EDUCATIONAL MEDIA COMPANY AT VIRGINIA TECH, INCORPORATED; Cavalier Daily, Incorporated, The Cavalier Daily, Incorporated, Plaintiffs-Appellants, v. J. Neal INSLEY, Commissioner, Virginia Alcoholic Beverage Control Commission; Sandra C. Canada, Commissioner, Virginia Alcoholic Beverage Control Commission; W. Curtis Coleburn, III, Chief Operating Officer Virginia Department of Alcoholic Beverage Control; Frank Monahan, Director, Law Enforcement Bureau of the Virginia Department of Alcoholic Beverage Control; Bryan M. Rhode, Commissioner, Virginia Alcoholic Beverage Control Commission, Defendants-Appellees. Washington Legal Foundation; Student Press Law Center; College Newspaper Business and Advertising Managers; Thomas Jefferson Center for the Protection of Free Expression; The Media Institute, Amici Supporting Appellants, 731 F.3d 291

Reversed by published opinion. Judge THACKER wrote the majority opinion, in which Judge KING joined. Judge SHEDD wrote dissenting opinion. THACKER, Circuit Judge: The Virginia Alcoholic Beverage Control Board (the “ABC”) prohibits college student newspapers from printing alcohol advertisements. Appellants Educational Media and The Cavalier Daily (hereinafter “Appellants” or the “College Newspapers”) are non-profit corporations that own student newspapers at Virginia Polytechnic Institute and State University (‘Virginia Tech”) and the University of Virginia (“UVA”), respectively.

In this action, the College Newspapers challenge the ABC ban on alcohol advertisements as violative of the First Amendment, as applied to them. The district court granted

summary judgment in favor of the ABC, concluding that the challenged regulation is a constitutionally appropriate restriction of commercial speech given Virginia’s substantial interest in combatting underage and abusive drinking on college campuses. However, in this as-applied challenge, because the advertising ban is not appropriately tailored to Virginia’s stated aim, we reverse the judgment of the district court. I. Virginia precludes college student newspapers from printing alcohol advertisements. See 3 Va. Admin. Code § 5-20-40(A)(2) (2010) (the “challenged regulation”).

The challenged regulation provides: Advertisements of alcoholic beverages are not allowed in college student publications unless in reference to a dining establishment, except as provided below. A “college student publication” is defined as any college or university publication that is prepared, edited or published primarily by students at such institution, is sanctioned as a curricular or extracurricular activity by such institution and which is distributed or intended to be distributed primarily to persons under 21 years of age. 3 Va. Admin. Code § 5-20-40(A)(2).

In 2006, the College Newspapers brought suit in the Eastern District of Virginia, contending that the challenged regulation was violative of the First Amendment. The College Newspapers made three distinct arguments. First, they argued that the challenged regulation impermissibly discriminates against a narrow segment of the media — college student newspapers — thus subjecting the regulation to the exacting strict scrutiny standard, which, they argued, it cannot withstand.

Second, they argued that, even if strict scrutiny is inapplicable, the challenged regulation fails, on its face, to satisfy the Supreme Court’s Central Hudson test, which subjects non-misleading commercial speech to intermediate scrutiny. See Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n of N.Y., 447 U.S. 557, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980). Finally, the College Newspapers argued that, even if the challenged regulation could withstand a facial challenge under Central Hudson, the regulation fails Central Hudson as-applied. A.

The ABC asserts that the purpose of the challenged regulation is to combat underage and abusive college drinking. During discovery, each party proffered expert testimony on the question of whether the challenged regulation was effective in this regard. Specifically, the ABC offered a declaration of Dr. Henry Saffer, an economics professor at Kean University in New Jersey. Dr. Saffer testified that, while the vast majority of studies found that alcohol advertising bans do not, in fact, reduce the overall market demand for alcohol, those studies are inapplicable here. Notably, Dr.

Saffer contends that, while most scholars assume that a prohibition on alcohol advertising in one forum simply pushes alcohol advertising to other forums, according to him, this assumption only holds true where a reasonable substitute for the regulated forum exists. Dr. Saffer testified that this assumption does not hold true in the context of college student newspapers, because “[a] college newspaper is a very targeted, specific kind of media,” and there is “nothing else that can replace that kind of targeted media that’s specifically oriented towards and reaches college students.” J.A. 313. According to Dr.

Saf-fer’s reasoning, in the unique instance of college newspapers, alcohol advertising bans actually do have a significant effect on market demand despite the vast majority of studies that show otherwise outside of this particular context. In contrast, the College Newspapers offered the testimony of Dr. Jon P. Nelson, an economics professor at Pennsylvania State University. Based on his research, Dr. Nelson testified that “[ajdvertising bans, partial or comprehensive, do not reduce the demand for alcohol.” J.A. 484.

Rather, he explains, “[i]n a ‘mature market,’ such as alcohol beverages, the primary effect of advertising is to create and maintain brand loyalty[,]” as opposed to expanding overall market demand. Id. He also notes that college students are continually exposed to alcohol advertisements in a variety of forums — including television, radio, and the internet — which “will totally offset any possible temperance effect of the ABC regulation.” Id. at 487. In a declaration filed in response to Dr. Saffer’s testimony, Dr. Nelson noted that Dr.

Saffer did not present any specific evidence in support of his proposition that targeted advertising bans in college student publications actually achieve the desired goal, that is, reduced drinking. Moreover, Dr. Saffer conceded that, in addition to the lack of empirical support for selective bans on alcohol advertising in college student publications, other methods of combatting alcohol consumption on college campuses have been proven more effective. Specifically, as the district court noted, “Dr.

Saffer also admits that increased taxation has been shown to reduce underage consumption in a more effective manner than advertising bans and that counter-advertising has effectively reduced levels of alcohol consumption.” J.A. 586-87. The College Newspapers also established, without any counter-argument from the ABC, that a majority of their readers are over the age of 21. Specifically, the Collegiate Times — the Virginia Tech student newspaper owned by Appellant Educational Media — has a daily readership of roughly 14,000 readers.

According to a 2004 survey, persons age 21 or over constituted roughly 60% of the Collegiate Times’ total readership and about 59% of the Collegiate Times’ total student readership. Similarly, the Cavalier Daily has a daily readership of about 10,000 readers. As of January 1, 2007, persons age 21 or over comprised approximately 64% of UVA’s total student population. B. In 2008, following the close of discovery, the district court granted the College Newspapers’ motion for

summary judgment, declaring that the challenged regulation was facially unconstitutional under Central Hudson. Having determined that the regulation failed Central Hudson on its face, the district court declined to reach the College Newspapers’ remaining arguments regarding strict scrutiny or an as-applied challenge. On appeal, a panel of this court reversed, holding that, on its face, the ban does not violate the First Amendment. See Educ. Media Co., et al. v. Swecker, 602 F.3d 583 (4th Cir.2010).

Because the underlying district court opinion was premised solely on the College Newspapers’ facial challenge, Sweeker was similarly confined to that aspect of the case. See Swecker, 602 F.3d at 587 n. 2. (“The district court did not reach the college newspapers’ alternative arguments.... Though the college newspapers reiterate these alternative arguments on appeal, we decline to address them in the first instance.”).

Accordingly, the panel remanded the case to the district court for consideration of the two remaining challenges: (1) the College Newspapers’ argument that the challenged regulation is subject to, and fails, strict scrutiny and (2) the College Newspapers’ argument that, as-applied to them, the challenged regulation fails Central Hudson. The parties then filed cross motions for

summary judgment without supplementing the record. On September 7, 2012, the district court granted the ABC’s motion and, simultaneously, denied the College Newspapers’ motion. In so doing, the district court first rejected the College Newspapers’ contention that strict scrutiny applied, opting instead to analyze the regulation under Central Hudson. The district court then held that, in light of our opinion in Sweeker, it was constrained to conclude that the challenged regulation did not violate Central Hudson as applied to the College Newspapers. Thus, the district court granted

summary judgment in favor of the ABC on all remaining claims. The College Newspapers timely noted this appeal. II. We review a district court order granting

summary judgment de novo, viewing the evidence in the light most favorable to the non-moving party. Lansdowne on the Potomac Homeowners Ass’n, Inc. v. OpenBand at Lansdowne, LLC, 713 F.3d 187, 195 (4th Cir.2013). In reviewing a challenge to a restriction on commercial speech, “[i]t is well established that the party seeking to uphold a restriction on commercial speech carries the burden of justifying it.” Edenfield v. Fane, 507 U.S. 761, 770, 113 S.Ct. 1792, 123 L.Ed.2d 543 (1993) (internal quotation marks omitted). III. A.

While commercial speech is protected by the First Amendment, there is a “commonsense distinction” between commercial speech and other varieties of speech. Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n of New York, 447 U.S. 557, 562, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980). Thus, “[t]he Constitution ... accords a lesser protection to commercial speech than to other constitutionally guaranteed expression.” Id. at 562-63, 100 S.Ct. 2343. Accordingly, under Central Hudson, a restriction on commercial speech must withstand “intermediate scrutiny” in order to survive a First Amendment challenge.

Id. at 573, 100 S.Ct. 2343 (Blackmun, J. concurring); Fla. Bar v. Went For It, Inc., 515 U.S. 618, 623, 115 S.Ct. 2371, 132 L.Ed.2d 541 (1995). The parties agree that the challenged regulation impacts only commercial speech. However, the College Newspapers argue that, despite this, Central Hudson “intermediate scrutiny” does not apply.

Instead, they argue that the challenged regulation is subject to strict scrutiny because it allegedly engages in both content-based and speaker-based discrimination by singling out a narrow segment of the media, college student publications, and subjecting that segment to heightened regulation. In so arguing, the College Newspapers primarily rely on Sorrell v. IMS Health Inc., — U.S. -, 131 S.Ct. 2653, 180 L.Ed.2d 544 (2011). In Sorrell, the Supreme Court invalidated a Vermont law that prevented pharmacies from revealing “prescriber-identifying information” for marketing purposes.

Vermont argued that this law satisfied Central Hudson as it was appropriately tailored to the important governmental aims of ensuring medical privacy and reducing the price of prescription drugs. The Supreme Court invalidated the ban, concluding that, by only prohibiting the disclosure of prescriber-identifying information for marketing purposes, Vermont had engaged in both content-based and speaker-based discrimination. Accordingly, the Court concluded that the Vermont law needed to withstand “heightened scrutiny” in order to survive a First Amendment challenge.

See Sorrell, 131 S.Ct. at 2664 (“The First Amendment requires heightened scrutiny whenever the government creates a regulation of speech because of disagreement with the message it conveys.... Commercial speech is no exception.”) (internal quotation marks and citations omitted). However, the Court also concluded that Vermont’s ban could not even withstand intermediate scrutiny under Central Hudson. Thus, the Court did not actually apply “heightened scrutiny,” striking the ban under Central Hudson alone.

Id. at 2667 (“[T]he outcome is the same whether a special commercial speech inquiry or a stricter form of judicial scrutiny is applied.”) The College Newspapers and their amici argue that, like the regulation at issue in Sorrell, the challenged regulation here involves both content-based and speaker-based discrimination. Based on this alleged discrimination, the College Newspapers argue that strict scrutiny applies.

However, like the Court in Sorrell, we need not determine whether strict scrutiny is applicable here, given that, as detailed below, we too hold that the challenged regulation fails under intermediate scrutiny set forth Central Hudson, B. Next, we consider whether the challenged regulation violates Central Hudson as applied to the College Newspapers.

Under Central Hudson, a regulation of commercial speech will be upheld if (1) the regulated speech concerns lawful activity and is not misleading; (2) the regulation is supported by a substantial government interest; (3) the regulation directly advances that interest; and (4) the regulation is not more extensive than necessary to serve the government’s interest. Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343. Central Hudson applies to both facial and as-applied challenges. See Swecker, 602 F.3d at 588. However, the type of challenge dictates the state’s burden of proof. Id.

In an as-applied challenge, which we address here, the state must justify the challenged regulation with regard to its impact on the plaintiffs. Id. Applying Central Hudson to the facts at hand, we conclude that the district court erred in holding that, as applied to the College Newspapers, the challenged regulation is constitutionally permissible. Rather, we conclude that, as applied to the College Newspapers, the challenged regulation violates the First Amendment. We address each of the four Central Hudson prongs in turn. 1.

Lawful Activity All parties are in agreement that the first prong of Central Hudson, i.e. whether the regulated speech concerns lawful activity and is not misleading, is satisfied. Specifically, the challenged regulation regulates lawful activity, as alcohol advertisements — even those that reach a partially underage audience — concern the lawful activity of alcohol consumption. See Swecker, 602 F.3d at 589. Additionally, the ABC has not presented any evidence that the advertisements implicated by this regulation are misleading. Thus, this prong is clearly satisfied. 2.

Substantial Government Interest “Next, we ask whether the asserted governmental interest is substantial.” Central Hudson, 447 U.S. at 566, 100 S.Ct. 2343. As with the first prong, the parties are in agreement that Virginia’s stated interest in combatting underage and abusive drinking on college campuses represents a substantial governmental interest. See, Swecker, 602 F.3d at 589. Accordingly, the second Central Hudson prong is satisfied. 3. Direct and Material Advancement a.

Under Central Hudson’s third prong, the ABC must prove that the challenged regulation directly advances the government’s asserted interest. “This burden is not satisfied by mere speculation or conjecture; rather, a governmental body seeking to sustain a restriction on commercial speech must demonstrate that the harms it recites are real and that its restriction will in fact alleviate them to a material degree.” Edenfield v. Fane, 507 U.S. 761, 770-71, 113 S.Ct. 1792, 123 L.Ed.2d 543 (1993).

To be sure, under this prong, we do not require a government to produce empirical data “accompanied by a surfeit of background information[.]” Lorillard Tobacco Co. v. Reilly, 533 U.S. 525, 555, 121 S.Ct. 2404, 150 L.Ed.2d 532 (2001) (internal quotation marks omitted). Rather, the Supreme Court has “permitted litigants to justify speech restrictions by reference to studies and anecdotes pertaining to different locales altogether, or even, in a case applying strict scrutiny, to justify restrictions based solely on history, consensus, and simple common sense.” Id. (internal quotation marks and citations omitted).

Nonetheless, the burden remains on the state to establish a commercial speech regulation’s efficacy, and a regulation cannot be sustained “if there is little chance that the restriction will advance the State’s goal[.]” Id. at 566,121 S.Ct. 2404 (internal quotation marks and citations omitted).

In Swecker, we concluded that the challenged regulation satisfies the third prong because, given the general correlation between advertising of a product and demand for that product, it follows that a decrease in alcohol advertising on college campuses will necessarily result in a decrease in alcohol consumption by college students. Swecker, 602 F.3d at 590 (“Though the correlation between advertising and demand alone is insufficient to justify advertising bans in every situation, here it is strengthened because ‘college student publications’ play an inimitable role on campus.” (internal citations omitted)).

Additionally, we concluded that the efficacy of the regulation was further substantiated by the fact that “alcohol advertisers want to advertise in college student publications. It is counterintuitive for alcohol vendors to spend their money on advertisements ... if they believed that these ads would not increase demand by college students.” Id. (emphasis in original). b. On remand following Swecker, the district court concluded that it was compelled to follow Swecker’s analysis on this prong. We agree.

While it is true, as the College Newspapers argue, that Swecker dealt with a facial challenge and the present action involves an as-applied challenge, this distinction is immaterial under Central Hudson’s third prong. See United States v. Edge Broad., 509 U.S. 418, 427, 113 S.Ct. 2696, 125 L.Ed.2d 345 (1993). Specifically, in Edge Broad., the Supreme Court was faced with the question of whether a series of federal statutes that prohibited the broadcast of lottery advertisements violated the First Amendment as applied to a broadcaster licensed to do business in a state that allowed lotteries.

In analyzing the statute under Central Hudson’s third prong, the Court noted that it was “readily apparent that this question cannot be answered by limiting the inquiry to whether the governmental interest is directly advanced as applied to a single person or entity.” Edge Broad., 509 U.S. at 427, 113 S.Ct. 2696. According to the Court, “[e]ven if there were no advancement as applied in that manner— in this case, as applied to Edge — there would remain the matter of the regulation’s general application to others —in this case, to all other radio and television stations in North Carolina and countrywide.” Id.

Thus, while an as-applied challenge ordinarily compels a reviewing court to take the challenger’s individual circumstances into account when assessing the constitutionality of a particular restriction, that general practice is inapplicable when analyzing a provision under the third Central Hudson prong. Accordingly, Swecker’s conclusion that the challenged regulation directly and materially advances the State’s asserted interest is dispositive of our analysis under the third Central Hudson prong. 4.

Regulation More Extensive Than Necessary The individual circumstances of the College Newspapers are nonetheless relevant in this appeal. Per Edge Broadcasting, those circumstances are relevant under Central Hudson’s fourth prong, to which we now turn. See Edge Broad., 509 U.S. at 427, 113 S.Ct. 2696 (“This is not to say that the validity of the statutes’ application to Edge is an irrelevant inquiry, but that issue properly should be dealt with under the fourth factor of the Central Hudson test.”).

Under this prong of Central Hudson, the party defending the regulation “must demonstrate narrow tailoring of the challenged regulation to the asserted interest — a fit that is not necessarily perfect, but reasonable; that represents not necessarily the single best disposition but one whose scope is in proportion to the interest served.” Greater New Orleans Broad. Ass’n, Inc. v. United States, 527 U.S. 173, 188, 119 S.Ct. 1923, 144 L.Ed.2d 161 (1999).

The ABC argues that the challenged regulation does not fail under this fourth prong because it is “reasonably tailored” to the stated aim of reducing underage and abusive alcohol consumption. Specifically, the ABC notes that the challenged regulation does not prohibit all alcohol advertisements and is but one facet of its multi-faceted approach to combat the problem of underage drinking. Given the dual purpose of the regulation to combat both underage and abusive drinking, the district court agreed.

Specifically, the district court held that, while the challenged regulation did have the effect of preventing the dissemination of truthful information to legal adults, this was not unduly out-of-proportion to the second half of the government’s stated aim: reducing abusive alcohol consumption by college students who are 21 years of age or older. We disagree. Instead, we conclude that in this as-applied challenge, “based on ... the application of [the challenged regulation] to [these] specific” plaintiffs, Richmond Med. Ctr. for Women v.

Herring, 570 F.3d 165, 172 (4th Cir.2009) (en banc), the challenged regulation fails under the fourth Central Hudson prong because it prohibits large numbers of adults who are 21 years of age or older from receiving truthful information about a product that they are legally allowed to consume. In Lorillard, the Supreme Court struck down a Massachusetts statute prohibiting tobacco advertising within 1,000 feet of a school.

In concluding that the statute was not adequately tailored to pass muster under Central Hudson’s fourth prong, the Supreme Court indicated that, while it was true that Massachusetts had a substantial interest in preventing underage tobacco use, it was also true that “the sale and use of tobacco products by adults is a legal activity. We must consider that tobacco retailers and manufacturers have an interest in conveying truthful information about their products to adults, and adults have a corresponding interest in receiving truthful information about tobacco products.” Lorillard, 583 U.S. at 564, 121 S.Ct. 2404.

Here, a majority of the College Newspapers’ readers are age 21 or older. Specifically, roughly 60% of the Collegiate Times’s readership is age 21 or older and the Cavalier Daily reaches approximately 10,000 students, nearly 64% of whom are age 21 or older. Thus, the College Newspapers have a protected interest in printing non-misleading alcohol advertisements, just as a majority of the College Newspapers’ readers have a protected interest in receiving that information. Accordingly, the challenged regulation is unconstitutionally overbroad. In Pitt News v.

Pappert, the Third Circuit reached a similar conclusion on similar facts. See 379 F.3d 96 (3d Cir.2004). That case involved a Pennsylvania Liquor Control Board regulation prohibiting alcohol advertisements in college or university publications. The Pitt News, the University of Pittsburgh’s student newspaper, brought an action under 42 U.S.C. § 1983, alleging that the regulation violated the First Amendment as applied.

The Third Circuit agreed, invalidating the prohibition under Central Hudson’s fourth prong and noting that “[m]ore than 67% of Pitt students and more than 75% of the total University population is over the legal drinking age and, in Lorillard, the Supreme Court held that a restriction on tobacco advertising was not narrowly tailored in part because it prevented the communication to adults of truthful information about products that adults could lawfully purchase and use.” Id. at 108.

In analyzing the challenged regulation under Central Hudson’s fourth prong, the district court concluded that the fact that a majority of the College Newspapers’ readers were age 21 or older did not render the challenged regulation overbroad because the regulation was designed, in part, to prevent abusive drinking by persons who are of legal age to drink. However, regardless of the importance of this interest, the ABC’s approach remains over-broad under Supreme Court precedent.

In Sorrell, Vermont attempted to justify its prohibition on the commercial disclosure of prescriber-identifying information based on the fact that the use of such information by pharmaceutical representatives “undermines the doctor-patient relationship by allowing detailers to influence treatment decisions.” Sorrell, 131 S.Ct. at 2670. The Supreme Court rejected this argument, observing that “ ‘[t]he First Amendment directs us to be especially skeptical of regulations that seek to keep people in the dark for what the government perceives to be their own good.’ ” Id. at 2671 (quoting 44 Liquormart, Inc. v.

Rhode Island, 517 U.S. 484, 503, 116 S.Ct. 1495, 134 L.Ed.2d 711 (1996)). Thus, the Court held that states may not “seek to remove a popular but disfavored product from the marketplace by prohibiting truthful, non-misleading advertisements.” Id. at 2670. Accordingly, the Court invalidated the Vermont ban under Central Hudson. Here, the portion of the challenged regulation seeking to prevent the dissemination of alcohol advertisements to readers age 21 or older does exactly what Sorrell prohibits: it attempts to keep would-be drinkers in the dark based on what the ABC perceives to be their own good.

Therefore, the district court erred in concluding that the challenged regulation is appropriately tailored to achieve its objective of reducing abusive college drinking. IV. Because a regulation of commercial speech must satisfy all four Central Hudson prongs in order to survive an as-applied challenge, and the regulation in question here does not satisfy the fourth prong, the challenged regulation violates the First Amendment as applied to the College Newspapers. Therefore, the judgment of the district court is REVERSED. .

In 2008, at the time of the initial district court disposition in this case, the relevant regulatory language was codified at 3 Va. Admin. Code § 5-20-40(B)(3) (2008). This

section was amended in 2010. The only substantive difference between the version of the regulation in place in 2008 and the version in place at present is that the current version of the regulation does not enumerate the phrases permissible for use in an alcohol advertisement by a “dining establishment/' while the former version did. This difference is immaterial to our resolution of the present appeal. . Citations to the “J.A.” refer to the Joint Appendix filed by the parties in this appeal. . "Prescriber-identifying information” is information that identifies doctors and the drugs that they prescribe.

It is often used as a marketing tool by pharmaceutical sales representatives. See Sorrell, 131 S.Ct. at 2659-60. . While Sorrell spoke in terms of “heightened scrutiny” as opposed to "strict scrutiny,” the College Newspapers nonetheless argue that strict scrutiny applies. To be sure, the question of whether Son-ell’s “heightened scrutiny” is, in fact, strict scrutiny remains unanswered. However, because we conclude that the challenged regulation in this case fails Central Hudson, we need not attempt to answer that question here. .

The difference between a facial challenge and an as-applied challenge lies in the scope of the constitutional inquiry. Under a facial challenge, a plaintiff may sustain its burden in one of two ways. First, a plaintiff asserting a facial challenge “may demonstrate 'that no set of circumstances exists under which the law would be valid, or that the law lacks any plainly legitimate sweep.' ” Greater Balt. Ctr. for Pregnancy Concerns, Inc. v. Mayor and City Council of Balt., 721 F.3d 264, 282 (4th Cir.2013) (en banc)(alterations omitted) (quoting United States v.

Stevens, 559 U.S. 460, 130 S.Ct. 1577, 1587, 176 L.Ed.2d 435 (2010)). Second, a plaintiff asserting a facial challenge may also prevail if he or she "show[s] that the law is 'overbroad because a substantial number of its applications are unconstitutional, judged in relation to the statute’s plainly legitimate sweep.' ” Id. (alterations omitted)(quoting Stevens, 130 S.Ct. at 1587). Under either scenario, a court considering a facial challenge is to assess the constitutionality of the challenged law "without regard to its impact on the plaintiff asserting the facial challenge.” Swecker, 602 F.3d at 588.

In contrast, an as-applied challenge is "based on a developed factual record and the application of a statute to a specific person[.]” Richmond Med. Ctr. for Women v. Herring, 570 F.3d 165, 172 (4th Cir.2009) (en banc).