. at 54; see also United States v. Frost, 125 F.3d 346, 360 (6th Cir. 1997) ("[E]ven the cases which have held that convictions may rest upon the deceit of a person other than the ultimate victim contemplated that the deception was causally related to the scheme to obtain property from the victim."). In that case, we found "the causal connection between the deception and the loss of property" to be "obvious." Christopher, 142 F.3d at 54. This characterization was justified by the facts of the case at hand, which were very different from those at issue here. The defendant was convicted for making misrepresentations to insurance regulators in connection with his acquisition of two companies. More specifically, he assured regulators "that the collateral liens would be paid off by the time of closing, and that assets of the acquired companies would not be used for the purchase." Id. at 49. But, after closing, the acquired companies "loaned" money to satisfy the liens. Id. Similarly, cash belonging to one company was used to pay its purchaser. Id. The wire fraud charges were predicated on these two categories of payments. See id. at 50-51. Thus, the transfers through which the defendant realized the required monetary benefit directly contradicted his prior statements to regulators. In the parlance of Loughrin, the defendant's lies "naturally induc[ed]" the resulting benefit. 134 S.Ct. at 2393. As discussed above, the causal connection is much more attenuated in the present case. The other precedents cited by the dissent on this point, like Christopher, deal primarily with the convergence issue, not the required causal connection between the fraud and the obtainment of money or property. In any event, to the extent that these out-of-circuit decisions could be read to be inconsistent with Loughrin, we are bound to follow the standard imposed by the Supreme Court.6
Contrary to the dissent's suggestion, Loughrin's interpretation of "by means of" did not impose a convergence requirement like the one we rejected in Christopher. Indeed, the Court expressly recognized the possibility of bank fraud convictions in cases where the fraud never actually reaches a bank. It explained that "the clause covers property `owned by' the bank but in someone else's custody and control...; thus, a person violates § 1344(2)'s plain text by deceiving a non-bank custodian into giving up bank property that it holds." Loughrin, 134 S.Ct. at 2389. To be sure, the Loughrin Court ultimately held that, in the specific context of bank fraud, "by means of" requires a "false statement [that] will naturally reach [a federally insured] bank (or a custodian of the bank's property)." Id. at 2394 n.8. But, in contending that our application of Loughrin to § 1341 imposes a convergence requirement, the dissent overlooks its own warning that "what relationships count as close enough to satisfy the phrase `by means of' will depend almost entirely on context." Id. Generally speaking, "by means of" requires "an inquiry into the directness of the relationship between means and ends." Id. In its initial discussion of the phrase, the Loughrin Court relied upon dictionary definitions. After citing these generally applicable sources of meaning, the Court concluded that, in order to satisfy the bank fraud statute, the false statement must be "the mechanism naturally inducing a bank (or custodian of bank property) to part with money in its control." Id. at 2393. It is the specification that the fraud target bank property, not the widely applicable "naturally inducing" standard, that makes context relevant. The Court's rejection of the "Little Bobby" hypothetical posited by Justice Scalia illustrates the point. In the hypothetical, "Bobbly falsely tells his mother that he got an A on his weekly spelling test and so deserves an extra cookie after dinner." Id. at 2396 (Scalia, J., concurring). Bobby's mother, who will not be home for dinner, leaves a note for Bobby's father indicating that Bobby should get an extra cookie. According to Justice Scalia, when Bobby receives his cookie, he has done so "by means of the fib to his mother." Id. We agree that, in these circumstances, Bobby's false statement "naturally induc[ed]" his father to give him an extra cookie. And the Loughrin majority did not dispute the point. Rather, it responded by citing the importance of context and concluded that, in the bank fraud statute, "by means of" is "best read" to include only those frauds in which the false statement will reach a bank, either directly or indirectly. Id. at 2394 n.8. Of course, this specific application could not possibly apply to all uses of the phrase "by means of."
In light of the above analysis, there was insufficient evidence to support the defendants' convictions for money or property mail fraud. Because we find the government's theory legally deficient, we must reverse these convictions.
B. Honest-Services Mail Fraud Conspiracy
The government also charged the defendants with conspiracy to commit honest-services mail fraud. See 18 U.S.C. §§ 371, 1341, 1346. The four defendants convicted on these counts now appeal. Because we find sufficient evidence to support the convictions, we affirm.
At its core, a conspiracy is "an agreement between two or more persons to accomplish an unlawful purpose." United States v. Dellosantos, 649 F.3d 109, 115 (1st Cir. 2011). A conviction for general conspiracy, 18 U.S.C. § 371, "requires proof that the defendant agreed to commit an unlawful act and voluntarily participated in the conspiracy, and that an overt act was committed in furtherance of the conspiracy." United States v. McDonough, 727 F.3d 143, 156 (1st Cir. 2013). While we have described the presence of an agreement as the "sine qua non of a conspiracy," Dellosantos, 649 F.3d at 115, "[t]he agreement itself need not be express, but may consist of no more than a tacit understanding," United States v. Echeverri, 982 F.2d 675, 679 (1st Cir. 1993) (citation omitted). And to meet its burden, "[t]he government need not show that each conspirator knew of or had contact with all other members. Nor need it show that the conspirators knew all of the details of the conspiracy or participated in every act in furtherance of the conspiracy." United States v. Soto-Beníquez, 356 F.3d 1, 19 (1st Cir. 2003).
Here, the government charged a conspiracy to commit honest-services mail fraud, a specific type of mail fraud involving a scheme "to deprive another of the intangible right of honest services." 18 U.S.C. § 1346. The Supreme Court has held that this statute only criminalizes schemes involving bribes or kickbacks. Skilling v. United States, 561 U.S. 358, 412, 130 S.Ct. 2896, 177 L.Ed.2d 619 (2010); see also United States v. Urciuoli, 613 F.3d 11, 17 (1st Cir. 2010) ("[T]hose who bribe public officials take part in a scheme to deprive the public of the honest services of those they attempt to influence."). In the case at hand, the government specifically alleged that the defendants conspired to deprive the Board of Rodríguez's honest services.
The defendants' sufficiency of the evidence challenges to the honest-services fraud convictions need not detain us long. The trial evidence was sufficient for the jury to find the following facts. Julio Castro, Pacheco, and Davila all failed at least one of the required admissions exams. They later knowingly provided something of value to Rodríguez (through an intermediary) in exchange for falsified passing scores. Finally, use of the mails (e.g., mailing of letters indicating that the licenses were ready for pick-up) was a foreseeable element of the scheme. See, e.g., United States v. Pimental, 380 F.3d 575, 589 (1st Cir. 2004) (holding that a defendant need not personally mail anything, so long as the "`use of the mails' in the course of the scheme" is "reasonably foreseeable" (collecting cases)).7
Pacheco argues that the alleged overt act, namely, the act of physically receiving medical licenses from the Board, fell outside the scope of the conspiracy. This argument fails because "[a] conspiracy endures as long as the co-conspirators endeavor to attain the `central criminal purposes' of the conspiracy." United States v. Upton, 559 F.3d 3, 10 (1st Cir. 2009) (quoting Grunewald v. United States, 353 U.S. 391, 401, 77 S.Ct. 963, 1 L.Ed.2d 931 (1957)). The government maintained throughout trial that the conspiratorial object of each defendant was fraudulently to gain a medical license, not merely to obtain a passing exam score. This is hardly a stretch, as a passing score on the threshold tests alone has only incremental value. We decline to overturn the jury's finding that Pacheco's receipt of her medical license was an act within the scope and timeframe of the conspiracy. See id. at 11 ("Determining the contours of the conspiracy ordinarily is a factual matter entrusted largely to the jury.").
C. Aggravated Identity Theft
Berroa and Geraldo Castro also challenge their convictions for aggravated identity theft. In order to meet its burden on this charge, the government was required to show that each defendant "knowingly transfer[red], possesse[d], or use[d], without lawful authority, a means of identification of another person" and did so "in relation" to one or more specified crimes, including mail fraud. 18 U.S.C. § 1028A(a)(1), (c). The term "means of identification" is defined broadly to include names. Id. § 1028(d)(7). Where the necessary showing is made, "a term of imprisonment of 2 years" is added to the punishment for the underlying offense. Id. § 1028A(a)(1). In the present case, because we find insufficient evidence that the defendants "used" a means of identification within the meaning of the statute, we reverse the identity theft convictions.
During the course of their medical practices (utilizing fraudulently obtained licenses), both defendants issued prescriptions that their patients would then fill at various pharmacies in Puerto Rico. The government alleges that the use of patient names and addresses on the prescriptions constituted use without lawful authority of the identification of another person.
In support of this argument, the government focuses on the absence of "lawful authority." We have held that this statutory element does not "require that the means of identification be stolen, or otherwise taken without permission of the owner." United States v. Ozuna-Cabrera, 663 F.3d 496, 498 (1st Cir. 2011). In reaching this result, we drew a distinction between the terms "authorized" and "lawful" to conclude that "regardless of how the means of identification is actually obtained, if its subsequent use breaks the law ... it is violative of § 1028A(a)(1)." Id. at 499. Here, the government reasons, the patients' consent to the inclusion of their names in the prescriptions is not dispositive. Such use was rendered unlawful by the fact that the defendants' medical licenses were obtained by fraud. Moreover, the patients could not provide knowing consent because they were unaware of the underlying fraud. Ultimately, we need not resolve these issues. The government's proof was insufficient to satisfy the separate requirement that the defendants knowingly "used" the patient information.
The Supreme Court has recognized that the word "use" is fraught with "interpretational difficulties because of the different meanings attributable to it." Bailey v. United States, 516 U.S. 137, 143, 116 S.Ct. 501, 133 L.Ed.2d 472 (1995), superseded by statute on other grounds, 18 U.S.C. § 924(c)(1). "Use" has been "variously defined as `[t]o convert to one's service,' `to employ,' `to avail oneself of,' and `to carry out a purpose or action by means of.'" Id. at 145, 116 S.Ct. 501 (alteration in original). The statute at issue here fails to provide a specific definition. And, in context, "use" cannot be given its broadest possible meaning, which would subsume the separate statutory terms "transfer" and "possess." See United States v. Miller, 734 F.3d 530, 541 (6th Cir. 2013). Accordingly, we find the statutory language ambiguous. The Sixth Circuit has reached the same conclusion. See id. at 540-41.
We turn next to legislative history. The relevant House Report makes clear that the legislation was intended to address "the growing problem of identity theft." See H.R. Rep. No. 108-528, at 3, reprinted in 2004 U.S.C.C.A.N. 779. The report goes on to provide several examples of identity theft. Notably, each of these examples involved the defendant's use of personal information to pass him or herself off as another person, or the transfer of such information to a third party for use in a similar manner. See id. at 5-6, 2004 U.S.C.C.A.N. at 781-82 (e.g., submission of "bogus Federal income tax returns" in others' names; use of "stolen identity to apply for and receive Social Security benefits" and "establish credit"). The facts of Ozuna-Cabrera, where the defendant presented another person's expired passport in an attempt to obtain a valid passport under that person's name, falls comfortably within this understanding of identity theft. See 663 F.3d at 497. By contrast, the purported "use" of patient information alleged here strays far afield from the conduct targeted by Congress. While, in a colloquial sense, Berroa and Castro could be said to have "used" their patients' names in writing prescriptions, they certainly did not attempt to pass themselves off as the patients.
The government's reading of the statute is virtually unlimited in scope. Indeed, if, as the government implies, "use" of a "means of identification" is to be given its broadest possible meaning, it could encompass every instance of specified criminal misconduct in which the defendant speaks or writes a third party's name. See United States v. Spears, 729 F.3d 753, 756 (7th Cir. 2013) (warning, in interpreting the statutory language "another person," that the government's reading would "require a mandatory two-year consecutive sentence every time a tax-return preparer claims an improper deduction"). We will not lightly presume that Congress intended this extreme result.
In light of § 1028A's legislative history, as well as the limitless nature of the government's alternative construction, we read the term "use" to require that the defendant attempt to pass him or herself off as another person or purport to take some other action on another person's behalf.8 Our holding on this point is consistent with that of the only other circuit to have addressed the issue. See Miller, 734 F.3d at 541 (reversing convictions where the defendant "did not steal or possess [others'] identities, impersonate them or pass himself off as one of them, act on their behalf, or obtain anything of value in one of their names" (footnote omitted)); United States v. Medlock, 792 F.3d 700, 705 (6th Cir. 2015) (rejecting the government's argument that the defendants "`used' the name and Medicare Identification Numbers of Medicare beneficiaries" when they submitted claims containing false statements).
III. Other Claims of Error
Pacheco raises two challenges to her indictment: (1) that the district court erred by refusing to strike surplusage; and (2) that the court erred by amending the indictment.
Prior to trial, Pacheco moved, pursuant to Fed. R. Crim. P. 7(d), to strike surplusage from her indictment. Rule 7(d) serves to "protect the defendant `against immaterial or irrelevant allegations in an indictment, ... which may ... be prejudicial.'" United States v. Lewis, 40 F.3d 1325, 1346 (1st Cir. 1994) (alteration in original) (quoting Fed. R. Crim. P. 7(d), advisory committee note). Pacheco sought to strike any language related to her prior failures of Board exams and information about where and when she graduated from medical school.
The contested language was "neither irrelevant nor unfairly prejudicial." Id. The prosecution had the burden of showing that Pacheco acted knowingly and intentionally in order to secure a conviction for conspiracy to commit honest-services mail fraud. And to that end, Pacheco's prior exam failures and graduation date were relevant to her motive. See United States v. Pires, 642 F.3d 1, 11 (1st Cir. 2011) ("[E]vidence that bears on the question of motive ordinarily has some probative value in a criminal case."). Pacheco's difficulty in gaining a medical license on her own merit was appropriate information for the jury to consider in finding that she turned to illicit means. Accordingly, any prejudice resulting from the inclusion of this information in the indictment was not unfair.9
Pacheco next claims that the district court improperly allowed an amendment to her indictment. At issue is the date that the government alleged that Pacheco committed an overt act, namely, receipt of her regular medical license. The indictment initially identified the overt act as occurring on March 18, 2005, the date that Pacheco received her acupuncture license. The correct date should have been January 13, 2004. Over Pacheco's objection, the district court amended the date on the indictment. We review her preserved challenge de novo. See United States v. Hernández, 490 F.3d 81, 83 (1st Cir. 2007).
As an initial matter, the error that Pacheco complains of is not a constructive amendment, but rather a direct amendment. See United States v. Dowdell, 595 F.3d 50, 66-67 (1st Cir. 2010). While an amendment that broadens the charges in an indictment runs afoul of a defendant's rights and is, accordingly, prohibited, "this prohibition does not extend to alterations that are `merely a matter of form.'" Id. at 67 (quoting Russell v. United States, 369 U.S. 749, 770, 82 S.Ct. 1038, 8 L.Ed.2d 240 (1962)). Pursuant to this exception, courts may "allow ministerial corrections of clerical errors in names, dates, and citations, so long as the change [will] not deprive the defendant of notice of the charges against him." Id. at 68.
Here, there can be no argument that Pacheco was deprived of such notice. When identifying the overt act necessary for the conspiracy charge, the indictment explicitly stated that Pacheco "received a regular license to practice medicine in Puerto Rico." Further, the evidence and theory throughout trial consistently maintained that the overt act was receipt of a regular medical license, not a specialized acupuncture license. In discovery, the government turned over Pacheco's entire Board file, including her medical license dated January 13, 2004. In light of these facts, the amendment was a proper clerical change.
B. Other Evidentiary Claims
As part of its case, the government introduced evidence about non-defendant applicants who obtained false passing scores identical to the defendants'. On appeal, Pacheco and Davila challenge the admissibility of this evidence. Our review is for abuse of discretion. See United States v. Anthony, 545 F.3d 60, 66 (1st Cir. 2008).
The fact that the defendants obtained exactly the same scores as other applicants bore directly on the existence of a scheme to defraud. It also served to connect Rodríguez to those falsified scores that she did not specifically remember creating. Moreover, the evidence presented was less prejudicial than that in other cases where we have found no abuse of discretion. See, e.g., United States v. McGauley, 279 F.3d 62, 72 (1st Cir. 2002) (upholding admission of 217 fraudulently obtained refund checks not charged in indictment). Given the probative value of the evidence, the defendants fall well short of establishing "extraordinarily compelling circumstances" such that we will, "from the vista of a cold appellate record, reverse a district court's on-the-spot judgment concerning the relative weighing of probative value and unfair effect." United States v. Williams, 717 F.3d 35, 41 (1st Cir. 2013) (citation omitted).10
C. Alleged Judicial Bias
The defendants next contend that the district court exhibited bias in favor of the prosecution. "Under the usual framework for judicial bias claims, a party must ... show (1) that the [judge's] comments were improper and (2) that there was serious prejudice." United States v. Lanza-Vázquez, 799 F.3d 134, 143 (1st Cir. 2015) (second alteration in original) (citation omitted). In controlling the courtroom, "[i]t is well-established that a judge `is not a mere moderator, but is the governor of the trial for the purpose of assuring its proper conduct.'" Id. (quoting Quercia v. United States, 289 U.S. 466, 469, 53 S.Ct. 698, 77 S.Ct. 1321 (1933)). Accordingly, "`remarks during the course of trial that are critical or disapproving of, or even hostile to, counsel, the parties, or their cases' are usually insufficient to prove bias." United States v. Caramadre, 807 F.3d 359, 375 (1st Cir. 2015) (quoting Liteky v. United States, 510 U.S. 540, 555, 114 S.Ct. 1147, 127 L.Ed.2d 474 (1994)).
Here, Davila points to a number of instances during the cross-examination of three government witnesses where the judge sustained objections and commented in open court on the nature of the defense questions.11 Berroa separately takes issue with the court's admonishing his counsel not to interrupt the prosecutor's closing argument and its indication that counsel was "misstating" the law. We conclude that, at most, the district court indicated moderate displeasure with defense counsel. This conduct was more benign than that found insufficient to warrant reversal in previous cases. See, e.g., Lanza-Vázquez, 799 F.3d at 143-45. The statements identified by the defendants represent "the few instances of the trial court's frustration, cherry-picked from the voluminous record" and, as a result, "do not reveal judicial bias, let alone the serious prejudice" required for reversal. United States v. Ofray-Campos, 534 F.3d 1, 34 (1st Cir. 2008).
D. Defense Witness Immunity
Berroa, Davila, and Pacheco assert that the district court erred by refusing to grant use immunity to former Board employee Dr. Rafael Jiménez-Méndez. The defendants claim that Jiménez's testimony was necessary to contradict the government's case and attack the credibility of the government's witnesses.
The power and discretion to immunize witnesses lies primarily with the prosecution. Trial courts have the ability to grant immunity to a witness upon a showing that the government's refusal to provide said immunity violated the defendant's due process rights. See United States v. Angiulo, 897 F.2d 1169, 1190 (1st Cir. 1990). In order to prevail on such a claim, a defendant must establish that the government intentionally attempted to present a distorted version of the facts at trial. See Curtis v. Duval, 124 F.3d 1, 9 (1st Cir. 1997). "This type of deliberate distortion can occur in two ways: if the government attempts to intimidate or harass a potential witness, or if the prosecutor purposefully withholds use immunity to hide exculpatory evidence from the jury." United States v. Castro, 129 F.3d 226, 232 (1st Cir. 1997).
Jiménez entered into a pre-trial diversion agreement containing a promise by the government not to bring any federal charges against him in relation to the scandal at the Board. Upon being called to testify, Jiménez invoked his Fifth Amendment right not to incriminate himself. See U.S. Const. amend. V. The defense objected, citing the pre-trial diversion agreement. Rejecting this argument, the district court found that Jiménez was only protected from federal charges and, thus, there was a cognizable risk that his testimony could be used against him in a Puerto Rico court. See United States v. Jiménez-Bencevi, 788 F.3d 7, 15 (1st Cir. 2015) ("Informal immunity agreements ... are shaped ... by the language of the contract conferring the immunity." (citation omitted)).
In her brief, Pacheco points to Jiménez's presence on a witness list for an upcoming trial as proof that the government planned to offer immunity to him. When asked at a status conference in the subsequent case about potential trial witnesses, the prosecutor responded that Jiménez could potentially testify but no final decision had been made at that time. While this response may have been ambiguous, it is not the type of "affirmative government misconduct" required to show a due process violation and compel the government to grant immunity. United States v. Mackey, 117 F.3d 24, 27 (1st Cir. 1997).12
Moreover, the purported value of Jiménez's testimony was minimal at best and cumulative of other witnesses. The bulk of Jiménez's purported value was to respond to Rodríguez's testimony that she was unaware of any score-fixing schemes at the Board other than her own. Pacheco avers that Jiménez would testify otherwise and would offer further evidence about misconduct and sloppy record-keeping at the Board. Assuming that such information would have been admissible, Pacheco had other witnesses available (as noted by the district court) who could offer similar testimony. This fact further undercuts any assertion that the government sought to "withhold use immunity to hide exculpatory evidence from the jury." Castro, 129 F.3d at 232.13
E. Jury Instructions
Pacheco and Davila argue that the district court erred by not including certain jury instructions in its charge for conspiracy to commit honest-services mail fraud. Both defendants assert that they were entitled to a gifts instruction, and Davila argues that the district court also should have issued a good faith instruction.
We assess a district court's refusal to give a requested instruction in two steps. First, we look de novo to see whether the evidence, taken in a light most favorable to the defendant, supports the instruction. United States v. Baird, 712 F.3d 623, 627 (1st Cir. 2013). If the defendant makes this showing, we proceed to the second step where "refusal to give a requested instruction is reversible error only if the instruction was (1) substantively correct; (2) not substantially covered elsewhere in the charge; and (3) concerned a sufficiently important point that the failure to give it seriously impaired the defendant's ability to present his or her defense." United States v. Callipari, 368 F.3d 22, 32 (1st Cir. 2004) (citation omitted), vacated on other grounds, 543 U.S. 1098, 125 S.Ct. 985, 160 L.Ed.2d 998 (2005).
First, relying upon our decision in United States v. Sawyer, 85 F.3d 713, 741 (1st Cir. 1996), the defendants argue that the jury was required to consider whether the payments to Rodríguez were gifts (and not bribes). We need not move beyond the first stage of our inquiry because the facts — even in a light most favorable to the defendants — fail to sufficiently support a gifts instruction. In Sawyer, we advised that a cautionary instruction is needed "where ... the line between the merely unattractive and actually criminal conduct is blurred." Id. Such blurring occurs when the conduct in question (e.g., "payments for entertainment, lodging, golf, sports events, and the like") could colorably be an attempt to "cultivate a business or political `friendship.'" Id. Here, there was no evidence to suggest that the payments by the defendants were designed to cultivate such a friendship. Rather, they were made in exchange for fraudulent scores.
With respect to the second requested instruction, "we have held that [a] separate instruction on good faith is not required ... where the court adequately instructs on intent to defraud." Christopher, 142 F.3d at 55 (first alteration in original) (citation omitted). This is because intent to defraud is "essentially the opposite of good faith." United States v. Dockray, 943 F.2d 152, 155 (1st Cir. 1991). Here, a review of the transcript makes clear that the district court provided adequate instructions. The court told the jury that "[t]o establish specific intent, the Government must prove that the defendant knowingly did an act which the law forbids, purposely intending to violate the law." Accordingly, no separate good faith instruction was required.14
F. Purported Prosecutorial Misconduct
Davila identifies several instances of purported prosecutorial misconduct during opening, closing, and rebuttal arguments. Her admittedly unpreserved claims that the prosecutor misstated certain evidence fail because, even assuming that the relevant statements were improper, they did not "so poison the well" as to require a new trial. See United States v. Henderson, 320 F.3d 92, 107 (1st Cir. 2003) (citation omitted). Davila's preserved argument that the prosecutor improperly shifted the burden of proof, however, merits further discussion.
We review the propriety of the prosecutor's comments de novo. United States v. Glover, 558 F.3d 71, 76 (1st Cir. 2009). If we find that the statements were improper, we must determine whether the error was harmless. Id. With respect to the first step in the analysis, "a prosecutor may cross the line [into impermissibility] by arguing to the jury that the defendant is obligated to present evidence of his innocence." Id. at 77 (alternation in original) (citation omitted). With that said, the government is afforded some latitude "to discuss competing inferences from the evidence on the record" and "to comment on the plausibility of the defendant's theory." Id. at 77-78. In doing so, prosecutors must focus "on the evidence itself and what the evidence shows or does not show, rather than on the defendant and what he or she has shown or failed to show." Id. at 78.
Here, Davila takes issue with excerpts of the prosecutor's rebuttal argument referring to defense counsel's failure to provide "innocent explanations" for or otherwise address (1) evidence that the defendants' false test results were identical to those of other applicants and (2) Davila's request for a certification of her score. At trial, the defendants sought to use evidence of Board irregularities to suggest that they might not have been complicit in the falsification of their own test scores. Accordingly, the prosecutor was entitled to respond by contesting the plausibility of that defense theory.
It is at least arguable, however, that the prosecutor exceeded the bounds of propriety by specifically referring to defense counsel's failure to address aspects of the prosecution's case. See United States v. Wihbey, 75 F.3d 761, 770 (1st Cir. 1996) (referring to "how-does-counsel-explain" argument as "an impermissible shift of [the] burden of proof" (citation omitted)). Assuming (without deciding) that the specific invocation of defense counsel's omissions was improper, we turn to the issue of prejudice.
Reversal is warranted only if the possibly improper statements were "harmful." Glover, 558 F.3d at 78. In making this assessment, "we consider the totality of the circumstances, including the severity of the misconduct, the prosecutor's purpose in making the statement ..., the weight of the evidence supporting the verdict, jury instructions, and curative instructions." Id. (citation omitted). Here, Davila points to three isolated passages from the government's rebuttal argument. Moreover, the prosecutor appears to have acted with the permissible goal of rebutting the defense theory, rather than intentionally shifting the burden to the defendants. To cinch the matter, the court clearly instructed the jury:
[I]t is a cardinal principle of our system of justice that every person accused of a crime is presumed to be innocent unless and until his or her guilt is established beyond a reasonable doubt. The presumption is not a mere formality, it is a matter of the most important substance. The Defendants do not have to prove their innocence at any time or to produce any evidence.... The burden of proof never shifts to a defendant. It is always the Government's burden to prove each of the elements of the crimes charged beyond a reasonable doubt....
In light of the significant evidence against the defendants, the context of the prosecutor's statements, and the court's strong and unequivocal instruction on the presumption of innocence, any impropriety in the prosecutor's rebuttal argument was harmless.
Finally, Pacheco challenges the district court's application of a twelve-level sentencing enhancement pursuant to United States Sentencing Guidelines § 2C1.1(b)(2)(A) and § 2B1.1(b)(1)(G).15 We review the district court's fact-finding for clear error and its interpretation of the Guidelines de novo. United States v. Souza, 749 F.3d 74, 85 (1st Cir. 2014). Relying on the pre-sentence investigation report, the court calculated the benefit Pacheco received in return for her bribe to be more than $200,000 but less than $400,000. Pacheco now seeks to vacate her sentence, arguing that the phrase "benefit ... to be received in return for the payment" in § 2C1.1(b)(2)(A) does not include her medical license or earnings realized therefrom.
Contrary to Pacheco's contention, the sentencing court may apply the § 2C1.1(b)(2)(A) enhancement so long as the defendant actually received or expected to receive the requisite benefit. See, e.g., United States v. Vázquez-Botet, 532 F.3d 37, 67 (1st Cir. 2008). Here, the district court relied on the actual benefit Pacheco received from practicing under the aegis of her fraudulent license. To the extent the court also considered expected benefit, we see no error (let alone clear error) in its finding that Pacheco intended to receive a financial benefit from obtaining her medical license.16
For the foregoing reasons, we affirm the convictions and sentences of Pacheco and Davila (both of whom were convicted only of honest-services mail fraud conspiracy). By contrast, we reverse the convictions and vacate the sentence of Geraldo Castro (who was convicted only of mail fraud and aggravated identity theft).
With respect to the remaining two defendants, Berroa and Julio Castro, we reverse their convictions for money or property mail fraud and affirm their convictions for honest-services mail fraud conspiracy. Finally, we reverse Berroa's convictions for aggravated identity theft. Under our caselaw, the partial reversal of Berroa's and Julio Castro's convictions may "alter the dimensions of the sentencing `package.'" United States v. Genao-Sánchez, 525 F.3d 67, 71 (1st Cir. 2008). We therefore vacate the sentences of these defendants and remand for resentencing. We take no view as to whether the new sentences should be equal to or less than the sentences previously imposed. That is a matter committed, in the first instance, to the sound discretion of the sentencing court.
LIPEZ, Circuit Judge, concurring in part and dissenting in part.
I join the majority's thoughtful opinion, with the exception of its novel interpretation of the mail fraud statute. On that issue I respectfully dissent.
Three terms ago, in Loughrin v. United States, the Supreme Court interpreted the phrase "by means of" in the bank fraud statute, 18 U.S.C. § 1344. ___ U.S. ___, 134 S.Ct. 2384, 189 L.Ed.2d 411 (2014). Bank property, the Court held, is obtained "by means of" a false or fraudulent statement only when that statement "is the mechanism naturally inducing a bank (or custodian of bank property) to part with money in its control." Id. at 2393. With federalism concerns in mind, the Loughrin Court sought to prevent the bank fraud statute from being used to prosecute conduct beyond the scope of the specific problem addressed by that statute's enactment in 1984. Citing those federalism concerns, the majority applies the "naturally inducing" limitation to the mail fraud statute, 18 U.S.C. § 1341, a measure originally enacted in 1872 to address a much broader problem. We are the first court of appeals to extend Loughrin in this manner. Because this ruling ignores the Supreme Court's own cautionary language in Loughrin, as well as differences in the text and purpose of the two statutes, I would not incorporate the Loughrin limitation into the mail fraud statute.
The majority bolsters its reliance on federalism concerns by citing Cleveland v. United States, 531 U.S. 12, 121 S.Ct. 365, 148 L.Ed.2d 221 (2000), a case interpreting the term "property" in the mail fraud statute. Although Cleveland addresses the same statute at issue here, the concerns raised by the government's use of the mail fraud statute in that case are inapplicable to this case. Consequently, I find the majority's reliance on that precedent inapt.
The majority insists that the Supreme Court's limiting interpretation of the bank fraud statute must be applied to the mail fraud statute because the words interpreted by the Court in Loughrin — "by means of" — also appear in the mail fraud statute. In so concluding, the majority relies heavily on a presumption that the same federalism concerns which drove the Loughrin Court to interpret that phrase narrowly in the bank fraud statute also apply to the mail fraud statute. To reach these conclusions, however, one must ignore both significant differences between the two statutes and critical parts of the Loughrin opinion.
To be sure, the threshold argument that the same words should be construed the same way is intuitively appealing. Indeed, the Supreme Court has noted that such a presumption can be a fair starting point for a statutory analysis "when Congress uses the same language in two statutes having similar purposes, particularly when one is enacted shortly after the other." Smith v. City of Jackson, 544 U.S. 228, 233, 125 S.Ct. 1536, 161 L.Ed.2d 410 (2005). Importantly, however, the equivalent language here is not the product of contemporaneous drafting. The bank fraud statute was enacted more than a century after the original version of the mail fraud statute, and more than thirty years after the "by means of" language was added to the mail fraud statute. See Jed Rakoff, The Federal Mail Fraud Statute, 18 Duq. L. Rev. 771, 772 (1980) (original mail fraud statute enacted in 1872); 62 Stat. 763 (1948) (modern mail fraud statute); Pub. L. No. 98-473, Title II, § 1108(a), 98 Stat. 2147 (1984) (bank fraud statute). While not dispositive, this substantial gap in time raises a threshold doubt as to the applicability of this presumption. See Smith, 544 U.S. at 233, 125 S.Ct. 1536.
Here, however, we have more than mere doubts about the applicability of the presumption. The Supreme Court in Loughrin explicitly instructed that applying such a presumption to the words at issue would be inappropriate:
[W]hat relationships count as close enough to satisfy the phrase "by means of" will depend almost entirely on context.... Language like "by means of" is inherently elastic: It does not mean one thing as to all fact patterns — and certainly not in all statutes, given differences in context and purpose.
134 S.Ct. at 2394 n.8. Hence, in determining whether the "naturally inducing" formulation articulated in Loughrin applies to the mail fraud statute, we must compare the contexts, purposes, and language of the two statutes.
A. The Bank Fraud Statute
1. Context and Scope
The bank fraud statute was enacted in 1984 "to provide an effective vehicle for the prosecution of frauds in which the victims are financial institutions that are federally created, controlled or insured." S. Rep. No. 98-225, at 377 (1983). According to the Senate Report, the need for such a law became apparent when the Supreme Court vacated the mail fraud conviction of a defendant who used a stolen credit card to purchase food and lodging. See id. (citing United States v. Maze, 414 U.S. 395, 396, 398, 94 S.Ct. 645, 38 L.Ed.2d 603 (1974)). To satisfy the mailing element, the government in Maze relied on the post-purchase mailing of sales slips from the merchants to the bank that had issued the credit card. 414 U.S. at 397, 94 S.Ct. 645. The Court held that those mailings were not related closely enough to the fraudulent scheme to sustain a conviction for mail fraud. Id. at 405, 94 S.Ct. 645. Because the federal mail and wire fraud statutes require the use of either the mail or an interstate wire, federal prosecutors were unable to pursue charges when one of those forms of communication was not used "for the purpose of" committing a fraud against a bank. See id. at 404, 94 S.Ct. 645. In the wake of the Maze holding, Congress recognized that a "serious gap now exist[s] in federal jurisdiction over frauds against banks." S. Rep. No. 98-225, at 377. Accordingly, it enacted a new statute to fill this gap.
Rather than focusing on mail or interstate wire communications, the bank fraud statute bases its jurisdiction "on the fact that the victim of the offense is a federally controlled or insured [banking] institution."17 Id. at 378. Therefore, while the text of the bank fraud statute is "modeled on the present wire and mail fraud statutes," id., it differs in a key respect. Unlike those statutes, it specifies a victim that the fraud must harm, a bank. Rather than requiring only a "scheme or artifice to defraud," 18 U.S.C. § 1341, the first prong of the statute requires a "scheme or artifice... to defraud a financial institution," id. § 1344(1). Likewise, in the second prong of the statute, the scheme is not simply "for obtaining money or property," id. § 1341, but "to obtain any of the money... or other property owned by, or under the custody or control of, a financial institution," id. § 1344(2).
In multiple respects, therefore, the bank fraud statute was written narrowly, reflecting its limited purpose — protecting banks.
2. The Loughrin Decision
Loughrin involved "a scheme to convert altered or forged checks into cash." 134 S.Ct. at 2387. The defendant went door-to-door in Salt Lake City, pretending to be a Mormon missionary, and stole checks from residents' mailboxes. Id. He altered those checks or, if he happened to find blank checks, completed them, and used them to purchase items at a Target store. Id. He then immediately returned the purchased items for cash. Id. Loughrin challenged his bank fraud conviction for this scheme, arguing that the provision of the statute under which he was convicted, § 1344(2), requires a specific intent to deceive a bank, which he did not possess. Id. at 2389. He posited that such an element must be read into § 1344(2) to prevent it from being applied to mundane frauds that did not target a bank, but happened to involve a check. Id. at 2392.
Acknowledging the legitimacy of Loughrin's point about federalizing ordinary frauds, the Court gave the example of a common fraudster who "passes off a cheap knock-off as a Louis Vuitton handbag." Id. According to Loughrin's theory, if the victim who purchases the handbag happens to pay with a check rather than with cash, this fortuity would make the crime bank fraud because the fraudster has made "false or fraudulent ... representations" and has carried out a scheme to obtain money "`under the custody or control of' [a] bank (the money in the victim's checking account)." Id. (quoting 18 U.S.C. § 1344(2)). That approach would turn the statute into "a plenary ban on fraud, contingent only on the use of a check (rather than cash)." Id. Aware of the ubiquity of checks in our financial system, the Court was wary of interpreting the statute in a manner that would "approve a sweeping expansion of federal criminal jurisdiction in the absence of a clear statement by Congress," id. (quoting Cleveland, 531 U.S. at 24, 121 S.Ct. 365), by "federalizing frauds that are only tangentially related to the banking system," id. at 2393 (quotation omitted).
The Court disagreed with Loughrin, however, that to foreclose such an application of the bank fraud statute it had to read a specific intent requirement into § 1344(2).18 Instead, the Court explained, Loughrin's argument "fail[ed] to take account of a significant textual limitation on § 1344(2)'s reach." Id. at 2393. Under that provision, "it is not enough that a fraudster scheme to obtain money from a bank and that he make a false statement." Id. The clause also includes "a relational component," id., a requirement that the criminal obtain the bank property "by means of" the false statement, id. (quoting 18 U.S.C. § 1344(2)). That language "typically indicates that the given result (the `end') is achieved, at least in part, through the specified action, instrument, or method (the `means'), such that the connection between the two is something more than oblique, indirect, and incidental." Id.
The Court explained that the "relational component" is satisfied when "the defendant's false statement is the mechanism naturally inducing a bank (or custodian of bank property) to part with money in its control."19 Id. Accordingly, the fraud must reach the bank or the custodian of the bank's property, either directly or indirectly. Id. The standard is most clearly met when the misrepresentation or false statement is made to the bank itself, for example when a fraudster attempts to cash a forged check at a teller's window. Id. But a counterfeit check can also be the "means" by which a fraud is accomplished when it is presented to a third party, as Loughrin presented his forged checks to Target. Id. Predictably, the merchant will forward the check to the bank in an attempt to receive reimbursement in the form of bank funds. Id. In either scenario, the check itself is a fraudulent representation, and it "naturally induc[es]" the bank to part with money.
What the "naturally inducing" standard precludes, the Court explained, is the prosecution of cases like that of the handbag fraudster. Id. 2394. In that scenario, the fraudster's misrepresentation never reaches the bank, either directly or indirectly. Id. Unlike in Loughrin's scheme, the check itself does not convey the misrepresentation because it is a "perfectly valid" check. Id. Nor does the purchaser pass the misrepresentation about the quality of the handbag on to the bank or any custodian of the bank's property. Id. The bank fraud statute, the Court explained, "draw[s] a line at frauds that have some real connection to a federally insured bank — namely, frauds in which a false statement will naturally reach such a bank (or a custodian of the bank's property)." Id. at 2394 n.8. Because the handbag fraudster's misrepresentation never reaches the bank (or a custodian), the bank fraud statute does not encompass that kind of fraud.
The Loughrin Court thus viewed the "by means of" language through the lens of its federalism concerns about the scope of the bank fraud statute. Construing those words to require that the misrepresentation or false statement reach the bank or its custodian, either directly or indirectly, ensures that the "deceptions ... have some real connection to a federally insured bank, and thus implicate the pertinent federal interest." Id. at 2394-95.
B. The Mail Fraud Statute
With respect to every concern prompting the Loughrin Court's interpretation of the bank fraud statute, the mail fraud statute is distinguishable. The text of the statute, its purpose, and the federalism concerns implicated by its application are all manifestly different. Consequently, the majority has done exactly what the Loughrin Court cautioned against — transferred the Court's construction of the phrase "by means of" to an inapt context.
1. The Text
At the heart of the majority's argument is the contention that the same language should be construed the same way in related statutes. They emphasize that "other circuits have consistently applied precedents construing [the mail fraud statute] to the bank fraud statute." This assertion of consistency, however, is flawed in two respects. First, the majority's contention that courts treat the mail and bank fraud statutes uniformly is belied by the Loughrin decision itself. Second, even if it were true that mail fraud precedents generally have been applied in bank fraud cases, it does not follow that every bank fraud precedent should apply in the mail fraud context.
Loughrin attempted to support his argument for reading the specific intent requirement of § 1344(1) of the bank fraud statute into § 1344(2) of the statute by citing McNally v. United States, 483 U.S. 350, 107 S.Ct. 2875, 97 L.Ed.2d 292 (1987), a case that analyzed the mail fraud statute.20 Id. at 2390-91. As noted above, the mail fraud statute criminalizes schemes "to defraud, or for obtaining money or property." 18 U.S.C. § 1341. McNally held that Congress did not use the word "or" to convey a disjunctive meaning. 483 U.S. at 358-59, 107 S.Ct. 2875. Rather than describing a separate type of scheme, the Court held, the phrase "or for obtaining money or property" simply clarifies that the criminalized scheme "to defraud" must involve money or property. Id.
Loughrin argued that, accordingly, the comparable provision in the bank fraud statute ("to obtain any of the money ... or other property" of a bank) may not be read as a separate prong, but must be understood as a clarification of the statute's first provision, which prohibits schemes "to defraud a financial institution." 134 S.Ct. at 2390-92. Hence, like the first provision, it would include a specific intent requirement. The Supreme Court rejected this argument and held that "or" in the bank fraud statute, unlike "or" in the mail fraud statute, is disjunctive. Id. The Court thus demonstrated, in the very opinion relied upon by the majority for much of its argument, that a construction of one of the statutes does not always apply to the other.
The Loughrin Court, of course, based its decision in part on differences in language and structure between the two statutes. The Court pointed out that, unlike the mail fraud provision, the clauses before and after "or" in the bank fraud statute were arranged so as to "indicat[e] that they have separate meanings." Id. at 2391. Here, the majority argues that the "by means of" language is identical in both statutes, and, hence, the phrase should be construed consistently. Yet, in this instance, too, a significant difference between the two provisions counsels against a common interpretation.
The text of the mail fraud statute does not include any reference to the person or entity harmed by the alleged fraud; the provision criminalizes frauds that use the mails without regard to the victim. Although the bank fraud statute was "modeled on" the mail fraud statute, see S. Rep. No. 98-225, at 378, its most distinctive feature is the inclusion of a specific victim — i.e., a bank. That difference was at the core of the Loughrin Court's interpretation of the "by means of" language. Emphasizing that the crime of bank fraud requires a connection between the fraudulent statement and the victim bank, the Court concluded that Congress must have intended "by means of" to provide that "relational component." 134 S.Ct. at 2393. Because no similar link between the misrepresentation and the victim is part of the crime of mail fraud, it does not inevitably follow that "by means of" must be construed in the same way.21
Admittedly, when Congress incorporated language from the mail fraud statute into the bank fraud statute (and likely the wire fraud statute as well), it anticipated some commonality in interpretation. See H.R. Rep. No. 901, 98th Cong., 2nd Sess., at 4 (1984) (endorsing the courts' "current interpretations of the language" taken from 18 U.S.C. §§ 1341 & 1343); United States v. Bonallo, 858 F.2d 1427, 1432 (9th Cir. 1988) ("[T]he House Judiciary Committee, in considering the proposed bank fraud statute, expressly endorsed the broad reading courts have given the mail and wire fraud provisions."). We cannot assume, however, that Congress intended to incorporate mail fraud precedents that had not yet been decided. See Loughrin, 134 S.Ct. at 2391 ("Loughrin's reliance on McNally encounters a serious chronological problem. Congress passed the bank fraud statute in 1984, three years before we decided that case."); United States v. Saks, 964 F.2d 1514, 1520-21 (5th Cir. 1992) (questioning the applicability of McNally to the bank fraud statute because it was decided after the statute's enactment). Nor can we expect that Congress, in enacting the mail fraud statute a century and a half ago, would have anticipated the Loughrin Court's recent interpretation of the "by means of" language in the context of a different statute that did not yet exist. Contrary to the majority's assertion, the chronology of the statutes' enactment matters.
As my textual comparison has intimated, the purpose of the mail fraud statute also does not support the imposition of Loughrin's limiting interpretation. The Loughrin Court's interpretation of the "by means of" language was premised on the narrow objective of the bank fraud statute to protect banks from fraud. Unlike the bank fraud statute, the mail fraud statute was not enacted to protect a narrow class of victims. Rather, its purpose was to protect the public at large from a wide range of fraudulent schemes. See Cong. Globe, 41st Cong., 3d Sess. 35 (1870) (remarks of Rep. Farnsworth) ("[A]ll through this country thousands of innocent and unsophisticated people ... are continually fleeced and robbed, and the mails are made use of for the purpose of aiding them in their nefarious designs.").
The broad purpose of the statute has led our circuit to hold that the mail fraud statute encompasses many different kinds of frauds, including those where the false statement does not reach the victim:
Nothing in the mail and wire fraud statutes requires that the party deprived of money or property be the same party who is actually deceived. The phrase "scheme or artifice ... for obtaining money or property by means of false or fraudulent pretenses, representations, or promises," 18 U.S.C. § 1341, is broad enough to include a wide variety of deceptions intended to deprive another of money or property.... We see no reason to read into the statutes an invariable requirement that the person deceived be the same person deprived of the money or property by the fraud. If, for example, the role of a government regulator is to protect the monetary interests of others, a scheme to mislead the regulator in order to get at the protected funds will affect "property rights"....
United States v. Christopher, 142 F.3d 46, 54 (1st Cir. 1998) (involving scheme to obtain assets of insurance companies by deceiving state insurance regulators).
Strikingly, this convergence requirement that we rejected in Christopher for the mail fraud statute ("an invariable requirement that the person deceived be the same person deprived of the money or property by the fraud") is the essence of the "naturally inducing" standard adopted by the Supreme Court for the bank fraud statute in Loughrin. The Loughrin Court required that the false statement reach the victim bank, directly or indirectly. See 134 S.Ct. at 2394 n.8 ("drawing a line at ... frauds in which a false statement will naturally reach ... a bank (or a custodian of the bank's property)"). In other words, the Loughrin standard requires that the bank or a custodian of its property be deceived, and that the bank or the custodian lose the bank's money or property. That requirement was met in Loughrin because checks altered by the fraudster — i.e., the misrepresentations — reached the bank when Target submitted them for payment.
The majority argues that Loughrin's standard does not require convergence because it specifically contemplates that bank property covered by the statute may be held (and thus lost) by a non-bank custodian as well as by a bank. See id. at 2389 ("[A] person violates § 1344(2)'s plain text by deceiving a non-bank custodian into giving up bank property that it holds."); 18 U.S.C. § 1344(2) (stating that property "owned by" a bank, as well as property "under the custody or control of" a bank, is within the statute's scope). Therefore, an entity other than the bank itself may be deceived and may relinquish the bank's property as a result of a covered fraudulent scheme. In such cases, however, the custodian that holds the bank's property acts as a surrogate for the bank, standing in the bank's place. Indeed, the Loughrin court referred to such custodians interchangeably with the bank itself. See id. at 2393 ("Section 1344(2)'s "by means of" language is satisfied when, as here, the defendant's false statement is the mechanism naturally inducing a bank (or custodian of bank property) to part with money in its control."). Whether the property is under the custody or control of the bank or of a non-bank custodian, it is still the bank's property. The custodian is effectively acting as the bank, and the bank is victimized by the fraud. Hence, even if the surrogate custodian is the deceived party and/or the party relinquishing bank money, "the party deprived of money or property" is still "the same party who is actually deceived." Christopher, 142 F.3d at 54.
In another attempt to avoid the convergence reading of the Loughrin standard, the majority invokes Justice Scalia's hypothetical about Little Bobby's attempt to obtain an extra cookie. Contrary to the majority's suggestion, Justice Scalia used the example to explain that, under Loughrin's "naturally inducing" standard, Little Bobby would not have obtained the cookie "by means of" his fib because "the lie did not make its way to the father." Id. at 2396 (Scalia, J., concurring). So, although he deceived his mother, and his deception worked, Little Bobby could not be charged with cookie fraud. The Loughrin majority responded without indicating whether they agreed with this depiction of their holding. Id. at 2394 n.8. Instead, they reemphasized that their interpretation of "by means of," as it is used in the bank fraud statute, requires convergence. Id. ("All we say here is that the phrase, as used in § 1344(2), is best read, for the federalism-related reasons we have given ... as drawing a line at frauds ... in which a false statement will naturally reach such a bank (or a custodian of the bank's property)."). With respect to other contexts, however, the majority refused to decide whether convergence is required. Id. ("Language like `by means of' is inherently elastic: It does not mean one thing as to all fact patterns — and certainly not in all statutes, given differences in context and purpose."). Thus, the majority explicitly acknowledged, as we have already noted, that it may not be appropriate to apply the convergence standard to other statutes.
Having rejected the convergence reading of Loughrin's "naturally inducing" standard, the majority must try to explain what that standard means. To this end, it says that Loughrin requires a "causal nexus" between the deception and the loss of property. Oddly, this phrase appears nowhere in the Loughrin decision. Instead, the majority seems to have invented the phrase by drawing on our own Christopher decision, where, having rejected the convergence requirement for the mail fraud statute, we acknowledged that, in both the mail and wire fraud statutes, "the deception must in fact cause the loss."22 Christopher, 142 F.3d at 53. The critical question, then, is what does the causal nexus standard imputed by the majority to Loughrin, and now incorporated by the majority into the mail fraud statute, mean if it does not signify convergence?
The majority says that the causal nexus standard is something "like proximate causation." But they never define this standard, stating instead that it is "flexible." Indeed, the standard is so flexible that we cannot predict what it will mean in any future case. We know that in Loughrin it meant that the false statement had to "reach the bank," 134 S.Ct. at 2394 n.8; in Christopher, the majority says, it was met because the defendant's actions "directly contradicted" his prior statements. Here, the majority suggests critically that the standard was not met because only "in the ensuing years" after receiving their fraudulent medical licenses did the doctors obtain funds from consumers and medical insurers. Is the majority imposing some temporal requirement on the causal connection between the deception and the loss of property? Would it have mattered in this case if the doctors had only used their fraudulent licenses to obtain money from consumers and insurers immediately after receiving those licenses? Does the passage of time somehow mean in this case that the fraud of the defendants did not "naturally induce" healthcare consumers to part with their money?
These unanswered questions, and the resulting uncertainty for the future application of the mail fraud statute, highlight the majority's threefold mistake in its reading of Loughrin. First, without justification, they have incorporated Loughrin's standard for bank fraud into the mail fraud statute. In an effort to minimize the consequences of that mistake, they deny that the "naturally inducing" standard of Loughrin is a convergence requirement — the party deceived must lose the property. Then, in place of that convergence standard, the majority offers a vague "causal nexus" standard that apparently means something more than the present causation requirement of the mail fraud statute. In so doing, the majority has unwisely circumscribed the broad protective purpose of the mail fraud statute.
3. Federalism Concerns
The Loughrin Court's federalism concerns about the bank fraud statute also do not apply to the mail fraud statute. Underlying the mail fraud statute is a federal interest in ensuring that the national mail system is not used to further fraudulent schemes. The bank fraud statute, by contrast, is based upon an interest in preventing federally regulated and insured banks from being victimized by fraud. The Loughrin Court's federalism concerns focused on the defendant's contention that the bank fraud statute could be expanded beyond its narrow focus to be used as "a plenary ban on fraud, contingent only on the use of a check (rather than cash)." Loughrin, 134 S.Ct. at 2392.
The Court resolved this concern by finding a textual limitation in the "by means of" language. By requiring that the fraudulent statement reach the bank that relinquishes money, the Court excluded from the scope of the bank fraud statute those frauds that are "only tangentially related" to the underlying federal interest in protecting banks. Id. at 2393. To be sure, the "naturally inducing" standard of Loughrin could function the same way in the mail fraud statute as it does in the bank fraud statute. But imposing that restriction on the mail fraud statute would not exclude from prosecution those frauds that are "only tangentially related" to the federal interest in preventing the mails from being used in fraudulent schemes. The Loughrin standard has nothing to do with the mailing element. Hence, the federalism concerns of the Loughrin Court do not justify the narrowing interpretation adopted by the majority here.
Indeed, the mail fraud statute has been used for decades to prosecute frauds similar to the case at hand, where the fraudulent representations did not reach the entity whose property the scheme sought to obtain (or the custodian of that property). See, e.g., United States v. Greenberg, 835 F.3d 295 (2d Cir. 2016) (scheme to obtain money from credit card holders by deceiving payment processor and other intermediaries); United States v. Seidling, 737 F.3d 1155 (7th Cir. 2013) (scheme to obtain money from opposing parties by deceiving state small claims court); United States v. McMillan, 600 F.3d 434 (5th Cir. 2010) (scheme to obtain money from insureds by deceiving state health insurance regulators); United States v. Blumeyer, 114 F.3d 758 (8th Cir. 1997) (scheme to defraud insurance policyholders and brokers by deceiving state insurance regulator); United States v. Granberry, 908 F.2d 278 (8th Cir. 1990) (scheme to obtain job as a school bus driver by deceiving state driver's licensing authority); United States v. Brownlee, 890 F.2d 1036 (8th Cir. 1989) (scheme to obtain proceeds of automobile theft insurance policies by deceiving state authority certifying automobile titles). Permitting the mail fraud statute to continue covering such frauds would not, therefore, effect "a sweeping expansion of federal criminal jurisdiction" as the majority suggests. Loughrin, 134 S.Ct. at 2392 (quoting Cleveland, 531 U.S. at 24, 121 S.Ct. 365).
The majority also finds support for its federalism concerns in Cleveland v. United States, 531 U.S. 12, 121 S.Ct. 365, 148 L.Ed.2d 221 (2000). Just as the federalism concerns expressed in Loughrin are not applicable here, however, neither are those raised by the Cleveland Court. In Cleveland, the government asked the Court to adopt an expansive and unsupported definition of the term "property" in the mail fraud statute so that federal authorities could prosecute the defendant for a scheme to obtain a license issued by a state. Here, the government does not request such a textual expansion, nor has it charged a scheme directed against a state.
Cleveland involved a scheme to fraudulently obtain a video poker operating license by making false statements on an application to the state licensing authority. Id. at 15, 121 S.Ct. 365. Noting that the mail fraud statute is "limited in scope to the protection of property rights," id. at 18, 121 S.Ct. 365 (quoting McNally, 483 U.S. at 360, 107 S.Ct. 2875), the Supreme Court held that licenses do not constitute property when they are in the hands of the licensing agency, id. at 20, 121 S.Ct. 365. The idea of licenses as government property, the Court noted, "stray[s] from traditional concepts of property." Id. at 24, 121 S.Ct. 365. Regardless of any fees that the state may obtain during the licensing process, "the [s]tate's core concern is regulatory," id. at 20, 121 S.Ct. 365, "implicat[ing] the [g]overnment's role as sovereign, not as property holder," id. at 24, 121 S.Ct. 365.
The Court also noted that there was no evidence Congress had intended to override the traditional definition of "property" by including licenses in the "property" protected by the mail fraud statute. Id. Moreover, the state had already created penalties for false statements made on license applications. Accordingly, defining licenses as property would "`significantly change the federal-state balance' in the prosecution of crimes." Id. at 25, 121 S.Ct. 365 (quoting Jones v. United States, 529 U.S. 848, 858, 120 S.Ct. 1904, 146 L.Ed.2d 902 (2000)). Because Congress had not clearly stated that the term "property" included licenses, the Court declined to accept such an interpretation with its resultant policy implications.
Although the majority acknowledges that the government has not charged a scheme in which a state license was treated as property, it asserts that "the very same federalism concerns" underlie this case. In so doing, it misconstrues the federalism problem identified by the Cleveland Court. The Court hesitated to adopt a novel definition of "property" in Cleveland that could change the balance of federal-state power in the prosecution of crimes without a clear statement of Congressional intent. It did not imply, however, that the government is forbidden from bringing a case that otherwise fits within the boundaries of the mail fraud statute simply because the case begins with an abuse of the state's licensing authority. Cf. Pasquantino v. United States, 544 U.S. 349, 357, 125 S.Ct. 1766, 161 L.Ed.2d 619 (2005) (distinguishing Cleveland and upholding wire fraud conviction for scheme to defraud Canadian government of tax revenue); McMillan, 600 F.3d at 434 (upholding mail and wire fraud convictions for scheme deceiving state health insurance regulators).
Here, the government does not focus its mail fraud charge, as it did in Cleveland, on the use of false statements to obtain a license. Instead, it focuses on the use of the fraudulently obtained licenses to secure the money of patients (and their insurers and other payers) who received services from the improperly licensed defendant physicians. Contrary to the majority's implication, the money of the patients and payers was not "far removed from" the defendants' fraudulent scheme. Indeed, the object of the charged scheme and its fraudulent representations was the money the physicians could earn with their fraudulently obtained medical licenses. Their continued use of their fraudulent licenses to practice medicine and obtain patients' money years after deceiving the Board of Medical Examiners demonstrates that the payments, not the license, were the object of the scheme. The use of the mail fraud statute to protect the property of members of the public in situations such as this fits comfortably within the intended scope of the mail fraud statute.
The majority argues that the mail fraud statute must be narrowed, or else "virtually any false statement in an application for a medical license could constitute a federal crime." Permitting the statute to have such a broad scope, it asserts, would "impermissibly infringe on the states' `distinctively sovereign authority to impose criminal penalties for violations of' licensing schemes." (Quoting Cleveland, 531 U.S. at 23, 121 S.Ct. 365). The fact that a federal statute may be used to prosecute conduct that also implicates state concerns does not, however, mean that the federal government has "impermissibly infringe[d]" on the state's sovereign authority. The federal and state governments often bring parallel prosecutions to vindicate their separate interests. See United States v. Lanza, 260 U.S. 377, 382, 43 S.Ct. 141, 67 S.Ct. 314 (1922) ("We have here two sovereignties, deriving power from different sources, capable of dealing with the same subject-matter within the same territory.... [A]n act denounced as a crime by both national and state sovereignties is an offense against the peace and dignity of both and may be punished by each."). Congress has determined that there is a significant federal interest in ensuring that the mails are not used in furtherance of fraudulent schemes. The Commonwealth might have invoked its own criminal statute to prosecute the defendants in this case for false statements in their license applications. But the Commonwealth's interest does not prevent the federal government from also prosecuting them for a different crime that involves a distinct federal interest. See United States v. Volungus, 595 F.3d 1, 9 (1st Cir. 2010) ("When the federal government exercises any of the powers granted to it by the Constitution, it is not a valid objection that the exercise may bring with it some incidents of the [state's] police power.").
The government's charging decision in this case was not an evasion of Cleveland. It was a recognition of Cleveland's limited scope. The federalism concerns raised in that case are inapplicable here, and significant federal interests support the government's application of the mail fraud statute in this case.
The convictions of the defendants in this case comport with the traditional understanding of the scope of the mail fraud statute applied by courts for decades.23 The majority's decision to incorporate into the mail fraud statute the Loughrin Court's limiting interpretation of the bank fraud statute's "by means of" language will unnecessarily and unwisely constrain the federal government in its prosecution of fraud cases. Because the text, purpose, and long-standing application of the mail fraud statute do not support the novel limitation that the majority imposes on it, I respectfully dissent.