OPINION

GRABER, Circuit Judge:

Petitioner Salomon Ledezma-Cosino, a native and citizen of Mexico, petitions for review of a final order of the Board of Immigration Appeals ("BIA"), which affirmed an immigration judge's ("IJ") denial of Petitioner's application for cancellation of removal. We deny the petition.1

FACTUAL AND PROCEDURAL BACKGROUND

Petitioner entered the United States from Mexico, without admission or inspection, in 1987. On May 7, 2008, police in Carlsbad, California, arrested him on charges of driving under the influence of intoxicants and driving with a suspended license. A few days later, the Department of Homeland Security issued a notice to appear, charging Petitioner with removability under 8 U.S.C. § 1182(a)(6)(A)(i) because he was an alien present in the United States without having been admitted or paroled.

Petitioner appeared, with counsel, before an IJ, admitted all the factual allegations in the notice to appear, and conceded removability. But, as now relevant, he applied for cancellation of removal pursuant to 8 U.S.C. § 1229b(b)(1). To qualify for cancellation of removal, Petitioner had to demonstrate, among other things, that he was "a person of good moral character" during the 10-year period preceding his application for cancellation of removal. Id. § 1229b(b)(1)(B). Congress has defined the term "good moral character" to exclude anyone who has been a "habitual drunkard" during the relevant period. Id. § 1101(f)(1).

After a hearing on the merits, the IJ denied Petitioner's application for cancellation of removal. The IJ found that Petitioner had not met his burden of establishing that he was "a person of good moral character" because, during the requisite 10-year period, he had been a "habitual drunkard." The BIA affirmed that ground of decision and dismissed the appeal. A timely petition for review to this court followed. We have jurisdiction pursuant to 8 U.S.C. § 1252.

A three-judge panel granted the petition, vacated the BIA's decision, and remanded the matter for further proceedings on the ground that the "habitual drunkard" provision violates equal protection principles. Ledezma-Cosino v. Lynch, 819 F.3d 1070 (9th Cir. 2016). Upon grant of rehearing en banc, the panel's opinion was vacated. Ledezma-Cosino v. Lynch, 839 F.3d 805 (9th Cir. 2016) (order).

STANDARDS OF REVIEW

We review the agency's factual findings for substantial evidence. Angov v. Lynch, 788 F.3d 893, 898 (9th Cir. 2015). We must uphold the findings unless the record compels a contrary conclusion. Id. We review de novo whether a statutory provision is constitutional. Vilchez v. Holder, 682 F.3d 1195, 1198 (9th Cir. 2012).

DISCUSSION

To qualify for cancellation of removal, Petitioner had the burden of establishing that he:

8 U.S.C. § 1229b(b)(1). Congress has defined the term "good moral character" in the following way:

For the purposes of this chapter —

Id. § 1101(f).

In his opening brief to this court, Petitioner argued that substantial evidence does not support the agency's finding that he was a "habitual drunkard." He also argued that, under due process principles, the statutory "habitual drunkard" provision is unconstitutionally vague. The three-judge panel ordered supplemental briefing on additional constitutional issues, including whether the statutory provision violates equal protection principles. We address those three issues in turn.2

A. Substantial evidence supports the finding that Petitioner was a "habitual drunkard."

The immigration statutes do not define the term "habitual drunkard." "When a statute does not define a term, we generally interpret that term by employing the ordinary, contemporary, and common meaning of the words that Congress used." Arizona v. Tohono O'odham Nation, 818 F.3d 549, 556 (9th Cir. 2016) (internal quotation marks omitted). The ordinary meaning of "habitual drunkard" is a person who regularly drinks alcoholic beverages to excess. See, e.g., Black's Law Dictionary 587 (4th ed. 1951) (defining "habitual drunkard" as "[h]e is a drunkard whose habit it is to get drunk; whose ebriety has become habitual," citing a case that refers to a person who has been proved to be repeatedly drunk within a limited period); Black's Law Dictionary 607, 827 (10th ed. 2014) (defining "habitual drunkard" as, among other things, "[s]omeone who habitually consumes intoxicating substances excessively; esp., one who is often intoxicated").

Notably, not all alcoholics are habitual drunkards. As the government emphasizes in its brief to us, the statute asks whether a person's conduct during the relevant time period meets the definition; the person's status as an alcoholic, or not, is irrelevant to the inquiry. We know that Congress did not intend to equate "habitual drunkard" with "alcoholic" because, elsewhere in the statute, Congress used the term "alcoholic." See 8 U.S.C. § 1101(f)(1) (1952) (defining those who lack "good moral character" for certain purposes to include "habitual drunkard[s]"); 8 U.S.C. § 1182(a)(5) (1952) (defining excludable aliens to include "[a]liens who are narcotic drug addicts or chronic alcoholics"); SEC v. McCarthy, 322 F.3d 650, 656 (9th Cir. 2003) ("It is a well-established canon of statutory interpretation that the use of different words or terms within a statute demonstrates that Congress intended to convey a different meaning for those words.").

Here, the record amply supports the agency's finding that Petitioner was a habitual drunkard. In 2010, treating doctors recorded a "more than ten year history of heavy alcohol abuse," during which time Petitioner drank "1 liter of tequila per day on the average." In 2008, he was convicted of driving under the influence. During Petitioner's removal proceedings, Petitioner's daughter testified that he had "a drinking problem" and that his liver had failed because of "[t]oo much alcohol," "[t]oo much drinking." At a minimum, the evidence does not compel the conclusion that Petitioner was not a habitual drunkard.3

The dissenting opinion argues that the term "habitual drunkard" encompasses only those who "cause[] harm to other persons or the community." Dissent at 1058. We need not decide whether "public harm" is a necessary component of the "habitual drunkard" definition. In making its determination that Petitioner was a habitual drunkard, the BIA expressly noted that Petitioner had been convicted of driving under the influence. Driving under the influence is, self-evidently, a public harm. At a minimum, the record does not compel the contrary result. We therefore disagree with the dissenting opinion that further proceedings are necessary in this case, even if public harm is required.

B. The statutory "habitual drunkard" provision is not unconstitutionally vague.

A statute is unconstitutionally vague if it "is so standardless that it authorizes or encourages seriously discriminatory enforcement" or if it "fails to provide a person of ordinary intelligence fair notice of what is prohibited." United States v. Williams, 553 U.S. 285, 304, 128 S.Ct. 1830, 170 L.Ed.2d 650 (2008). As just noted, the term "habitual drunkard" readily lends itself to an objective factual inquiry. And whatever uncertainty the term "habitual drunkard" may raise in borderline cases, a person of ordinary intelligence would have fair notice that the term encompasses an average daily consumption of one liter of tequila for a 10-year period, leading to a conviction for driving under the influence. Because Petitioner has engaged in conduct that is clearly covered, he "cannot complain of the vagueness of the law as applied to the conduct of others." Holder v. Humanitarian Law Project, 561 U.S. 1, 19, 130 S.Ct. 2705, 177 L.Ed.2d 355 (2010) (internal quotation marks omitted). Because the statute is not unconstitutionally vague under the criminal law standard, it necessarily satisfies any lesser vagueness standard that might apply in a non-criminal context. See Village of Hoffman Estates v. Flipside, Hoffman Estates, Inc., 455 U.S. 489, 498-99, 102 S.Ct. 1186, 71 L.Ed.2d 362 (1982).

C. The statutory "habitual drunkard" provision does not violate equal protection principles.

"Where, as here, the Congress has neither invaded a substantive constitutional right or freedom, nor enacted legislation that purposefully operates to the detriment of a suspect class, the only requirement of equal protection is that congressional action be rationally related to a legitimate governmental interest." Harris v. McRae, 448 U.S. 297, 326, 100 S.Ct. 2671, 65 L.Ed.2d 784 (1980). "A legislative classification must be wholly irrational to violate equal protection." De Martinez v. Ashcroft, 374 F.3d 759, 764 (9th Cir. 2004) (internal quotation marks omitted). Petitioner bears the burden "to negate every conceivable basis which might have supported the [legislative] distinction." Angelotti Chiropractic, Inc. v. Baker, 791 F.3d 1075, 1086 (9th Cir. 2015), cert. denied, ___ U.S. ___, 136 S.Ct. 2379, 195 L.Ed.2d 274 (2016) (internal quotation marks omitted).

Congress reasonably could have concluded that, because persons who regularly drink alcoholic beverages to excess pose increased risks to themselves and to others, cancellation of removal was unwarranted. We see nothing irrational about that legislative choice, which furthers the legitimate governmental interest in public safety. Nor does it matter that Congress has permitted cancellation of removal for other groups who may pose similar risks. "[I]n `the area of economics and social welfare, a State does not violate the Equal Protection Clause merely because the classifications made by its laws are imperfect.' A legislature may address a problem `one step at a time,' or even `select one phase of one field and apply a remedy there, neglecting the others.'" Jefferson v. Hackney, 406 U.S. 535, 546, 92 S.Ct. 1724, 32 L.Ed.2d 285 (1972) (quoting Williamson v. Lee Optical of Okla., Inc., 348 U.S. 483, 489, 75 S.Ct. 461, 99 S.Ct. 563 (1955)); see also McDonald v. Bd. of Election Comm'rs, 394 U.S. 802, 809, 89 S.Ct. 1404, 22 L.Ed.2d 739 (1969) ("[A] legislature need not run the risk of losing an entire remedial scheme simply because it failed, through inadvertence or otherwise, to cover every evil that might conceivably have been attacked.").

Petitioner does not seriously dispute the foregoing analysis. Instead, he asserts that it is irrational to classify habitual drunkards as persons who lack good moral character. Petitioner misunderstands the nature of the equal protection inquiry.

The constitutional inquiry is limited to assessing congressional action. "[T]he only requirement of equal protection is that congressional action be rationally related to a legitimate governmental interest." McRae, 448 U.S. at 326, 100 S.Ct. 2671. "Where there are plausible reasons for Congress' action, our inquiry is at an end." FCC v. Beach Commc'ns, Inc., 508 U.S. 307, 313-14, 113 S.Ct. 2096, 124 L.Ed.2d 211 (1993) (internal quotation marks omitted).

Here, Congress' action was the denial of cancellation of removal to habitual drunkards. It is irrelevant, for purposes of analyzing the equal protection claim, whether habitual drunkards lack good moral character. Congress achieved its result by using an intermediate category of persons who lack "good moral character" and by then defining that category to include habitual drunkards, among others. But the specific term, "good moral character," has no significance under rational basis review, which does not require a court to account for all of a statute's text, just whether the statute is rationally related to a legitimate governmental interest. Congress could have chosen any phrase for the intermediate category — "special class of persons not eligible for cancellation of removal," for example — and the effect would be the same. Or Congress could have eliminated the intermediate label altogether and simply listed behaviors that would disqualify applicants from obtaining cancellation of removal — and again the effect would be the same. The intermediate label is therefore of no constitutional moment, even if we were to agree that the label is unfortunate, outdated, or inaccurate.

The Supreme Court's decision in Beach Communications is instructive on this point. Congress required persons to obtain a franchise if they wished to operate a "cable system," 47 U.S.C. § 541, and Congress defined that term to encompass some facilities but not others, 47 U.S.C. § 522(7). The Supreme Court addressed an equal protection challenge to the statutory scheme by asking whether the congressional action — requiring operators of some facilities to obtain a franchise but not requiring operators of other facilities to obtain a franchise — was irrational. Beach Commc'ns, 508 U.S. at 317-20, 113 S.Ct. 2096. The Court did not ask whether, in the abstract, it was rational for Congress to define the term "cable system" in the manner that Congress had chosen.

That same approach applies here. We must ask whether the operative congressional action is rational, not whether the mere definition of a statutory term is rational. Because the denial of cancellation of removal to habitual drunkards is rationally related to the legitimate governmental interest in public safety, Petitioner's equal protection argument fails.

Judge Kozinski's concurring opinion faults us for applying ordinary rational basis review; the concurrence asserts that an even more deferential standard applies. But we have consistently held, citing the same cases that the concurrence cites, that ordinary rational basis review is the appropriate standard in the immigration context. See, e.g., Hernandez-Mancilla v. Holder, 633 F.3d 1182, 1185 (9th Cir. 2011) ("We review equal protection challenges to federal immigration laws under the rational basis standard...."); Ablang v. Reno, 52 F.3d 801, 804 (9th Cir. 1995) (holding that the deferential test described by the Supreme Court "is equivalent to the rational basis test typically applied in equal protection cases").4 Our sister circuits agree.5 Because Petitioner's equal protection fails under the ordinary rational basis test, this case provides no reason to question that longstanding approach.

Petition DENIED.

KOZINSKI, Circuit Judge, with whom Circuit Judges BEA and IKUTA join, concurring.

The majority analyzes this case as if it involved governmental conduct in the domestic sphere, but it doesn't. The President and Congress have excluded an alien pursuant to their plenary power over immigration. The Supreme Court "has firmly and repeatedly endorsed the proposition that Congress may make rules as to aliens that would be unacceptable if applied to citizens." Demore v. Kim, 538 U.S. 510, 522, 123 S.Ct. 1708, 155 L.Ed.2d 724 (2003). We thus owe far more deference here than in an ordinary domestic context. See Fiallo v. Bell, 430 U.S. 787, 792, 97 S.Ct. 1473, 52 L.Ed.2d 50 (1977).

For well over a century, the Court has sharply curtailed review of laws governing the admission or exclusion of aliens under the plenary power doctrine. The Court has said that "over no conceivable subject is the legislative power of Congress more complete than it is over the admission of aliens." Kleindienst v. Mandel, 408 U.S. 753, 766, 92 S.Ct. 2576, 33 L.Ed.2d 683 (1972) (quotation marks and alteration omitted). This is because the power to exclude or expel is "an inherent and inalienable right of every sovereign and independent nation." Fong Yue Ting v. United States, 149 U.S. 698, 711, 13 S.Ct. 1016, 37 S.Ct. 905 (1893). Such power is inherent because the very idea of nationhood requires the drawing of thorny lines — between members and non-members, between admitted and excluded. Our Constitution is the organizing document of a well-defined polity, not an international Golden Rule.

In recent years, the federal courts have been less than consistent in articulating the strength and scope of the plenary power doctrine. See Stephen H. Legomsky, Ten More Years of Plenary Power: Immigration, Congress, and the Courts, 22 Hastings Const. L. Q. 925, 930 (1995). Today's opinion adds to the uncertainty by applying the domestic equal protection test to the sphere of foreign relations. That our circuit has made this error before, as the majority notes, is of no moment when we are sitting en banc. Our principal duty as an en banc court is to correct our circuit law when it has gone astray. I would overrule our precedent and hold that the government's burden is even lighter than rational basis: We approve immigration laws that are facially legitimate without probing or testing possible justifications. See Fiallo, 430 U.S. at 799, 97 S.Ct. 1473 (citing Mandel, 408 U.S. at 770, 92 S.Ct. 2576).1

One reason for the confusion in this area may be that courts have had difficulty articulating a standard below ordinary rational basis review, even though the existence of such a standard — call it "minimally rational basis" — ineluctably follows from the Supreme Court's repeated insistence that Congress can "make rules as to aliens that would be unacceptable if applied to citizens." Demore, 538 U.S. at 522, 123 S.Ct. 1708. What could the Supreme Court mean if not that something less than ordinary rational basis applies?2

The majority interprets section 1101(f)(1) as applying solely to conduct rather than medical status, and it reads the statute's "good moral character" language to mean nothing. Such interpretive gerrymandering may be necessary to preserve the constitutionality of a statute that operates in the domestic sphere. But there's no need to nip and tuck the text here. Congress can exclude Ledezma on account of a medical condition or it can do so because it considers him immoral. This is a facially legitimate exercise of Congress's plenary power, and we have no business passing judgment on it.

Judge Watford's deft concurrence shows why the difference in standards matters. I agree with him that the statute draws distinctions between aliens based on moral judgment.3 But the discrimination here would be far more problematic if a legislature allocated public housing or Medicare only to those citizens with "good moral standing" — and conclusively excluded "habitual drunkards" from the eligible list. In my view, it's the near limitless power of the political branches over immigration and foreign affairs that puts the statute here beyond cavil.

Untold masses were turned away at Ellis Island — or prevented from boarding ships for America — for medical reasons, my grandfather among them. This was a misfortune for those turned away, but excluding aliens for reasons Congress believes sufficient to serve the public welfare is a nigh-unquestioned power of a sovereign nation. I'm aware of no country that fails to adhere to this precept. Nor has the Supreme Court stepped back from it. Until and unless it does, we have no business applying domestic equal protection law to political judgments — even foolish ones — made in the sphere of foreign relations. I would deny the petition summarily with a citation to Fiallo.

WATFORD, Circuit Judge, joined by McKEOWN and CLIFTON, Circuit Judges, concurring:

We took this case en banc to decide whether the Immigration and Nationality Act's "habitual drunkard" provision, 8 U.S.C. § 1101(f)(1), is facially unconstitutional on the theory that it violates the equal protection component of the Fifth Amendment's Due Process Clause. I think the majority rightly rejects that challenge, but my reasons for reaching that conclusion differ.

The statutory classification at issue does not implicate a fundamental right or target a suspect class, so it is subject to rational basis review. FCC v. Beach Communications, Inc., 508 U.S. 307, 313, 113 S.Ct. 2096, 124 L.Ed.2d 211 (1993); Fiallo v. Bell, 430 U.S. 787, 793 n.5, 97 S.Ct. 1473, 52 L.Ed.2d 50 (1977). No one disputes that it is perfectly rational for Congress to deny cancellation of removal to those who lack good moral character. See 8 U.S.C. § 1229b(b)(1)(B). Congress' judgment on that score is entitled to considerable deference, given the breadth of its authority to regulate the admission and exclusion of non-citizens. Fiallo, 430 U.S. at 792, 97 S.Ct. 1473. The only question, then, is whether Congress had a rational basis for establishing a conclusive presumption, not subject to rebuttal, that habitual drunkards lack good moral character, which is what § 1101(f)(1) does.

Conclusive presumptions of this sort are, by their nature, blunt instruments. No doubt there are individuals who, if given the opportunity to do so, could establish that they possess good moral character notwithstanding the fact that they are or were an habitual drunkard. It may well be that the petitioner in this case, Salomon Ledezma-Cosino, is one of those people. But the Supreme Court has long held that conclusive presumptions survive rational basis review even when the presumption established is both over- and underinclusive. See, e.g., Massachusetts Board of Retirement v. Murgia, 427 U.S. 307, 316-17, 96 S.Ct. 2562, 49 L.Ed.2d 520 (1976) (per curiam) (upholding mandatory retirement age of 50 for police officers); Weinberger v. Salfi, 422 U.S. 749, 776-77, 95 S.Ct. 2457, 45 L.Ed.2d 522 (1975) (upholding provision denying Social Security benefits to surviving spouse of wage earner married less than nine months at time of wage earner's death). Here, if Congress could rationally conclude that a substantial number of those found to be habitual drunkards would also be found, upon examination, to lack good moral character, then it could establish the conclusive presumption created by § 1101(f)(1) simply to avoid the administrative costs that individual determinations of good moral character would entail. See Salfi, 422 U.S. at 777, 95 S.Ct. 2457.

I think Congress could rationally conclude that most habitual drunkards would be found to lack good moral character if individual determinations were permitted. That could be true, of course, only if habitual drunkards may in some sense be deemed morally blameworthy for acquiring their condition, for it would be irrational to brand someone as lacking in good moral character due to a medical condition developed through no fault of their own. In my view, Congress could rationally deem habitual drunkards to be at least partially responsible for having developed their condition. Habitual drunkards are those who have allowed themselves to become so addicted to alcohol that they can no longer control their habit of drinking to excess. That loss of control does not come about overnight; it is acquired as a result of frequent, repetitive acts of excessive drinking. See, e.g., Bouvier's Law Dictionary 489 (William Edward Baldwin ed., Baldwin's Century ed. 1948) (defining the term "habitual drunkard" in this way: "A person given to inebriety or the excessive use of intoxicating drink, who has lost the power or the will, by frequent indulgence, to control his appetite for it."). Drinking to excess with such frequency that it leads to the loss of one's ability to refrain from excessive drinking in the future is conduct that Congress could rationally view as volitional, and therefore the proper subject of moral blame.

None of this is to say that Congress' decision to regard habitual drunkards as morally blameworthy for their condition is a wise one. We know considerably more about alcohol addiction today than we did back in 1952, when Congress enacted § 1101(f)(1). Scientists tell us, for example, that some people are much more prone to becoming addicted to substances like alcohol than others, with genetic factors accounting for 40 to 70 percent of individual differences in the risk for addiction. U.S. Department of Health and Human Services, Office of the Surgeon General, Facing Addiction in America: The Surgeon General's Report on Alcohol, Drugs, and Health 2-22 (2016). In addition, there is a high correlation between alcohol abuse and post-traumatic stress disorder (PTSD), a condition that virtually no one could be blamed for acquiring. As the Surgeon General's report notes, "[i]t is estimated that 30-60 percent of patients seeking treatment for alcohol use disorder meet criteria for PTSD, and approximately one third of individuals who have experienced PTSD have also experienced alcohol dependence at some point in their lives." Id. at 2-22 to 2-23 (footnotes omitted).

Still, as a result of advances in our understanding of the neurobiology underlying addiction, we know that substance use disorders (including addiction, the most severe form) "typically develop gradually over time with repeated misuse" of the substance in question, and that one of the key factors in determining whether a person develops an addiction is "the amount, frequency, and duration of the misuse." Id. at 1-6 to 1-7. Modern science thus confirms that, at least to some extent, there is indeed a volitional component to developing an addiction to alcohol, even if many other factors outside an individual's control also contribute. Whether the volitional component is weighty enough to warrant treating habitual drunkards as morally blameworthy for their condition is a policy question for Congress to resolve. Under rational basis review, it is not for us "to judge the wisdom, fairness, or logic" of Congress' decision in that regard. Beach Communications, 508 U.S. at 313, 113 S.Ct. 2096.

It has been suggested that Congress' decision to treat habitual drunkards as lacking in good moral character is irrational because Congress has not classified individuals suffering from other chronic medical conditions, such as diabetes, heart disease, and bipolar disorder, as morally blameworthy for their conditions. The mere fact that a classification drawn by Congress may be underinclusive, however, is not sufficient to render it invalid under rational basis review. Salfi, 422 U.S. at 776, 95 S.Ct. 2457. In any event, Congress could rationally conclude that habitual drunkards are not similarly situated to those suffering from other medical conditions. Even if there is arguably a volitional component involved in developing medical conditions like diabetes and heart disease (say, consuming excessive amounts of sugar or red meat), Congress could rationally view that conduct as less morally blameworthy than consuming excessive amounts of alcohol to the point of losing control over the ability to abstain. Plus, the well-documented connection between alcohol addiction and harm to others (in the form of drunken driving, domestic violence, and the like) distinguishes alcohol addiction from other medical conditions that pose a risk primarily to the health of the individual sufferer, rather than to the safety of others. These differences provide a rational basis for Congress' decision to classify habitual drunkards as lacking in good moral character, but not those suffering from other medical conditions.

THOMAS, Chief Judge, with whom CHRISTEN, Circuit Judge, joins, dissenting:

The government proposes to bar the petitioner from immigration relief simply because is he a recovering alcoholic. It reasons that, because he was diagnosed with the disease during the qualifying period, he categorically must be labeled a "habitual drunkard," and is per se ineligible for cancellation of removal as someone who lacks good moral character. But the terms "alcoholic" and "habitual drunkard" are not synonymous, either as a matter of immigration law, or as a matter of fact. For that reason, I would grant the petition for review and remand for the Board of Immigration Appeals ("BIA") to reconsider the case under a proper construction of the law. Therefore, I must respectfully dissent.

I

A

As we observed more than four decades ago, "[t]he proposition that chronic acute alcoholism is itself a disease, `a medically determinable physical or mental impairment,' is hardly debatable today." Griffis v. Weinberger, 509 F.2d 837, 838 (9th Cir. 1975). It has been recognized as a disease by the American Medical Association since 1956. American Medical Association, Manual on Alcoholism for Physicians (American Medical Association, 1957). Alcoholism is a neurobiological medical condition, and an individual's risk of becoming alcoholic depends on a number of factors beyond volitional choice, including genetics and environmental influences. See, e.g., U.S. Dep't of Health & Human Servs., Office of the Surgeon General, Facing Addiction in America: the Surgeon General's Report on Alcohol, Drugs, and Health (2016) [hereinafter Surgeon General's Report]; World Health Org., Neuroscience of Psychoactive Substance Use and Dependence (2004). Indeed, the Surgeon General has rejected the notion that alcoholism and other addictions are moral failings; instead, they are chronic illnesses "that we must approach with the same skill and compassion with which we approach heart disease, diabetes, and cancer." Surgeon General's Report at v, 1-2.

Despite this near-universal medical consensus, the government urges that a diagnosis of the disease of alcoholism must categorically mean that someone lacks good moral character and is therefore per se ineligible for cancellation of removal. This view is not supported by the statute, and certainly not by common sense. Perhaps, as some suggest, the phrase "habitual drunkard" is purely anachronistic. That well may be so, as evidenced by the fact that, aside from this case, there is only one reported BIA decision-from more than a half century ago-discussing it. Matter of H, 6 I & N Dec. 614 (1955). But it is still part of the statute and, if the government now intends to invoke it, a more definitive explanation of its meaning is required.

B

Our analysis must begin with the acknowledgment that this case presents serious constitutional questions as to the vagueness of the statute and whether it violates the Equal Protection Clause. In such circumstances, we are instructed to avoid constitutional issues "where an alternative interpretation of the statute is `fairly possible.'" INS v. St. Cyr., 533 U.S. 289, 299-300, 121 S.Ct. 2271, 150 L.Ed.2d 347 (2001) (citation omitted). Fortunately, we need not confront those constitutional questions, because an examination of the statute confirms that a diagnosis of the disease of alcoholism does not, as a matter of immigration, mean that a petitioner lacks good moral character as a "habitual drunkard."

Employing the familiar tools of statutory construction, and mindful of the need to avoid constitutional questions, we look first at the plain words of the statute, "particularly to the provisions made therein for enforcement and relief." Middlesex Cnty. Sewerage Auth. v. Nat'l Sea Clammers Ass'n, 453 U.S. 1, 13, 101 S.Ct. 2615, 69 L.Ed.2d 435 (1981). "[W]hen deciding whether the language is plain, we must read the words `in their context and with a view to their place in the overall statutory scheme.'" King v. Burwell, ___ U.S. ___, 135 S.Ct. 2480, 2489, 192 L.Ed.2d 483 (2015) (quoting FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 133, 120 S.Ct. 1291, 146 L.Ed.2d 121 (2000)). In addition, we examine the legislative history, the statutory structure, and "other traditional aids of statutory interpretation" in order to ascertain congressional intent. Middlesex Cnty., 453 U.S. at 13, 101 S.Ct. 2615. As part of statutory analysis, "[w]e also look to similar provisions within the statute as a whole and the language of related or similar statutes to aid in interpretation." United States v. LKAV, 712 F.3d 436, 440 (9th Cir. 2013).

The present "good moral character" definition was enacted as part of the Immigration and Naturalization Act of 1952 ("INA"), which defined certain categories of individuals who were, per se, lacking in good moral character, including "habitual drunkard[s]," adulterers, gamblers, persons who gave false testimony for the purpose of obtaining immigration benefits, murderers, and those who had been convicted of a crime and confined to a penal institution for an aggregate of at least 180 days. Public L. 82-414 § 101(f), 66 Stat. 163, 172 (1952).

So, did Congress mean to include in the term "habitual drunkard" all individuals who had been diagnosed with alcoholism, or did it intend to distinguish between the two concepts? The text and history of the INA lead to the conclusion that Congressional intent was to create a distinction.

First, Congress well knew how to use the terms "alcoholism" and "alcoholic" in immigration law. In the Immigration Act of 1917, Congress added "persons with chronic alcoholism" to the classes of aliens excluded from admission to the United States. Immigration Act of 1917, Pub. L. No. 64-301, § 3, 39 Stat. 874, 875 (1917) (repealed 1952). With enactment of the INA, Congress repealed many of the provisions of the 1917 Act relating to categories of excludable aliens, but explicitly modified the exclusion provisions to include "[a]liens who are narcotic drug addicts or chronic alcoholics." 66 Stat 163, 172-73 (1952). "[W]hen Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion." Barnhart v. Sigmon Coal Co., 534 U.S. 438, 452, 122 S.Ct. 941, 151 L.Ed.2d 908 (2002) (internal quotation marks omitted); see also Center for Community Action and Environmental Justice v. BNSF R.R. Co., 764 F.3d 1019, 1024 (9th Cir. 2014). Therefore, the structure and context of the INA indicate a Congressional intent to distinguish the phrases. In addition, during the period when the INA was enacted, common public understanding was that the concepts were distinct. For example, Webster's New World Dictionary — published four years after the passage of the INA — distinguishes between a drunkard and an alcoholic: a "drunkard" is "a person who often gets drunk; inebriate," whereas an "alcoholic" is "one who has chronic alcoholism." Webster's New World Dictionary 17, 231 (1956). Webster's Collegiate Dictionary, published in 1947, defined "alcoholism" as "a diseased condition caused by excessive use of alcoholic liquors" and a "drunkard" as a "toper or "sot." Webster's Collegiate Dictionary (5th ed. 1947). And, as previously observed, the American Medical Association recognized alcoholism as a disease in 1956.

Second, the statutory context of the phrase "habitual drunkard" is critical. It is contained in the definition of "good moral character," as one of the listed categories of character attributes that preclude relief. 8 U.S.C. § 1101(f)(1). The general concept of "good moral character" as a prerequisite to obtaining immigration benefits dates back to the adoption of the first naturalization statute in 1790, Act of Mar. 26, 1970, ch. 3, § 1, 1 Stat. 103, and grounded in the notion that an applicant should have spent some time as a resident and then "be able to bring testimonials of a proper and decent behavior."1 It deals with one's character, not one's medical afflictions.

According to the present statute's terms, its purpose is to define which individuals necessarily lack good moral character. See 8 U.S.C. § 1101(f) ("No person shall be regarded as, or found to be, a person of good moral character who ... is, or was ... a habitual drunkard."). Other noncitizens who necessarily lack good moral character — and are therefore categorically barred from receiving discretionary relief — under Section 101(f) are (1) individuals engaged in prostitution, the smuggling of illegal immigrants into the country, or polygamy; (2) individuals "whose income is derived principally from illegal gambling activities" or who have "been convicted of two or more gambling offenses"; (3) individuals who have "been convicted of an aggravated felony"; or (4) individuals engaged in conduct relating to "assistance in Nazi persecution, participation in genocide, or commission of acts of torture or extrajudicial killings." 8 U.S.C. § 1101(f).

Every other category in Section 101(f) describes conduct that results in public harm or harm to others. Under the doctrine of noscitur a sociis, according to which "a word is known by the company it keeps," S.D. Warren Co. v. Maine Bd. of Env'l Protection, 547 U.S. 370, 378, 126 S.Ct. 1843, 164 L.Ed.2d 625 (2006), "habitual drunkard" should apply only to individuals who engage in certain types of harmful conduct. Therefore, an individual's status as suffering from the disease of alcoholism cannot be sufficient to trigger the "habitual drunkard" provision; being an alcoholic does not necessarily result in public harm or harm to others.

Third, not only does the context of "good moral character" suggest analysis of conduct, rather than a disease, but the statutory provisions as to the avenue of relief afforded strengthens that conclusion. The phrase "good moral character" is employed in various immigration contexts including naturalization, see 8 U.S.C. § 1427(a)(3), becoming a lawful permanent resident, see 8 U.S.C. § 1255b, adjustment of status, see 8 U.S.C. § 1154, grant of voluntary departure, see 8 U.S.C. § 1229c, and cancellation of removal, 8 U.S.C. § 1229b.

For our contextual purposes, the relevant provision is cancellation of removal and its predecessor statute, suspension of deportation. Relief via suspension of deportation was established by the INA. One of the eligibility requirements for suspension of deportation was that the applicant be someone whose deportation would result in "exceptional and extremely unusual hardship" to the alien or an immediate family member who was a citizen or lawful permanent resident. 8 U.S.C. § 1254(a)(1) (repealed).2 Among the critical factors in determining the requisite hardship was "health, especially tied to inadequate medical care in the home country." Urban v. INS, 123 F.3d 644, 648 (7th Cir. 1997); see also In re Anderson, 16 I. & N. Dec. 596, 597-98 (1978) (noting among the relevant factors to be "condition of health" and "severe illness"); In re Louie, 10 I. & N. Dec. 223, 225 (1963) (granting suspension of deportation based on medical condition of father). In 1962, Congress replaced the "exceptional and extremely unusual hardship" standard with "extreme hardship."

In 1997, Congress eliminated the remedy of suspension of deportation, and replaced it with "cancellation of removal," which is the operative statute in this case. 8 U.S.C. § 1229b. Eligibility for relief was still predicated on hardship, with a stricter standard returning to the original language of "exceptional and extremely unusual hardship," and limiting the hardship determination to qualifying relatives. Id. Medical condition continued to be an important factor in determining eligibility for relief. See, e.g., Fernandez v. Gonzales, 439 F.3d 592, 601-02 (9th Cir. 2006) (noting it would be error for the BIA to refuse to hear evidence of a life-threatening medical condition in the context of cancellation of removal); see also In re Gonzalez Recinas, 23 I. & N. Dec. 467, 470 (2002) (noting that the new hardship standard "is not so restrictive that only a handful of applicants, such as those who have a qualifying relative with a serious medical condition, will qualify for relief."). In short, medical condition and health have always been important considerations in determining hardship, either through suspension of deportation or cancellation of removal. Thus, it would be inconsistent with the statute, when considered in context, to construe it to mean that the disease of alcoholism, by itself, would per se disqualify a petitioner from relief when the establishment of a serious medical condition can be a qualifying factor.

Fourth, in a different section of the INA, Section 212(a)(1)(A)(iii), Congress demonstrated a more nuanced understanding of alcohol dependence. There, it established that a noncitizen is inadmissible if he or she "is determined (in accordance with regulations prescribed by the Secretary of Health and Human Services in consultation with the Attorney General) to have a physical or mental disorder and behavior associated with the disorder that may pose, or has posed, a threat to the property, safety, or welfare of the alien or others." See 8 U.S.C. § 1182(a)(1)(A)(iii)(I); see also 8 U.S.C. § 1182(a)(1)(A)(iii)(II). An implementing regulation treats alcoholics as having a "physical or mental disorder" for the purpose of inadmissibility under this statute. See 42 C.F.R. 34.2(n); Am. Psychiatric Ass'n, supra; Center for Disease Control & Prevention, Technical Instructions for Physical or Mental Disorders with Associated Harmful Behaviors and Substance-Related Disorders for Civil Surgeons (2013), available at http://tinyurl.com/jqaggoo. It is important to emphasize, however, that this statute only refuses admissibility to alcoholics whose alcohol-related behavior "pose[s] ... a threat to the property, safety, or welfare of the alien or others." 8 U.S.C. § 1182(a)(1)(A)(iii)(I). It does not exclude alcoholics based on an outdated stigma that they are categorically immoral.

Finally, as we have noted, the assessment of good moral character in the immigration context requests the agency to "weigh and balance the favorable and unfavorable facts or factors, reasonably bearing on character, that are presented in evidence." Torres-Guzman v. INS, 804 F.2d 531, 534 (9th Cir. 1986). Although this general concept does not construe the "habitual drunkard" provision, it reinforces the idea of the general purpose of the statute, and the need for a case-by-case determination. Consistent with this approach, courts have declined to find a lack of good moral character based on isolated alcohol-related conduct. For example, several district courts have held that a single conviction for driving under the influence ("DUI") — and sometimes more than one — cannot render someone a person of bad moral character under Section 101(f). See, e.g., Rangel v. Barrows, No. 4:07-cv-279, 2008 WL 4441974, at *4 (E.D. Tex. Sept. 25, 2008) ("[T]he applicable law is unanimous in support of the proposition that, in the absence of aggravating factors, a single [DUI] conviction is insufficient to deny an application for naturalization on the basis that the applicant lacks good moral character."); Ragoonanan v. U.S. Citizenship & Immigration Servs., No. 07-3461 PAM/JSM, 2007 WL 4465208, at *5 (D. Minn. Dec. 18, 2007) (holding that "a single [DUI] conviction resulting in probation" is insufficient to establish bad moral character); Yaqub v. Gonzales, No. 1:05-cv-170, 2006 WL 1582440, at *4 (S.D. Ohio June 6, 2006) (concluding that "two DUI arrests" are insufficient to find petitioner lacks good moral character); Le v. Elwood, No. Civ.A. 02-CV-3368, 2003 WL 21250632, at *3 (E.D. Pa. 2003) (concluding that two DUI convictions, did not, standing alone, "amount to a finding of `habitual drunkard'").

In addition, the statutory construct of "good moral character" has also embraced the concept of redemption. See, e.g., Santamaria-Ames v. INS, 104 F.3d 1127, 1132 (9th Cir. 1996) ("Whether the petitioner can establish that he has reformed and rehabilitated from this prior conduct is germane to the determination of whether he has established good moral character...."); Yuen Jung v. Barber, 184 F.2d 491 (9th Cir. 1950) (noting that if a prior bad act precluded one from establishing good moral character, "would require a holding that Congress had enacted a legislative doctrine of predestination and eternal damnation.") Here, however, the government's proposed holding would mean that all sober, recovering alcoholics who were diagnosed during the seven year qualifying period would necessarily be considered "habitual drunkards" and categorically ineligible for relief. The construction more consistent with the statute would be to allow the agency to consider and balance the equities of each individual circumstance on a case by case basis.

In short, when we consider the plain language of the statute, its structure, and its legislative history, we must conclude that the phrase "habitual drunkard" is not synonymous with "alcoholic." Thus, a diagnosis of the disease of alcoholism is insufficient to trigger the "habitual drunkard" provision, and render a petitioner categorically ineligible for discretionary cancellation of removal relief.

Instead, the phrase "habitual drunkard" is best understood in the context of its statutory setting of "good moral character," which has commonly been understood to reflect, as Judge Learned Hand put it, the "common conscience" of the community. Johnson v. United States, 186 F.2d 588, 590 (2d Cir. 1951). To that end, courts have generally focused on whether the challenged conduct is harmful to the public, or whether it is purely private. See Nemetz v. INS, 647 F.2d 432, 436 (4th Cir. 1981) (noting that the appropriate analysis "is whether the act is harmful to the public or is offensive merely to a personal morality."); In re Labady, 326 F.Supp. 924, 927 (S.D.N.Y. 1971) ("The most important factor to be considered is whether the challenged conduct is public or private in nature.").

Thus, in context, the best construction of "habitual drunkard" within the "good moral character" definition is one who habitually abuses alcohol and whose alcohol abuse causes harm to other persons or the community. This interpretation is consistent with the statutory language, structure, and context, and avoids any constitutional infirmity.

To be sure, an alcoholic may also fit the definition of "habitual drunkard" by conduct that causes harm to others or the public. But to say that status of being diagnosed an alcoholic always means that one is a "habitual drunkard," is not consistent with the statute.

II

A proper construction of the phrase "habitual drunkard" is critical to the outcome of this petition. Mr. Ledezma-Cosino has been in the United States for twenty years. He works in construction, specializing in cement masonry and concrete finishing. He and his wife have eight children, five of whom are United States citizens. At his first immigration hearing, he admitted removability, but applied for cancellation of removal and voluntary departure under 8 U.S.C. § 1229b-c. In support of his application for cancellation of removal, he contended that his removal would result in exceptional and extremely unusual hardship to his children because of economic disadvantage, the difficulties of adjusting to life in Mexico, and his youngest daughter's asthma. At the first hearing, the immigration judge ("IJ") found, on the merits, that Ledezma-Cosino had not established the statutory hardship requirement. The IJ denied cancellation of removal, but granted voluntary departure.

Ledezma-Cosino appealed to the BIA. The BIA remanded the case because the trial transcript was defective because there was no record of the last witness, Ledezma-Cosino's daughter, Yadira Ledezma. The BIA instructed the IJ to complete the record.

At the first post-remand hearing, the government attested to the IJ that there had been no negative developments as to Ledezma-Cosino between the hearings. At the scheduled second hearing, counsel requested a continuance because Ledezma-Cosino had been hospitalized with a liver ailment. At a subsequent scheduling hearing, counsel presented medical records to show that his client had been hospitalized.

At the hearing on the merits, the judge placed the medical records into the record himself. Yadira Ledezma testified, as well as Ledezma-Cosino. The judge questioned both about Ledezma-Cosino's drinking because it had been reflected in the medical records. Ledezma-Cosino testified that he had been sober since the hospitalization. At the conclusion of the testimony, the government argued that he had not satisfied the hardship requirement and questioned whether he had been truthful on his application. The government did not argue that he was categorically ineligible for relief because he was a "habitual drunkard."

The IJ then, sua sponte, declared Ledezma-Cosino ineligible for relief because he was a "habitual drunkard," and also denied voluntary departure for that reason. The BIA dismissed his appeal on the sole basis that he is a "habitual drunkard" and therefore failed to "me[e]t the requisite period of good moral character" required for discretionary relief. In reaching this conclusion, the BIA considered only that Ledezma-Cosino (1) was hospitalized for a serious liver condition in 2010 and drank alcohol excessively for a year leading up to his hospital visit, (2) had a decade-long alcohol dependency, (3) was an alcoholic according to his daughter's testimony, and (4) was convicted of a DUI in 2008.

Aside from the DUI conviction, there was no evidence in the record of any harm to the public or others. Indeed, the only evidence was that he was an excellent worker. Thus, if the evidence pertaining to his diagnosis of alcoholism is set aside, there was not sufficient evidence to sustain the determination of ineligibility for cancellation or voluntary departure based on the "habitual drunkard" clause.

Therefore, I would remand this petition to the BIA for application of the correct statutory standard or, to the extent there is remaining statutory ambiguity, for it to determine the meaning of the phrase "habitual drunkard" in a way that does not make the phrase synonymous with "alcoholic."

Whether or not the agency would ultimately grant relief in this case is a separate question. In the end, the decision as to whether an applicant is afforded discretionary cancellation of removal is committed solely to the executive branch, not subject to our review. But legal eligibility for relief is subject to our review, and it is important for future cases of those who seek relief, and the attorneys who represent them, that the law is accurately defined. Given the government's new reliance on what had been considered an antediluvian phrase, resolution of its meaning is particularly critical.

Because I would resolve the petition on the basis of statutory interpretation, or remand, I would not reach the constitutional questions raised in this case.

For these reasons, I respectfully dissent.