United States Court of Appeals For the First Circuit

No. 20-1694

JONALSON DOR,

Petitioner,

v.

MERRICK B. GARLAND,* Attorney General,

Respondent.

PETITION FOR REVIEW OF AN ORDER OF THE BOARD OF IMMIGRATION APPEALS

Before

Thompson, Howard, and Gelpí, Circuit Judges.

Edward Crane, with whom Philip L. Torrey, Crimmigration Clinic, Harvard Law School, Shaiba Rather, Lena Melillo, and Katie Quigley, Law Student Advocates, Crimmigration Clinic, Harvard Law School, were on brief, for petitioner. Andrew B. Insenga, Trial Attorney, U.S. Department of Justice, Civil Division, with whom Jeffrey B. Clark, Acting Assistant Attorney General, Civil Division, and Matthew B. George, Senior Litigation Counsel, Office of Immigration Litigation, were on brief, for respondent.

* Pursuant to Fed. R. App. P. 43(c)(2), Attorney General Merrick B. Garland has been substituted for former Attorney General William P. Barr. August 19, 2022 THOMPSON, Circuit Judge. Petitioner, Jonalson Dor

("Dor"), seeks judicial review of a Board of Immigration Appeals

("BIA") decision affirming an immigration judge's ("IJ") decision

to deny Dor's applications for relief from removal based on two

marijuana offenses that the IJ and BIA found, for different

reasons, to be "particularly serious" pursuant to

(b)(2)(A)(ii) and 1231(b)(3)(B)(ii). Before us, Dor argues

that the BIA failed to conduct the appropriate analysis to reach

its particularly-serious-crime conclusion. The government urges

that we shouldn't even reach Dor's substantive challenges due to

his petition's jurisdictional defects, but even if we can find our

way to the merits, the government maintains the BIA's decision was

correct.

We find we have jurisdiction to review the petition.

And, having undertaken that review, we remand to the BIA for

further proceedings consistent with this opinion.

BACKGROUND

We begin our work by recounting the relevant parts of

Dor's story and reciting the procedural history that brought him

to us, pulling all relevant facts from the administrative record.

See Adeyanju v. Garland,

(1st Cir. 2022) (citing

Martínez-Pérez v. Sessions,

, 37 n.1 (1st Cir. 2018)).

Dor is a native and citizen of Haiti who was admitted to

the United States as a legal permanent resident back in 2007. But

- 3 - in April 2019, the Department of Homeland Security initiated

removal proceedings against Dor, charging him with removability as

a non-citizen convicted of a criminal offense relating to a

controlled substance.

After a hearing, the IJ found Dor removable based on two

2016 Massachusetts state court convictions: one for distribution

of $20 worth of marijuana, on May 20, and one for possession of

what a police report says was "a large amount" (25 grams) of

marijuana with the intent to distribute, on June 1. Dor then filed

for various forms of relief, including applications for asylum,

statutory withholding of removal, and withholding of removal

pursuant to the United Nations Convention Against Torture.1

In a written decision denying Dor's applications for

relief, the IJ found Dor ineligible for the relief he was seeking

because his convictions were particularly serious crimes. By way

of explanation, the IJ said that Dor had a "large amount of

marijuana," and, under Matter of Y-L-, "all drug trafficking

offenses are per se 'particularly serious crimes.'"

(A.G. 2002). The IJ observed that sometimes a drug-

1 Dor also filed applications for cancellation of removal and voluntary departure. Both were denied by the IJ, and Dor is not appealing those decisions. Rather, as we've touched on and will explain, Dor's arguments target what he sees as the BIA's flawed analysis of the particularly-serious-crime bar to asylum (

(b)(2)(A)(ii)) and withholding of removal (

(b)(3)(B)(ii)).

- 4 - trafficking offense is not a particularly serious crime -- this is

true when the following rare and extraordinary circumstances are

shown: (1) "a very small quantity of controlled substance"; (2)

"a very modest amount of money paid for the drugs"; (3) "peripheral

involvement . . . in the criminal activity"; (4) absence of any

violence or threat thereof; (5) absence of organized crime; and

(6) "absence of any adverse or harmful effect of the activity or

transaction on juveniles."

(providing this list and

instructing that all criteria must be met for a court to scrutinize

the default setting that all drug-trafficking felonies are

particularly serious crimes). In the IJ's view, though, Dor could

not even satisfy the first of those required factors since he was

found with a "large amount of loose green leafy vegetable matter

believed to be marijuana." Thus the IJ ordered Dor removed to

Haiti.

Dor timely appealed to the BIA, arguing that the

presumption set forth in Matter of Y-L- that "all drug trafficking

offenses are per se 'particularly serious crimes'" applies only

when the offenses are aggravated felonies, and his convictions

were not. So, according to Dor, the BIA should remand the matter

to the IJ to instead apply the multi-factor test for convictions

that are not aggravated felonies, as set forth in Matter of

Frentescu, to determine whether either of his convictions amounted

to a particularly serious crime. See

, 247

- 5 - (BIA 1982) [hereinafter "Frentescu"] (listing "the nature of the

conviction, the circumstances and underlying facts of the

conviction, the type of sentence imposed, and, most importantly,

whether the type and circumstances of the crime indicate that the

alien will be a danger to the community" as the relevant factors

to consider for this analysis).

The BIA agreed with Dor that the presumption in Matter

of Y-L- that "all drug trafficking offenses are per se

'particularly serious crimes'" should not have been applied to his

case because his convictions were not aggravated felonies. But

instead of remanding to the IJ on this basis, the BIA observed

that "whether an offense is a particularly serious crime is a

question of law [it would] review de novo." The BIA, citing

Frentescu (laying out relevant factors to make the particularly-

serious-crime determination) and Matter of N-A-M-,

(BIA 2007) (same, and also instructing the immigration

agencies to look at the elements of the statute of conviction to

see whether the elements of the offense bring the crime into the

particularly-serious-crime category), then indicated that

"[w]here, as in the instant case, a conviction is not for an

aggravated felony . . . , [the BIA] examine[s] the nature of the

conviction, the type of sentence imposed, and the circumstances

and underlying facts of the case." "If the elements of the offense

are found to potentially bring it within the ambit of a

- 6 - particularly serious crime," the BIA went on, again citing Matter

of N-A-M-, "all reliable information that is relevant to the

determination may be considered."

In the paragraph critical to Dor's appeal to this court

(we'll lay it out in full later), the BIA stated that "[t]he

nature, type, and circumstances of [Dor's] offenses are all

indicative of a particularly serious crime," recited a number of

the IJ's findings regarding both the May 20 and June 1 offenses,

and then, "[f]or these reasons," upheld "the [IJ's] determination

that [Dor was] ineligible for asylum and withholding of removal

for having been convicted of a particularly serious crime."

Believing the BIA committed a variety of errors, Dor

filed this timely petition for review, and we granted his

subsequent motion to stay his removal.

JURISDICTION

We start by examining our jurisdiction, meaning we'll

save for later our sum-up of Dor's merits arguments regarding the

BIA's missteps. Our jurisdiction over petitions for review of BIA

decisions is constrained by statute. And indeed, the government

maintains we shouldn't reach any of Dor's contentions because

(a)(2)(C) precludes judicial review and, what's more,

Dor failed to exhaust all administrative remedies. Below, we take

these -- and Dor's responses to each -- in turn.

- 7 - Section 1252 and Our Jurisdiction

Generally, we lack jurisdiction to review a final order

of removal when the respondent has committed a criminal offense,

see

§ 1252(a)(2)(C) (the so-called "criminal bar"), but, that

provision notwithstanding, we do have jurisdiction over petitions

that raise "constitutional claims or questions of law," id.

§ 1252(a)(2)(D) (the "limited review provision"). The phrase

"questions of law," as used in § 1252(a)(2)(D), is not limited to

pure questions of law -- it also "includes the application of a

legal standard to undisputed or established facts." Guerrero-

Lasprilla v. Barr,

(2020); Valerio-

Ramirez v. Sessions,

(1st Cir. 2018) (concluding

that the court had jurisdiction, in removal proceedings, to address

questions of law raised by petitions for review, "includ[ing] what

standard governs 'particularly serious crime' determinations for

non-aggravated felons in deportation proceedings"). And not only

can we exercise jurisdiction when a petition raises an argument

about such application-of-a-legal-standard questions of law, but

we can do so when a petition challenges the sufficiency of that

application -- i.e., the adequacy of the reasoning. See, e.g.,

Berhe v. Gonzales,

(1st Cir. 2006) (instructing

that "the adequacy of the [BIA]'s reasoning is a legal question

that we may review").

- 8 - In asserting that the criminal bar precludes

jurisdiction here, the government nods at our precedent (Valerio-

Ramirez, specifically) and agrees that we would not lack

jurisdiction over "alleged legal errors raised about a

particularly serious crime[]" analysis. But it posits that Dor

only "provides a scattershot of allegations that do not identify

the nature of the error," so really, according to the government,

Dor is just asking us to reweigh facts to find his crime was not

particularly serious, and reweighing facts, the government urges,

is not something we can do.2

A close review of his arguments confirms that this is

not what Dor is doing. Indeed, this is not a situation in which

a factual challenge is masquerading as a legal one. See, e.g.,

Tacuri-Tacuri v. Garland,

(1st Cir. 2021)

(observing, in a cancellation of removal case, that "styling a

factual challenge as a constitutional or legal error . . . does

not 'transform an unreviewable issue of fact into a reviewable

issue of law'" (quoting Alvarado v. Holder,

(1st

2 In its opening brief, the government leans on Bare v. Barr,

(9th Cir. 2020), to support its criminal-bar jurisdictional arguments. As the government later acknowledges in a Federal Rule of Appellate Procedure 28(j) submission, that case concerns the application of the discretionary bar, § 1252(a)(2)(B). After recognizing its mistaken reliance on that case, though, the government uses the Rule 28(j) moment to reiterate its argument that Dor is basically asking us to reweigh facts.

- 9 - Cir. 2014))). When Dor points, for instance, to the relative

leniency of his sentence, or the fact that the "large amount" of

marijuana as described in the police report really only amounted

to 25 grams, he is not asking us to reweigh these facts to then

find his crime wasn't particularly serious. Rather, Dor's

reference to such facts and circumstances is designed to buttress

his argument that the BIA's analysis -- its reasoning and

application of the Frentescu test to the facts to which Dor is

pointing -- was either entirely absent or, at a minimum,

deficiently explained. Dor says this all comes together to present

a classic example of a question of law over which we retain

jurisdiction to consider even when the criminal bar would otherwise

apply.

We agree. In view of guiding precedent, and with Dor

clearly seeking our review of the BIA's Frentescu application and

adequacy of its discussion, the petition presents questions of law

for our determination. See Guerrero-Lasprilla,

;

Valerio-Ramirez,

; Berhe,

. Thus,

the criminal bar does not limit our jurisdiction.

Exhaustion

And then there's exhaustion, which is yet another piece

of the jurisdictional puzzle.

We "may review a final order of removal only if . . .

the alien has exhausted all administrative remedies available to

- 10 - the alien as of right."

(d)(1). "This exhaustion

requirement is jurisdictional; that is, it constitutes a

limitation on our power of review" -- a petitioner "who neglects

to present an issue to the BIA fails to exhaust his administrative

remedies with respect to that issue and, thus, places it beyond

our jurisdictional reach." Mazariegos-Paiz v. Holder,

(1st Cir. 2013). In other words, "theories not advanced

before the BIA may not be surfaced for the first time in a petition

for judicial review of the BIA's final order." Makhoul v.

Ashcroft,

(1st Cir. 2004).

The government tells us that "essentially every

argument" Dor advances is unexhausted. This is so, according to

the government, because Dor now makes "legalistic claims that the

[BIA] did not and could not address" because his "brief to the

[BIA] failed to raise any of the arguments he now raises."

Instead, the government argues, he "asked [the BIA] for remand

without presenting any arguments on the merits," leaving the BIA

"to guess" at what his argument would be relative to how

Frentescu's analysis should go in his case. "The consequence of

[this] failure to raise to the [BIA] any [of] the merits of the

particular [sic] serious crime issue is that he did not exhaust

most of his claims."

Dor rejoins that, based on the IJ's misanalysis under

Matter of Y-L-, he asked the BIA to remand so the IJ could conduct

- 11 - a proper Frentescu-factor assessment to determine whether he had

committed a particularly serious crime. But the BIA undertook its

own Frentescu discussion sua sponte, which he says he cannot be

expected to have foreseen, and the BIA's choice to tackle the

matter on the merits exhausts the issue anyway.

"[A]n issue is exhausted when it has been squarely

presented to and squarely addressed by the agency, regardless of

which party raised the issue (or, indeed, even if the agency raised

it sua sponte)." Mazariegos-Paiz,

; see also García-

Cruz v. Sessions,

(1st Cir. 2017) ("Even if an issue

was not raised by a party, the issue is exhausted if the BIA

addresses the issue on the merits."); Velerio-Ramirez v. Lynch,

(1st Cir. 2015) ("We reject the government's

position that the petitioner has precluded remand because she

failed to exhaust the issue of applicable law; the BIA itself

raised the issue, and that suffices."); Xin Qiang Liu v. Lynch,

(1st Cir. 2015) ("The exhaustion requirement is

satisfied where . . . the agency chooses to address the merits of

a particular issue, regardless of whether the alien raised that

issue." (alteration in original) (quoting Meng Hua Wan v. Holder,

(1st Cir. 2015))).

Indeed,

by addressing an issue on the merits, an agency is expressing its judgment as to what it considers to be a sufficiently developed issue. When a court defers to

- 12 - that exhaustion-related judgment, it avoids judicial intrusion into the domain that Congress has delegated to the agency. We think it follows that if the BIA deems an issue sufficiently presented to warrant full-dress consideration on the merits, a court should not second- guess that determination but, rather, should agree that such consideration exhausts the issue.

Mazariegos-Paiz,

(citations omitted).

In view of these important carveouts from the usual

exhaustion constraints and the policy driving it all, and on the

facts of Dor's case, it is clear exhaustion does not operate as a

bar to our review. We're aware that Dor did raise the issue of

the IJ's use of the wrong legal test -- remember, Dor appealed the

IJ's decision, arguing that the IJ applied the wrong legal

standard, then identifying the correct one -- but he sought remand

so the IJ could reassess his claims for relief using the proper

legal framework. In specifically seeking remand as the remedy to

the IJ's misstep, Dor reasonably followed the pathway set forth in

the BIA's own precedent, arguing that the BIA had remanded when

confronted with identical misanalysis. See Brief for Respondent

on Appeal to the BIA at 2-3, In re Dor (Mar. 9, 2020) (relying on

Matter of J-F-B- (BIA Sept. 13, 2018) (unpublished), where the IJ

incorrectly found a per se particularly serious crime under Matter

of Y-L-, 23 I. & N. Dec at 270, and the BIA remanded because Matter

of Y-L- was inapplicable (there had been no aggravated felony

conviction) and the IJ needed to conduct its particularly-serious-

crime analysis by applying Matter of N-A-M- and Frentescu). The

- 13 - BIA, though, sua sponte opted to forgo remand and took it upon

itself to tackle the appropriate particularly-serious-crime

discussion. In other words, the BIA considered the issue

sufficiently developed such that it could handle it itself, giving

it de novo consideration on the merits. This "exhausts the issue."

Mazariegos-Paiz,

.

And so, where, as here and in Mazariegos-Paiz, "the BIA

undertook a developed discussion of the merits-related issues that

the petitioner now seeks to raise," "this court has jurisdiction

to consider those issues notwithstanding the fact that" the

petitioner isn't the one who asked the BIA to decide them.

Equipped with jurisdiction, we move to the substance of

Dor's appellate contentions.

THE MERITS

No one disputes that the BIA was right to identify

Frentescu's case-by-case inquiry as the test to be applied to tease

out whether Dor's crimes were particularly serious as a matter of

law. It's what happened after the BIA identified Frentescu as the

appropriate legal test that forms the basis for today's dispute.

Dor says the BIA erred when it didn't actually apply

Frentescu or adequately explain its reasoning when it addressed

the particularly-serious-crime determination -- it just listed

some IJ factual findings and concluded "particularly serious,"

never getting around to the "why" of that conclusory determination.

- 14 - In particular, Dor posits that the BIA committed error and/or

abused its discretion (we'll get to the standard of review in a

minute) by: cumulatively conflating the circumstances of his two

convictions to find a particularly serious crime when it "made no

attempt to independently examine the underlying allegations of

each offense"; failing to consider whether either offense as a

standalone could bring it within the particularly-serious-crime

category; neglecting to weigh the lenient sentence imposed on Dor;

and ignoring critical facts relative to the underlying

circumstances of his offense conduct (no intent to harm anyone,

for example) while impermissibly looking at unrelated offenses

(considering the June 1 offense when scrutinizing the May 20

offense and vice versa).

The government responds by insisting that the BIA

reviewed the offenses separately and sufficiently analyzed both to

support a particularly-serious-crime determination as to each.

The government also urges that the BIA did consider the relevant

factors (like the nature of Dor's crimes), its emphasis was on the

nature of Dor's trafficking as particularly serious, and any

confusion about the "large amount" of marijuana is "a red herring"

that presumes a faulty reading of the BIA decision.3

3 The government's brief focused almost exclusively on jurisdiction and exhaustion, and these buzzwords pop up even in the government's short merits discussion as reasons we should not be taken in by the substance of Dor's arguments. But we're not

- 15 - In the normal course, we review only the decision of the

BIA (unless the BIA defers to or adopts the IJ's reasoning, that

is, in which case we'd review those portions of the IJ's analysis,

too). See, e.g., Hasan v. Holder,

(1st Cir. 2012).

Here, as we know, the BIA rejected the IJ's reasoning and supplied

its own, so we're just looking at the BIA's decision today. As we

undertake that examination, we're mindful that "[w]e review for

abuse of discretion the BIA's assessment and weighing of the

Frentescu factors, including its conclusion that the crime of

conviction was 'particularly serious.'" Valerio-Ramirez,

. "Under this deferential standard, we will uphold the

determination unless it was made without a rational explanation,

inexplicably departed from established policies, or rested on an

impermissible basis."

(cleaned up). "We also review legal

questions de novo, while affording deference to the BIA's

interpretations of the statutes and regulations it administers."

Dor is right that we need only be convinced on one of

his appellate contentions to send it back to the BIA for further

proceedings. More than one of his arguments has teeth, but there's

no need to explore the nuances of each when a particular path to

remand is readily apparent to us: The BIA's decision did not apply

persuaded. For the reasons we've already explained, we view Dor's appellate contentions as ripe for our discussion.

- 16 - Frentescu to the facts, and even to the extent it could be argued

the BIA did conduct such an application by listing off facts, the

conclusion that followed was bereft of any meaningful or rational

explanation.

Let's begin with what Frentescu demands. "In judging

the seriousness of a crime" under Frentescu, the agency must "look

to such factors as the nature of the conviction, the circumstances

and underlying facts of the conviction, the type of sentence

imposed, and, most importantly, whether the type and circumstances

of the crime indicate that the alien will be a danger to the

community."

. Additionally, "[c]rimes

against persons are more likely to be categorized as 'particularly

serious crimes.'"

"Nevertheless, we recognize that there may

be instances where crimes (or a crime) against property will be

considered as such crimes."

The BIA singled out Frentescu's "the nature of the

conviction, the circumstances and underlying facts of the

conviction, [and] the type of sentence imposed" language. To

appreciate precisely what the BIA did next -- and what it did not

do -- we reproduce the paragraph in its entirety (cleaned up a

little, with exhibit and IJ decision citations omitted):

The nature, type, and circumstances of the respondent's offenses are all indicative of a particularly serious crime. The [IJ] cited the language of the statute of conviction in her decision. See Mass. Gen. Laws Ch. 94C § 32C(a) (2016). The [IJ] noted that the respondent was

- 17 - found to have a "large amount of loose green leafy vegetable matter believed to be marijuana contained in a large plastic powder drink container." During the same incident on June 1, 2016, the [IJ] stated that police officers also found a digital scale and "a large amount of US Currency with various denominations bundled together." Police officers had observed the respondent smoking inside a park, and when the officers approached the respondent, the respondent "quickly stood up and began to walk away." The police officers caught up to the respondent and "observed a brown colored tobacco leaf in his hand and an odor of burnt marijuana emanating from it." In a prior incident on May 20, 2016, a police officer observed the respondent "reach into a tan color bag and retrieve an item with his right hand." The officer then observed the respondent "place the item in his left hand and have a hand to hand exchange with [a buyer] as [the buyer] was holding and gave [the respondent] US currency with his right hand." The officers then approached the buyer, who admitted to buying marijuana from the respondent. On December 12, 2016, the respondent pled guilty to possession with intent to distribute a class D controlled substance and distribution of a class D controlled substance in violation of Massachusetts law.

This is immediately followed by a new paragraph that states: "For

these reasons, we uphold the [IJ's] determination that the

respondent is ineligible for asylum and withholding of removal for

having been convicted of a particularly serious crime," and "[t]he

respondent's conviction for a particularly serious crime also

precludes him from being granted withholding of removal under the

Convention Against Torture." The decision then segues to deferral

of removal (not on appeal).

This is an insufficient and unclear assessment and

weighing of the Frentescu factors that amounts to a deficient

application of the legal standard. Merely identifying the

- 18 - appropriate legal test and some of the relevant factors to be

assessed, saying the test is indicative of a certain conclusion,

then listing certain facts as found by the IJ, does not constitute

an application of law to facts. Compare Valerio-Ramirez,

(affirming the agencies' decisions denying relief when

the IJ and BIA carefully applied Frentescu, weighing the nature

and circumstances that brought the crime within the ambit of

particularly serious crimes, examining the mandatory minimum

sentence imposed and bases for its imposition, and conducting a

detailed inquiry into how the crime showcased the danger to the

community posed by the petitioner).4 There is no way to discern

which facts go to which element, no hint as to what amount of

weight was attributed to each factor, and no indication whether

the BIA considered Frentescu's directive to consider the type of

sentence Dor got or ("most importantly") whether Dor would pose a

4 Compare also Nethagani v. Mukasey,

(2d Cir. 2008) (upholding the BIA's particularly-serious-crime determination on a reckless endangerment conviction because it had properly applied Frentescu, considering: that the conviction involved behavior that "could end a human life," that firing a pistol into the air (as the petitioner had done) involves "high potential for serious or fatal harm to the victim or an innocent bystander," the "not insignificant" sentence the petitioner received, and the petitioner's own version of events); Arbid v. Holder,

(9th Cir. 2012) (per curiam) (upholding the BIA's particularly-serious-crime conclusion pursuant to its application of the Frentescu guideposts when the BIA had highlighted the petitioner's "substantial" sixteen-month imprisonment term and apparent lack of remorse, on top of the $650,000 restitution order and complex nature of the petitioner's scheme).

- 19 - danger to the community given the nature of his crime. And an

overarching issue here, as Dor points out throughout his papers,

is that this factual recitation bleeds from one offense's fact

pattern straight into the other, leaving us without any clarity as

to what role (if any) any given fact was playing in the BIA's

purportedly Frentescu-guided determination.

We are cognizant that Frentescu directs IJs and the BIA

to "look" to these factors. That does not amount to a hard-and-

fast requirement that the IJ or BIA conduct an exhaustive analysis

of every facet of every factor as applied to a petitioner's case.

But it does prescribe a case-specific inquiry that demands some

application and analysis, and here, the application of the factors

to the facts of Dor's case is entirely deficient.

Even if this constituted an acceptable application of

Frentescu, we could not conclude the BIA provided an explanation

of its conclusion, let alone a rational one. See Valerio-Ramirez,

(finding the decision as to the particularly-

serious-crime determination was not an abuse of discretion when

the agencies analyzed comparable case law alongside the

petitioner's case, and the BIA highlighted the IJ's detailed

description of the crime's complexity, scope, and duration); see

also Dominguez v. Barr,

(9th Cir. 2020)

(upholding the BIA's individualized particularly-serious-crime

analysis when it "provided a reasoned explanation" in view of the

- 20 - Frentescu factors, specifically in that it considered several

facts underlying the conviction, such as the marijuana-growing

operation setup in the petitioner's home (replete with special

lights and approximately fifty potted plants), plus the BIA's

notation of the IJ's adverse credibility finding and its

consideration of the nature of the conviction (observing generally

that drug-trafficking crimes can have "devastating effects")). In

lieu of any explanation as to how the Frentescu elements guided

the BIA to its outcome, what we have is a list of facts from the

record kicked off with a confusing combination of singular and

plural bound up together in the BIA's conclusory statement that

the nature, type, and circumstances of Dor's "offenses are all

indicative of a particularly serious crime." (Emphases added.)

Then there's the reference to what the police report

described as the "large amount" of marijuana Dor had in his

possession during the June 1 offense. This may have been intended

to be an explanation by the BIA as to why this was a particularly

serious crime. But it's unclear, and it does not even seem to be

an accurate description. Indeed, compare the actual drug quantity

(25 grams) to the 30 grams the BIA itself has offered up as the

baseline "small amount" of marijuana, at least in other contexts.

See Matter of Castro Rodriguez,

(BIA 2012)

(reasoning that "30 grams or less may, in general, serve as a

useful guidepost in determining whether an amount is 'small'");

- 21 - see also Moncrieffe v. Holder,

, 194 n.7 (2013)

(recognizing that the BIA has used 30 grams as a guide to determine

whether an amount of marijuana is small). It is problematic that

the BIA would choose to rely on the assessment of the on-scene

police officer that Dor was caught with a "large amount" of

marijuana when the actual 25-gram figure was readily available in

the same police report. It is similarly troubling that the BIA

would thus conclude that 25 grams constitutes a "large amount" of

marijuana when it has indicated 30 grams or less is the guidepost

for discerning a "small" amount. See Adeyanju,

(stating that the BIA "abuses its discretion if it 'inexplicably

departs from established policies,' including its own precedents"

(quoting Benitez v. Wilkinson,

(1st Cir. 2021))).

To the extent the trafficking component of Dor's

offenses (as opposed to the amount in Dor's possession) is what

drove the BIA's particularly-serious-crime conclusion, that is

likewise unclear and insufficiently explained. True, the BIA's

discussion was not solely focused on the drug quantity involved.

But by the same token, it also wasn't solely focused on trafficking

-- the 25-gram quantity is the very first fact the BIA cited. And

we're mindful that, as a practical matter, drug quantity can play

a role when it comes time to evaluate the severity of a trafficking

offense. See, e.g., Marquez v. Garland,

, 723

(9th Cir. 2021) (affirming the IJ and BIA's particularly-serious-

- 22 - crime determination under Frentescu when the

trafficking conviction was for a "staggering" amount of marijuana

(200 pounds) that would be distributed "across broad swaths of the

American landscape"); and see generally Matter of Y-L-,

(explaining (in seminal case in aggravated-felony

context) that drug trafficking is particularly serious, but that

"particularly serious" presumption can be rebutted by showing,

among other things, that only "a very small quantity of controlled

substance" was involved). But again, we simply cannot know what

the reasoning was here because it was not explained.

Given our familiarity with the record at this point, we

are prompted to note that it is not at all apparent to us how an

application of the Frentescu factors to Dor's case would lead to

a particularly-serious-crime determination. For instance,

consider again the June 1 incident -- the BIA relied on a police

officer's assessment that Dor had a "large amount" of marijuana on

him, but this on-the-scene appraisal by an officer is largely

irrelevant to an immigration-law-driven determination that a crime

is particularly serious pursuant to the guiding statutes,

especially when the actual amount (25 grams, a small amount) is

available. See Matter of Castro Rodriguez, 25 I. & N. at 703;

Moncrieffe,

n.7. Consider, too, that while the

BIA identified the type of sentence imposed as a Frentescu factor

but never mentioned (or weighed) Dor's sentences, we observe that

- 23 - Dor received lenient sentences with respect to both offenses (a

two-year probation and a one-year suspended sentence that never

went into effect since Dor completed a violation-free probation

period).

As to Dor's involvement in trafficking as part of the

calculus here, based on the amount in question, and again on the

face of this record, this characterization seems ambitious. The

May 20 offense officers observed Dor sell "20 bucks[' worth]" of

marijuana to another individual; the June 1 incident revealed Dor

had in his possession a digital scale, a large amount of U.S.

currency, and 25 grams of marijuana.

Bottom line: The BIA's particularly-serious-crime

conclusion is devoid of any actual application of the Frentescu

factors, and even if we considered it a solid application of the

law to Dor's case, we still do not have a sufficiently rational

explanation of the BIA's particularly-serious-crime conclusion as

to Dor's minor marijuana offenses, and a rational explanation is

necessary to ensure Dor was appropriately precluded from obtaining

the humanitarian relief he seeks.

CONCLUSION

For the foregoing reasons, we grant Dor's petition and

remand for further proceedings consistent with this opinion.

- 24 -