The hate crime enhancement in the federal sentencing guidelines, U.S.S.G. § 3A1.1(a), increases a defendant’s offense level by three if they “intentionally selected any victim . . . because of the actual or perceived” protected characteristic “of any person.”1 Though the enhancement is in relevant part titled “Hate Crime Motivation,” its text does not mention hate,2 and courts differ on how to interpret the guideline and similar statutory language.3 Recently, in United States v. Patterson,4 the Ninth Circuit held that section 3A1.1(a) applies only when the defendant was motivated by hatred or animus.5 However, rather than requiring hate motivation, a better reading of the enhancement would require only that the protected trait be a conscious reason supporting the defendant’s selection of the victim. This reading would include cases like Patterson, where the victim’s protected characteristic was a deliberate proxy for another trait.
Derrick Patterson robbed a series of gay men.6 Each time, he would find his victim on the gay-focused dating app Grindr, “lure[]” him “by promising gay sexual conduct,” and then rob the victim while alone with him.7 At sentencing for one such deed before the U.S. District Court for the Central District of California, the government argued that section 3A1.1(a) applied because it did not require animus and Patterson strategically targeted gay men for his robberies.8 Patterson argued that the guideline required hatred and that he did not target gay men out of hate but because, as he was gay himself, Grindr was how he met people.9
The district court applied the enhancement, holding that it did not require hate — only that “the defendant intentionally selected any victim on the basis of” a protected trait, no matter the “underlying motivation.”10 The court found that Patterson purposefully “targeted gay men” so that he could “exploit[] his victims’ sexual interest in him and his familiarity with gay culture to make his robberies easier.”11 It found that Patterson “belie[ved] that gay men shared characteristics that made them . . . vulnerable robbery targets,” like willingness to meet alone with strangers and susceptibility to blackmail.12
The Ninth Circuit “vacat[ed] the sentence and remand[ed].”13 Writing for the panel, Judge Tashima14 held that section 3A1.1(a) “require[d] a finding that the defendant was motivated by hate or animus.”15 He explained that the text was “ambiguous,” noting that the defense identified multiple viable meanings of intentionally choosing a victim “because of” a protected trait, including that the trait was a “‘but for’ cause . . . or . . . the primary impetus for the selection.”16 Given the ambiguity, Judge Tashima turned to other interpretive tools; he reasoned that, “though the word ‘motivation’ is” not in the text, “the title, history, and purpose of the guideline” show that it requires hate motivation.17
Judge Tashima relied on the title to contextualize the plain text; he pointed out that “the title, Hate Crime Motivation, has a focused meaning based on the ordinary understanding of a hate crime,” which can prevent an interpretation based “on the broadest imaginable definitions of [the text’s] component words.”18 He analogized to Dubin v. United States,19 where a statute titled “aggravated identity theft” was read not to apply to incidental uses of another person’s identification in the course of a crime.20 Just as the ordinary meaning of “identity theft” constrained the meaning of “in relation to” to refer to misuses of identification that specifically further the offense’s criminality, the ordinary meaning of “hate crime motivation” limited the potentially broad text of section 3A1.1(a) to require an actual motivation of hatred.21
The opinion also cited extratextual sources of meaning, including the Sentencing Commission’s commentary on the guideline and the legislative history of the statute that directed the Commission to promulgate section 3A1.1(a).22 The commentary states that the statute called for a sentencing enhancement that applied when “the defendant had a hate crime motivation.”23 And the House Report for the statute’s predecessor bill stated that the victim selection “must result from the defendant’s hate or animus toward any person for bearing” a protected “characteristic[].”24 The court also noted that the Sentencing Guidelines Manual describes the enhancement as involving “crimes motivated by hate” in its stated reasons for an amendment to the guideline.25 To validate his reading, Judge Tashima identified a district court case that read section 3A1.1(a) as requiring hatred,26 as well as circuit court opinions construing similar “because of” language in other statutes as requiring animus.27
Judge Tashima concluded that Patterson did not choose his victims out of hatred but used their sexuality as “a proxy for other information about them — their willingness to agree to meetings where he knew they would be alone”; absent a finding of “hatred or animus,” this strategic motive for picking gay victims was “insufficient” for the enhancement.28
Finally, Judge Tashima explained why In re Terrorist Bombings of U.S. Embassies in East Africa,29 a case cited by the district court, was compatible with his analysis.30 In Terrorist Bombings, the defendant stated that he chose American victims for “political” reasons rather than out of hate.31 The Second Circuit held that the defendant’s “underlying motivation [was] simply beside the point” so long as he “intentionally selected” the victims based on their nationality.32 Though this holding seemed to reject a hate requirement, Judge Tashima explained that the “distin[ction] between purely political and purely hate motives” was rendered “irrelevant to the application of the enhancement” in Terrorist Bombings because “there was no question that the defendants targeted Americans when they chose to bomb the United States embassies.”33 Thus, he did not see this reasoning “as controlling” in Patterson’s case.34
Judge Christen concurred, arguing that the text of section 3A1.1(a) was unambiguous but still did not apply to Patterson.35 She explained that both the guideline and the statute directing its promulgation “unambiguously require[d] . . . that the defendant selected the victim because of the victim’s protected characteristic,” and that the only way to read this language was as “bak[ing] in the requirement that the protected characteristic . . . was the defendant’s motivation for selecting the victim,” with “nothing more . . . required.”36
She further argued that since the enhancement required that the protected trait be “a motivation for the defendant’s selection of victims,” it would not apply when the trait served as a strategic “proxy.”37 In cases like Patterson’s, or that of a hypothetical woman whose “modus operandi” for theft involved luring men into her home, it was not the victims’ protected attribute that “motivated [the defendant] to steal from them”;38 rather, the defendant had “some other motivation that the defendant may have correlated with the protected class, such as vulnerability.”39 In contrast, section 3A1.1(a) would apply to cases like Terrorist Bombings because the defendant chose his targets not “as a convenient proxy” but directly “because of their status as U.S. nationals,” regardless of whether the motive was purely political or hate based.40
The best reading of the enhancement would not require animus, and it would cover “proxy” cases so long as the proxy was deliberate. While courts often look to context and history to clarify meanings, the Patterson court used these tools to reach a meaning unsupported by the plain text itself. Instead of requiring hate motivation, section 3A1.1(a) should be read as requiring that the protected trait be an “intentional[]”41 reason the victim was chosen, including the conscious use of the trait as a proxy.
The guideline’s text does not support imposing a hate requirement. It requires only that the defendant “intentionally selected” the victim “because of” the protected trait — with no mention of animus or any other motivation.42 Relying on Dubin, Judge Tashima justified the departure from the plain text by explaining that “even a common phrase such as ‘because of’ can be ambiguous and require reliance on statutory context and history to interpret a statute.”43 But statutory context in Dubin was meant to provide clarity when “[t]he parties’ competing readings both fall within the range of meanings of [the statute’s words] taken alone.”44 In Dubin, “key terms” of the statute — “use” and “in relation to” — were “indeterminate” and “context-dependent.”45 To “use[] a means of identification” can have a broad meaning46 that includes “‘passive,’ ‘passing,’ or ancillary employment in a crime,”47 or a more narrow meaning that “impl[ies] action and implementation”48 that is “at ‘the locus of [the criminal] undertaking.’”49 Likewise, “in relation to [a predicate offense]” can be read broadly as denoting “a relationship . . . of some kind,” with essentially “no limits” to its scope,50 or narrowly as referring to acts “at the crux of the underlying criminality.”51 The Dubin court used the statute’s title to choose between these “competing” options by “point[ing] to a narrower reading”52 and “reinforc[ing] what the text’s nouns and verbs independently suggest[ed].”53
In contrast, a hate requirement does not “fall within the range of meanings” of the enhancement’s text “taken alone.”54 Taking the text alone, a defendant may choose a victim “because of” a protected trait for reasons other than hatred, from fetishization to stirring up support for the targeted group.55 In Dubin, the ambiguity was a matter of degree — context helped determine the right “strength,”56 “central[ity],”57 and “genuine[ness]” of the “nexus” between the use of identification and the criminality.58 But whether hate is required for section 3A1.1(a) is a matter of kind. Instead of determining the specific strength of the causal link between the protected trait and the victim selection, inserting a hate requirement would limit the enhancement to causal links of a certain substantive character, in a way not naturally understood from the text itself.59 In fact, when describing the linguistic ambiguity of “because of,” Judge Tashima did not claim that hate motivation was in the range of possible meanings and instead listed two nonhate readings of the phrase: “that the protected [trait] is the ‘but for’ cause of . . . or . . . the primary impetus for the selection.”60 If the contextual tools were meant to isolate or corroborate an existing plausible reading, then they could only help to decide among these and similar notions of “because of,” which mainly vary in the strength of the causal link.61 Meanwhile, the title cannot “supplant the actual text” by imposing an interpretation, like one requiring hate, that was not itself among the viable readings.62
While Judge Tashima’s hate requirement is unsupported by the text, Judge Christen’s broader account of “because of” raises problematic line-drawing issues. Under her reading, the victim’s identity must have been a direct motivation for the selection, rather than a “proxy” for some associated trait.63 But reasons for action can be described at different levels of proximity to the act. For instance, a person buying apples may believe that green apples are generally juicier and that juicy apples are usually tastier. If they choose a green Granny Smith over a red McIntosh, one could say that they selected the apple because of its color, because of its presumed juiciness (for which its color was a proxy), or even because of its expected tastiness (for which its juiciness was a proxy). Judge Christen does not specify when a feature becomes a “motivation” for selection and when it remains a “convenient proxy.”64
Professor Joseph Raz’s distinction between operative and auxiliary reasons for action is one intelligible, but ultimately flawed, way to draw that line. He explains that an operative reason is one whose adoption “entails having [a] practical critical attitude,”65 or an “attitude towards” different “behaviour[s]” that then “manifests itself in action,”66 while an auxiliary reason “transmits . . . the force of the operative reason to the particular act.”67 So in the apple case, the buyer’s operative reason for the choice is a desire to eat tasty food, and the auxiliary reasons are the beliefs that the apple is green, green apples are juicier, and juicy apples taste better. Through this lens, the guideline would apply when the defendant’s operative reason for selection is a desire to victimize those with the protected trait, or a belief that targeting them is itself good; and it would not apply when the operative reason is a wish to do some other thing — like acquire money — that is then joined by an auxiliary belief identifying the victim as suitable for achieving this goal due to the trait.
The problem, though, is that many victim selections that we recognize as being “because of” a protected trait can be reframed to evade this definition. Instead of believing that targeting the protected class is itself good, the offender’s operative reason might be a desire to victimize those with a trait they associate with the class, and the protected characteristic would, in their mind, act only as a “convenient proxy” for the associated trait. For instance, an anti-LGBT attacker might seek to thwart “predators” and (falsely) believe that LGBT people are likely to fit this profile;68 or someone targeting Muslims may be driven by an urge to fight “terrorists” and (falsely) view Islam as a reliable proxy for this trait.69 For them, the proximate motivation for choosing the victim is not the sexuality or religion per se but rather the victim’s likelihood of being a “predator” or “terrorist,” which the attacker thinks is indicated by the protected trait. Or consider “gay panic” cases, where the defendant attacks a victim of the same sex in response to a romantic or sexual advance.70 There, the operative reason may be a desire to avoid humiliation or association with gay people,71 and the belief that targeting the victim serves this aim would be an auxiliary reason for picking them. In all these cases, the attacker can argue that the victim’s protected trait was not what “motivated” the selection and that it instead only singled out the victim as a suitable target in fulfilling “some other motivation.”72
Thus, it is hard to find a principled distinction between when the protected trait itself is a motivation for choosing the victim and when it is a proxy or auxiliary reason for the selection. One way to avoid this confusion is to interpret the enhancement straightforwardly: Selecting a victim “because of” a protected trait means simply that the trait was a reason that the defendant employed for selecting the victim, regardless of whether the reason was auxiliary or operative, direct or proxy.73
Crucially, the word “intentionally” in section 3A1.1(a) constrains the enhancement so that it applies only to deliberate, rather than instinctive or incidental, reasons for victim selection.74 If “intentionally” modifies the full phrase of “selected any victim . . . because of the [protected trait],”75 then the defendant must have intended to select the victim because of the protected trait. As Professor Gideon Yaffe argues, acting “intention[ally]” as to a circumstantial element of a crime requires “occurrently” recognizing it as a “salient fact.”76 Applied here, that means the defendant must have actively recognized the fact that the protected trait was serving as a reason for his selection of the victim.77
This formulation would thus exclude cases where the defendant instinctively selects victims from his own romantic community, or where the “victims happen[] to be [in the protected class] as a consequence of the [defendant’s] particular modus operandi.”78 It would also exclude cases of implicit bias, where the trait’s influence on the defendant’s reasoning is unknown to them.79 But it would cover cases where the defendant, like Patterson, consciously relies on stereotypes about a group as justification for targeting its members, whether or not the rationale is hate based or only instrumental. As well as realizing the best reading of the text, focusing on stereotyping might then help the guideline realize the goal of deterring crimes that “transcend their immediate victims and cast a shadow of fear and terror throughout entire communities.”80