United States v. Faust, 853 F.3d 39

Lead Opinion

TORRUELLA, Circuit Judge.

Todd Faust (“Faust”) entered a conditional guilty plea to being a felon in possession of ammunition under 18 U.S.C. § 922(g)(1). He contends on appeal that his conviction must be overturned because police obtained the ammunition in violation of the Fourth Amendment and because the statements he made to police during his station house interview were obtained as part of a two-step interrogation technique in violation of Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966). Faust also appeals his sentence on the ground that the district court erroneously concluded that his prior convictions for resisting arrest, Mass. Gen. Laws ch. 268, § 32B(a), and assault and battery on a police officer (“ABPO”), Mass. Gen. Laws ch. 265, § 13A, were violent felonies under the Armed Career Criminal Act, 18 U.S.C. § 924(e)(2)(B) (“ACCA”). For the reasons that follow we affirm in part and remand in part.

I. Background

On May 6, 2011, the Massachusetts Palmer District Court issued a warrant for police to search the premises at 220 Pearl Street, Apt. 4-R in Springfield, Massachusetts (“Apartment 4-R”), as well as the persons of Faust and Kristina Leighty (“Leighty”). The warrant application included an affidavit from Sergeant Boucher of the Monson Police Department and several other documents, including several police reports from the Chicopee Police Department.

These documents set forth that on April 22, 2011, the home of Joseph Barrett was broken into in Monson, Massachusetts (the “Monson robbery”), which resulted in the theft of several items, including a wristwatch, a laptop computer, ten rolls of pennies, a Leatherman tool, and two women’s wallets. That same morning, a Monson police officer pulled over a vehicle that Leighty was driving as part of a traffic stop. Faust was a passenger in the vehicle. The officer had the vehicle towed due to a lack of insurance and a revoked registration.¹

According to the submitted documents, Leighty and Faust remained in the area until Gregory Charbonneau (“Charbon-neau”) picked them up and drove them to Apartment 4-R. Charbonneau told police that Leighty gave him $6.50 in rolled pennies as payment for the ride. Additionally, Leighty gifted him a watch that was later identified as one of the items stolen during the Monson robbery.²

Sergeant Boucher’s affidavit concluded that based on the timing of Leighty’s gift to Charbonneau, he believed that Leighty and Faust were in possession of additional items stolen during the Monson robbery. The search warrant was granted on May 6, 2011.

Pursuant to the warrant, officers from the Massachusetts State Police, the Springfield Police Department, the Mon-son Policé Department, the Chicopee Police Department, and the Bureau of Alcohol, Tobacco, Firearms, and Explosives (collectively, the “officers”) executed a search of Apartment 4-R, as well as the persons of Leighty and Faust.

When the officers entered Apartment 4-R, Leighty told the officers that Faust had fled upstairs. The officers followed Faust into Apartment 5-R. Upon entering the apartment, two officers observed Faust run towards the back door of the apartment and drop a black backpack on the floor. Faust was apprehended as he attempted to flee out the backdoor and was placed in handcuffs. The officers did not advise Faust of his Miranda rights.

Officer Richard, a Springfield police officer, recovered Faust’s backpack, in which he found a loaded 9mm pistol and ammunition for other weapons. Officer Richard asked Faust if he had a right to be in Apartment 5-R and Faust responded that he did not. Faust was held in the back porch of Apartment 5-R while the officers assessed the scene.

Although the sequence of events is not entirely clear from the record, it is evident that Faust spoke with Officer Richard, Detective Dion of the Chicopee Police Department, and Agent Meehan of the Bureau of Alcohol, Tobacco, Firearms and Explosives while he was on the back porch. According to Officer Richard’s testimony, Faust offered, without prompting, to show him where other stolen guns were located after Faust overheard Officer Richard tell other officers at the scene that the gun in Faust’s backpack matched one of the guns stolen during the Chicopee robbery.³ Officer Richard’s testimony suggests that this exchange took place before Agent Meehan spoke with Faust. However, Officer McNally, a Springfield police officer at the time, testified that Faust began talking to Officer Richard after Detective Dion and Agent Meehan had spoken with Faust.

At some point, Detective Dion approached Faust and informed him that the gun found in the backpack matched one of the guns reported stolen during the Chico-pee robbery. Detective Dion told Faust that he did not have to speak with him, but inquired if Faust would like to do so. Faust told Detective Dion that he was willing to speak with him.

Agent Meehan also approached Faust and asked his name, address, and whether he would like to speak with investigators. Agent Meehan told Faust that he was interested in recovering other guns that had been stolen during the Chicopee robbery. Faust again indicated that he was willing to speak with the investigators.

Subsequently, Officer Richard and Officer McNally took Faust in a police cruiser and drove past the location where Faust claimed that additional guns stolen during the Chicopee robbery were located. Officer Richard and Officer McNally then drove Faust to the Springfield police station.

Once at the station, Faust was advised of his Miranda rights. Faust confirmed that he understood his rights and that he wished to speak with the interviewing officers. During his interview, Faust admitted to his involvement, alongside Leighty and two other individuals, in the Chicopee robbery. Faust also admitted to handling the 9mm pistol found in his possession and stated that his fingerprints would likely be found on it.

Faust sought to suppress the statements he made during his police station interrogation, a motion which the district court denied. In accordance with the ACCA, the district court sentenced Faust to 180 months of imprisonment. During sentencing, Faust objected to his classification as an armed career criminal on the ground that neither his conviction for resisting arrest nor for ABPO qualify as ACCA predicates.

This timely appeal followed.

II. Probable Cause to Search

When reviewing the denial of a motion to suppress, “we view the facts in the light most favorable to the district court’s ruling on the motion, and we review the district court’s findings of fact and credibility determinations for clear error.” United States v. Fermin, 771 F.3d 71, 76-77 (1st Cir. 2014) (quoting United States v. Camacho, 661 F.3d 718, 723 (1st Cir. 2011)) (internal citations and quotation marks omitted). “[W]e review conclusions of law de novo, giving plenary review to the district court’s application of law to facts, reasonable-suspicion determinations, and ultimate decision to deny the motion.” Id. at 77 (citing Camacho, 661 F.3d at 724). However, “we afford an ample amount of deference to the issuing magistrate’s finding of probable cause” when reviewing if an affidavit supports the issued warrant. United States v. Dixon, 787 F.3d 55, 58 (1st Cir. 2015) (citations omitted). As a result, we will reverse a finding of probable cause “only if we see no substantial basis for concluding that probable cause existed.” Id at 59 (quoting United States v. Ribeiro, 397 F.3d 43, 48 (1st Cir. 2005)). Probable cause is present if “the facts and circumstances as to which police have reasonably trustworthy information are sufficient to warrant a person of reasonable caution in the belief that evidence of a crime will be found.” United States v. Silva, 742 F.3d 1, 7 (1st Cir. 2014) (quoting Robinson v. Cook, 706 F.3d 25, 32 (1st Cir. 2013)).

“A warrant application must demonstrate probable cause to believe that (1) a crime has been committed — the ‘commission’ element, and (2) enumerated evidence of the offense will be found at the place to be searched — the so-called ‘nexus’ element.” United States v. Rodrigue, 560 F.3d 29, 32-33 (1st Cir. 2009) (quoting Ribeiro, 397 F.3d at 48). To satisfy the nexus element, the warrant application “must give someone of ‘reasonable caution’ reason to believe that evidence of a crime will be found at the place to be searched.” Ribeiro, 397 F.3d at 49 (citation omitted).

Faust contends that Sergeant Boucher’s affidavit failed to establish probable cause to search his person. Specifically, he argues there was no probable cause to believe that the items listed in the warrant could be found on him.⁴

As the district court observed, Sergeant Boucher’s affidavit stated that: (1) Faust and Leighty had been detained for a traffic violation on April 22, 2011, the same day as the Monson robbery, which resulted in their vehicle being towed; (2) on that same day Charbonneau picked up both Faust and Leighty after the vehicle was towed; (3) Leighty gave Charbonneau a wristwatch that was later identified as stolen during the Monson robbery; (4) the items stolen during the Monson robbery were “the sort of items that a person would carry on their person.” Moreover, Leighty gave Charbonneau $6.50 in rolled pennies as payment for giving her and Faust a ride. We find that the affidavit adequately satisfied the commission element. The fact that Leighty and Faust were detained on the day of the Monson robbery and had items from that robbery in their possession supports a conclusion that evidence of the Monson robbery would be found on Faust’s person. The facts presented in a search warrant application must be such that they permit a man of reasonable caution to conclude that “evidence of a crime will be found.” See United States v. Soto, 799 F.3d 68, 84 (1st Cir. 2015) (quoting United States v. Feliz, 182 F.3d 82, 86 (1st Cir. 1999)). Accordingly, we find that there was ample probable cause to search Faust.

To the extent Faust is arguing that the officers could not follow him into Apartment 5-R, we note that there is a recognized exception to the warrant requirement when police are faced with the “threatened escape by a suspect.” Bilida v. McCleod, 211 F.3d 166, 171 (1st Cir. 2000) (citing McCabe v. Life-Line Ambulance Serv., Inc., 77 F.3d 540, 545 (1st Cir. 1996)). Here, the police were executing the warrant when Leighty informed them that Faust had fled upstairs. The police immediately pursued Faust into Apartment 5-R where they observed him attempting to escape through the backdoor. Thus, even if the search warrant could not establish that the fruits of the Monson robbery could be found inside Apartment 5-R, the police could enter Apartment 5-R in pursuit of Faust.⁵

Accordingly, we find that the district court properly denied Faust’s motion to suppress.

III. Miranda

Faust posits that his post-Miranda statements at the police station are also subject to suppression because he was subjected an impermissible two-step interrogation tactic.

A failure to administer Miranda warnings, “unaccompanied by any actual coercion or other circumstances calculated to undermine the suspect’s ability to exercise his free will, [does not] so taint[] the [later] investigatory process that a subsequent voluntary and informed waiver is ineffective for some indeterminate period.” United States v. Jackson, 544 F.3d 351, 360 (1st Cir. 2008) (quoting Oregon v. Elstad, 470 U.S. 298, 309, 105 S.Ct. 1285, 84 L.Ed.2d 222 (1985)) (alteration in original). Thus, in the absence of coercion or improper tactics by law enforcement in obtaining an initial statement, a subsequent statement is admissible if the defendant was advised of his Miranda rights and knowingly and voluntarily waived those rights. Elstad, 470 U.S. at 318, 105 S.Ct. 1285.

However, suppression may be proper when police deliberately employ a two-step interrogation tactic designed to circumvent Miranda warnings. United States v. Verdugo, 617 F.3d 565, 574-75 (1st Cir. 2010) (citing Missouri v. Seibert, 542 U.S. 600, 605-06, 124 S.Ct. 2601, 159 L.Ed.2d 643 (2004) (plurality opinion)). The Supreme Court addressed the use of such a tactic in Seibert. The facts at issue in that case involved a defendant who was taken to the police station and subjected to questioning for thirty to forty minutes without the benefit of Miranda warnings. Seibert, 542 U.S. at 604-05, 124 S.Ct. 2601. Once the defendant made a crucial admission, she was given a twenty-minute break, after which she was provided Miranda warnings. The defendant was then confronted with her pre-warning statements in order to get her to repeat her confession. Id. at 605, 124 S.Ct. 2601.

A four justice plurality of the Court held that, under these circumstances, the Miranda warnings were ineffective, thereby rendering the defendant’s post-Miranda statements inadmissible. Id. at 611-14, 124 5.Ct. 2601. The plurality focused on the circumstances surrounding the contested statements. Specifically, the plurality considered: (1) “the completeness and detail of the questions and answers in the first round of interrogation”; (2) “the overlapping content of the two statements”; (3) “the timing and setting of the first and the second [interrogations]”; (4) “the continuity of police personnel”; and (5) “the degree to which the interrogator’s questions treated the second round as continuous with the first.” Id. at 615,124 S.Ct. 2601.

Justice Kennedy, who provided the fifth vote in favor of the judgment, advanced a narrower test.⁶ Under Justice Kennedy’s approach the deliberate use of a two-step interrogation creates a presumptive taint. Id. at 622, 124 S.Ct. 2601 (Kennedy, J., concurring). However, when police do not employ a two-step tactic, “[t]he admissibility of postwarning statements should continue to be governed by the principles of Elstad....” Id.

Here, the result is the same under either approach. Faust’s contention withers under the plurality’s analysis. The testimony provided at the evidentiary hearing does not support a conclusion that the questions posed in Apartment 5-R formed part of a larger continuous investigation that persisted at the police station. The record illustrates that Faust was read his rights at the police station and voluntarily waived them. Further, there is no evidence that police leveraged Faust’s post-Miranda statements by utilizing any of his pre-Miranda responses. In sum, we agree with the district court’s conclusion that Faust was not submitted to a two-step interrogation because the timing and settings of the pre-Miranda and post-Miranda questioning were different; the type of questions and their degree of detail varied greatly; and the pre-Miranda and post-Miranda questioning were not part of the same continuum.

We reach an identical result under Justice Kennedy’s test. Nothing in the record reveals the use of a deliberate two-step strategy geared towards leveraging Faust’s confession. In the absence of such a tactic, Faust’s station house statements are admissible if he was provided with Miranda warnings and he voluntarily and knowingly waived his rights. See Elstad, 470 U.S. at 318, 105 S.Ct. 1285. As mentioned above, there is no indication that the officers employed a two-step tactic designed to circumvent Miranda or leverage Faust’s confession. Moreover, the record confirms that Faust was advised of his Miranda rights and that he knowingly and voluntarily waived them. As such, his post-warning statements are not subject to suppression.

Faust also contends that the district court erred when it found that Faust offered to speak to Officer Richard before Detective Dion and Agent Meehan approached him. As we pointed out earlier, the record is unclear as to the precise sequence of events after Faust’s apprehension in Apartment 5-R. Due to this lack of clarity, we cannot conclude that the district court’s conclusion was clear error. See In re Brady-Zell, 756 F.3d 69, 72 (1st Cir. 2014) (“[I]t is apodictic that where the facts can support two plausible but conflicting interpretations of a body of evidence, the factfinder’s choice between them cannot be clearly erroneous.” (citations omitted)).⁷

IV. Sentence Enhancement Under the ACCA

The district court found that Faust was subject to a fifteen-year enhancement of his sentence under the ACCA. The ACCA enhancement applies when a defendant is convicted of being a felon in possession of a firearm and “has three previous convictions ... for a violent felony or a serious drug offense, or both.” 18 U.S.C. § 924(e)(1). For our purposes, the relevant definition of “violent felony” under the ACCA is “any crime punishable by imprisonment for a term exceeding one year ... that — (i) has as an element the use, attempted use, or threatened use of physical force against the person of another.” Id. § 924(e)(2)(B).

Faust has two uncontested predicate convictions that qualify as “serious drug offense[s]” and two possible violent felonies that are at issue in this appeal: a conviction for resisting arrest and two convictions arising out of the same conduct for assault and battery on a police officer (ABPO).⁸

The district court found that Faust’s convictions for ABPO and resisting arrest both counted as predicate convictions. In so finding the judge explicitly stated that he felt bound by First Circuit precedent: United States v. Carrigan, 724 F.3d 39 (1st Cir. 2013) and United States v. Weekes, 611 F.3d 68 (1st Cir. 2010) with respect to resisting arrest, and United States v. Dancy, 640 F.3d 455 (1st Cir. 2011) for ABPO. Applying the ACCA enhancement, the district court sentenced Faust to fifteen years of imprisonment, though he specifically stated that he believed “the 15-year sentence in this case is excessive” and that he “wouldn’t be imposing a 15-year sentence if [he] wasn’t required to.”

Because Carrigan and Weekes have been called into question by the Supreme Court’s recent case of Mathis v. United States, — U.S. -, 136 S.Ct. 2243, 195 L.Ed.2d 604 (2016), and Dancy relied upon a portion of 18 U.S.C. § 924(e) that has since be deemed unconstitutionally vague, Johnson v. United States (“Johnson II”), - U.S. -, 135 S.Ct. 2551, 2563, 192 L.Ed.2d 569 (2015), we must return to the questions previously determined by those cases: do the Massachusetts offenses of resisting arrest and ABPO qualify as violent felonies under the ACCA? We review de novo “[w]hether a prior conviction qualifies as an ACCA predicate.” United States v. Whindleton, 797 F.3d 105, 108 (1st Cir. 2015) (citing Carrigan, 724 F.3d at 48).⁹

A. Applicable Law

1. Step One: Categorical Approach

The basis of all of our ACCA decisions can be found in Taylor v. United States, 495 U.S. 575, 110 S.Ct. 2143, 109 L.Ed.2d 607 (1990), which held that in determining whether a particular conviction would count as an ACCA predicate courts must take a “categorical” approach. Id at 602, 110 S.Ct. 2143. This means that a prior conviction will either count or not based solely on the fact of conviction rather than on facts particular to the individual defendant’s case. Id. (finding that “the only plausible interpretation” of the ACCA is that “it generally requires the trial court to look only to the fact of conviction and the statutory definition of the prior offense”); see also Whindleton, 797 F.3d at 108 (same). The Taylor court based its decision on the statutory language of the ACCA, its legislative history, and the practical difficulties as well as potential unfairness of a factual approach. Id. at 600-01, 110 S.Ct. 2143. The Court’s evolving Sixth Amendment jurisprudence has made clear, however, that this “potential unfairness” actually envelops constitutional limitations dictating the categorical approach. Des-camps, 133 S.Ct. at 2288 (pointing to the “categorical approach’s Sixth Amendment underpinnings” and reasoning that a finding by the sentencing court that “went beyond merely identifying a prior conviction ... to ‘make a disputed’ determination about what the defendant and state judge must have understood as the factual basis of the prior plea” would raise “serious Sixth Amendment concerns” (internal citation omitted)); see also Mathis v. United States, - U.S. -, 136 S.Ct. 2243, 2252, 195 L.Ed.2d 604 (2016) (citing “serious Sixth Amendment concerns” if a sentencing judge were permitted to go beyond the elements of the statute in determining whether a previous conviction qualifies as a predicate).

In applying the 'categorical approach to offenses that involve the force clause, the Supreme Court has made clear that the term physical force “means violent force — that is, force capable of causing physical pain or injury to another person.” Johnson I, 559 U.S. at 140, 130 S.Ct. 1265 (emphasis in the original). Thus, the question is whether the predicate offense contains as an element violent force capable of causing physical pain or injury.

Therefore, the first question a sentencing court must answer when applying the ACCA is whether all of the conduct covered by the statute categorically requires violent force. If the answer is yes, then a conviction under the statute will always count as a predicate under the ACCA. If the answer is no, then the court must move to step two and determine whether the statute is divisible.

2. Step Two: Divisibility

Taylor recognized that the categorical approach would face difficulties in states where the state statute defined the predicate offense “more broadly” than was found in the ACCA. 495 U.S. at 599, 110 S.Ct. 2143. In those situations, the Taylor court held, the categorical approach “may permit the sentencing court to go beyond the mere fact of conviction in a narrow range of cases where a jury was actually required to find all the elements of [the ACCA defined offense].” Id. at 602, 110 S.Ct. 2143 (providing as an example a situation where a “burglary statute! ] include[s] entry of an automobile as well as a building, if the indictment or information and jury instructions show that the defendant was charged only with a burglary of a building, and that the jury necessarily had to find an entry of a building to convict, then the Government should be allowed to use the conviction for enhancement.”).

These “narrow range of cases” are now referred to as employing, somewhat misleadingly, the modified categorical approach. Whindleton, 797 F.3d at 108 (citing Descamps, 133 S.Ct. at 2281). The term is somewhat misleading because the framework is still categorical in the sense that the elements of the offense of conviction are compared with the elements of the statutory offense and only if they align may the offense count as a violent felony. Descamps, 133 S.Ct. at 2285 (“[T]he modified approach merely helps implement the categorical approach. ... And it preserves the categorical approach’s basic method: comparing those elements with the generic offense’s.”). Because the statute in question sweeps more broadly than the definition provided by Congress, however, it is necessary to separate out those offenses listed in the statute that align with Congress’s definition from those that do not and to determine which offense formed the basis of the defendant’s prior conviction. The modified categorical approach thus involves a two-stage process: determine if the statute contains discrete offenses that can be separated from each other (termed “divisibility”) and determine under which the defendant was convicted. Id. at 2281; see also United States v. Serrano-Mercado, 784 F.3d 838, 843 (1st Cir. 2015) (discussing divisibility and the application of the modified categorical approach).

While this appeal was pending the Supreme Court handed down Mathis v. United States, — U.S. -, 136 S.Ct. 2243, 195 L.Ed.2d 604 (2016), which purports to clarify prior case law on when a statute may be deemed divisible. Mathis reiterates a focus on “the elements of the crime of conviction,” which it defines as

the “constituent parts” of a crime’s legal definition — the things the “prosecution must prove to sustain a conviction.” At a trial, they are what the jury must find beyond a reasonable doubt to convict the defendant, and at a plea hearing, they are what the defendant necessarily admits when he pleads guilty.

Id. at 2248 (citations omitted) (quoting Black’s Law Dictionary 634 (10th ed. 2014)). Mathis distinguishes between statutes that “list[ ] multiple elements disjunc-tively” and ones “that enumerate[ ] various factual means of committing a single element.” Id. at 2249. The Court used a hypothetical from Descamps to illustrate this distinction:

suppose a statute requires use of “a deadly weapon” as an element of a crime and further provides that the use of a “knife, gun, bat, or similar weapon” would all qualify. Because that kind of list merely specifies diverse means of satisfying a single element of a single crime- — or otherwise said, spells out various factual ways of committing some component of the offense — a jury need not find (or a defendant admit) any particular item.

Id. (citation omitted) (quoting Descamps, 133 S.Ct. at 2289).

Mathis thus directs that when a sentencing court is faced with a statute that lists alternatives, it must first determine “whether its listed items are elements or means.” Id. at 2256. If they are elements then the court proceeds to apply the modified categorical approach and determine which “of the enumerated alternatives played a part in the defendant’s prior conviction, and then compare that element (along with all others) to those of the generic crime.” Id. If they are means, however, then the court’s inquiry is at an end and the sentencing court may not delve into the facts of the case to determine which means this particular defendant used to commit the offense. Id.

As to the “threshold inquiry- — elements or means?,”, Mathis states that this need not be difficult. Id. Mathis itself relies on a state court decision that specified that the statute in question listed alternative means of committing a single offense.¹⁰ Id. Mathis also directs the sentencing court to the statute itself. “If statutory alternatives carry different punishments, then under Apprendi they must be elements.” Id. (citing Apprendi v. New Jersey, 530 U.S. 466, 490, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000)). “[I]f state law fails to provide clear answers,” then Mathis directs the sentencing court to “ ‘peek at the [record] documents’ ... for ‘the sole and limited purpose of determining whether [the listed items are] elements] of the offense.’ ” Id. at 2256-57. (quoting Rendon v. Holder, 782 F.3d 466, 473-74 (9th Cir. 2015) (Kozinkski, J., dissenting from denial of reh’g en banc)). For example, if the indictment and jury instructions reiterate “all the terms of [the] law” then “[t]hat is as clear an indication as any that each alternative is only a possible means of commission, not an element that the prosecutor must prove to a jury beyond a reasonable doubt.” Id. at 2257. Similarly, “if those documents use a single umbrella terms like ‘premises’: Once again, the record would then reveal what the prosecutor has to (and does not have to) demonstrate to prevail.” Id. If, at the end of this review “such record materials” do not “speak plainly,” then “a sentencing judge will not be able to satisfy ‘Taylor’s demand for certainty’ when determining whether a defendant was convicted of a generic offense.” Id. (quoting Shepard, 544 U.S. at 21, 125 S.Ct. 1254).

3. Step Three: Determining the Offense of Conviction

If a statute is found to be divisible, then the sentencing court must proceed to the third and final step and determine which offense the defendant was actually convicted of. This step is guided by the Supreme Court’s opinion, Shepard v. United States, 544 U.S. 13, 125 S.Ct. 1254, 161 L.Ed.2d 205 (2005). There the Supreme Court specified what documents the sentencing court could reference in order to determine which of the multiple offenses listed in the statute was the crime committed by the defendant: “the statutory definition, charging document, written plea agreement, transcript of the plea colloquy, and any explicit factual finding by the trial judge to which the defendant assented.” Id. at 16, 125 S.Ct. 1254. These became known as Shepard documents. See, e.g., Sauceda v. Lynch, 819 F.3d 526, 531 (1st Cir. 2016). Even at this stage, however, the district court’s task is not to fit the facts of the individual defendant’s conduct into one of the divisible offenses. Mathis, 136 S.Ct. at 2251 (“How a given defendant actually perpetrated the crime — what we have referred to as the ‘underlying brute facts or means’ of commission — makes no difference.” (quoting Richardson v. United States, 526 U.S. 813, 817, 119 S.Ct. 1707, 143 L.Ed.2d 985 (1999))). Rather, the question is “whether the plea had ‘necessarily’ rested on the fact identifying the [offense] as generic.” Shepard, 544 U.S. at 21, 125 S.Ct. 1254 (quoting Taylor, 495 U.S. at 602, 110 S.Ct. 2143).

B. Massachusetts Resisting Arrest

1. Step One: Categorical Approach

Resisting Arrest in Massachusetts is defined by Mass. Gen. Laws ch. 268, § 32B(a), which states:

A person commits the crime of resisting arrest if he knowingly prevents or attempts to prevent a police officer, acting under color of his official authority, from effecting an arrest of the actor or another by:

(1) using or threatening to use physical force or violence against the police officer or another; or

(2) using any other means which creates a substantial risk of causing bodily injury to such police officer or another.

While the first form of resisting arrest was held to “fit[ ] squarely within the definition of a crime of violence,” the second qualified under the so-called residual clause. United States v. Almenas, 553 F.3d 27, 33, 35 (1st Cir. 2009). This clause defined as a violent felony any offense which “involves conduct that presents a serious potential risk of physical injury to another.” 18 U.S.C. § 924(e)(2)(B)(ii). It was invalidated as unconstitutionally vague by the Supreme Court in Johnson II, 135 S.Ct. at 2563. The government concedes that the second form of resisting arrest, which could be accomplished by merely stiffening one’s arm to avoid being handcuffed, does not survive as a violent felony. Commonwealth v. Grandison, 433 Mass. 135, 741 N.E.2d 25, 35 (2001) (stiffening arms to avoid being handcuffed is identified as one of “any other means” that creates a “substantial risk of bodily injury”). Because a portion of the statute therefore defines offenses that do not meet the requirements of 18 U.S.C. § 924(e)(2)(B), resisting arrest cannot categorically qualify as a predicate under the ACCA. We therefore turn to step two to determine whether the statute is divisible.

2. Step Two: Divisibility

This offense was previously found by this court to be divisible. Carrigan, 724 F.3d at 50. Following Mathis and its requirement that for a statute to be divisible the listed alternatives must be different elements rather than merely different means of committing an offense, the government now concedes that state law sources are equivocal as to whether resisting arrest is divisible. Mathis, 136 S.Ct. at 2256. The government bases its concession on the following: the lack of clear state court decisions parsing the elements of the offense; the lack of distinct penalties for the two “types” of resisting listed in the statute; Faust’s indictment, which charges both variants; Faust’s plea colloquy, which does not recite the elements of the offense; the model complaint, which charges both forms; and the model jury instructions, which list both alternatives within a single element of “resisting.”¹¹

These pieces of evidence persuasively establish that the Massachusetts offense of resisting arrest is not divisible. When there is no state law ruling “definitively answering] the question,” we are called upon to examine the statute. Id. “If statutory alternatives carry different punishments, then under Apprendi they must be elements.” Id. Here the punishment for resisting arrest is not impacted by whether the first or second method of committing the offense is used. Mass. Gen. Laws ch. 268, § 32B(d) (“Whoever violates this section shall be punished by imprisonment in a jail or house of correction for not more than two and one-half years or a fine of not more than five hundred dollars, or both.”). The model charging language used by district courts in Massachusetts list both types of resisting arrest. District Court Complaint Language Manual 378 (2016). Faust’s own indictment followed this model language. The model jury instructions used for resisting arrest list both subsections (1) and (2) as alternatives under a single element of resistance.¹² Massachusetts Criminal Model Jury Instructions for Use in the District Court, Instruction 7.460, at 1-2 (2009); see also Mathis, 136 S.Ct. at 2257 (finding that an indictment and “correlative jury instructions” that reiterate all the terms of the statute “[are] as clear an indication as any that each alternative is only a possible means of commission, not an element that the prosecutor must prove to a jury beyond a reasonable doubt.”). Finally, Faust’s plea colloquy with the sentencing judge did not specify the elements of resisting arrest and when entering his plea Faust was simply asked “as to count ten, charging resisting arrest, how do you wish •to plead, sir?” to which he replied “Cg]uilty.”

All of these pieces of evidence point to the same conclusion: sections (1) and (2) of the Massachusetts resisting arrest statute merely list two different means of committing a single element of “resisting.” Because resisting arrest is not categorically a violent felony and the offense is not divisible, Faust’s Massachusetts conviction for resisting arrest cannot be counted as a “violent felony” under the ACCA.¹³ Given that Faust’s resisting arrest conviction does not qualify him for an ACCA enhancement, we turn now to his ABPO convictions and ask the same questions again: does the state statute categorically require violent force, and, if not, is it divisible?

C. Massachusetts ABPO

1. Step One: Categorical Approach

The ABPO statute provides that a person commits ABPO when he “commits an assault and battery upon any public employee when such person is engaged in the performance of his duties at the time of such assault and battery.” Mass. Gen. Laws ch. 265, § 13D. The Massachusetts’s statute for assault and battery, Mass. Gen. Laws ch. 265, § 13A, encompasses the common-law variants of assault and battery. Commonwealth v. Eberhart, 461 Mass. 809, 965 N.E.2d 791, 798 (2012) (citations omitted). Under the common law, there are two theories of assault and battery: intentional battery and reckless battery. Id at 798 n.13 (citing Commonwealth v. Porro, 458 Mass. 526, 939 N.E.2d 1157, 1162 (2010)). Intentional assault and battery includes both harmful battery and offensive battery. Harmful battery is defined as “[a]ny touching ‘with such violence that bodily harm is likely to result.’ ” Eberhart, 965 N.E.2d at 798 (quoting Commonwealth v. Burke, 390 Mass. 480, 457 N.E.2d 622, 624 (1983)). Offensive battery is defined as any uncon-sented touching that constitutes an “affront to-the victim’s personal integrity.” Id. (quoting Burke, 457 N.E.2d at 624). In contrast, reckless battery involves the “wilful, wanton and reckless act which results in personal injury to another.” Id. (quoting Commonwealth v. Welch, 16 Mass.App.Ct. 271, 450 N.E.2d 1100, 1102 (1983)).

The government concedes that offensive assault and battery does not qualify as requiring “violent force” under Johnson I. Indeed, this has been recognized in this circuit long before Johnson I and ever since the first case to apply assault and battery as an ACCA predicate, United States v. Bregnard, 951 F.2d 457, 459 (1st Cir. 1991). Because the statute is thus overbroad, we must turn to the question of whether ABPO is divisible among the different types of assault and battery.

2. Step Two: Divisibility

On its face this statute does not appear to list potential crimes in the alternative (listing, for example, harmful, offensive, or reckless battery). It simply states “assault and battery.” Similarly, the punishment for ABPO is in no way altered by whether it involved harmful, offensive or reckless battery. Mass. Gen. Laws ch. 265, § 13D (providing that anyone guilty of “an assault and battery upon any public employee ... shall be punished by imprisonment for not less than ninety days nor more than two and one-half years in a house of correction or by a fine of not less than five hundred nor more than five thousand dollars”); see Mathis, 136 S.Ct. at 2256 (stating that “[i]f statutory alternatives carry different punishments, then under Apprendi they must be elements”). Because the statute itself fails to make clear whether the different forms of assault and battery are separate elements or merely distinct means of committing a single element (the element being assault and battery) we turn to other sources to determine divisibility.

a. Commonwealth v. Eberhart

If state law is clear that the alternatives are elements or means then that can resolve the question of divisibility. Mathis, 136 S.Ct. at 2256. In its brief to us the government argued that the question of divisibility was determined by the Massachusetts Supreme Judicial Court case Commonwealth v. Eberhart, 461 Mass. 809, 965 N.E.2d 791, 798 (2012). Because the government backed away from this position at oral argument wé will not tarry over it here, except to say that it was wise for them to have done so. Eberhart defines the elements of the different types of assault and battery under Massachusetts law and holds that assault and battery is divisible between these types under Massachusetts’s similarly worded sentence enhancement statute. This would appear to control in this case except that, in defining those elements, at no point does Eberhart state that any of them must be found beyond a reasonable doubt by a jury or necessarily admitted to by a defendant when he or she pleads guilty to an ABPO. This question is crucial because that is how Mathis defines the term “element” in the ACCA. 136 S.Ct. at 2248. Thus, it is not enough that Eber-hart states that there are three types of assault and battery under state law and then lists their elements or even that Eberhart found assault and battery to be divisible under Massachusetts’s sentencing enhancement statute. 965 N.E.2d at 796, 800. The foundational question is what are the elements of ABPO and Mathis directs that this is answered by determining what a jury has to find beyond a reasonable doubt or a defendant must necessarily admit when pleading guilty. Because Eber-hart does not answer this threshold question for us, it cannot be dispositive in this case.

b. United States v. Tavares

At oral argument the government switched gears and argued that our decision is dictated by this court’s recent decision, United States v. Tavares, 843 F.3d 1 (1st Cir. 2016). Tavares examined the divisibility of assault and battery in Massachusetts by examining recent state court decisions as well as model jury instructions. In particular, it noted that the recent Appeals Court of Massachusetts case Commonwealth v. Mistretta, 84 Mass.App.Ct. 906, 995 N.E.2d 814 (per curiam), rev. denied, 466 Mass. 1108, 996 N.E.2d 881 (2013) suggests that assault and battery is not divisible. 843 F.3d at 15. In that case, the court found that the intentional and reckless forms of assault and battery “are closely related subcategories of the same crime,” and therefore “[s]pecific unanimity is not required, because they are not ‘separate, distinct, and essentially unrelated ways in which the same crime can be committed.’ ” Mistretta, 995 N.E.2d at 815-16 (citation omitted). The Tavares opinion states that “[biased on Mistretta” the 2016 model jury instructions no longer required a “verdict slip or specific unanimity instruction^” where both forms of assault and battery were alleged. 843 F.3d at 14.

Because Mistretta is an Appeals Court decision, however, Tavares proceeds to “predict how the SJC would decide whether a specific unanimity instruction is required,” using Mistretta as a piece of relevant evidence but not determinative of the outcome. Id. at 15. In particular, Tavares looks to an earlier SJC case, Commonwealth v. Santos, 440 Mass. 281, 797 N.E.2d 1191 (2003), which concluded that a jury did not need to be unanimous on the theories underlying the “ ‘assault’ element of armed robbery.” Id at 1194. The SJC found instead that the different theories are actually “overlapping subcategories of a single element into separate ‘theories.’ ” Id. at 1196. The SJC reasoned that in order to require unanimity the different methods of committing the offense have to be “substantively distinct or dissimilar.” Id. at 1197-98. It then cited two offenses having different mens rea requirements (voluntary and involuntary manslaughter) as being an example of “substantively distinct or dissimilar” offenses that would require juror unanimity. Id. at 1197. Based in part on this argument, Tavares “predict[s] that the SJC would not follow Mis-tretta” and holds that assault and battery is divisible between its intentional and reckless forms. Id. at 17.

Tavares involved the career offender provision of the Sentencing Guidelines, though we have previously found that “the terms ‘crime of violence’ under the career offender guideline and ‘violent felony’ under the ACCA are nearly identical in meaning, so that decisions construing one term inform the construction of the other.” United States v. Willings, 588 F.3d 56, 58 n.2 (1st Cir. 2009); see also United States v. Fields, 823 F.3d 20, 35 (1st Cir. 2016) (same). This would suggest that Tavares’s determination that assault and battery in Massachusetts is divisible between intentional assault and battery and reckless assault and battery precludes any other determination by this court.

The defendant argues, however, that he should not be bound by Tavares’s holding specifically because that case involved the Sentencing Guidelines whereas his involves the ACCA. He argues that because the Sentencing Guidelines are non-binding, whereas the ACCA creates a mandatory minimum, both due process as well as the Sixth Amendment dictates that we cannot take an informed prophecy approach to state law but must, rather, determine what the law was at the time of his conviction.

This argument is supported by the Supreme Court’s decision of McNeill v. United States, which makes clear that when applying the ACCA the task for the sentencing court is to determine the defendant’s “previous conviction” and “[t]he only way to answer this backward-looking question is to consult the law that applied at the time of that conviction.” 563 U.S. 816, 820, 131 S.Ct. 2218, 180 L.Ed.2d 35 (2011). In making this argument the unanimous Court in McNeill pointed to its previous ACCA cases, which looked to the versions of state law that were current at the time of the defendant’s convictions, not at the time of the Court’s decision. Id. at 821-22, 131 S.Ct. 2218 (calling them “the historical statute of conviction”).

The approach that McNeill dictates that we take in ACCA cases thus conflicts with the “informed prophecy” approach in Ta-vares, 843 F.3d at 14. We can avoid this conflict, however, because Faust was convicted of ABPO in 2009, prior to the Mis-tretta decision but following Santos. We thus find the “informed prophecy” approach unnecessary for the resolution of this case in which the defendant was clearly convicted of ABPO prior to Mistretta.¹⁴ At that time, the model jury instructions provided that

If the evidence would warrant a guilty verdict for the offense of assault and battery on more than one theory of culpability, the judge must provide the jury with a verdict slip to indicate the theory or theories on which the jury bases its verdict and, on request, instruct the jurors that they must agree unanimously on the theory of culpability.

Massachusetts Criminal Model Jury Instructions for Use in the District Court, Instruction 6.210, at 8 n.l (May 2011) (citations omitted). This requirement of juror unanimity between the intentional and reckless forms of ABPO is strong evidence that Faust’s conviction for ABPO was divisible between those two forms.

We acknowledge, however, that there are considerations that point in the opposite direction. As with Faust’s resisting arrest conviction, Faust’s indictment merely states that he did “assault and beat” a police officer while the model complaint language similarly dictates the simple “did assault and beat” language. Finally, Faust’s plea colloquy fails to include anywhere the terms “intentional assault and battery” or “reckless assault and battery.” Rather, the judge simply asked Faust “[h]as your attorney explained to you what the Commonwealth would have to prove in order for you to be found guilty of these charges?” to which Faust answered “[y]es,” and then the judge asked “[attorney Bernard, have you explained to your client the elements?” to which Faust’s attorney responded “[a]ll the essential elements, yes.” Following this exchange the clerk asked Mr. Faust “as to counts five and six, each charging assault and battery on a police officer ... how do you wish to plead, sir?” to which Faust replied “[gjuilty.” The plea colloquy thus leaves equivocal what, exactly, the essential elements were believed to be.

Given Mathis’s specific reference to what a jury “necessarily” must find in order to convict, however, we find the model jury instructions to be particularly persuasive and hold that at the time of his guilty plea Faust’s ABPO conviction was divisible between the intentional and reckless forms. 136 S.Ct. at 2255. We do not find, however, any indication that the intentional form was further divisible between offensive and harmful assault and battery.¹⁵ Because offensive assault and battery does not require violent force, the intentional form of ABPO is therefore overbroad and categorically cannot count as a predicate for ACCA purposes.

3. Step Three: Determining the Offense of Conviction

The district court below found that the offense of ABPO qualified as an ACCA predicate under then controlling First Circuit precedent, Dancy, 640 F.3d at 470. Having so found, it did not need to turn to the question of which type of ABPO Faust committed.¹⁶ As outlined above, we now hold that ABPO is divisible. Once a sentencing court determines that a defendant has a prior conviction under a divisible statute, it must proceed to step three and determine which among the multiple crimes listed was the offense of conviction. The Supreme Court provided guidance on how to conduct this inquiry in Shepard. There it enumerated the documents that the sentencing court might consult at this stage: “the terms of the charging document, the terms of a plea agreement or transcript of colloquy between judge and defendant in which the factual basis for the plea was confirmed by the defendant, or ... some comparable judicial record of this information.” Shepard, 544 U.S. at 26, 125 S.Ct. 1254. If, at sentencing, the district court is faced with a predicate conviction that is based on a divisible offense then it must consult approved Shepard documents in order to determine which of the offenses the defendant “necessarily” pled guilty to. Id. at 21, 125 S.Ct. 1254.¹⁷

In its supplemental brief the government included all of Faust’s Shepard documents and urged us to affirm Faust’s conviction because the facts he admitted as part of his plea colloquy were consistent with harmful battery. At oral argument the government switched gears and urged us to affirm Faust’s sentence because the facts he admitted as part of his plea colloquy were consistent with reckless battery. The government therefore appears to all but concede that the facts admitted by Faust during his plea could be consistent with either intentional battery or with reckless battery.

As we stated earlier, if “such record materials” do not “speak plainly,” then “a sentencing judge will not be able to satisfy ‘Taylor’s demand for certainty’ when determining whether a defendant was convicted of a generic offense.” Mathis, 136 S.Ct. at 2257 (quoting Shepard, 544 U.S. at 21, 125 S.Ct. 1254). Facts that are as consistent with intentional ABPO as they are with reckless ABPO can hardly be said to “speak plainly.” More to the point, however, although step three invites the district court to examine the documents of the convicting court, the question to be answered remains concentrated on the elements of the actual offense of conviction rather than on the specific facts of Faust’s conduct. The enquiry posed in step three is no broader than that posed under step two: what did Faust “necessarily admit[] when he [pled] guilty”? Mathis, 136 S.Ct. at 2248; see also Shepard, 544 U.S. at 26, 125 S.Ct. 1254 (defining the “enquiry” to be “whether a plea of guilty to burglary defined by a nongeneric statute necessarily admitted elements of the generic offense”). The aim therefore remains on determining the elements that Faust pled guilty to, not how he committed the particular crime. This focus is particularly important in situations such as here where numerous of the Shepard documents (indictment, model charging language and the clerk’s statement of what offense Faust was pleading guilty to) fail to specify the type of assault and battery at issue and particularly where the punishment received was in no way impacted by the type of assault and battery to which Faust pled.¹⁸

We therefore remand to the district court for it to determine which form of ABPO Faust necessarily pled guilty to. If it was the intentional form, then his conviction for ABPO categorically cannot count as a violent felony under the ACCA. If it is plain that Faust pled guilty to the reckless form, however, then the district court must determine whether reckless conduct qualifies as the “use” of force under the ACCA. At this time we decline to take a position on this question.¹⁹ Finally, if the district court finds that it is unclear which form of ABPO Faust pled guilty to, then his ABPO conviction cannot be used as a predicate under the ACCA because the court’s step three inquiry utilizing Shepard documents will not have made plain what the offense of conviction actually was.

V. Conclusion

For the foregoing reasons, we affirm the district court’s denial of Faust’s motion to suppress and we vacate and remand for resentencing proceedings consistent with this opinion.

Affirmed in part, Vacated and Remanded in part.

Notes

1 . Sergeant Boucher’s affidavit noted that Leighty and Faust had two large black bags that could have "easily conceal[ed]’’ the items stolen during the Monson robbery.

2 . On May 2, 2011, Charbonneau's home in Chicopee, Massachusetts was broken into (the "Chicopee robbery"), which resulted in the theft of nine firearms of varying calibers. Charbonneau identified pictures of Leighty and Faust, and told police that he suspected that Leighty was involved in the robbery of his home. Charbonneau identified Leighty as a woman he had been involved with for several weeks and whom he knew as "Cory.” Char-bonneau also identified Faust, whom he believed to be Leighty's brother, as "Jay.”

3 . The district court found that, once he was secured outside of Apartment 5-R, Faust repeatedly asked Officer Richard what he could do to help himself.

4 . Arguably, Faust has waived his challenge to the validity of the search warrant by failing to properly develop his argument. See United States v. Zannino, 895 F.2d 1, 17 (1st Cir. 1990). Because we find that his argument fails on the merits, we opt to discuss the reasons why the search warrant was valid.

5 . Faust also claims that the search of his person was the fruit of his unlawful arrest for breaking and entering into Apartment 5-R, and thus any evidence found in the course of the search must be excluded. Specifically, he argues that police lacked probable cause to arrest him for breaking and entering under Massachusetts law because he lacked the necessary mens rea to commit that crime. Whether arresting officers had probable cause is not measured by the offense invoked at the time of the arrest, but rather by examining if "the facts known at the time of the arrest objectively provided probable cause to arrest.” United States v. Jones, 432 F.3d 34, 41 (1st Cir. 2005) (citation omitted). Probable cause to arrest hinges on whether police, "relying on reasonably trustworthy facts and circumstances, have information upon which a reasonably prudent person would believe the suspect had committed or was committing a crime.” Id. (quoting United States v. Young, 105 F.3d 1, 6 (1st Cir. 1997)). As a result, it is irrelevant if Faust was arrested for breaking and entering. Here, the arresting officers had probable cause to arrest Faust because they were aware that he was connected to the Monson robbery and attempted to flee when police sought to execute the search warrant. Thus, a prudent person could conclude that Faust had committed a criminal offense. In any event, the officers recovered the backpack containing the gun and ammunition after Faust discarded it and before he was apprehended. Thus, Faust abandoned any expectation of privacy in the backpack, the recovery of which did not constitute a wrongful seizure. See United States v. Soto-Beníquez, 356 F.3d 1, 36 (1st Cir. 2003).

6 . We note that some Circuits have held that Seibert's reach is limited by Justice Kennedy’s vote. See United States v. Jackson, 608 F.3d 100, 103-04 (1st Cir. 2010) (collecting cases). In contrast, we have not settled on a definitive reading of Seibert. Compare Jackson, 608 F.3d at 103-04 (applying both the plurality’s and Justice Kennedy’s test), with United States v. Rogers, 659 F.3d 74, 79 (1st Cir. 2011) (referring to Justice Kennedy’s opinion as controlling). Because we find that Faust’s argument fails under either approach, there is no need to address this question here.

7 . Faust also argues that his post-Miranda statements should be suppressed because he was searched in violation of the Fourth Amendment. However, we have already found that Faust was properly searched in accordance with the warrant and that police had probable cause to arrest Faust.

8 . Because they were committed during a single course of conduct, only one of the ABPO convictions may count as a predicate offense. 18 U.S.C. § 924(e)(1) (predicate convictions must be for crimes “committed on occasions different from one another”).

9 . The government claims that Faust waived his arguments under the ACCA because they allege he based his argument below on Johnson v. United States (“Johnson I”), 559 U.S. 133, 130 S.Ct. 1265, 176 L.Ed.2d 1 (2010), rather than on the question of divisibility as it is being presented to us now. This argument is without merit. In his objections below Faust’s counsel specifically referenced Descamps v. United States, - U.S. -, 133 S.Ct. 2276, 186 L.Ed.2d 438 (2013), the then-most-current authority on divisibility, and the district court specifically stated that Faust had preserved his objections under Descamps.

10 . In deciding the case on this basis, Mathis answered a question left open by Descamps as to whether a state court decision can provide guidance in determining if something is an element or a means. See Descamps, 133 S.Ct. at 2291.

11 . Among the pieces of evidence cited by the government both for resisting arrest and ABPO are Faust’s Shepard documents. These were not produced below. Rather, the government provided them for the first time as part of its supplemental briefing.

12 . We note that Mathis does not make clear whether model jury instructions may always be used. Mathis permits a sentencing court to "peek at the [record] documents" — the documents being the Shepard documents. 136 S.Ct. at 2256. In a case such as this one where the defendant pled guilty there are no jury instructions included in the Shepard documents. Using model jury instructions in such a situation might suggest greater clarity to the question than the guilty plea documents might give (indeed, as we will see below, they might make a dispositive difference). While we note these concerns, we opt to use model jury instructions as one of the many pieces of evidence we will consult to determine divisibility. This is in keeping with Taylor's admonishment that an offense should count or not based solely on the fact of conviction, suggesting that the outcome should not depend on how the defendant was found guilty. 495 U.S. at 602, 110 S.Ct. 2143. Moreover, the concerns raised above will still be relevant in, and perhaps better addressed at, step three, when the sentencing court has to determine the actual offense of conviction. A determination that a statute is divisible does not absolve the sentencing court from having to make the further finding of what this particular defendant necessarily pled guilty to, and if the answer to that question is not clear then the conviction cannot count as a predicate, even if the statute is divisible and some convictions under it categorically are predicates. Mathis, 136 S.Ct. at 2257 (citing the need for the sentencing court to meet "Taylor's demand for certainty” (quoting Shepard, 544 U.S. at 21, 125 S.Ct. 1254)).

13 . A recent case in this court examined the question of whether the first “type” of resisting arrest is categorically a violent felony, but because the defendant in that case did not contest divisibility the court's opinion did not examine that question. United States v. Tavares, 843 F.3d 1, 10 (1st Cir. 2016). We thus now answer a question left open by the defendant in that case.

14 . We note that apparently neither the government nor the defendant in Tavares argued whether a historical approach should be taken in determining the offense of conviction even though Tavares's own conviction also predated Mistretta. United States v. Tavares, 843 F.3d 1, 3 (1st Cir. 2017)

15 . As with resisting arrest above, all evidence indicates that intentional assault and battery is not divisible between the harmful and offensive forms. In particular, the model jury instructions for intentional ABPO state that the jury must find "That the touching was either likely to cause bodily harm to [the alleged victim], or was done without his (her) consent.” Massachusetts Criminal Model Jury Instructions for Use in the District Court, Instruction 6.210, at 8 n.l (May 2011) (emphasis in original).

16 . In part for this reason we will subject Faust’s argument that the government has not proven which type of ABPO offense he committed to de novo review. Additionally, although the government argues that Faust's divisibility argument is waived, see supra n.9, it does not argue that the question of which form of ABPO Faust committed should be subject to plain error review.

17 . Previous of our cases have grappled with the question of whether it is plain error to, at this stage, rely on the presentence report’s ("PSR”) categorization of a prior conviction as a predicate offense when there has been no objection by the defendant. See, e.g., Serrano-Mercado, 784 F.3d at 846-48 (summarizing cases and concluding that whether or not there was plain error, in nope of the cases was it prejudicial to have relied on an unob-jected to PSR). The court in those cases faced a markedly different situation from the one with which we are presented. Here Faust did object to the PSR's categorization of his prior offenses. In that situation "a presentence report in a subsequent case ordinarily may not be used to prove the details of the offense conduct that underlies a prior conviction.” United States v. Dávila-Félix, 667 F.3d 47, 57 (1st Cir. 2011) (quoting United States v. Turbides-Leonardo, 468 F.3d 34, 39 (1st Cir. 2006)). Moreover, Mathis makes clear that the elemental approach dictates that the question at this stage of the inquiry is what the defendant "necessarily admitted] when he [pled] guilty.” 136 S.Ct. at 2248. The fáctual account of Faust's prior conviction that is included in his PSR simply cannot provide an answer to this inquiry.

18 . In light of this latter fact, we would do well to remember the caution the Supreme Court in Mathis and Descamps has given against an overreliance on the factual statements contained in a plea colloquy: "Statements of non-elemental fact' in the .records of prior convictions are prone to error precisely because their proof is unnecessary.” Mathis, 136 S.Ct. at 2253 (quoting Descamps, 133 S.Ct. at 2288). During a plea hearing "a defendant may have no incentive to contest what does not matter under the law; to the contrary, he 'may have good reason not to'— or even be precluded from doing so by the court.” Id. (quoting Descamps, 133 S.Ct. at 2289).

19 . Tavares likewise left this question unanswered. 843 F.3d at 18-20.

---

Concurrence Opinion

LYNCH, Circuit Judge,

concurring.

I join that portion of Judge Torruella’s fine opinion affirming the judgment of conviction and the order remanding the case, but not the discussion of the sentencing issues concerning the two predicate state crimes. I do agree that the case must be remanded to the district court for a review of the Shepard documents newly produced on appeal by the government. But I believe we should not address issues beyond that remand for consideration of the Shepard documents by the district court in the first instance.

The average person on the street would ordinarily think that the state crime of assault and battery on a police officer would meet the ACCA definition of crime of violence, that is “the use, attempted use, or threatened use of physical force against the person of another.” 18 U.S.C. § 924(e)(2)(B)©. Other circuits have noted that tension. See United States v. Harris, 844 F.3d 1260, 1262 (10th Cir. 2017) (noting that question of whether robbery is a “violent felony” for ACCA “should be [] obvious,” but that under the governing analysis, “the obvious may not be so plain”). It is perfectly clear Congress intended enhanced punishment to apply to recidivists whose prior crimes met that definition. But in recent years the Supreme Court has issued a series of decisions restructuring the analysis from one directly addressed to Congressional intent, into a number of other tests. I do not question here that the sentencing portion of Judge Torruella’s opinion is faithful to those judicially created tests. I do have doubts about the majority’s holding as to several issues I consider not to be necessary to the opinion. My doubts about these matters do not prevent me from agreeing that the remand is the correct outcome.

My concern is that use of these tests can lead courts to reach counterintuitive results, and ones which are not what Congress intended. Respected circuit judges share this concern. See, e.g., United States v. Doctor, 842 F.3d 306, 313 (4th Cir. 2016) (Wilkinson, J., concurring) (observing that “the categorical approach can serve as a protracted ruse for paradoxically finding even the worst and most violent offenses not to constitute crimes of violence” and that “too aggressively applied, [the categorical approach] eviscerates Congress’s attempt to enhance penalties for violent recidivist behavior”), petition for cert. filed, 85 U.S.L.W. - (U.S. Mar. 17, 2017) (No. 16-8435).

The concern has also been expressed most eloquently by several Justices of the United States Supreme Court. As Justice Kennedy has said, “arbitrary and inequitable results produced by applying an elements based approach ... could not have been Congress’ intent.” Mathis v. United States, - U.S. -, 136 S.Ct. 2243, 2258, 195 L.Ed.2d 604 (2016) (Kennedy, J., concurring). In Mathis, Justice Breyer, joined by Justice Ginsburg, in dissent stated “[t]he elements/means distinction that the Court draws should not matter for sentencing purposes. I fear that the majority’s contrary view will unnecessarily complicate federal sentencing law, often preventing courts from properly applying the sentencing statute that Congress enacted.” Id at 2259 (Breyer, J., dissenting). And, also dissenting in Mathis, Justice Alito, discussing the “modified categorical” approach culminating in Descamps v. United States, — U.S. -, 133 S.Ct. 2276, 186 L.Ed.2d 438 (2013), stated “[p]rogrammed in this way, the Court set out on a course that has increasingly led to results that Congress could not have intended.” Id. at 2268 (Alito, J., dissenting).

I also agree with Justice Kennedy’s hope for Congress to “amend[ ] the ACCA to resolve these concerns,” id. at 2258 (Kennedy, J., concurring), as well as other concerns, including racially disparate results, which have-been highlighted in efforts at sentencing reform. See Seung Min Kim, Senators Plan to Revive Sentencing Reform Push, Politico (Jan. 4, 2017, 5:13 AM), http://www.politico.com/story/2017/ 01/senate-criminal-justiee-sentencing-reform-233071; Bill Keller, Will 2017 Be the Year of Criminal Justice Reform?, N.Y. Times (Dec. 16, 2016), http://www. nytimes.com/2016/12/16/opinion/will-2017-be-the-year-of-criminal-justice-reform. html.

---

Concurrence Opinion

BARRON, Circuit Judge,

with whom TORRUELLA, Circuit Judge, joins, concurring.

I fully agree with this excellent majority opinion. I write separately to add these thoughts about the growing sentiment that Congress could not have intended for us to apply the increasingly complicated framework that the Supreme Court now requires us to apply in order to determine the scope of the Armed Career Criminal Act (ACCA). In my view, this line of criticism rests on a mistaken premise.

I do not doubt that Congress may have failed to foresee how hard it would be to identify those offenders who, by dint of their past violence, must automatically be subject to the severe sentencing enhancement that ACCA imposes. I do doubt that Congress wanted us to ignore the reality that such a task is a hard one if, in reality, it is. And, experience has shown, that task is hard. For that reason, I think it is important to review Congress’s own role in leading us down this complicated jurisprudential path, which has lately come in for such criticism. That review highlights to me a number of too-easily overlooked problems that the Supreme Court’s current framework for applying ACCA manages to avoid but that otherwise might arise.

I.

First, I realize that the counterin-tuitive results sometimes produced by the categorical approach may appear to be ones that Congress could not possibly have intended, especially given what often appears to be the violent nature of the underlying conduct of the defendant. We should not lose sight of the fact, however, that often “[t]he best indication of Congress’s intentions ... is the text of the statute itself.” S. Port Marine, LLC v. Gulf Oil Ltd. P’ship, 234 F.3d 58, 65 (1st Cir. 2000). And the text of ACCA indicates that Congress did want us to focus on the offense of conviction rather than on the conduct that the defendant engaged in while committing the offense for which he was ultimately convicted. As the Court has observed, ACCA expressly makes the defendant’s “conviction” the trigger for the enhancement and makes no reference to “conduct.” 18 U.S.C. § 924(e)(1) (“In the case of a person who violates section 922(g) of this title and has three previous convictions by any court referred to in section 922(g)(1) of this title for a violent felony or a serious drug offense.... ” (emphasis added)); Mathis v. United States, - U.S. -, 136 S.Ct. 2243, 2252, 195 L.Ed.2d 604 (2016).

Second, and relatedly, ACCA’s text expressly tells us that, in focusing on the “conviction,” we should not focus on the label that the state legislature may have given to the underlying criminal offense — no matter how violent-sounding that label may be. In denominating crimes of violence, Congress has said nothing about the labels that states have given to offenses. Congress instead has referred only to the actual elements of the offense. Specifically, Congress, in the definition of “violent felony,” expressly orients us to the elements of the predicate crime: “any crime ... that [ ] has as an element the use, attempted use, or threatened use of physical force against the person of another[.]” 18 U.S.C. § 924(e)(2)(B) (emphasis added).

In light of these aspects of ACCA’s plain text, it seems to me that the case is actually quite strong for concluding that Congress, by writing ACCA the way that it did, intended for us to adopt the focus reflected in the extended line of Supreme Court cases that establishes the categorical approach. After all, there are good reasons to privilege the “conviction” and the “elements” that comprise that conviction over either the label that a state legislature may have given to the offense or the nature of the defendant’s conduct in committing the underlying offense. As the Court has observed, such a categorical— some might say, abstract — focus for the inquiry avoids some very real problems that otherwise would arise.

It must be remembered in this regard that the label that a state legislature gives to a criminal offense is not necessarily a good guide for determining whether a con- . viction is for a violent act. That label is, in many ways, no different from a statute’s title. As such, that label is often a poor proxy for the substance of the underlying legislation. For that reason, there is good reason to be wary of giving an offense’s violent-sounding label much interpretive weight — at least when the label does not reflect the fact that the elements of the offense clearly encompass non-violent conduct. See Bhd. of R.R. Trainmen v. Baltimore & O.R. Co., 331 U.S. 519, 528-29, 67 S.Ct. 1387, 91 L.Ed. 1646 (1947) (“[T]he title of a statute and the heading of a section cannot limit the plain meaning of the text.”).

To be sure, Congress, in focusing on the “conviction” rather than on the defendant’s conduct and on the “elements” of the conviction rather than on its label, may not have expected for there to be as much of a mismatch as there often turns out to be between the label that a state gives to a crime and the elements of that crime. For example, Congress apparently envisioned that the generic offense of “assault” would qualify as a violent felony under what is known as ACCA’s force clause, on the thought that the underlying conduct would invariably be violent. See H.R. Rep. No. 99-849, at 3 (1986) (suggesting that the list of “violent felonies under Federal or State law ... would include such felonies involving physical force against a person such as murder, rape, assault, robbery, etc.”).

But, I do not think Congress’s possible misapprehension about the actual elements of a crime that bears an indisputably violent-sounding moniker supplies a reason for us to conclude that Congress wanted us to ignore the reality that the elements of the crime of conviction often actually cover non-violent conduct. See Popal v. Gonzales, 416 F.3d 249, 254 n.5 (3d Cir. 2005) (“In short, we recognize that our holding today might seem to conflict with Congress’s intent as expressed in the legislative history of § 16. But we believe that this apparent conflict is a mirage, caused by the fact that Pennsylvania has chosen to classify as simple assault offenses that the congressional drafters were unlikely to have had in mind.”). In fact, the crimes at issue in this case demonstrate this mismatch between the level of violence implied by the title of the crime and the level of violence required to commit the crime. Stiffening one’s arm to avoid arrest qualifies as resisting arrest, Commonwealth v. Grandison, 433 Mass. 135, 741 N.E.2d 25, 35 (2001), and an offensive touching, even if slight, qualifies as assault and battery, Commonwealth v. Cohen, 55 Mass.App.Ct. 358, 771 N.E.2d 176, 177 (2002). And, even the government appears to accept that such conduct is not “violent” within the meaning of ACCA.

A state’s choice to expand the scope of its crimes to encompass both violent and non-violent conduct may make great sense in terms of state policy. It certainly makes it easier for state and local prosecutors to make their cases. But a state, in so acting, is not attempting to ensure what Congress is attempting to ensure through ACCA— that armed, repeat, violent offenders are treated especially harshly. Rather, a state that enacts a broad assault or robbery statute is just attempting to ensure that anyone who falls with the scope of that state-defined crime may be punished for that particular crime, broadly defined as it is. As a result, there is no reason to think that Congress would have wanted us to rely on state criminal law labels — when they oversell the actual nature of the prohibited conducted that juries must actually find the defendant to have engaged in — to implement Congress’s attempt through ACCA to single out that class of criminals who are repeatedly violent and thus must be specially punished in consequence. And thus there is no reason to think that, in directing us to look to the elements of state law crimes, Congress, in writing ACCA, did not mean what it said.

Of course, a direct focus on a defendant’s actual conduct could, in theory, solve the “problem” that arises from the frequent mismatch between a crime’s name and a crime’s elements. Sometimes, perhaps often, assaults are violent, even though a jury need not find them to be so in order to convict. But, it must also be remembered that, if the defendant’s conduct alone mattered, then some very real problems still would arise.

For one thing, as the Court has pointed out, a characterization of the defendant’s conduct by a federal judge made well after the conviction could violate a defendant’s Sixth Amendment rights if that characterization would increase the maximum penalty faced by the defendant. Mathis, 136 S.Ct. at 2252. Why, then, should we assume that Congress wished for us to focus on that conduct in applying ACCA, given that it would increase that maximum penalty? Haven’t we long construed statutes to avoid constitutional problems rather than to create them, on the understanding (however fictional it may sometimes be) that Congress is not in the habit of pushing constitutional bounds?

There is also another reason to think that Congress directed us to focus on the “conviction” and its “elements” rather than on the defendant’s underlying conduct in order to avoid a problem that might arise if the focus were on conduct. As the Court has observed, a federal judge’s characterization of a crime as a violent one could be quite unfair to a defendant if it were based only on a years-later review of a defendant’s conduct that is, in turn, based only on agreed-upon facts that were adduced at, say, a plea colloquy.

After all, at the time that a defendant decides to plead guilty to a state crime, the defendant would likely have no real notice of the potentially severe federal consequences of a decision not to challenge the-state’s characterization of the defendant’s underlying conduct. Id. at 2253. For example, the first-time offender cannot be presumed to know that he will later commit two additional state crimes and a federal firearms offense, but it is only after those crimes are committed that the first crime’s violent nature becomes relevant. Thus, a federal district court’s reliance on the description of the conduct from the plea colloquy, despite that lack of notice, could therefore' raise serious due process concerns if the district court did not also give the defendant an opportunity to contest that description of the conduct.

Nor would that notice problem be efficiently solved by affording the defendant an opportunity to contest that description in the federal sentencing proceedings. That approach would invite a mini-trial about the facts of the prior offense years or decades after the fact. See Moncrieffe v. Holder, — U.S. -, 133 S.Ct. 1678, 1690, 185 L.Ed.2d 727 (2013) (“The categorical approach ... promotes judicial and administrative efficiency by precluding the relitigation of past convictions in minitrials conducted long after the fact.”). It is hard to make the categorical approach look like an efficient way of determining which offenses qualify as predicate offenses. A conduct-based approach that, in order to protect the due process interests of the defendant, necessitated searching retrospective inquiries into long-since-passed state court proceedings might well accomplish that seemingly impossible feat.

I suppose that an alternative, more conduct-based framework could, in time, become well known enough to defendants that, in response, they would begin routinely to take care in their state court proceedings to protect themselves from the potential future impact of ACAA. But, even if that quite speculative possibility came to pass, it would give rise to a serious problem of its own.

Such a reaction by defendants to a revised, conduct-focused framework could seriously disrupt the dynamics of the state criminal process, both in plea-bargaining and in trials. The concern would be that, under such a revised legal framework, defense counsel, in developing a record for plea or trial, would no longer focus solely on what would appear to be the state’s only concern: the guilt or innocence of the defendant for the crime for which the defendant was then being prosecuted. Instead, the focus would also be on the federal government’s hypothetical future interest in assessing how dangerous the defendant’s underlying conduct was.

But that focus would interfere with the state’s goals in legislating crimes to contain elements that sweep in non-violent conduct. After all, a state may have chosen to group together both violent and nonviolent conduct under a particular label precisely because it did not want the criminal proceedings to focus on whether the conduct in which the defendant engaged was violent or not. In this way, then, a concern solely of the federal government’s — identifying repeat violent offenders — would distort state court criminal proceedings that, by design, seek to make the violent nature of the conduct irrelevant.

Indeed, a version of this federalism concern seemed to be one that Congress took quite seriously in drafting ACCA. Congress rejected, on federalism grounds, initial legislative plans to formulate an alternative to this enhancement for fear that— by creating a federal crime of robbery and burglary and the like — the federal government would displace the state criminal process. See H.R. Rep. No. 99-849, at 3 (noting that the House Subcommittee on Crime “delet[ed] ... specific predicate offenses ... and added as predicate offenses ... violent felonies under Federal or State law if the offense has as an element the use, attempted use, or threatened use of physical force against a person” (emphasis in original)). Thus, the Court’s current categorical approach — by attending to the textual directives to focus on the “conviction” and the “elements” of the offense— has the virtue of ensuring that ACCA does not give rise to a version of the federalism-based concern about respecting the primacy of state criminal law over a broad range of crimes that we know Congress took quite seriously in crafting ACCA as it did.

II.

Ordinarily, there is good reason to assume that Congress does not intend to enact criminal punishment schemes that are overly complicated. Such complicated schemes may fail to provide defendants with sufficient notice of the consequences of their actions. But the body of jurisprudence that has emerged from Taylor through Mathis does not give rise to this notice-based concern.

The complexity of the categorical approach, as it has developed, is undeniable. This case makes that perfectly clear. But, as this case also demonstrates, in practice, that complexity serves to narrow the scope of this mandatory sentencing enhancement, at least as compared to the breadth that it might otherwise have. Thus, the current categorical approach respects— rather than violates — the notice-protecting principle of lenity that we have long presumed that Congress has in mind when it imposes severe criminal punishment.

Perhaps, insofar as such an enhancement remains in place, there is some alternative to the categorical framework that Congress might adopt that would be superior to the one that the Supreme Court, based on the text of ACCA, requires us to administer. I do not deny that such a framework might be discovered. I do think, though, that we should not discount the very legitimate concerns that led Congress to direct us to focus on the “conviction” and its “elements” — rather than on the defendant’s conduct or on statutory labels — in drafting ACCA. And I think it important, as well, to highlight the related and quite sensible assumptions .that have led the Court to interpret ACCA — and the directive Congress wrote into it — as it has.

The categorical approach is difficult to apply. And its application may in many cases seem to exclude from ACCA’s reach defendants whose past records appear to be violent ones. But, the simple fact is that it is hard to devise a system for identifying those individuals who as a class — and thus regardless of particular circumstances that could be evaluated through individualized sentencing — must be sentenced very harshly because of the violence that they have perpetrated in the past. It is thus important, it seems to me, not to lose sight of the significant ways in which the categorical approach, for all of its faults, reflects respect both for due process and federalism. Otherwise, we may find ourselves, in time, administering a revised framework that, though supposedly improved, actually creates problems more serious than those that it sought to solve.

Case Details

Case Name: United States v. Faust

Court Name: Court of Appeals for the First Circuit

Date Published: Apr 5, 2017

Citations: 853 F.3d 39; 2017 WL 1244844; 2017 U.S. App. LEXIS 5918; 14-2292P

Docket Number: 14-2292P

Court Abbreviation: 1st Cir.