WILLIAMS v. ATTORNEY GENERAL, No. 16-3816 (3d Cir. 1/19/2018)




Nos. 16-3816 & 17-1705


RAMON ANDREW WILLIAMS  a/k/a Andrew Denton Williamson  a/k/a Ramon Williams,






On Petitions for Review of Orders of the Board of Immigration Appeals

Agency No: A018-687-061

Immigration Judge:  Leo A. Finston


Argued November 8, 2017

Before: SMITH, Chief Judge, HARDIMAN, Circuit Judge, and BRANN, District Judge[1]

(Opinion Filed:  January 19, 2018)

Thomas H. Lee, II

Argia J. DiMarco

Ryan M. Moore

Christopher J. Mauro                                   [ARGUED]


2929 Arch Street

18th Floor, Cira Centre

Philadelphia, PA 19104


Seymour James, Jr.

Adriene Holder

Maria E. Navarro

Hasan Shafiqullah

Ward Oliver

Sarah Gillman

Whitney W. Elliott

Legal Aid Society

Immigration Law Unit

199 Water Street

3rd Floor

New York, NY 10038

            Counsel for Petitioner


Chad A. Readler

Acting Assistant Attorney General

Terri J. Scadron

            Assistant Director

Shahrad Baghai

Christina Greer                                             [ARGUED]

United States Department of Justice

Office of Immigration Litigation

P.O. Box 878

Ben Franklin Station

Washington, DC 20044

Counsel for Respondent




SMITH, Chief Judge.

In this consolidated proceeding, Ramon Williams asks us to consider whether a prior conviction under Georgia’s forgery statute, Ga. Code Ann. § 16-9-1(a) (2006), constitutes an aggravated felony conviction for purposes of the Immigration and Naturalization Act (“INA”).  See 8 U.S.C. § 1227(a)(2)(A)(iii).  Because we conclude that the Georgia conviction is an offense “relating to . . . forgery,” 8 U.S.C. § 1101(a)(43)(R), Williams is properly subject to removal as an aggravated felon, and we will therefore deny the petitions for review.


Williams, a citizen of Guyana and a lawful permanent resident of the United States, immigrated to this country in 1970, when he was thirteen months old.  He has no family in Guyana; his parents, grandparents, siblings, and children are all United States citizens.  In 2006, he pleaded guilty in Georgia state court to five counts of first degree forgery pursuant to section 16-9-1(a) of the Georgia Code.  He initially received a sentence of two years in prison, which later was reduced to one year.

In 2013, Williams received a notice to appear charging him as removable as a result of having been convicted of an aggravated felony.  See 8 U.S.C. § 1227(a)(2)(A)(iii).  Appearing before an Immigration Judge (“IJ”) in New Jersey, he contested removability.[2]  The IJ determined that the Georgia forgery conviction rendered Williams deportable as an aggravated felon and otherwise denied relief.  Williams appealed to the Board of Immigration Appeals (“BIA”).

Before the BIA, he argued, inter alia, that the Georgia forgery statute is broader than generic forgery because it criminalizes the use of a fictitious name when signing a document and because the statute does not require a showing of prejudice.  The BIA rejected these arguments, upheld the IJ’s decision, and dismissed the appeal.

Williams timely filed a petition for review, and also sought reconsideration before the BIA in light of the Supreme Court’s decision in Mathis v. United States, 136 S. Ct. 2243 (2016).  In his motion for reconsideration, Williams argued that Georgia’s forgery statute is indivisible under Mathis and is overbroad because it criminalizes some conduct that does not relate to forgery, namely, false agency endorsements.  The BIA denied the reconsideration motion, and Williams timely filed a second petition for review.

The petitions have been consolidated.  We have jurisdiction over them pursuant to 8 U.S.C. § 1252(a).


The issue of whether Williams’s conviction under the Georgia forgery statute qualifies as an aggravated felony is a question of law over which we have jurisdiction.  Id. § 1252(a)(2)(D).  We conduct a de novo review of the BIA’s determination.  Denis v. Atty. Gen., 633 F.3d 201, 209 (3d Cir. 2011); Bobb v. Atty. Gen., 458 F.3d 213, 217 (3d Cir. 2006).


The INA provides for the deportation of an alien “who is convicted of an aggravated felony.”  8 U.S.C. § 1227(a)(2)(A)(iii).  The INA’s definition of an “aggravated felony” encompasses an extensive list of various types of offenses, see id. § 1101(a)(43)(A)–(U), but for current purposes, only one definition is pertinent: an “aggravated felony” is “an offense relating to . . . forgery . . . for which the term of imprisonment is at least one year.”  Id. §1101(a)(43)(R). In his petitions for review, Williams calls upon us to consider whether the BIA was correct when it determined that his 2006 conviction under Georgia’s forgery statute, for which he was imprisoned for a year, is an “offense relating to forgery.”


At the time of Williams’s conviction, Georgia’s forgery statute provided:

A person commits the offense of forgery in the first degree when with intent to defraud he knowingly makes, alters or possesses any writing in a fictitious name or in such manner that the writing as made or altered purports to have been made by another person, at another time, with different provisions, or by authority of one who did not give such authority and utters or delivers such writing.

Ga. Code Ann. § 16-9-1(a) (2006).  The Georgia legislature’s decision to denote this offense as “forgery” does not dictate whether it comes within the meaning of forgery as Congress intended it in the INA.  Drakes v. Zimski, 240 F.3d 246, 248 (3d Cir. 2001) (“The language of a federal statute must be construed to have the meaning intended by Congress, not the [state] legislature.”).  To make that assessment, we employ what is known as the “categorical approach.”2   See Moncrieffe v. Holder, 569 U.S. 184, 190 (2013).


Under the categorical approach, we look to the substance of the statute of conviction to determine whether it categorically fits within the “generic” federal definition of the corresponding aggravated felony, without considering the facts of the particular case.  Id.; see also Mathis, 136 S. Ct. at 2249 (“The comparison of elements that the categorical approach requires is straightforward . . . .  The court . . . lines up that crime’s elements alongside those of the generic offense and sees if they match.”).  We thus compare the state and federal offenses “in the abstract,” consulting only their respective elements to determine whether the state conviction “necessarily involved facts equating to the generic federal offense.”  Moncrieffe, 569 U.S. at 190 (internal quotation marks and alterations omitted) (quoting Shepard v. United States, 544 U.S. 13, 24 (2005)).

Before we may conduct this comparison, we must consider what constitutes the “generic federal offense” of forgery.  See id.  As we observed in Drakes v. Zimski, Congress has not articulated a specific definition for the term.  240 F.3d at 249.  “Where federal criminal statutes use words of established meaning without further elaboration, courts typically give those terms their common law definition.”  Id.  As we have long held, the traditional common law definition of forgery has three elements:  “(a) The false making or material alteration (b) with intent to defraud (c) of a writing which, if genuine, might be of legal efficacy.”  United States v. McGovern, 661 F.2d 27, 29 (3d Cir. 1981).  Moreover, because the INA applies not only to forgery, but also extends to offenses “relating to” forgery, we have concluded that Congress intended to define forgery “in its broadest sense.”  Drakes, 240 F.3d at 249.

Here, however, the definition of the term “forgery” is not enough, on its own, to answer the question of whether the crime defined in section 16-9-1 of the Georgia Code is “an offense relating to forgery” within the meaning of the INA.  Accordingly, in comparing the generic federal offense to the Georgia statute, we employ a “looser categorical approach.” Flores v. Atty. Gen., 856 F.3d 280, 286 (3d Cir. 2017) (applying this approach in analyzing whether an alien had been convicted of offenses “relating to obstruction of justice”).  Under this looser approach, we do not require a precise match between the elements of the generic federal crime and those of the Georgia offense.  Id. at 291.  Instead, we “survey the[ir] interrelationship” and consider whether there is “a logical or causal connection” between them.  Id. (quoting Denis, 633 F.3d at 212).  We may conclude that the crimes are logically connected if they both “target the same, core criminal conduct such that they are ‘directly analogous.’”  Id.  And, we may conclude that the crimes are causally connected where there is a “link between the alien’s offense and a listed federal crime: without the listed federal offense, the alien’s offense could not have occurred.”  Id.  Because the parties agree that there is no “causal connection” between the federal and state crimes under discussion here, our focus is the “logical connection” between them.


Williams’s primary claim is that the Georgia forgery statute is broader than the federal common law definition of forgery because it punishes the possession of certain “genuine” documents, namely, documents that “purport[ ] to have been made . . . by authority of one who did not give such authority.”  Ga. Code Ann. § 16-9-1(b).  Williams refers to this aspect of the Georgia statute as “false agency endorsement,” which, he argues, is conduct that does not fall within the traditional common law definition of forgery.

As a threshold matter, the Government argues that, although false agency endorsement may technically fall within the Georgia statute’s language, Georgia does not actually prosecute false agency endorsement as forgery.  Accordingly, the Government claims, Williams has established no more than a “theoretical possibility” that Georgia would apply its statute to conduct falling outside the federal definition of forgery.  See Singh, 839 F.3d at 278.

Williams responds that there is Georgia case law demonstrating that the State actually prosecutes false agency endorsement as forgery, citing Warren v. State, 711 S.E.2d 108 (Ga. Ct. App. 2011).  In Warren, a business manager of a medical practice used business checks to pay personal credit card debt without the authorization of her employer.  The Court of Appeals of Georgia concluded that the evidence was sufficient to support her conviction of first degree forgery because it showed that, “with intent to defraud the doctor and his medical practice, she knowingly possessed fifty-two checks drawn on the practice’s bank account without authority, and uttered the checks as payment on her personal credit card debt.”  Id. at 109.

In a supplemental appendix, Williams provided a copy of the Warren indictment as further support for his position that the Georgia forgery statute is employed to prosecute false agency endorsement.  See S.A.  1–18.  The indictment indicates that the defendant was charged with first degree forgery for signing her own name to a check, “purportedly on behalf of [the medical practice] as an authorized signatory of [the medical practice], but having not been written and signed with the authority of [the medical practice and the doctor], and did utter said check.”  S.A. 2, Count 5.  In other words, the Warren defendant’s signature was her own but was made without authorization of the principal, and therefore was a false agency endorsement.  Inasmuch as both the Warren indictment and the opinion of the Court of Appeals of Georgia support Williams’ view, and the Government has not offered anything to rebut that evidence, we conclude that Williams has established a sufficiently “realistic probability” that Georgia would apply its forgery statute to false agency endorsement.  See Singh, 839 F.3d at 278 (quoting Moncrieffe, 569 U.S. at 191).

Next, Williams contends that, in contrast to the Georgia statute, the federal common law definition of forgery does not extend to false agency endorsement, making the Georgia statute broader than the federal version of the crime.  But whatever the scope of forgery under federal law, Congress expressly extended its coverage to offenses “relating to” forgery.  We must therefore resolve the question of whether the Georgia statute’s inclusion of false agency endorsement extends so far beyond the traditional common law definition that it criminalizes conduct that is unrelated to forgery.  Employing the looser categorical approach, we conclude it does not.  See Flores, 856 F.3d at 286.

Williams argues that false agency endorsements do not relate to forgery because they do not involve a “false instrument,” and a false instrument, he contends, is at the core of the federal common law crime.  See Pet. Br. at 29.  Although a false instrument is an integral aspect of common law forgery, we cannot agree with Williams’s argument that the falsity of the instrument must appear on the face of the document in order for an offense to “relate to” forgery.  Even if facial falsity is viewed as an essential element of common law forgery that is missing from the provision of Georgia’s statute prohibiting false agency endorsement, the omission of an essential element simply does not resolve whether the conduct is “related” for purposes of the INA.  See Bobb, 458 F.3d at 219.

In conducting the necessary survey of the interrelationship between common law forgery and false agency endorsement, we are satisfied that, although their elements do not line up with precision, the crimes share a logical connection.  See Flores, 856 F.3d at 291.  We offer a series of related examples to demonstrate the analogous nature of the two crimes, taken from the example in Warren, the Georgia case to which Williams refers.

First, we consider the most straightforward scenario: if the defendant in Warren had signed her employer’s name rather than her own name on the business checks, then we may uncontroversially conclude that her act would fall within the quintessential common law definition of forgery.  The forged instrument would, on its face, reflect the sort of falsity in execution that Williams proposes is integral to the definition of the common law crime—the use of a false name.  Next, we consider a slightly different but related scenario, in which the defendant signs her true name to the employer’s checks under the handwritten phrase “by authority of” her employer— authority that was never granted.  Much like the signing of a false name, the falsity of the authorization would appear on the face of the document, and we therefore presume that Williams would agree that the defendant’s hypothetical act would be related to forgery.[3]  Finally, consider the conduct for which the Warren defendant actually was prosecuted: she signed her true name to a check, implicitly under the authority of her employer, but without writing the phrase “by authority of.”  By Williams’s logic, this third example would not be a “false instrument” because there is no falsity on the document’s face, and the defendant’s act therefore would no longer relate to forgery.  But, in our view, each of these incrementally different acts is logically related to the same underlying core conduct.  See Flores, 856 F.3d at 291.  Each example gives rise to essentially the same concerns about the inauthenticity and unauthorized nature of the written instrument.  Thus, we think Williams’s proposed approach of drawing a bright line at facial falsity presents too fine a distinction given the broad “relating to” language that Congress chose to employ in the INA.  In short, common law forgery and false agency endorsement share a logical connection because they “target the same, core criminal conduct such that they are ‘directly analogous.’”  See id.

The Government also points us to the Model Penal Code and a number of state statutes employing the Model Penal Code’s provisions as a source for a “broad minority definition” of forgery that extends to false agency endorsement.  See Model Penal Code § 224.1(1)(b) (Am. Law Inst. 1985) (“A person is guilty of forgery if, with purpose to defraud or injure anyone, or with knowledge that he is facilitating a fraud or injury to be perpetrated by anyone, the actor . . . makes, completes, executes, authenticates, issues or transfers any writing so that it purports to be the act of another who did not authorize that act . . . .”); see also Iowa Code § 715A.2(1)(b) (1996); N.J. Stat. Ann. § 2C:21-1(a)(2) (West 2002); 18 Pa. Cons. Stat. § 4101(a)(2) (2003); Wyo. Stat. Ann. § 6-3602(a)(ii) (2007).  We have observed that conduct falling within a “broad minority definition” can be an appropriate source of information for discerning whether conduct “relates to” an offense for INA purposes.  Drakes, 240 F.3d at 250.  We conclude that the Model Penal Code’s broad minority definition of forgery buttresses our conclusion that false agency endorsement shares a logical connection with common law forgery.

To our knowledge, only one other Court of Appeals has published a decision interpreting § 1101(a)(43)(R) in the context of a state statute that, like the Georgia statute, defines forgery in a manner that encompasses false agency endorsement.  In Vizcarra-Ayala v. Mukasey, 514 F.3d 870 (9th Cir. 2008), the Ninth Circuit noted, as we have, that the Supreme Court and other Circuits have concluded that forgery requires “falsification of a document or instrument,” making it “clear that an essential element of the generic offense of forgery is the false making or alteration of a document, such that the document is not what it purports to be.”  Id. at 875.  The Ninth Circuit then cited several cases in which California had prosecuted individuals for “possession or use of a genuine instrument with intent to defraud but not to forge,” in other words, conduct falling outside the generic federal definition of forgery.  Id. at 876.  And, it described one particular case in which California prosecuted false agency endorsement under its forgery statute: a defendant office manager took checks from her employer and, rather than use them for purchasing office supplies, made the checks out to “cash” and used the proceeds for herself.  Id. at 876–77.  In that case, the checks were genuine and the defendant’s signatures on them were true.  Id. at 877.  According to the Ninth Circuit, this “made her conduct fraud, but not forgery under the generic definition.”  Id.

Up to this point, we concur with the Ninth Circuit’s analysis.  But, we diverge from its ultimate conclusion: that the California statute’s application to “genuine instrument[s],” including false agency endorsements, means that it extends to conduct that does not “relate to” forgery.  Id. at 876–77.  Relying on a facial falsity premise much like the one Williams proposes, the Ninth Circuit summarily concluded that “[e]xpanding the definition of offenses ‘relating to’ forgery to include conduct where documents are not altered or falsified [would] stretch[] the scope too far.”  Id. at 877.  Because we conclude that concerns about the inauthenticity or unauthorized nature of a written instrument establish a logical relationship between common law forgery and false agency endorsement, we respectfully disagree with the premise that the falsity of the instrument must be reflected on its face in order for conduct to “relate to” forgery.

Another important distinction is the degree to which the “relating to” language of § 1101(a)(43)(R) affects the analysis.  While in this Circuit it triggers the application of the “looser categorical approach” and its “logical or causal connection” test, Flores, 856 F.3d at 286, 291, the Ninth Circuit in Vizcarra-Ayala acknowledged only that offenses with a causal connection can “relate to” forgery (e.g., possession of a forged document), but declined to afford logical connections like the one at issue there the same treatment, see 514 F.3d at 877.  The Ninth Circuit’s test for whether a state offense “relat[es] to . . . forgery” is more restrictive than this Court’s, and here, that difference was outcome-determinative.


Williams presents a secondary argument as to the purported overbreadth of the Georgia forgery statute, contending that the Georgia statute is missing a necessary element of federal common law forgery: a requirement that the forged instrument be “capable of effecting a fraud” or have “legal efficacy.”  Relatedly, Williams argues that the absence of the “legal efficacy” element extends Georgia’s statute beyond the commercial realm into merely “personal” acts and such personal acts do not relate to forgery for purposes of the INA.[4]

The Government retorts that Williams waived this argument because he did not present it to the BIA.  Williams disagrees, pointing out that he argued before the BIA that the Georgia statute “lacked a prejudice requirement,” an argument that the BIA considered and rejected.  See J.A. 43–44.  While acknowledging that he used different terminology, Williams contends that his argument to the BIA—that the Georgia statute lacks an element of “prejudice [to] another”—is sufficiently similar to his argument to this Court—that the Georgia statute lacks an element of being “capable of prejudicing another’s rights.”  Reply Br. at 19.  He argues that precision is not required and that his prejudice argument below was sufficient to put the BIA on notice of the issue.

We accept that the claim presented on appeal is sufficiently similar to the argument presented to the BIA to satisfy the exhaustion requirement.  Nonetheless, we are not persuaded that it has merit.  In Bobb, we observed that, “[a]t their core, all common law forgery offenses contain as an element an intent to deceive.”  458 F.3d at 218.  Analogously, the Georgia forgery statute expressly sets forth an “intent to defraud” as one element of the crime.  Ga. Code Ann. § 16-91(a).  These intent elements are “directly analogous” and target the “same core criminal conduct.”  See Flores, 856 F.3d at 291.

To the extent Williams contends that the common law definition goes a step further by looking to the potential effect of the forged instrument on its victim rather than the intent of the forger alone, such a minute distinction does not carry the day.  Even apart from our skepticism about this claim, in Bobb, we observed that a state criminal statute can relate to forgery even where it “encompasses conduct beyond the traditional definition of forgery, and includes criminal conduct that is causally connected to forgery, but may lack as an essential element an intent to defraud or deceive.”  458 F.3d at 219.  Thus, so long as the Georgia statute covers conduct that is logically or causally connected to forgery—which, undoubtedly, it does—it is of no moment that a supposed additional “essential element” of the generic federal definition, such as the ability of the forged instrument to cause harm to a victim, might theoretically be absent from the Georgia statute.  See id.

Moreover, there is no basis for concluding that the Georgia statute lacks this element and therefore extends to conduct that is purely “personal” or “non-commercial” in nature.  Notably, Williams offers no evidence to support his contention that there is a “personal” version of forgery that would be subject to prosecution in Georgia and yet be exempted from the federal common law conception of fraud.

We see no “realistic probability” that the State would apply its forgery statute in this manner.  Singh, 839 F.3d at 278 (quoting Moncrieffe, 569 U.S. at 191).  Accordingly, Williams’s argument fails.


For the foregoing reasons, these consolidated petitions for review will be denied.[5]


[1] The Honorable Matthew W. Brann of the United States District Court for the Middle District of Pennsylvania, sitting by designation.

2. While this is the general rule, certain disjunctivelyworded statutes that set forth a number of separate crimes warrant a departure from the categorical approach known as the “modified categorical approach.”  Mathis v. United States, 136 S. Ct. 2243, 2249 (2016); see Moncrieffe v. Holder, 569 U.S. 184, 191 (2013).  When it applies, the modified categorical approach permits a court to consult a limited set of documents, such as an indictment, guilty plea, or jury instructions, to determine which specific offense is at issue in the case.  Moncrieffe, 569 U.S. at 191.  In Mathis, 136 S. Ct. at 2248–50, the Supreme Court provided guidance regarding how to determine whether a disjunctively-worded state statute sets forth an “indivisible” set of elements defining a single crime—but listing several different means of satisfying the elements of the crime—to which the strict categorical approach applies, or a “divisible” list of elements in the alternative—thereby defining multiple separate crimes— to which the modified categorical approach applies.  Georgia’s forgery statute is undoubtedly disjunctive and, initially, the issue of its divisibility or indivisibility under Mathis was hotly contested.  At oral argument, however, the Government conceded its agreement with Williams’s view that the statute defines a single crime and is therefore indivisible, warranting the application of the categorical approach.  Accordingly, for our purposes, we assume without deciding that the parties are correct that the categorical approach applies.

[2] Williams also sought asylum, withholding of removal, and relief under the Convention Against Torture.  The IJ denied these forms of relief and the Board of Immigration Appeals (“BIA”) affirmed.  Williams does not challenge the denial of those claims in his petitions before this Court.

[3] We note that Williams has never contended that use of a false name is the only manner in which a document can qualify as a “false instrument” for purposes of common law forgery.  At oral argument, Williams contended that, for instance, mimicking a company’s logo to create a false check would qualify as the false making of a document and therefore is a form of forgery.

[4] Upon being questioned about the contours of this claim at oral argument, Williams all but abandoned it, candidly acknowledging that this was not his “primary argument” and that the false agency endorsement claim presented a “much closer question.”  Oral Argument at 5:49–8:58, Williams v. Atty. Gen., Nos. 16-3816, 17-1705 (3d Cir. Nov. 8, 2017).  Williams’s response was sufficiently equivocal to leave us with some doubt as to whether this argument has been waived, so we proceed to address its merits.


[5] In his brief, Williams offers a final argument, contending that he should not be subject to removal because removal is a punishment disproportionate to his crime and should be set aside as unconstitutional under the Fifth or Eighth Amendments of the United States Constitution.  He stated, however, that he offered the argument solely for the purpose of preserving the issue for future review, acknowledging that we are bound by our precedent holding that removal is not a punishment and is therefore not subject to challenge as a disproportionate punishment under the Fifth or Eighth Amendment.  Sunday v. Atty. Gen., 832 F.3d 211, 218–19 (3d Cir. 2016).  Relying upon Sunday, we will not further address the proportionality argument.