SYKES, Circuit Judge.

Nathaniel Capler pleaded guilty to two drug offenses and was sentenced to 141 months' imprisonment based in part on an Illinois conviction for unlawful restraint that the district court deemed to be a "crime of violence" within the meaning of U.S.S.G. § 4B1.2. See 720 ILL. COMP. STAT. 5/10-3. On appeal he challenges only the district court's conclusion that his prior conviction is a crime of violence.1

We affirm. The district court's ruling finds direct support in our decisions in United States v. Wallace, 326 F.3d 881, 887 (7th Cir.2003), and United States v. Billups, 536 F.3d 574, 584 (7th Cir.2008). Wallace analyzed the Illinois statute in question and concluded that the offense is a crime of violence. Billups reached the same conclusion about the closely analogous crime of false imprisonment in Wisconsin. Notwithstanding Capler's efforts to challenge the validity of Wallace and Billups, these decisions remain sound and together control the outcome of this appeal.

I. Background

Capler sold crack cocaine to a police informant. He was arrested and pleaded guilty to two counts of distribution. See 21 U.S.C. § 841(a)(1). The district court sentenced him as a career offender to a total of 141 months' imprisonment. See U.S.S.G. § 4B1.1. The career-offender designation is based on two Illinois convictions, one for delivery of a controlled substance and the other for unlawful restraint. Capler's appointed lawyer initially cast the appeal as frivolous and moved to withdraw under Anders v. California, 386 U.S. 738, 87 S.Ct. 1396, 18 L.Ed.2d 493 (1967), but we denied that motion after concluding that an appeal asking us to revisit Wallace would not be frivolous. Capler's attorney has now briefed that argument on the merits.

In deciding if a conviction is for a crime of violence, we look to the statutory elements and the manner in which the offense ordinarily is committed. United States v. Sonnenberg, 628 F.3d 361, 364 (7th Cir.2010); United States v. Dismuke, 593 F.3d 582, 591 (7th Cir.2010), petition for cert. filed, 79 U.S.L.W. 3062 (U.S. July 19, 2010) (No. 10-109). As relevant here, a "crime of violence" is any offense punishable by imprisonment exceeding one year that "has as an element the use, attempted use, or threatened use of physical force against the person of another," U.S.S.G. § 4B 1.2(a)(1), or "is burglary of a dwelling, arson, or extortion, involves use of explosives, or otherwise involves conduct that presents a serious potential risk of physical injury to another," id. § 4B1.2(a)(2). In Illinois a person "commits the offense of unlawful restraint when he or she knowingly without legal authority detains another." 720 ILL. COMP. STAT. 5/10-3(a) (2010). The use or threatened use of force is not an element, see People v. Brials, 732 N.E.2d 1109, 1119 (2000); People v. Bowen, 241 Ill.App.3d 608, 182 Ill.Dec. 43, 609 N.E.2d 346, 361 (1993), so the crime as defined in Illinois does not satisfy subsection (a)(1). Our analysis thus shifts to subsection (a)(2).

Unlawful restraint is not one of the specific crimes listed in the first part of § 4B1.2(a)(2), so to count as a violent felony, the crime must fall within the residual clause, which covers offenses that "otherwise involve[] conduct that presents a serious potential risk of physical injury to another." This requires a determination that the offense is "roughly similar" to those that are specifically enumerated. See Begay v. United States, 553 U.S. 137, 143, 128 S.Ct. 1581, 170 L.Ed.2d 490 (2008); United States v. Woods, 576 F.3d 400, 404 (7th Cir.2009). Before the Supreme Court decided Begay, we had analyzed the residual clause by asking only whether the crime in question typically encompasses conduct creating a risk of injury similar to the enumerated crimes. See, e.g., United States v. Mathews, 453 F.3d 830, 836-37 (7th Cir.2006); United States v. Franklin, 302 F.3d 722, 723 (7th Cir.2002). Begay added a second level of analysis, a similar-in-kind inquiry, which asks whether the crime as ordinarily committed reflects the same "purposeful, violent, and aggressive" conduct as the listed crimes. 553 U.S. at 144-45, 128 S.Ct. 1581. Only if a crime is similar both in kind and degree of risk posed can it qualify as a crime of violence under § 4B 1.2(a)(2). Id.

After Begay we held in Woods that when the statute in question is divisible—when it defines alternative means of committing a crime, some violent and some not—the district court may expand its categorical inquiry by examining a limited set of additional materials to determine whether the defendant was convicted of the violent version of the crime. See Woods, 576 F.3d at 403. The Illinois statute does not define alternative means of committing the crime of unlawful restraint and thus is not "divisible" as we explained the concept in Woods, id. at 404-07. The offense is a crime of violence, then, only if it is categorically similar to burglary, arson, extortion, and crimes involving the use of explosives. See Begay, 553 U.S. at 144-45, 128 S.Ct. 1581; Dismuke, 593 F.3d at 594; United States v. Hart, 578 F.3d 674, 680-81 (7th Cir.2009); United States v. Evans, 576 F.3d 766, 768-69 (7th Cir. 2009). This similarity was found to be present in Wallace, 326 F.3d at 887, a pre-Begay decision concluding that a violation of the Illinois unlawful-restraint statute generally will carry a risk of violence. We have not reevaluated that position since Begay, but we have concluded post-Begay that the analogous crime of false imprisonment as defined by Wisconsin statute is categorically violent. See WIS. STAT. § 940.30; Billups, 536 F.3d at 584.

II. Discussion

The sole issue in this appeal is whether unlawful restraint as defined by Illinois law falls within § 4B1.2(a)'s residual clause for crimes roughly similar to the enumerated offenses. Capler argues that the statute encompasses too much nonviolent conduct for it to be categorically labeled as a violent felony. Although Capler concedes that Wallace and Billups are on point, he contends that both decisions are incorrect and urges us to reconsider them in light of more recent opinions from this court. Because Capler preserved this argument for appeal, our review is plenary. See United States v. Clinton, 591 F.3d 968, 972 (7th Cir.2010), cert. denied, ___ U.S. ___, 131 S.Ct. 246, 178 L.Ed.2d 164, 79 U.S.L.W. 3202 (2010).

In Wallace we held that unlawful restraint as defined in Illinois is a violent felony for purposes of the Armed Career Criminal Act, see 18 U.S.C. § 924(e)(2)(B), which tracks the language in § 4B1.2(a). Wallace, 326 F.3d at 887. Applying the pre-Begay categorical approach, we reasoned that unlawful restraint generally precipitates a risk of violence, whether in the initial restraint or a resulting confrontation between victim and assailant. Id. Although we did not expressly conclude that unlawful restraint is "similar in kind" to the crimes enumerated before the residual clause (that requirement was added in Begay), our analysis of the risk of injury strongly implies that this sort of categorical similarity is present. Id.

Our decision in Wallace, then, is not a dead letter. If anything it anticipated Begay even though we did not articulate the analysis in precisely the same terms as the standard later announced by the Supreme Court. Like any precedential decision, Wallace stands until we are given good reason to overrule it. See United States v. Sykes, 598 F.3d 334, 338 (7th Cir.2010), cert. granted, ___ U.S. ___, 131 S.Ct. 63, 177 L.Ed.2d 1152, 79 U.S.L.W. 3194 (2010); Tate v. Showboat Marina Casino P'ship, 431 F.3d 580, 583 (7th Cir.2005). We invited Capler to identify a reason, but his response is simply to write off Wallace on the ground that the opinion was issued five years before Begay. Instead, he mostly focuses on trying to overcome Billups, but that strategy is shortsighted. Capler has not offered a principled basis to overturn Wallace, and as we discuss below, his lengthy effort to indirectly undermine Wallace by challenging Billups is unpersuasive.

In Billups we made explicit what we left to inference in Wallace: Restraining another against his will, apart from carrying a serious risk of injury, is an aggressive act categorically similar to the crimes enumerated in § 4B 1.2(a)(2). The Wisconsin statute analyzed in Billups provides that "[w]hoever intentionally confines or restrains another without the person's consent and with knowledge that he or she has no lawful authority to do so is guilty of a Class H felony." WIS. STAT. § 940.30. A separate chapter of the code, § 939.22(48), defines "without consent" to mean "no consent in fact" or consent given because the assailant threatened violence, or because the victim did not give knowing consent, either by reason "of ignorance or mistake of fact or of law other than criminal law or by reason of youth or defective mental condition, whether permanent or temporary." We examined the elements of § 940.30 and the definition of "without consent" in concluding that the conduct encompassed by the statute is categorically violent. Billups, 536 F.3d at 578-79.

Billups read the Wisconsin statute to reach five types of victims. The first four variations of the offense—a nonconsenting victim, a victim who consented to the confinement out of fear, a victim who consented because she believed the assailant had authority to confine her, and a victim who was tricked into consenting—carry a substantial risk of physical injury because each scenario sets the stage for a violent confrontation between victim and assailant. Id. at 580-81. In the fifth variation where the victim is legally incapable of giving consent, we recognized that the risk of confrontation, while lower, is not insubstantial because the potential remains that the victim could be injured trying to escape. We never decided whether the fifth variation involves a serious risk of injury because we were satisfied that a crime with four out of five dangerous modes of commission is categorically violent. Id. at 582.

Moreover, we distinguished a subsection of Indiana's criminal-confinement statute that had been held to be a nonviolent felony. See United States v. Gilbert, 464 F.3d 674, 676-82 (7th Cir.2006). The Indiana offense criminalizes removal from one place to another by trick, see IND. CODE § 35-42-3-3(a)(2), and in Billups we were unwilling to assume—at least without empirical evidence—that deceiving someone into changing locations poses the same risk as involuntary confinement. Billups, 536 F.3d at 582. We went on to observe that the Wisconsin statute criminalizes conduct that is "inherently more coercive" than the conduct within the sweep of the Indiana statute, since in Wisconsin the victim must be confined, not merely tricked into moving. On this point our opinion in Billups cites approvingly to Wallace and notes that the crime of unlawful restraint in Illinois is "akin" to false imprisonment as defined by Wisconsin. Id.

Because Billups analyzes a statute that is closely analogous to the Illinois crime of unlawful restraint, the reasoning in that decision provides post-Begay support for our holding in Wallace. The Wisconsin statute analyzed in Billups is substantively indistinguishable from its Illinois counterpart. Compare 720 ILL. COMP. STAT. 5/10-3(a) (2010) ("A person commits the offense of unlawful restraint when he or she knowingly without legal authority detains another."), with WIS. STAT. § 940.30 ("Whoever intentionally confines or restrains another without the person's consent and with knowledge that he or she has no lawful authority to do so is guilty of a Class H felony."). One minor difference is that Wisconsin makes nonconsent an express requirement and defines the phrase by statute, while the Illinois offense is silent about consent. Illinois caselaw, however, shows that nonconsent is an implicit requirement, its contours shaped by state common law that appears to adopt a common-sense understanding of the term. See People v. Hunt, 150 Ill.App.3d 874, 104 Ill.Dec. 201, 502 N.E.2d 464, 466-67 (1986) (reversing conviction for unlawful restraint where State failed to prove defendant's knowledge that victim was too intoxicated to effectively consent); People v. Kittle, 140 Ill.App.3d 951, 95 Ill.Dec. 260, 489 N.E.2d 481, 484-85 (1986) (concluding that unlawful restraint is a lesser-included crime of kidnapping by fraud or deceit); People v. Warner, 98 Ill.App.3d 433, 53 Ill.Dec. 956, 424 N.E.2d 747, 749-50 (1981) (upholding conviction for unlawful restraint where restraint of children was achieved by threat of physical punishment); cf. 720 ILL. COMP. STAT. 5/12-17(a) (defining consent in context of sexual assault as "freely given agreement"). But if less concrete, the definition of consent in Illinois does not appear to be any broader than that in Wisconsin; nor does the crime of unlawful restraint appear to cover any forms of nonviolent conduct that would not also be covered by the Wisconsin statute.

Furthermore, Billups stands alone as the only post-Begay decision from this court involving criminal confinement, and its holding is consistent with previous decisions interpreting similar crimes. See United States v. Franco-Fernandez, 511 F.3d 768, 771-72 (7th Cir.2008) (distinguishing nonviolent state offense of child abduction by putative father from violent offense of unlawful restraint); Gilbert, 464 F.3d at 676-82 (declining to apply Wallace to subsection of Indiana's criminal-confinement statute involving action not necessarily taken without victim's consent); United States v. Hagenow, 423 F.3d 638, 644 (7th Cir.2005). The holding in Billups also falls in line with the weight of authority from other circuits. Before Begay several circuits had examined similar statutes and uniformly characterized them as violent crimes. United States v. Riva, 440 F.3d 722, 723-25 (5th Cir.2006) (holding that Texas crime of unlawfully restraining person under age 17 is crime of violence under residual clause of § 4B 1.2(a)(2) because offense involves serious potential risk of physical injury); Dickson v. Ashcroft, 346 F.3d 44, 49-50 (2d Cir.2003) (holding that New York offense of unlawful imprisonment of competent adult is crime of violence within meaning of Immigration and Nationality Act); United States v. Zamora, 222 F.3d 756, 764-65 (10th Cir.2000) (holding that New Mexico crime of false imprisonment is violent felony under § 4B 1.2(a)(2) because of substantial risk of violence); United States v. Kaplansky, 42 F.3d 320, 324 (6th Cir.1994). After Begay no out-of-circuit published opinion has evaluated a statute comparable to the Illinois crime of unlawful restraint, but the Third Circuit in two unpublished opinions has kept with the trend of characterizing criminal confinement as a categorically violent crime. See United States v. McMiller, 376 Fed.Appx. 199, 202-03 (3d Cir. 2010); United States v. Smith, 284 Fed. Appx. 943, 944-45 (3d Cir.2008).

Capler, not surprisingly, seeks to overturn Wallace by arguing that Billups too was wrongly decided, albeit after Begay. Specifically, he contends that Billups should be disregarded because we decided it on the incorrect assumption that the Wisconsin statute at issue was indivisible; that our opinion in Billups improperly applied the Begay analysis and offered only conclusory support for its holding; that subsequent decisions in this circuit applying the crime-of-violence inquiry call Billups into question; and that Billups should be reassessed in light of the Supreme Court's decision in Chambers v. United States, 555 U.S. 122, 129 S.Ct. 687, 692-93, 172 L.Ed.2d 484 (2009).

Before addressing Capler's challenges to Billups, we must resolve a threshold dispute over how to approach the question of what conduct typifies the "ordinary case" of unlawful restraint. Capler insists that the "ordinary case" of unlawful restraint should encompass only separately punishable instances of the offense—that is, it should exclude cases in which a judgment of conviction for unlawful restraint cannot be entered because the conduct at issue supports a separate conviction for a more serious crime, thus making unlawful restraint a lesser-included offense. See, e.g., Brials, 247 Ill.Dec. 777, 732 N.E.2d at 1119; Kittle, 95 Ill.Dec. 260, 489 N.E.2d at 485. This argument misunderstands the categorical approach, which asks whether the elements of the offense ordinarily involve conduct that falls within the scope of the residual clause. See James v. United States, 550 U.S. 192, 202, 127 S.Ct. 1586, 167 L.Ed.2d 532 (2007); Sonnenberg, 628 F.3d at 365. Whether in the ordinary case the offense conduct can support a separate punishment is of no consequence; courts look instead at the conduct encompassed within the statutory definition of the crime to determine what an ordinary case is like. See United States v. Goodpasture, 595 F.3d 670, 672 (7th Cir.2010) ("[T]he Supreme Court's question is what the offense as generally committed entails, not which offenses lead to criminal prosecutions.").

Capler holds little back in his attack on Billups. As a preliminary matter, he contends that the Billups analysis is flawed from the outset because in applying the categorical approach, we did not treat the false-imprisonment statute as divisible. His point has some appeal: Wisconsin law, which requires that false imprisonment occur without the consent of the victim, provides four definitions for "without consent" and so would appear to create several divisible means of committing false imprisonment. WIS. STAT. § 939.22(48). But even if Billups was mistaken in analyzing the Wisconsin statute as a whole instead of focusing on the particular subsections of the definition of "without consent," our holding in Billups—that false imprisonment as generally committed carries a risk of violence—can still inform our analysis. The Illinois crime of unlawful restraint is indivisible; that is, it resembles the Wisconsin false-imprisonment statute as we conceptualized it in Billups.

Capler contends that Billups bungled the Begay inquiry—both in terms of its assessment of the crime's similarity to the listed offenses in the residual clause and its assessment of the risks posed by unlawful restraint. With respect to the similar-in-kind inquiry, Capler contends that Billups offers only conclusory support for the holding that unlawful restraint typically involves violent and aggressive behavior. Not so. In Billups we concluded that restraining a person against his will generally creates a significant risk of violence, even where restraint is not accomplished by force. 536 F.3d at 580-81. That determination informed the similar-in-kind analysis just as it did the risk-of-injury analysis. See Dismuke, 593 F.3d at 594 (explaining that categorically violent felonies are those that carry a genuine risk of violence); United States v. Patterson, 576 F.3d 431, 442 (7th Cir.2009), cert. denied, ___ U.S. ___, 130 S.Ct. 1284, 175 L.Ed.2d 1078, 78 U.S.L.W. 3438 (2010) (explaining that risk of violence need only "attend" a violation of predicate crime for purposes of crime-of-violence inquiry). Describing the Wisconsin offense as "inherently. . . coercive," our decision in Billups goes on to draw parallels between false imprisonment and the listed crimes of extortion and burglary—crimes that can be committed in a nonviolent manner and yet still carry a substantial risk of harm. 536 F.3d at 583.

Capler insists that recent precedent in this circuit calls into question the similar-in-kind analysis. In support he relies on several recent cases in which we vacated crime-of-violence determinations because the predicate statutes covered nonviolent conduct. See Goodpasture, 595 F.3d at 672; Hart, 578 F.3d at 679; Evans, 576 F.3d at 769; United States v. Templeton, 543 F.3d 378, 384 (7th Cir.2008). None of these cases, however, addresses unlawful restraint; nor does any case purport to modify the legal inquiry followed in Billups. These cases do not, as Capler suggests, support the proposition that an indivisible statute can never be categorized as a violent crime if it encompasses some nonviolent conduct. Rather, the predicate offenses in these cases simply involved relatively passive conduct, conduct too dissimilar to warrant comparison with unlawful restraint. For example, Templeton and Hart involved escape offenses that cover such passive, nonaggressive conduct as walking away from nonsecure custody, see Templeton, 543 F.3d at 381; Hart, 578 F.3d at 677, and Goodpasture examined a statute prohibiting lewd and lascivious conduct involving a person under fourteen, see Goodpasture, 595 F.3d at 672. And though the aggravated-battery crime deemed nonviolent in Evans would appear closer to the mark, the outcome of that case is misleading because the government gave it away by failing to argue that the predicate crime of conviction (insulting or provoking physical contact with a pregnant woman) is characteristically violent and aggressive. See Evans, 576 F.3d at 769.

As for the second part of the Begay inquiry that focuses on the potential for physical injury, Capler argues that Billups and Wallace should be reconsidered in light of the Supreme Court's decision in Chambers, 129 S.Ct. at 692-93. Noting that Chambers places an emphasis on empirical data in the analysis of potential risk, Capler appears to contend that Chambers calls into question the validity of any evaluation conducted without benefit of statistics. But Capler can point to no language in Chambers requiring the use of empirical data; nor does he cite a federal decision interpreting Chambers as doing so. The Supreme Court in Chambers may have encouraged the use of statistical data in determining whether an offense carries a risk of injury, but the Court does not require it. See United States v. Alexander, 609 F.3d 1250, 1256-57 (11th Cir. 2010); see also Sonnenberg, 628 F.3d at 366 ("[I]n the absence of detailed statistical information that is unlikely to be available, the best we can do is use common sense and experience to determine the variety of crimes a statute would cover. . . ."); Welch v. United States, 604 F.3d 408, 424 (7th Cir.2010), petition for cert. filed, 79 U.S.L.W. 3129 (Sept. 1, 2010) (No. 10-314) (post-Chambers case conducting risk-of-injury analysis without data); Patterson, 576 F.3d at 442 (same).

Capler maintains that Chambers dislodged certain precedents underlying this court's holding in Wallace, casting further doubt on the decision. In Wallace we cited two escape cases, United States v. Franklin, 302 F.3d 722, 725 (7th Cir.2002), and United States v. Bryant, 310 F.3d 550, 554 (7th Cir.2002), as examples of categorically violent offenses that encompass potentially harmful conduct. See Wallace, 326 F.3d at 887. Our focus was on the crime of escape and the dangers attendant to fleeing confinement. Id. The holding in Chambers, however, concerned only the crime of failure to report—an offense the Supreme Court distinguished from traditional escape as being far less likely to result in harm. Chambers, 129 S.Ct. at 691.

As yet another basis for attacking Wallace, Capler contends that the decision is inconsistent with Hagenow, 423 F.3d at 644, which holds that one part of Indiana's criminal-confinement statute is not a violent felony. This argument has no merit. We addressed the apparent tension between Wallace and Hagenow in Gilbert, 464 F.3d at 682, and then again in Billups, 536 F.3d at 582. These decisions foreclose the need for further discussion.

In the end we are left with no reason to question the continuing validity of Wallace and Billups. As a faithful application of Begay to a closely analogous crime, Billups lends support to our holding here. But our lengthy defense of Billups should not distract focus from Wallace, the only decision in this circuit that has analyzed the Illinois crime of unlawful restraint, and accordingly, the decision that controls the outcome of this appeal. For the reasons we have explained, we are satisfied that Wallace survived Begay, that the decision's analysis is in harmony with the present-day approach of the Supreme Court, and that its holding remains sound.

The judgment of the district court is AFFIRMED.