FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
UNITED STATES OF AMERICA, No. 08-10560
Plaintiff-Appellee, D.C. No.
v. 2:05-CR-00923-
RAYFORD TERRELL, FJM-1
Defendant-Appellant. District of Arizona,
Phoenix
ORDER
Filed September 13, 2010
Before: A. Wallace Tashima, Susan P. Graber and
Jay S. Bybee, Circuit Judges.
Order;
Dissent by Judge M. Smith
ORDER
The panel has voted to deny the petition for panel rehear-
ing. Judges Graber and Bybee have voted to deny the petition
for rehearing en banc, and Judge Tashima has so recom-
mended.
The full court was advised of the petition for rehearing en
banc. A judge requested a vote on whether to rehear the mat-
ter en banc. The matter failed to receive a majority of the
votes of the nonrecused active judges in favor of en banc con-
sideration. Fed. R. App. P. 35.
The petition for panel rehearing and the petition for rehear-
ing en banc are denied. Judge M. Smith’s dissent from denial
of rehearing en banc is filed concurrently herewith.
13999
14000 UNITED STATES v. TERRELL
Judge M. SMITH, with whom Chief Judge KOZINSKI and
Judges PREGERSON and REINHARDT join, dissenting
from the denial of rehearing en banc:
Just one year ago, Chief Judge Kozinski presciently
observed that we had on our hands “a train wreck in the mak-
ing.” United States v. Mayer, 560 F.3d 948, 951 (9th Cir.
2009) (Kozinski, C.J., dissenting from the denial of rehearing
en banc). For years we had become accustomed to applying
uniformly the Taylor categorical approach in our immigration
and criminal sentencing jurisprudence. Mayer ended that
practice. It developed a unique categorical approach for cases
involving the Armed Career Criminal Act (ACCA) and the
Sentencing Guidelines career offender provision. It also
expanded those provisions’ “residual clause” to cover nearly
any crime. Chief Judge Kozinski warned of the growing num-
ber of casualties that would result in our circuit following our
decision in Mayer. See id. at 954. Add Rayford Terrell as the
next name on that list, and expect it to grow after the panel’s
decision here.
In this case the panel holds that we can find a categorical
match—and apply a substantial sentence enhancement—
without reference to a single state case or other objective mea-
sure of whether the state offense falls within the definition of
the generic federal offense that Congress intended to be the
basis for enhancement. Apparently, a panel’s gut instinct is all
that matters. It also holds that a state offense that lacks one
of the generic crime’s key elements—key because it makes
the generic offense inherently violent—is “almost always” the
same offense as the corresponding generic crime. As a result,
we have officially abandoned the categorical approach in “re-
sidual clause” cases. Because I believe the panel’s approach
is contrary to the intent of Congress and Supreme Court case
law, I respectfully dissent from the court’s denial of rehearing
en banc.
UNITED STATES v. TERRELL 14001
I
Determining whether an offense falls within the ACCA’s
residual clause involves two distinct questions. First, we
examine the degree of risk associated with the crime to deter-
mine whether the offense “involves conduct that presents a
serious potential risk of physical injury to another,” 18 U.S.C.
§ 924(e)(2)(B)(ii). See James v. United States, 550 U.S. 192,
203 (2007) (“ask[ing] whether the risk posed by [the offense
at issue] is comparable to that posed by its closest analog
among the enumerated offenses—[burglary, arson, extortion,
or crimes that involve the use of explosives]”); Begay v.
United States, 553 U.S. 137, 141 (2008) (assuming that New
Mexico’s DUI statute involves conduct presenting such a
risk). If we determine that the degree of risk posed by the
offense is such that it presents a serious potential risk of phys-
ical injury to another, we move on to the second question:
whether the offense is “roughly similar, in kind as well as in
degree of risk posed” to the enumerated offenses in that it
“typically involve[s] purposeful, violent, and aggressive con-
duct.” Begay, 553 U.S. at 143-45 (internal quotation marks
omitted).
The fatal flaw in the panel’s analysis is that it ignores the
Supreme Court’s repeated instructions that we answer these
questions under the categorical approach first articulated in
Taylor v. United States, 495 U.S. 575, 602 (1990). See Begay,
553 U.S. at 141 (citing Taylor for “adopting this ‘categorical
approach’ ” and explaining that “[i]n determining whether this
crime is a violent felony, we consider the offense generically,
that is to say, we examine it in terms of how the law defines
the offense and not in terms of how an individual offender
might have committed it on a particular occasion”).
The categorical approach focuses on “the elements of the
crimes of which the defendant was previously convicted.”
United States v. Selfa, 918 F.2d 749, 751 (9th Cir. 1990);
James, 550 U.S. at 202 (“[W]e consider whether the elements
14002 UNITED STATES v. TERRELL
of the offense are of the type that would justify its inclusion
within the residual provision, without inquiring into the spe-
cific conduct of this particular offender.”); United States v.
Riley, 183 F.3d 1155, 1158 (9th Cir. 1999) (recognizing that
under the categorical approach “we analyze only the elements
of the crime charged”). Under the residual clause, then, the
elements of the offense must satisfy: (1) James’s requirement
that the degree of risk associated with the crime presents a
serious potential risk of physical injury; and (2) Begay’s
requirement that the offense typically involves purposeful,
violent, and aggressive conduct.
As to the Arizona sexual assault offense at issue in Terrell,
this means that we first ask whether “sexual intercourse or
oral sexual contact . . . without consent,” Ariz. Rev. Stat.
§ 13-1406(A), presents a serious risk of physical injury. I
agree that our decision in Riley, 183 F.3d at 1159, governs
this part of the analysis, because we held in Riley that non-
consensual sexual intercourse presents such a risk.
At step two of the analysis, we must ask whether the ele-
ments of Arizona sexual assault typically involve purposeful,
violent, and aggressive conduct. They do not. Under Arizona
law, sexual assault does not require the use of force, the threat
of force, or any sexual penetration. See Ariz. Rev. Stat. § 13-
1401(5)(b)-(d). Because the statute’s elements cover instances
in which the offender acts neither violently nor aggressively,
it does not categorically fall under Begay. See United States
v. Grisel, 488 F.3d 844, 850 (9th Cir. 2007) (en banc)
(“Where . . . a state statute explicitly defines a crime more
broadly than the generic definition, no ‘legal imagination,’ is
required to hold that a realistic probability exists that the state
will apply its statute to conduct that falls outside the generic
definition of the crime.” (citing Gonzales v. Duenas-Alvarez,
549 U.S. 183, 193 (2007), which was cited with approval in
James, 550 U.S. at 208)).
The panel’s opinion attempts to explain that Arizona sexual
assault involves “violent” and “aggressive” conduct because
UNITED STATES v. TERRELL 14003
the Supreme Court told us in Coker v. Georgia, 433 U.S. 584,
597-98 (1977), that Georgia’s rape law is a violent crime. See
United States v. Terrell, 593 F.3d 1084, 1090 (9th Cir. 2010).
I concede that reasonable minds may disagree on how one
should go about applying Taylor’s categorical approach and
what exactly the Supreme Court has in mind when it repeat-
edly tells us that we are not to consider “how an individual
offender might have committed [the offense] on a particular
occasion.” Begay, 553 U.S. at 141. But surely, in examining
a particular state offense before us under the categorical
approach, no one thinks we may look to what the Supreme
Court has said about another state’s entirely different offense.
Yet that is precisely what the Terrell panel does. Arizona’s
sexual assault law covers conduct very different from the
Georgia rape statute in Coker. The Georgia statute defined
rape as “having carnal knowledge of a female, forcibly and
against her will. Carnal knowledge in rape occurs when there
is any penetration of the female sex organ by the male sex
organ.” 433 U.S. at 586 n.1 (emphases added) (internal quota-
tion marks omitted). The Arizona statute requires neither
force nor sexual penetration. See Ariz. Rev. Stat. § 13-
1401(3) (including “masturbatory contact” within the defini-
tion of “sexual intercourse”).
Nevertheless, relying on Coker and citing no Arizona cases,
the panel holds that while some cases prosecuted under the
Arizona law do not necessarily involve violence, the “typical”
case does, which is enough to create a categorical match with
Begay. Terrell, 593 F.3d at 1090. One is unable to discern
from the panel’s opinion how future panels, and more impor-
tantly future criminal defendants and their counsel, are to
determine when the “typical” offense involves “violent” and
“aggressive” conduct in the absence of an element in the state
offense requiring it.1 Mayer held that we determine the “typi-
1
Indeed, the panel’s interpretation of the role of “typicality” in this anal-
ysis draws the wrong lesson from Begay. “Typical” does not mean we can
14004 UNITED STATES v. TERRELL
cal” case by looking at whether “[m]ost of the cases” applying
the state statute involve such conduct. See Mayer, 560 F.3d at
961-62. Chief Judge Kozinski argued that Mayer’s “most of
the time” rule would prove difficult to administer. See id. at
952 (Kozinski, C.J., dissenting from the denial of rehearing en
banc). He was wrong in this case, but only because the panel
does not even bother to consult whether “most of the cases”
applying Arizona’s sexual assault offense involve “violent”
and “aggressive” conduct. It cites no Arizona cases whatso-
ever.
In any event, Mayer’s “most of the time” rule was not cor-
rect either. The Supreme Court has given us at least one
ignore the definition of the state offense. It relates to conduct covered by
the elements of the state offense “in the ordinary case,” James, 550 U.S.
at 208, rather than some hypothetical case “in which even a prototypically
violent crime might not present a genuine risk of physical injury,” id. One
would think that if the Supreme Court meant to depart in Begay from this
long-standing understanding of Taylor’s categorical approach, it would
stop relying on the same cases over and over again when explaining how
to apply it. See Begay, 553 U.S. at 141 (citing Taylor); id. (citing James,
550 U.S. at 208-09, for the proposition that an offense “is a violent felony
even if, on some occasions, it can be committed in a way that poses no
serious risk of physical harm”); James, 550 U.S. at 207-09 (explaining that
Taylor’s categorical approach does not “requir[e] that every conceivable
factual offense covered by a statute must necessarily present a serious
potential risk of injury before the offense can be deemed a violent felony”
and citing Duenas-Alvarez, 549 U.S. at 193); id. at 208 (“[T]he proper
inquiry is whether the conduct encompassed by the elements of the
offense, in the ordinary case, presents a serious potential risk of injury to
another.” (emphasis added)); Duenas-Alvarez, 549 U.S. at 193 (“[T]o find
that a state statute creates a crime outside the generic definition of a listed
crime in a federal statute requires more than the application of legal imagi-
nation to a state statute’s language. It requires a realistic probability, not
a theoretical possibility, that the State would apply its statute to conduct
that falls outside the generic definition of a crime. To show that realistic
probability, an offender, of course, may show that the statute was so
applied in his own case. But he must at least point to his own case or other
cases in which the state courts in fact did apply the statute in the special
(nongeneric) manner for which he argues.”).
UNITED STATES v. TERRELL 14005
approach for determining what type of conduct is involved in
the “typical” case that cannot be reconciled with either Mayer
or Terell. In Chambers v. United States, 129 S. Ct. 687, 692
(2009), the government attempted to show that typically, fail-
ure to report to a penal institution involves violence and
aggression by citing three cases clearly associating such vio-
lence and aggression with the offense. The Court rejected this
flimsy evidence, relying instead on a Sentencing Commission
report identifying every federal case over a period of two
years. Id.
Chambers teaches that we should rely on empirical data
when determining whether an offense “typically involve[s]
purposeful, violent, and aggressive conduct.” Begay, 553 U.S.
at 144-45 (internal quotation marks omitted).2 Without such
objective evidence we are left with nothing but a panel’s gut
instinct as to what constitutes a violent felony; or worse, com-
parison to inapposite statutes from other states. The law of our
circuit, when it comes to the ACCA, will turn on the arbitrary
composition of the panel hearing each case. This can’t be
what Congress intended when it passed a uniform sentencing
scheme for armed career offenders.
Terrell’s license to ignore objective indicia as to whether
the offense typically involves violent and aggressive conduct
has now been officially sanctioned by our court. This will
only lead to more decisions like it. Already, our court has held
that a theft statute encompassing pickpocketing constitutes a
“crime of violence.” See United States v. Alderman, 601 F.3d
949, 953-54 (9th Cir. 2010); see also State v. Johnson, 97
Wash. App. 1055 (Ct. App. 1999) (per curiam) (unpublished)
(upholding a conviction under the theft statute at issue in
Alderman where one teenager took twenty dollars from
another teenager without his permission); see also Thomas W.
2
At the very least, we should be pointing to an illustrative list of cases,
from the jurisdiction in question, to show that the state offense is the kind
of serious felony Congress intended to merit sentencing enhancement.
14006 UNITED STATES v. TERRELL
Hutchison et al., Federal Sentencing Law and Practice 1409
(2010) (questioning whether offenses such as pickpocketing
should be considered a “crime of violence” under the residual
clause absent empirical evidence that the offense will result in
physical injury).
The insidious nature of the panel’s treatment of the residual
clause is even clearer in the panel’s treatment of burglary
under Arizona and Missouri law. The panel recognized that
burglary under Arizona and Missouri law is defined more
broadly than generic burglary. Neither burglary statute limits
itself to unmovable structures, which is what the generic defi-
nition requires. See Grisel, 488 F.3d at 848-49. Arizona’s
statute does not even require unlawful entry. See Taylor, 495
U.S. at 598. Despite these categorical differences, the panel
holds that a state burglary offense, no matter how much it
may stray from the generic definition, is categorically “almost
always” a violent felony. See Terrell, 593 F.3d at 1095. Our
refusal to rehear this case means that we are developing a
“catch all categorical approach,” one which renders superflu-
ous the question whether the state offense meets the enumer-
ated offense’s generic definition. See also United States v.
Matthews, 374 F.3d 872, 876 (9th Cir. 2004) (“[E]ven though
certain conduct (whether defined categorically or by a modi-
fied categorical approach) may present a serious potential risk
of physical injury to another, it may still be excluded under
the [residual] clause if including it would render the inclusion
of the [enumerated offense] surplusage.”).
If the state offense of conviction is missing critical ele-
ments of the enumerated offense’s generic definition that
make the offense inherently violent, the state offense clearly
cannot “almost always” categorically involve the same pur-
poseful, violent, and aggressive conduct as the enumerated
offense. When the Supreme Court held in Begay that the enu-
merated offenses all typically involve purposeful, violent, and
aggressive conduct, it did so in light of those offenses’
generic definitions. See Begay, 553 U.S. at 145 (generically
UNITED STATES v. TERRELL 14007
defining burglary, arson, and extortion). In other words, a bur-
glar who commits generic burglary, i.e., who enters a building
unlawfully with intent to commit a crime, evinces “purpose-
ful,” “violent,” and “aggressive” conduct. That’s not necessar-
ily true for a burglar who enters lawfully. Nobody would
suggest that shoplifting is violent and aggressive, but the
panel decision will likely send some felons to prison for 15
years because they got caught shoplifting in the past. See
State v. Belcher, 776 P.2d 811, 812 (Ariz. Ct. App. 1989)
(“The requisite intent to commit burglary may be formed after
a person enters a store in all innocence.”).
The panel reasons that the distinction between lawful and
unlawful entry is inconsequential, since the Court in James
held that attempted burglary, which requires no entry at all,
posed the same risks as generic burglary. See Terrell, 593
F.3d at 1094-95. But the panel’s cursory analysis ignores the
fundamental role of this particular missing element. In James,
the Court noted that “[a]ll burglaries begin as attempted bur-
glaries,” and explained that attempted burglaries—that is,
attempts to enter a building unlawfully—where the offender
is apprehended before he completes the burglary, may present
an even greater risk than the typical completed burglary.
James, 550 U.S. at 204. For support, the Court relied on the
fact that the Sentencing Guidelines include attempt crimes in
its definition of a “crime of violence,” a “judgment . . . based
on the Commission’s review of empirical sentencing data and
presumably reflect[ing] an assessment that attempt crimes
often pose a similar risk of injury as completed offenses.” Id.
at 206.3 In other words, the Court in James reasoned that
3
The Court’s reliance on empirical data provides further support for the
notion that we should be employing such information in determining
which crimes should be categorized as violent felonies. Indeed, relying on
a 1992 opinion by then-Chief Judge Breyer, the James Court explained
that “the Commission, which collects detailed sentencing data on virtually
every federal criminal case, is better able than any individual court to
make an informed judgment about the relation between a particular
offense and the likelihood of accompanying violence.” 550 U.S. at 206
(internal quotation marks and brackets omitted).
14008 UNITED STATES v. TERRELL
attempting to enter unlawfully perhaps poses an even greater
risk of injury than entering unlawfully and seeing the burglary
through to its completion. It doesn’t support the panel’s the-
ory, which we’ve now endorsed by refusing to rehear this
case. Unlike the Court in James, the panel here provides no
“empirical sentencing data” to support its view that Arizona’s
burglary offense is “even more like generic burglary than the
statute in James.” Terrell, 593 F.3d at 1095. All we have is
the panel’s gut instinct.
II
The panel’s decision does not just allow future panels to
ignore the presence of the ACCA’s enumerated offenses. It
also does the same for those crimes the Sentencing Commis-
sion has listed as prototypical “crime[s] of violence” under
the Sentencing Guidelines.
The Sentencing Guidelines define a “crime of violence”
nearly identically to the ACCA’s definition of “violent felo-
ny.” Compare 18 U.S.C. § 924(e)(2)(B), with U.S.S.G.
§ 4B1.2(a). As a result, we have always interpreted the two
provisions “in a parallel manner,” see, e.g., United States v.
Jennings, 515 F.3d 980, 990 n.11 (9th Cir. 2008), and recently
held that “the same analysis the Court applied in Begay” con-
trols whether an offense constitutes a “crime of violence”
under Guideline § 4B1.2, United States v. Coronado, 603
F.3d 706, 710 (9th Cir. 2010). Indeed, as discussed above, the
Court has turned to the definition of “crime of violence” in
the Sentencing Guidelines for support in determining what
constitutes a “violent felony” under the ACCA. See James,
550 U.S. at 206-07.
In addition to the four enumerated crimes that appear at the
beginning of the residual clause, the Application Notes to sec-
tion 4B1.2 enumerate additional “crime[s] of violence.” See
U.S.S.G. § 4B1.2 cmt. n.1. The Guidelines provide:
UNITED STATES v. TERRELL 14009
“Crime of violence” includes murder, manslaughter,
kidnapping, aggravated assault, forcible sex offenses,
robbery, arson, extortion, extortionate extension of
credit, and burglary of a dwelling. Other offenses are
included as “crimes of violence” if (A) that offense
has as an element the use, attempted use, or threat-
ened use of physical force against the person of
another, or (B) the conduct set forth (i.e., expressly
charged) in the count of which the defendant was
convicted involved use of explosives (including any
explosive material or destructive device) or, by its
nature, presented a serious potential risk of physical
injury to another.
U.S.S.G. § 4B1.2 cmt. n.1 (emphasis added).
Following the Supreme Court’s ejusdem generis analysis in
Begay, we must read these offenses as “limiting the crimes
that [the residual clause] covers to crimes that are roughly
similar, in kind as well as in degree of risk posed, to the
examples themselves.” Begay, 553 U.S. at 143. The the list of
enumerated offenses “indicates that the statute covers only
similar crimes, rather than every crime that ‘presents a serious
potential risk of physical injury to another.’ ” Id. at 142.
In its list of offenses, the Sentencing Commission tells us
that “forcible sex offenses” are to be considered crimes of
violence. In United States v. Beltran-Munguia, 489 F.3d
1042, 1051 (9th Cir. 2007), we held that the term “forcible
sex offenses” in the Sentencing Guidelines requires the use of
force. There, we examined a conviction for sexual abuse
under Oregon law. We held that it was not a conviction for
a “crime of violence” because the Oregon offense, like the
Arizona offense at issue in Terrell, required merely a lack of
consent, which could be proven if the victim was mentally
defective, mentally incapacitated, or physically helpless. Id. at
1046. Although our court later held that “[f]orcible sex
offenses . . . require more force than that inherent to penetra-
14010 UNITED STATES v. TERRELL
tion but need not require violent force,” United States v.
Bolanos-Hernandez, 492 F.3d 1140, 1146 (9th Cir. 2007), we
still require a showing that the defendant has used or
attempted at least some level of force on the victim,” id. Ari-
zona requires neither force—violent or not—nor penetration.
The panel thus introduces a serious incongruity between
our circuit law on the ACCA’s residual clause and the Sen-
tencing Guidelines: To qualify as a “crime of violence,”
“forcible sex offenses” must involve at least some level of
force; yet as a result of the panel’s opinion, sex offenses that
do not require any force will somehow qualify as both “vio-
lent felonies” and “crimes of violence” because they “other-
wise involve[ ] conduct that presents a serious potential risk
of physical injury to another.” 18 U.S.C. § 924(e)(2)(B)(ii);
U.S.S.G. § 4B1.2(a)(2). Such a result is both irrational, and
irreconcilable with Begay. There the Court’s majority “re-
jected the approach of the dissent, which would have analyzed
the risk of harm posed by the behavior without regard for
whether the defendant’s crime involved ‘purposeful, violent,
and aggressive’ conduct.” United States v. Christensen, 559
F.3d 1092, 1095 (9th Cir. 2009).
In contrast, the holdings of Beltran-Munguia and Bolanos-
Hernandez are entirely consistent with Begay and James. For
a sexual offense to constitute a “crime of violence” it must
include “at least some level of force on the victim.” Bolanos-
Hernandez, 492 F.3d at 1146. Such a requirement ensures that
the offender engaged in “purposeful, violent, and aggressive
conduct,” and “effectuate[s] Congress’ purpose to punish only
a particular subset of offender, namely career criminals.”
Begay, 553 U.S. at 147. And as the Court explained in James
with respect to attempt crimes, the Commission’s decision to
include “forcible sex offenses” as the type of sexual offense
it considers to be a “crime of violence” reflects its considered
view that “forcible sex offenses” involve the requisite degree
of risk and accompanying violence. See James, 550 U.S. at
206-07.
UNITED STATES v. TERRELL 14011
While the residual clause covers crimes not enumerated in
the statute, Begay makes clear that the list of enumerated
offenses “indicates that the statute covers only similar
crimes.” 553 U.S. at 142. An offense that does not require
force, the threat of force, or sexual penetration, is simply not
similar to a “forcible sex offense.”
III
The panel has effectively rendered the residual clause a
catch all clause, potentially turning every recidivist into a vio-
lent, armed career criminal for sentencing purposes. The cate-
gorical approach we once used uniformly has been expanded
beyond recognition, without regard to the offenses meant to
limit its scope. The tail, according to the panel, now officially
wags the dog. I respectfully dissent.