Opinion
No. 23-1119 No. 23-1122 No. 23-1154
07-01-2024
ARGUED: Bryan Jeffrey Jones, BRYAN J. JONES, LLC, Charlottesville, Virginia; James Edward Kolenich, KOLENICH LAW OFFICE, Cincinnati, Ohio, for Appellants/Cross-Appellees. David E. Mills, COOLEY LLP, Washington, D.C.; Raymond P. Tolentino, KAPLAN HECKER & FINK LLP, Washington, D.C., for Appellees/Cross-Appellants. ON BRIEF: Joshua M. Siegel, Caitlin B. Munley, Robby Lee Ray Saldaña, Khary J. Anderson, Washington, D.C., Alan D. Levine, COOLEY LLP, New York, New York; Roberta A. Kaplan, Gabrielle E. Tenzer, KAPLAN HECKER & FINK LLP, New York, New York; Karen L. Dunn, Jessica E. Phillips, Washington, D.C., Yotam Barkai, Melina Maria Meneguin Layerenza, PAUL, WEISS, RIFKIND, WHARTON & GARRISON LLP, New York, New York, for Appellees/Cross-Appellants. Sarah Warbelow, Cynthia Cheng-Wun Weaver, JP Schnapper-Casteras, HUMAN RIGHTS CAMPAIGN FOUNDATION, Washington, D.C.; Elizabeth Littrell, Decatur, Georgia, Scott D. McCoy, SOUTHERN POVERTY LAW CENTER, Miami, Florida; Edward J. Jacobs, Michelle N. Tanney, Jonathan A. Forman, Robyn M. Feldstein, J'Naia L. Boyd, Shade I. Quailey, Sydney W. Park, New York, New York, Kendall C. Kash, BAKER & HOSTETLER LLP, Cleveland, Ohio, for Amici The Human Rights Campaign Foundation, The Southern Poverty Law Center, and Legal Aid Justice Center. Kyle McNew, David Thomas, MICHIEHAMLETT, Charlottesville, Virginia, for Amici Virginia Law Professors.
Appeals from the United States District Court for the Western District of Virginia, at Charlottesville. Norman K. Moon, Senior District Judge. (3:17-cv-00072-NKM-JCH) ARGUED: Bryan Jeffrey Jones, BRYAN J. JONES, LLC, Charlottesville, Virginia; James Edward Kolenich, KOLENICH LAW OFFICE, Cincinnati, Ohio, for Appellants/Cross-Appellees. David E. Mills, COOLEY LLP, Washington, D.C.; Raymond P. Tolentino, KAPLAN HECKER & FINK LLP, Washington, D.C., for Appellees/Cross-Appellants. ON BRIEF: Joshua M. Siegel, Caitlin B. Munley, Robby Lee Ray Saldaña, Khary J. Anderson, Washington, D.C., Alan D. Levine, COOLEY LLP, New York, New York; Roberta A. Kaplan, Gabrielle E. Tenzer, KAPLAN HECKER & FINK LLP, New York, New York; Karen L. Dunn, Jessica E. Phillips, Washington, D.C., Yotam Barkai, Melina Maria Meneguin Layerenza, PAUL, WEISS, RIFKIND, WHARTON & GARRISON LLP, New York, New York, for Appellees/Cross-Appellants. Sarah Warbelow, Cynthia Cheng-Wun Weaver, JP Schnapper-Casteras, HUMAN RIGHTS CAMPAIGN FOUNDATION, Washington, D.C.; Elizabeth Littrell, Decatur, Georgia, Scott D. McCoy, SOUTHERN POVERTY LAW CENTER, Miami, Florida; Edward J. Jacobs, Michelle N. Tanney, Jonathan A. Forman, Robyn M. Feldstein, J'Naia L. Boyd, Shade I. Quailey, Sydney W. Park, New York, New York, Kendall C. Kash, BAKER & HOSTETLER LLP, Cleveland, Ohio, for Amici The Human Rights Campaign Foundation, The Southern Poverty Law Center, and Legal Aid Justice Center. Kyle McNew, David Thomas, MICHIEHAMLETT, Charlottesville, Virginia, for Amici Virginia Law Professors. Before DIAZ, Chief Judge, and NIEMEYER and WYNN, Circuit Judges. Affirmed in part, vacated in part, and remanded with instructions by published opinion. Chief Judge Diaz wrote the opinion, in which Judge Niemeyer and Judge Wynn joined. DIAZ, Chief Judge:
In August 2017, a group of protesters traveled to Charlottesville, Virginia, to demonstrate against the City's decision to remove a statue of Robert E. Lee. The organizers called this event the "Unite the Right" rally, or the "Battle of Charlottesville." What followed was a harrowing weekend in our Nation's history.
As a later civil jury trial proved, these protesters (among whom were white nationalists, white supremacists, and neo-Nazis) conspired to commit racially motivated violence to, in part, "defend Western civilization and white men against perceived enemies—specifically, Jewish persons, Black persons, and their white gentile traitor allies." Sines v. Kessler, No. 3:17-cv-00072, 2022 WL 18026336, at *23 (W.D. Va. Dec. 30, 2022) (cleaned up). The protesters "sought violence, planned for violence, sparked violence, engaged in violence, and afterwards, glorified the violence" committed. Id. at *33.
The jury's damages award against the protester-defendants totaled over $26 million, split among compensatory damages the district court assigned jointly and severally against the defendants, and a historic $24 million in punitive damages. But the district court ultimately slashed those punitive damages to $350,000 by applying Virginia's so-called punitive damages cap across the eight plaintiffs who sought them, see Va. Code Ann. § 8.01-38.1. The compensatory and punitive damages form the bedrock of this consolidated cross-appeal.
Defendants Michael Hill, Michael Tubbs, League of the South, and Nathan Damigo, challenge the district court's decision to hold them jointly and severally liable for the jury's compensatory damages award. Meanwhile, the plaintiffs challenge the district court's decision to apply Virginia's punitive damages cap across all plaintiffs. This latter issue is one of first impression for both our court and the Supreme Court of Virginia.
The other defendants who appeared at trial, including Jason Kessler, are not parties to the portion of the appeal involving joint-and-several liability. All defendants other than Michael Enoch Peinovich join in the appeal involving punitive damages.
We affirm the district court's imposition of joint-and-several liability for the compensatory damages. But because we hold that Virginia's punitive damages cap applies on a per-plaintiff basis, we vacate the district court's ruling on that question and remand with instructions to apply the cap accordingly.
I.
A.
In June and July 2017, Jason Kessler began organizing the Unite the Right rally and inviting various Alternative Right (or Alt-Right) and white supremacist groups to attend. Among those attendees were Damigo (and the organization he founded, Identity Evropa), Hill (and the organization he founded, League of the South), and Tubbs.
Damigo had gained some renown within these groups after he was filmed punching a woman at a protest in Berkeley, California, a few months earlier. And Hill had published an online "Pledge of Allegiance," "in which he stated: 'I pledge to be a white supremacist, a racist, an anti-Semite, a homophobe, a xenophobe, an Islamophobe, and any sort of 'phobe that benefits my people.' " Sines, 2022 WL 18026336, at *22 (cleaned up).
Kessler's planning was extensive, and from the beginning contemplated violence. He and his codefendants "planned for months to provoke Antifa and its followers into a violent battle at Unite the Right." Id. at *3. Outfitted with "helmets, shields, and body armor, and armed with weapons," the defendant-protesters "strategized how any punch thrown by an Antifa supporter would give them a chance to respond with brutal and overwhelming violence." Id. Kessler sought to "rais[e] an army" and "encouraged attendees to bring signposts, shields, and other 'self-defense implements,' which could be transformed from 'free speech tools into self-defense weapons should things turn ugly.' " Id. at *9 (cleaned up).
Antifa—a portmanteau of the words, "anti" and "fascists"—has become known for opposing the views of the Alt-Right and for clashing with Alt-Right groups, often violently.
Damigo believed that "violence might be necessary to secure [a] white ethnostate." Id. at *3 (cleaned up). And he wasn't alone. Using the online communications platform Discord, the codefendants' servers were "replete" with messages "demonstrating that they shared expectations of, hoped for, planned for, and purposefully sought to instigate violence at Unite the Right—including discussing whether someone could drive a car through a crowd of demonstrators that might be blocking the street." Id. at *5.
"As numerous private social media messages reflect, Kessler undertook several steps ahead of time to spark violence at Unite the Right between the rally-goers and Antifa." Id. at *8 (cleaned up). For example, he asked that his followers "taunt[ ] [Antifa] a little on social media," and "tip off Antifa" about the rally location to ensure confrontations. Id. at *8; accord id. ("We definitely want to play these people into our hands Saturday in Charlottesville." (cleaned up)); id. at *19 (explaining that evidence showed codefendants "shared similar motives and sought various methods to spark violence at which point Unite the Right rallygoers would be prepared to respond with overwhelming force and violence").
Kessler and the codefendants planned and executed two events over the Unite the Right weekend. First, they "led hundreds wielding lit tiki torches and chanting[,] 'You will not replace us,' 'Jews will not replace us,' 'Blood and soil,' and 'Sieg Heil,' through University of Virginia grounds up to the Rotunda and down to the Thomas Jefferson statue." Id. at *6. There, the marchers encountered and surrounded counterprotesters (including several plaintiffs), where they "prevented [the counterprotesters] from escaping, pepper sprayed [them], and subjected [them] to racist insults," id., in a show "of dominance," id. at *10 (cleaned up).
The next day, the violence from the night before reached its "flashpoint," id. at *10, as protesters and counterprotesters clashed near and around the Lee statue. The defendants "le[d] the troops into battle and coordinat[ed their] attack," id. at *23, using weapons and words against both violent Antifa and peaceful counterprotesters alike, see id. at *24 (Hill conceding at trial that League of the South members "engaged in some violence that they initiated"). The mayhem ended with defendant James Alex Fields, Jr., driving his car into a group of counterprotesters, killing Heather Heyer, and injuring other plaintiffs.
The jury rejected the defendants' self-defense justification at trial. See Sines, 2022 WL 18026336, at *9.
After the rally, the protesters celebrated, saying that they " 'couldn't be happier' with the outcome" and that they "would not go back and change a thing." Id. at *25 (cleaned up). They reiterated their support for Fields, "expressing numerous times that Heather Heyer . . . was to blame for her own death," id. at *9, and even "refer[ing] to [her] grave as a urinal," id. at *12 (cleaned up). "Hail victory," they proclaimed. Id. at *6 (cleaned up). "This is just the beginning . . . the war goes on." Id. (cleaned up).
B.
Following these events, Elizabeth Sines and eight others sued twenty-three individual and organizational defendants, thirteen of whom appeared at trial. As relevant here, all the plaintiffs alleged against all the defendants a Virginia state-law civil conspiracy claim (Count III). A subset of the plaintiffs also alleged against a subset of the defendants a Virginia state-law claim for racial, religious, or ethnic harassment under Virginia's hate-crime statute (Count IV); and Virginia state-law tort claims for assault and battery (Count V), and intentional infliction of emotional distress (Count VI).
Plaintiffs also alleged federal civil rights and conspiracy claims (Counts I and II). The jury didn't reach a verdict on those claims, and they aren't at issue in this appeal.
The trial lasted over four weeks and saw the parties admit over 900 exhibits and call 35 witnesses.
In its final jury instructions, the district court explained for Count III that "[u]nder Virginia law, persons who conspired together to commit one or more unlawful acts may be held liable for the injuries that result from that conspiracy." J.A. 4706. It then defined the five "unlawful or tortious" predicate acts of the conspiracy:
• Subjecting persons to actions of intimidation or harassment, motivated by racial, religious, or ethnic animosity in violation of Virginia Code § 8.01-42.1 (often referred to as Virginia's hate[-]crime[ ] statute);J.A. 4706-07.
• Directing violence at another person, motivated by racial, religious, or ethnic animosity in violation of Virginia Code § 8.01-42.1;
• Committing an unwanted touching that was neither consent[ed] to, excused, or justified (battery);
• Engaging in an overt act intended to inflict bodily harm, or intended to place the victim in fear or apprehension of bodily harm (assault); and
• Causing a reasonable apprehension that force will be used unless a person willingly submits and causing him to submit to the extent that he is denied freedom of action (false imprisonment).
The district court emphasized that "Plaintiffs need only prove that Defendants conspired to commit one of these underlying acts to impose liability." J.A. 4707. And it said that it would instruct the jury "on the elements that plaintiffs must prove" for each of those five predicate acts. J.A. 4708.
The district court then provided those elements for Count IV (violations of the Virginia hate-crime statute) and Count V (assault and battery). It described those two counts as "standalone claim[s]," J.A. 4709, 4711, brought in addition to the plaintiffs' civil conspiracy claim that the defendants collectively violated the hate-crime statute and committed assault and battery.
But when the court instructed the jury on Count VI (intentional infliction of emotional distress), it didn't include such "standalone" language, nor did it relate that claim to the broader conspiracy charge.
C.
After three days of deliberation, the jury found for the plaintiffs on Counts III, IV, V, and VI.
On Count III, the jury awarded nominal damages to the plaintiffs (except for Sines and another plaintiff). It also awarded $500,000 in punitive damages against each individual defendant and $1,000,000 in punitive damages against each organizational defendant.
On Count IV, the jury awarded each of the two plaintiffs bringing that claim $250,000 in compensatory damages against Kessler, Richard Spencer, Elliot Kline, Robert "Azzmador" Ray, and Christopher Cantwell. It also awarded $200,000 in punitive damages against each of these defendants.
The jury found a fifth defendant, Fields, liable under Count IV, but it didn't assess any compensatory or punitive damages against him on that Count. Instead, as discussed more fully, infra pp. 347-48, the jury assessed a substantial punitive damages award against him on Counts V and VI.
On Count V, the jury awarded the five plaintiffs who brought that claim a total of $803,277 in compensatory damages against the one defendant, Fields, who it found liable. It then awarded $6,000,000 in punitive damages against Fields.
Finally on Count VI, the jury awarded the six plaintiffs who brought that claim a total of $701,459 in compensatory damages against the one defendant—again, Fields—who it found liable. And once more, it awarded $6,000,000 in punitive damages against Fields.
Because the jury found that all the defendants engaged in a conspiracy under Count III, the district court applied the collective compensatory damages for the conspiracy's predicate acts—Counts IV and V—jointly and severally. It didn't, however, apply the damages for Count VI jointly and severally because the intentional infliction of emotional distress wasn't one of the conspiracy's predicate acts.
D.
The parties exchanged a flurry of post-trial motions. As relevant here, Damigo, Hill, Tubbs, and League of the South, argued that Virginia's punitive damages cap should apply to the jury's punitive damages award, reducing them to $350,000, and that the damages were constitutionally excessive.
The plaintiffs, in turn, moved for default judgments against defendants Andrew Anglin, Moonbase Holdings, LLC, East Coast Knights of the KKK, Fraternal Order of the Alt-Knights, August Sol Invictus, The Loyal Knights of the KKK, and the Nationalist Front, none of whom appeared at trial. The plaintiffs maintained that the defaulted defendants should be jointly and severally liable with the trial defendants for the compensatory damage awards under Counts III and IV as coconspirators. The plaintiffs also sought punitive damages against the defaulted defendants.
E.
The district court addressed the parties' post-trial motions in two orders. Because Virginia's punitive damages cap figures prominently in the court's rulings, we recite it at length here:
In any action . . . including an action for medical malpractice under Chapter 21.1 (§ 8.01-581.1 et seq.), the total amount awarded for punitive damages against all defendants found to be liable shall be determined by the trier of fact. In no event shall the total amount awarded for punitive damages exceed $350,000 . . . . [I]f a jury returns a verdict for punitive damages in excess of the maximum amount specified in this section, the judge shall reduce the award and enter judgment for damages in the maximum amount provided by this section.
The district court first ruled on the defendants' combined post-trial motions, which it almost uniformly rejected, in a thoughtful nearly-90-page order. See Sines, 2022 WL 18026336, at *1. Critically, however, the district court agreed that the Virginia punitive damages cap applied.
In doing so, the district court rebuffed the plaintiffs' arguments that the punitive damages cap applied only in "run-of-the-mill" tort and insurance cases rather than in the hate-crime context. Id. at *27 (cleaned up). The plaintiffs had asserted that because the statute singles out "an action for medical malpractice" as a type of claim to which the cap applies, the legislature intended to limit recovery only in these garden-variety cases—not in a civil rights lawsuit for racially motivated crimes. Id. (cleaned up).
The district court also rejected the plaintiffs' fallback contention that the cap applies on a per-plaintiff basis, with each plaintiff receiving $350,000.
While it acknowledged that this court expressly "left th[is] [per-plaintiff] issue open" in Al-Abood v. El-Shamari, 217 F.3d 225, 237 n.10 (4th Cir. 2000), it found persuasive that in Al-Abood we had held that the cap didn't apply on a per-defendant basis. Id. at *28. And the district court reasoned that the statutory language mandated that the cap applied " '[i]n any action,' not 'a plaintiff's action,' or 'each plaintiff's right of action.' " Id. (cleaned up). Taken together, the district court found that our Al-Abood defendant-cap holding and the statutory language "strongly indicate[d]" the cap applied no matter how many plaintiffs. Id.
But the district court rejected the defendants' additional argument that the punitive damages award was unconstitutionally excessive given the purportedly outsized ratio between punitive and compensatory damages.
The defendants had claimed that the court should compare any such ratio on a claim-by-claim basis. The plaintiffs responded that, because the defendants were jointly and severally liable for Counts III, IV, and V, the court should compare the aggregate compensatory damages for those counts to the aggregate punitive damages. In his separate reply brief, Damigo dismissed this "argument regarding joint and several liability" as "miss[ing] the mark" for the appropriate punitive damages ratio because the plaintiffs hadn't sued Damigo under either Count IV or Count V. Defs.' Reply Br. as to Punitive Damages at 4, Sines, 2022 WL 18026336 (No. 3:17-cv-00072), ECF No. 1593.
The district court agreed with the plaintiffs' aggregate approach because "the jury found Defendants liable for a civil conspiracy," which "meant that Defendants were held jointly and severally liable for the damages award." Sines, 2022 WL 18026336, at *34 (cleaned up). The district court thus determined that the punitive damages award wasn't constitutionally excessive.
Then in a separate order, the district court granted the plaintiffs' default judgment motion, finding the defaulted defendants jointly and severally liable for the compensatory damages stemming from the unlawful conspiracy. But it held that the punitive damages cap applied to the punitive damages award against those defendants too. The court therefore reduced the jury's punitive damages from $24 million to a final $350,000.
This appeal followed.
II.
Hill, Tubbs, League of the South, and Damigo first challenge the district court's decision to impose the jury's compensatory damage awards for Counts IV and V jointly and severally. They contend that Virginia law and the Seventh Amendment foreclose doing so because none of these four defendants were charged under those two counts. Plaintiffs respond that defendants forfeited this argument by failing to raise it in the district court, and that the argument is otherwise meritless.
In their cross-appeal, plaintiffs ask us to reverse the district court's order reducing the jury's punitive damages to $350,000 under Virginia's punitive damages cap. They urge us to hold (1) that the punitive damages cap doesn't apply at all to awards remedying violations of Virginia's hate-crime statute, or (2) alternatively, that the cap applies on a per-plaintiff (not an all-plaintiff) basis.
A.
We turn first to the joint-and-several liability challenge. We're inclined to agree with plaintiffs that this argument was forfeited. As we've explained, "issues raised for the first time on appeal are generally not considered absent exceptional circumstances." Holly Hill Farm Corp. v. United States, 447 F.3d 258, 267 (4th Cir. 2006) (cleaned up). And in the waiver context, we've held that "tak[ing] a passing shot at [an] issue" cannot preserve that issue for appellate review. Grayson O Co. v. Agadir Int'l LLC, 856 F.3d 307, 316 (4th Cir. 2017) (cleaned up); see also Muhammad v. Giant Food Inc., 108 F. App'x 757, 764 (4th Cir. 2004) (holding that appellants forfeited issue on appeal after only a "passing mention" of that issue to the district court).
There are important distinctions between waiver and forfeiture not implicated here. But in sum, "forfeiture is the failure to make the timely assertion of a right, [while] waiver is the intentional relinquishment or abandonment of a known right." United States v. Olano, 507 U.S. 725, 733, 113 S.Ct. 1770, 123 L.Ed.2d 508 (1993) (cleaned up). Thus, for a waived argument, the parties may have exchanged briefing on the issue that the district court evaluated. See, e.g., Manning v. Caldwell for City of Roanoke, 930 F.3d 264, 271 (4th Cir. 2019). By contrast, a forfeited argument would have left no similar footprint in the record.
The best that defendants can muster is a single such "passing shot." They claim that "Damigo specifically made the joint and several liability argument in post-trial briefing." Appellants' Reply Br. at 7.
To be sure, Damigo wrote the words "joint and several liability" in his brief, but only as a rejoinder to plaintiffs' arguments on aggregating the ratio between the jury's compensatory and punitive damages awards for constitutional purposes. Defs.' Reply Br. as to Punitive Damages at 4, Sines, 2022 WL 18026336 (No. 3:17-cv-00072), ECF No. 1593. And neither Damigo nor any other defendant elaborated any further or developed any argument substantively challenging the court's joint-and-several liability finding. See M.D. Russell Constr., Inc. v. Consol. Staffing, Inc., No. 22-1420, 2023 WL 8798086, at *6 (4th Cir. Dec. 20, 2023) (holding that appellants forfeited issue on appeal even after raising argument in summary judgment briefing where it didn't "elaborate or provide any citations on the record or argument on this point").
Perhaps reading the tea leaves, defendants say that we should still consider their claim because "the equities require [review]." Appellants' Reply Br. at 7 (cleaned up). But even then, our standard is more demanding than defendants admit. Indeed, "[w]hen a party in a civil case . . . raises an argument for the first time before us, we may reverse only if the newly raised argument establishes fundamental error or a denial of fundamental justice." Tarashuk v. Givens, 53 F.4th 154, 167 (4th Cir. 2022) (cleaned up). And this "newly raised argument" doesn't simply fail to establish a fundamental error, it identifies no error at all. Nor could it.
That's because "a civil conspiracy [claim] is a mechanism for spreading liability among coconspirators for damages sustained as a result of the underlying act that is itself wrongful or tortious." La Bella Dona Skin Care, Inc. v. Belle Femme Enters., LLC, 294 Va. 243, 805 S.E.2d 399, 406 (2017) (cleaned up); see also Gelber v. Glock, 293 Va. 497, 800 S.E.2d 800, 821 (2017) ("The object of a civil conspiracy claim is to spread liability to persons other than the primary tortfeasor."). That's precisely what the district court did here: It jointly allocated the liability among the coconspirators for the conspiracy and its predicate acts.
In insisting that Virginia law permits joint-and-several liability only among joint tortfeasors who "produce a single injury," Appellants' Br. at 5 (cleaned up), defendants overlook that the jury found that they weren't simply joint tortfeasors, they were coconspirators, and so were liable for all damage stemming from that conspiracy.
Confoundingly, defendants concede that "[they] were found liable under the Count III state law conspiracy claim," id. at 6, yet they insist that the jury was somehow "specifically prohibited from finding them liable on Counts IV or V by the Court's instructions to the jury and the jury verdict form," id. at 5. Not so.
The court instructed the jury that "persons who conspired together to commit one or more unlawful acts may be held liable for the injuries that result from that conspiracy." J.A. 4706. The court then listed the five predicate acts, which included violations of Virginia's hate-crime statute and common law assault or battery. Count IV and Count V's "standalone" claims were those very predicate acts (and the objects of the conspiracy), which the jury found defendants committed. We need look no further for confirmation than the district court's own post-judgment ruling, where it affirmed that it "instructed the jury" that Counts IV and V "would constitute predicate unlawful acts of the conspiracy" in Count III. Sines, 2022 WL 18026336, at *34 n.31.
The verdict form required a liability finding on Count III for the conspiracy (and to all coconspirators) before assessing the specific damages for that conspiracy (and jointly to all coconspirators) in Counts IV and V. That Count III awarded nominal damages and Counts IV and V awarded compensatory (and punitive) damages ensured that the plaintiffs didn't receive duplicative damages for the conspiracy and its predicate acts.
Thus, the district court committed no fundamental error that would forgive a forfeiture, nor did it commit any error that would justify our disturbing its joint-and-several liability finding even absent forfeiture. We affirm.
B.
We turn next to the harder question about Virginia's punitive damages cap. Plaintiffs would have us not apply it at all or apply it on a per-plaintiff basis. Defendants would have us affirm the district court's application of the cap or certify the question to the Supreme Court of Virginia. As we did in Al-Abood, we choose instead to "predict what the Supreme Court of Virginia would decide were the [per-plaintiff] issue presented to it." 217 F.3d at 237.
The district court considered certifying this question to Virginia's high court but opted not to "given the challenges to the merits of the jury verdict and other federal and state law challenges to the damages award." Sines, 2022 WL 18026336, at *30 n.29.
As we explain, we hold that the cap applies generally in the hate-crime context but that it applies on a per-plaintiff basis.
1.
"The proper application of [Virginia's punitive damages cap] is a question of law" that we review de novo. Al-Abood, 217 F.3d at 237. And though we acknowledge, as did the district court, that the interpretative question before us is somewhat unsettled under Virginia law, we agree that we needn't begin on a "blank slate." Sines, 2022 WL 18026336, at *28.
Two cases guide our inquiry. First, in Wackenhut Applied Technologies Center, Inc. v. Sygnetron Protection Systems, Inc., we considered whether "the Virginia legislature intended that the punitive damages cap should apply only in unintentional tort actions." 979 F.2d 980, 984 (4th Cir. 1992). After reviewing the statute's text, and particularly its inclusion of the broad term "[i]n any action," we rejected this argument. Id. (cleaned up).
We reasoned that "[t]he drafters of the Virginia statute, in limiting punitive damages, chose the phrase 'any action' to define the class of cases to which the statute would apply—they did not further define 'any action.' " Id. We declared that phrase, therefore, "unequivocal," because there was simply "no definitional language indicating that . . . 'any action' is limited to unintentional tort actions." Id.
Nor were we convinced that "the legislative history or intent of the statute" changed that straightforward interpretation. Id. Instead, "[i]n our view, the history of this legislation indicate[d] that the Virginia legislature intended that the punitive damages cap be applied 'in any action.' " Id. So "in interpreting the statute," we declined to "deviate from [its] plain meaning." Id.
We also have our decision in Al-Abood. There, we addressed whether the district court erred by reducing a punitive damages award under the cap but applying it on a per-defendant basis. Al-Abood, 217 F.3d at 236-37. The defendants argued that the cap operated instead on a per-lawsuit basis, such that the award "should have been lowered to $350,000" total, because the cap applied to "the total amount awarded . . . against all defendants." Id. at 237 (cleaned up).
We agreed. Turning first to the statutory text (as we did in Wackenhut), we found "no ambiguity . . . and conclude[d] that its plain meaning dictate[d] that the cap on punitive damage awards applie[d] to the action as a whole, and not to each defendant." Id.
We treated as dispositive that the first sentence of the statute referred to both "the total amount" and "against all defendants," thus "mandat[ing] that the entirety of the punitive damages awarded in the action amount to no more than to $350,000." Id. (emphasis added). But our holding was narrow: We "expressed no opinion on how the cap would be applied in a case involving multiple plaintiffs." Id. at 237 n.10.
That's the question before us today.
But before reaching it, we reject plaintiffs' contention that the punitive damages cap doesn't apply at all. Plaintiffs reiterate that, in enacting the cap, the General Assembly was concerned with "run-of-the-mill tort claims, such as medical malpractice claims," Appellees' Br. at 22, not with "proceedings involving Virginia's hate-crime statute," or similar intentional torts, id. at 21. And they contend that the statute's text and legislative history supports this view.
We cannot agree. While the misconduct here was egregious, and was fueled by a repugnant desire to commit (and celebrate) racially motivated violence, our decision in Wackenhut forecloses plaintiffs' argument. We held there that neither the statute nor its legislative history brokered any ambiguity about a carve-out for intentional torts. Wackenhut, 979 F.2d at 984.
Plaintiffs nonetheless mobilize noscitur a sociis—a canon of construction by which one interprets an ambiguous term by the words around it—to argue that the statute's use of "action" "keeps very specific company: 'an action for medical malpractice.' " Id. at 22 (cleaned up). And because of this textual neighborhood, the General Assembly didn't intend "action" to apply to a civil rights lawsuit such as the one brought by plaintiffs here.
But that argument doesn't work because the statutory language isn't ambiguous. See United States v. Stevens, 559 U.S. 460, 474, 130 S.Ct. 1577, 176 L.Ed.2d 435 (2010) ("[A]n ambiguous term may be given more precise content by the neighboring words with which it is associated." (cleaned up)); Andrews v. Ring, 266 Va. 311, 585 S.E. 2d 780, 784 (2003) (using canons of construction to interpret "the meaning of doubtful words in a statute" (emphasis added)).
Equally unavailing is plaintiffs' invocation of the "mischief rule," which "dictates that every statute must be interpreted to promote the ability of the enactment to remedy the mischief at which it is directed." Appellees' Br. at 23 (cleaned up). Plaintiffs maintain that the General Assembly enacted the cap "to ensure predictability and stability for the insurance industry." Id. at 25. But Virginia passed its hate-crime statute only a year after passing its punitive damages cap. Compare Va. Code Ann. § 8.01-42.1 (1988), with id. § 8.01-38.1 (1987). It therefore had an opportunity to exclude hate crimes from the cap's purview.
But it didn't. The legislature's choice precludes the argument that the only mischief the cap addressed was in common tort and insurance claims. We have no mandate to second-guess that policy decision.
2.
With that issue settled, we return to plaintiffs' second argument: That even if the cap applies, it does so on a per-plaintiff basis. We agree that it does.
Defendants point us to a Virginia trial court decision that they say comes to the opposite conclusion. Bavely v. Geneva Enterprises, Inc., 112 Va. Cir. 323, 2023 WL 7708506 (Va. Cir. Ct. 2023). The court's central holding was that the punitive damages cap doesn't apply on a per-claim basis, id. at *13, a construction the district court found the plaintiffs didn't advance here, Sines, 2022 WL 18026336, at *29 n.24. And though the Bavely court casually remarked that the cap didn't allow a per-plaintiff interpretation, it did so merely by citing to our decision in Al-Abood and the district court's decision below. 2023 WL 7708506, at *5. Absent a more thorough engagement with this question, the case holds little weight.
First, recall the relevant statutory language:
In any action . . . including an action for medical malpractice under Chapter 21.1 (§ 8.01-581.1 et seq.), the total amount awarded for punitive damages against all defendants found to be liable shall be determined by the trier of fact. In no event shall the total amount awarded for punitive damages exceed $350,000 . . . . [I]f a jury returns a verdict for punitive damages in excess of the maximum amount specified in this section, the judge shall reduce the award and enter judgment for damages in the maximum amount provided by this section.
The district court was convinced that we had sent strong signals in Al-Abood that the cap should apply against all plaintiffs. Indeed, we did say that "even without the 'against all defendants' phrase, the statute refers to the 'total amount awarded' '[i]n any action,' " Al-Abood, 217 F.3d at 237, which suggests that the inclusion of that phrase alone doesn't change the "total amount awarded" for multiple plaintiffs.
But with the per-plaintiff question squarely in our sights, we'll take more than just a passing glance.
Our first reaction is that the statute doesn't include similar "all plaintiffs" language as it does "all defendants." We think that legislative drafting choice means something. And if we treated defendants and plaintiffs equally under the statute, even with the "all defendants" qualifier, we would render that term superfluous. See Commonwealth v. Squire, 278 Va. 746, 685 S.E.2d 631, 634 (2009) ("We do not consider actions of the General Assembly to be superfluous; instead, we seek to provide meaning to all words of a statute."); see also Northampton Cnty. Bd. of Zoning Appeals v. E. Shore Dev. Corp., 277 Va. 198, 671 S.E.2d 160, 162 (2009) ("It is the duty of the Court to read legislative enactments to give meaning to all the words used."). To give the "all defendants" phrase meaning (as we must), we would have to find that the statute applies to "all defendants" and not to all parties.
And we think there's good reason why the General Assembly didn't similarly single out "all plaintiffs," considering how it originally understood the word "action" when it enacted the punitive damages cap. This, in turn, requires a historical understanding of joinder in Virginia.
The common law largely didn't permit separate plaintiffs to jointly adjudicate their claims. See Thomas O. Main, Traditional Equity and Contemporary Procedure, 78 Wash. L. Rev. 429, 454 (2003) ("Single-issue pleading precluded the joinder of multiple claims or defenses, the joinder of multiple parties, and pleading in the alternative."); Baker v. Doe, 211 Va. 158, 176 S.E.2d 436, 437 (1970) ("Our research makes clear that at common law parties could not be joined unless their interests were joint."). So at common law, an "action" was typically one plaintiff prosecuting one claim. See generally id. at 437; Rasnick v. Pittston Co., 5 Va. Cir. 336, 1986 WL 740411, at *1 (Va. Cir. Ct. 1986); Dixon v. Robertson, 5 Va. Cir. 544, 1979 WL 201327, at *1-2 (Va. Cir. Ct. 1979).
Common law did, however, permit joinder in some cases. See Carufel v. Am. Isuzu Motors, Inc., 47 Va. Cir. 529, 1999 WL 33198, at *1 (Va. Cir. Ct. 1999) ("At common law and as a general rule of pleading, persons jointly affected by a tort are permitted to join as parties plaintiff in an action to recover the injury. For example, tenants in common should jointly sue a trespasser who has damaged their land; and co-owners of personal property should jointly sue for damages resulting from wrongful conversion."); see also Va. Hot Springs Co. v. Hoover, 143 Va. 460, 130 S.E. 408, 410 (1925) (allowing joinder where plaintiffs' "rights have been invaded by one and the same defendant and by one and the same means" so "they have a common remedy if they elect to pursue it").
Virginia adhered to this common-law approach until 1995, when the General Assembly permitted, for example, a court to join separate actions, or six or more plaintiffs to join in a single action. See Va. Code Ann. §§ 8.01-267.1,-267.5 (1995). These modern joinder principles were enacted after the punitive damages cap passed in 1987. As a matter of text and history then, the General Assembly could not have contemplated including a multi-plaintiff limitation in the punitive damages cap because single-plaintiff actions were the norm.
Virginia embraced these more modern joinder principles in the Multiple Claimant Litigation Act, see Va. Code. Ann. § 8.01-267.1 et seq., a state-law analogue to the Federal Rule of Civil Procedure 20. This Act also allowed the court to join a civil action when "[t]he common questions of law or fact predominate and are significant to the actions," and when the joinder "will promote the ends of justice and the just and efficient conduct . . . of the actions" without prejudicing the parties or violating their due process rights. Id. § 8.01-267.1(2)-(3).
3.
Beyond our historical understanding of joinder principles, we also agree that a per-plaintiff reading serves public policy interests by incentivizing plaintiffs to join their claims where appropriate. Were we to hold otherwise, any plaintiffs seeking punitive damages against a defendant (or defendants) for a single occurrence would have 350,000 reasons not to join their claims. And had that been the law here, the district court likely would have faced eight trials in a loop of repeat operative facts and players.
In fact, it was this policy concern that propelled other jurisdictions to interpret similar punitive damages caps as applying on a per-plaintiff basis. Both Georgia and North Carolina, for instance, have adopted a per-plaintiff construction on judicial economy principles. See Bagley v. Shortt, 261 Ga. 762, 410 S.E.2d 738, 739 (1991); Rhyne v. K-Mart Corp., 149 N.C.App. 672, 562 S.E.2d 82, 93 (2002), aff'd, 358 N.C. 160, 594 S.E.2d 1 (2004).
The Bagley court reasoned that capping the maximum award "without regard to the number of plaintiffs" "would likely produce a proliferation of case filings," thus "encourag[ing] the splitting of causes of action with sophistry and quibble that would rival medieval Schoolmen." 410 S.E.2d at 739. Then more recently, the Rhyne court noted that applying the cap as the district court did here "would discourage parties from joining" because "[p]laintiffs would not take the chance that their possible recoveries would be diluted, not by any defect in their claims, but solely because there was more than one plaintiff." 562 S.E.2d at 93.
We find these state court decisions persuasive and so hold that Virginia's punitive damages cap is best understood as applying on a per-plaintiff basis.
C.
We briefly address Defendants' request that we certify this per-plaintiff question to the Supreme Court of Virginia. We've opted not to, as we did in Al-Abood, because we find the statute's language and history clear enough to predict how Virginia's high court would rule. And while there are cases in which our notions of justice would support further delay, this is not one of them.
Over two years ago, the jury used its $24 million punitive damages award to send an unmistakable message to the defendants and to the public about the outrageous misconduct that took place in Charlottesville, Virginia. While the law compels us to reduce the award, it's long past time for that message to be delivered.
III.
For these reasons, we affirm the district court's order imposing joint-and-several liability for the compensatory damages in Counts IV and V. But we vacate the district court's order to the extent that it reduces the jury's punitive damages to $350,000 for all plaintiffs under the Virginia punitive damages cap. And we remand with instructions that the district court apply the cap instead on a per-plaintiff basis.
AFFIRMED IN PART, VACATED IN PART, AND REMANDED WITH INSTRUCTIONS