dissenting:
The majority has reconsidered its earlier decision and holds that the record before us does not establish that Malta-Espinoza pleaded guilty to conduct constituting a crime of violence, as defined by 18 U.S.C. § 16, when he pleaded guilty to stalking his female victim, Ms. Alma Espisito. In so holding, the majority has essentially decided that courts cannot rely on the words of a defendant’s guilty plea to mean exactly what they say. I respectfully dissent.
The Amended Felony Complaint against Malta-Espinoza charged, among of other things, that while under a prior restraining order prohibiting such conduct, Malta-Espinoza:
Maliciously and repeatedly follow[ed] and harassed] ALMA ESPISITO, and made a credible threat with the intent that she be placed in reasonable fear for her safety and the safety of her immediate family.
(Record on Appeal at 126)(emphasis added). This was the only count in the Amended Felony Complaint to which Malta-Espinoza pleaded guilty. The other counts were dismissed pursuant to his plea deal with the Government. An Abstract of the Judgment reflects the terms of his guilty plea. The majority holds that these documents, considered together, are not enough to establish that Malta-Espinoza both followed and harassed Ms. Espisito. Instead, they have found that when a defendant pleads guilty to a count he does not actually admit to the facts contained therein. They have in effect substituted their own findings of fact for those set forth in the record.
Defendants do not plead guilty to more than they must. It goes against well-established precedent to hold that a plea of guilty is not an admission to all of the facts contained in the charge. See United States v. Harris, 108 F.3d 1107, 1109 (9th Cir.1997) (citing United States v. Mathews, 833 F.2d 161, 164 (9th Cir.1987)(“a guilty plea conclusively proves the factual allegations contained in the indictment”)); see also United States v. Velasco-Medina, 305 F.3d 839, 851-53 (9th Cir.2002)(holding an Abstract of the Judgment and the charging papers, considered together, were sufficient to prove the facts alleged). The majority holds that this is not the case when the charge is phrased conjunctively. The authority the majority cites, however, is clearly distinguishable. The majority relies on United States v. Bonanno, 852 F.2d 434, 441 (9th Cir.1988), which holds that where a statute specifies two or more ways in which an offense may be committed, all may be alleged in the conjunctive in one count and a jury’s finding of proof of any one of those acts is sufficient to find guilt. This is different than a guilty plea, which is an “admission that [the defendant] committed the crime charged against him.” United States v. Broce, 488 U.S. 563, 570, 109 S.Ct. 757, 102 L.Ed.2d 927 (1989) (citations omitted).
By entering a plea of guilty, the accused is stating that he did the discrete acts described in the indictment and is admitting guilt of a substantive crime. See id. Because of the clear difference and its import, I cannot agree with the majority.1
*1086It is then necessary to determine if stalking by maliciously and repeatedly following and harassing a person, and making a credible threat with the intent that the person be placed in reasonable fear for her safety and the safety of her immediate family while a prior restraining order prohibiting such conduct is a crime of violence under 18 U.S.C. § 16(b).2
Subsection 16(b) defines a crime of violence as:
(b) any other offense that is a felony and that, by its nature, involves a substantial risk that physical force against the person or property of another may be used in the course of committing the offense.
18 U.S.C. § 16(b)(emphasis added). The majority’s analysis considers both the categorical and modified categorical approach. While the parties have agreed that stalking is not a crime of violence under the categorical approach, consideration of that approach is instructive when applying the modified categorical approach — particularly in light of the logic set forth in United States v. Becker, 919 F.2d 568, 573 (9th Cir.1990) (holding residential burglaries are categorically crimes of violence because “[t]he confluence of common sense and precedent lead to the conclusion that the unauthorized daytime entry of the dwelling of another with the intent to commit a larceny or any felony carries with it a substantial risk that force will be used against the person or property of another”).
In Becker, and the cases that follow it, this Circuit rightly considered the “substantial risk” requirement set forth in § 16(b). The Becker court did not require an absolute certainty of physical force, as the majority seems to require here. Even though physical force is not a necessary element of residential burglary, “[a]ny time a burglar enters a dwelling ... there is a risk that in the course of committing the crime he will encounter one of its lawful occupants, and use physical force against that occupant....” Becker, 919 F.2d at 571; see also United States v. M.C.E., 232 F.3d 1252, 1255-56 (9th Cir.2000) (holding that under the categorical approach residential burglary is a crime of violence because of the risk of a violent confrontation).
The majority notes that stalking can occur at a distance and implies that in those instances there is a small risk of physical force being used against the victim. The same is true of burglary. A burglar may enter an empty dwelling while its inhabitants are traveling far from home. In fact, it is generally the intention of most burglars to enter undetected and remove valuables without ever facing the occupants of the dwelling. Stalking is a far more personal crime. The stalker often seeks a sense of power or control over his victim, usually through fear. See National Institute of Justice and Centers for Disease Control and Prevention, Stalking in America: Findings From the National Violence Against Women Survey, Research in Brief (April 1998) at 8; see also U.S. Dep’t of Justice, Stalking and Domestic Violence, Report to Congress (May 2001) at 22 (stating, generally, the motive *1087for stalking is not necessarily sexual; stalkers are motivated by anger or hostility towards the victim and a desire to control the victim). And, because stalking, by definition, requires repeated victimization, it is intuitive that there is an increased opportunity for violence.
Research on stalking has found that violence against a person occurs in 30-50% of stalking cases. See Barry Rosenfeld, Violence Risk Factors in Stalking and Obsessional Harassment, 31 CRIMINAL Justioe ÁND Behavior 9, 31 (2004) (performing a meta-analysis of eight leading studies of stalking). It is likely these figures actually underestimate the rate of violence associated with stalking because cases in which stalking has turned violent are often identified solely by the more serious crime such as homicide, rape or assault. See id. at 11. The National Center for Victims of Crime reports that “46% of stalking victims experience one or more violent incidents by the stalker.” National Center for Victims of Crime Stalking Resource Center, Stalking Fact Sheet, http://www.ncvc. org/ src/AGP.Net/Components/Document-Viewer/Download. aspxnz?Documen-tlD=40616 (last visited Jan. 26, 2006). In addition to physically harming the victim, research has indicated that approximately 29% of stalkers vandalize the victim’s property, and 9% of stalkers kill or threaten to kill the victim’s family pets. See U.S. Dep’t of Justice, Stalking and Domestic Violence, Report to Congress (May 2001) at 22.
In California, the existence of a credible threat — whether direct or implied through conduct — is an element of the crime of stalking. See CaLPenal Code § 646.9(a). Studies have confirmed that there is a positive association between the existence of such threats and a stalker’s eventual use of violence. See Rosenfeld supra at 14-17. An analysis of six leading studies on stalking has shown the making of threats to be the second best indicator of a stalker’s potential to use violence in the future. See id. at 29-30. Because California does not recognize individuals who are obsessive followers and harassers to be stalkers unless they make credible threats, persons who meet the statutory definition presumably already present a greater risk of violence. As in this case, restraining orders are often not an effective deterrent; 69% of women said the stalker violated a restraining order. See National Center for Victims of Crime Stalking Resource Center, Stalking Fact Sheet, supra at 1.
The majority relies on the recent en banc decision, Fernandez-Ruiz v. Gonzales, 466 F.3d 1121 (9th Cir.2006)(en banc). This reliance is misplaced. Fernandez-Ruiz holds that the reckless use of force is not sufficient to support a finding of the commission of a crime of violence under 18 U.S.C. § 16(a). The opinion expressly declined to consider § 16(b), which is the only section of the statute at issue in this case. See id. at 1125 n. 6 (“Additionally, we need not consider 18 U.S.C. § 16(b) because Fernandez-Ruiz’s assault convictions were both misdemeanors.”).
Malta-Espinoza pleaded guilty to repeatedly following and harassing Ms. Espi-sito despite the existence of a restraining order. He pleaded guilty to making a credible threat with the intent to put her in fear for her safety and the safety of her immediate family. The majority somehow reaches the conclusion that they are unable to determine whether this created a substantial risk of physical force against Ms. Espisito or her property. They rest their conclusion on the “possibility that Malta-Espinoza engaged in ... long-distance harassing.” I cannot agree. The judicially noticeable documents show that, by repeatedly “following and harassing” Ms. Espisito, Malta-Espinoza was, to some degree, in physical proximity to her. Moreover, it is not without sad irony that I *1088note the lack of physical proximity of a stalker to his victim is not a reliable indicator of the risk of violence. California’s stalking laws are rooted in the case of a man who became obsessed with a young actress living hundreds of miles away. He sent her items through the mail until one day he traveled to her home, shot and killed her on her front step. See Gina Piccalo, The Safety Zone; Stalker’s Prey Tells of Terror; Cases like News Anchor Kelly Mack’s Show Police, Judges and Law Makers are Taking the Crime of Stalking More Seriously, L.A. Times, Dec. 4, 2000, at B6 (describing the stalking and murder of 21-year-old actress Rebecca Schaeffer, the co-star of the sitcom “My Sister Sam”).
I respectfully dissent.
. The Board of Immigrations Appeals (the "BIA”) examined the record of conviction and found that Malta-Espinoza was harassing Ms. Espisito, rather than following her. (Record on Appeal at 8.) The BIA did not identify the basis for its finding of fact. It may have reached this conclusion based on police reports presented to the immigration judge. Those police reports are not a part of this *1086Court's consideration. This is of no consequence. The BIA’s finding does not change this analysis. Surely, if Malta-Espinoza’s harassing conduct alone was sufficient to convince the BIA of the risk violence, the BIA's acknowledgment of his admission to following Ms. Espisito could only have supported its ultimate conclusion.
. Subsection 16(a) of Title 18 of the United States Code is not relevant to this case. It states:
The term "crime of violence" means—
(a) an offense that has an element the use, attempted use, or threatened use of physical force against the person or property of another....