Theresa Graham v. R.J Reynolds Tobacco Company

Case: 13-14590 Date Filed: 05/18/2017 Page: 1 of 284

[PUBLISH]

IN THE UNITED STATES COURT OF APPEALS

FOR THE ELEVENTH CIRCUIT
________________________

No. 13-14590
________________________

D.C. Docket No. 3:09-cv-13602-MMH-JBT

THERESA GRAHAM,
as PR of Faye Dale Graham, deceased,

Plaintiff-Appellee,

versus

R.J. REYNOLDS TOBACCO COMPANY,
individually and as successor by merger to the Brown and Williamson Tobacco
Corporation and the American Tobacco Company,
PHILIP MORRIS USA, INC.,

Defendants-Appellants.

________________________

Appeal from the United States District Court
for the Middle District of Florida
________________________

(May 18, 2017)

Before TJOFLAT, HULL, MARCUS, WILSON, WILLIAM PRYOR, MARTIN,
JORDAN, ROSENBAUM, JULIE CARNES, and JILL PRYOR, Circuit Judges.*

WILLIAM PRYOR, Circuit Judge:

*
Chief Judge Ed Carnes recused himself and did not participate in this decision.
 Case: 13-14590 Date Filed: 05/18/2017 Page: 2 of 284

This appeal presents the questions whether due process forbids giving a

jury’s findings of negligence and strict liability in a class action against cigarette

manufacturers preclusive effect in a later individual suit by a class member and, if

not, whether federal law preempts the jury’s findings. Florida smokers and their

survivors filed a class action against several tobacco companies, and after a year-

long trial designed to answer common questions concerning the companies’

tortious conduct against all members of the class, a jury found that each company

had breached its duty of care and sold defective cigarettes. The Florida Supreme

Court upheld the jury verdicts of negligence and strict liability in Engle v. Liggett

Group, Inc., 945 So. 2d 1246 (Fla. 2006) (Engle III), and decertified the class to

allow individual actions about the remaining issues of specific causation, damages,

and comparative fault. The Engle decision made clear that the jury findings of

negligence and strict liability had preclusive effect in the later individual actions,

and the Florida Supreme Court reaffirmed that ruling in Philip Morris USA, Inc. v.

Douglas, 110 So. 3d 419 (Fla. 2013). R.J. Reynolds Tobacco Company and Philip

Morris USA Inc. challenge a jury verdict against them in one of those individual

actions in the district court. They argue that giving the Engle findings preclusive

effect violates the Due Process Clauses, U.S. Const. Amends. V, XIV, and they

urge us to overrule our decision to the contrary in Walker v. R.J. Reynolds Tobacco

Co., 734 F.3d 1278 (11th Cir. 2013). They argue, in the alternative, that federal law

2
 Case: 13-14590 Date Filed: 05/18/2017 Page: 3 of 284

preempts giving preclusive effect to the Engle findings of negligence and strict

liability. Because we reaffirm our holding in Walker and conclude that federal law

does not preempt the Engle jury findings, we affirm the judgments against R.J.

Reynolds and Philip Morris.

I. BACKGROUND

In 1994, six individuals filed a putative class action in Florida court against

the major domestic cigarette manufacturers, including R.J. Reynolds and Philip

Morris, and two tobacco industry organizations. Id. at 1281. They alleged claims of

strict liability, negligence, breach of express warranty, breach of implied warranty,

fraud, conspiracy to commit fraud, and intentional infliction of emotional distress.

Id. The strict liability count alleged that the companies manufactured “cigarettes

containing nicotine,” “manufactured their defective tobacco products by

manipulating the levels of nicotine so as to addict the consuming public,” “failed to

design, manufacture, distribute and sell a safer alternative cigarette that would not

addict smokers,” and “failed to warn” members of the class of the dangers. The

negligence count alleged that the companies “breached their duty of reasonable

care” through several “acts and omissions,” including the “failure to design and

manufacture products that were not addictive,” the “failure to . . . adequately or

sufficiently reduce or remove the level of nicotine in cigarettes,” and the “failure to

warn the smoking consumers of the addictive nature of nicotine.” A Florida district

3
 Case: 13-14590 Date Filed: 05/18/2017 Page: 4 of 284

court of appeal approved the certification of the following class: all Florida citizens

and residents, “and their survivors, who have suffered, presently suffer or have

died from diseases and medical conditions caused by the addiction to cigarettes

that contain nicotine.” R.J. Reynolds Tobacco Co. v. Engle, 672 So. 2d 39, 40–42

(Fla. Dist. Ct. App. 1996) (Engle I).

The trial court in Engle divided the proceedings in three phases. Walker, 734

F.3d at 1281. In Phase I, a jury “decide[d] issues common to the entire class,

including general causation, the Engle defendants’ common liability to the class

members . . ., and the class’s entitlement to punitive damages.” Douglas, 110

So. 3d at 422. Phase I was a year-long trial on “common issues relating exclusively

to defendants’ conduct and the general health effects of smoking.” Liggett Grp.

Inc. v. Engle, 853 So. 2d 434, 441 (Fla. Dist. Ct. App. 2003) (Engle II). Phase I

required “hundreds of witnesses, thousands of documents and exhibits, and tens of

thousands of pages of testimony.” Douglas, 110 So. 3d at 431. In Phase II, the jury

determined the liability of the tobacco companies to three class representatives,

awarded them compensatory damages, and fixed the amount of class-wide punitive

damages. Walker, 734 F.3d at 1281. The trial court planned to have new juries

decide specific causation and damages for the remaining class members in Phase

III. Id.

4
 Case: 13-14590 Date Filed: 05/18/2017 Page: 5 of 284

In his opening statement in Phase I, the plaintiffs’ attorney stated, “The

evidence will show, ladies and gentlemen, that there is no dispute or controversy in

the medical and scientific communities but that cigarette smoking causes lung

cancer, heart disease, chronic obstructive pulmonary disease, emphysema and

many other diseases.” He stated that “the evidence will establish overwhelmingly”

that “[n]icotine is addictive.” And he explained that the tobacco companies “have

the technology to make a safer cigarette” but not one that is profitable. He also

stated that “the evidence will show that the tobacco companies have so

successfully misled the American people that many highly intelligent people, in

1998, are confused.”

The smokers presented a substantial body of evidence that all of the

cigarettes manufactured by the named defendants contained carcinogens that cause

disease, including cancer and heart disease, and that nicotine addicts smokers.

Douglas, 110 So. 3d at 423. They presented evidence that the tobacco companies

“failed to address the health effects and addictive nature of cigarettes, manipulated

nicotine levels to make cigarettes more addictive, and concealed information about

the dangers of smoking.” Id. For example, Dr. Julius Richmond, a former Surgeon

General of the United States and professor at the Harvard Medical School, testified

that cigarettes contain carcinogens and that cigarettes cause pulmonary disease,

emphysema, lung cancer, heart disease, and bladder disease. Dr. Ronald Davis, a

5
 Case: 13-14590 Date Filed: 05/18/2017 Page: 6 of 284

former director of the Office on Smoking and Health and former medical director

for the Michigan Department of Public Health, testified similarly that cigarette

smoking is addictive and that those who smoke have a heightened risk of stroke,

emphysema, cancer, and heart disease. Dr. David Burns, a professor of medicine at

the University of California, San Diego, School of Medicine, with a specialty in

pulmonary and critical care medicine, testified that nicotine is addictive and that

cigarette smoking causes cancers, lung disease, and heart disease. He was an

associate scientific editor of a 1981 Surgeon General’s Report, and he explained

that “the purpose of the report was to make it very clear to the public that there is

no safe cigarette and there is no safe level of consumption.” He testified, “[W]ith

the exception of the tobacco industry, no other scientific group in the last 30 years

has reviewed this evidence and reached a conclusion other than that cigarette

smoking causes disease.” Dr. John Holbrook, professor of medicine at the

University of Utah School of Medicine, who is board certified in the field of

internal medicine, testified that, in his experience, the tobacco industry “attempted

to confound and obfuscate science” in its funding of medical research. Dr. W.

Jarrard Goodwin, a professor at the University of Miami School of Medicine, with

a specialty in otolaryngology, testified that smoking causes cancer of the mouth,

larynx, and pharynx. Dr. Edward Staples, director of the artificial heart program at

the University of Florida, testified that cigarette smoking causes emphysema, lung

6
 Case: 13-14590 Date Filed: 05/18/2017 Page: 7 of 284

cancer, coronary artery disease, and atherosclerosis. Dr. Neal Benowitz, a doctor at

San Francisco General Hospital and professor of medicine, psychiatry, and

biopharmaceutical sciences at the University of California in San Francisco,

testified that 90 percent of individuals begin smoking before the age of 20 and,

within two or three years, those young people will become addicted to nicotine. He

stated that tobacco companies could reduce the level of nicotine in cigarettes to

nonaddictive quantities. Some of the evidence of design defects applied only to

some brands of cigarettes. For example, the smokers presented evidence that

people who smoke light cigarettes tend to smoke more and inhale more deeply. But

the common thrust of the smokers’ evidence was that all of the companies’

cigarettes cause disease and addict smokers.

The tobacco companies put on evidence to defend themselves against the

several theories of liability. For example, the companies repeatedly challenged the

evidence that cigarette smoking causes disease. Dr. George Hensley, a former

professor at the University of Miami School of Medicine with a specialty in

pathology, testified that smoking does not cause pancreatic cancer. Dr. Hugh

Gilmore, a cardiology professor at the University of Miami School of Medicine,

testified that smoking is not a risk factor for the development of aortic aneurysms

or congestive heart failure. And Dr. Alden Cockburn, a urologist and a clinical

7
 Case: 13-14590 Date Filed: 05/18/2017 Page: 8 of 284

professor at the University of South Florida, testified that smoking is a risk factor

for bladder cancer but was not definitively proven to be a cause of bladder cancer.

In closing argument, the smokers’ attorney explained that “[t]he common

issue trial has addressed the conduct of the tobacco industry.” He recounted some

of the expert testimony. He argued, without focusing on any specific brand or

manufacturer of cigarettes, that scientists agree that nicotine is addicting, and he

argued that there is no scientific debate as to whether cigarette smoking causes

certain diseases, including cancer and heart disease. He said, “None of them

qualified their answer one iota. Does cigarette smoking cause these diseases? Yes,

yes, yes. Clear, crisp and definitive.” He also referred the jury to a collection of

documents that discussed how the companies manipulated nicotine levels. He

mentioned different methods of manipulating nicotine levels but not different

brands.

In closing argument, the tobacco companies’ attorneys responded to the

smokers’ many arguments. The companies contended that cigarettes are not proven

to be addictive. They maintained that smokers can quit and that nicotine is a “far

cry from heroin or cocaine.” And the companies argued that they have tried to

make cigarettes safer. They argued that they have not “spiked” cigarettes with

nicotine but have reduced the level of nicotine in some cigarettes.

8
 Case: 13-14590 Date Filed: 05/18/2017 Page: 9 of 284

The trial court instructed the jury in Phase I about the claim of strict liability

and negligence without regard to specific brands of cigarettes. For the claim of

strict liability, the trial court explained that “the issues are whether one or more of

the defendants designed, manufactured and marketed cigarettes which were

defective and unreasonably dangerous to smokers.” For the claim of negligence,

the trial court instructed the jury as follows:

On the claim of negligence, the issues are whether one or more of the
defendants were negligent in manufacturing, designing, marketing,
selling and distributing cigarettes which defendants knew or should
have known would cause serious and fatal diseases, including lung
cancer, or dependence-producing substances; in negligently not
testing tobacco and commercial cigarettes to confirm that smoking
causes human disease; in failing to design and produce a reasonably
safe cigarette with lower nicotine levels; in negligently measuring and
. . . understating nicotine and tar levels in low-tar cigarettes; and in
failing to warn smokers of the dangers of smoking and the
addictiveness or dependence-producing effects of cigarettes prior to
July 1 of 1969.

The verdict form included a series of yes-or-no questions. The tobacco

companies requested a more detailed verdict form, in which the jury would be

asked to identify “specific defects and tortious actions,” but the trial court rejected

that proposal. Id. The jury returned its verdict after eight days of deliberation. The

first question on the verdict form asked whether smoking cigarettes causes a list of

enumerated diseases and medical conditions. The jury answered “yes” for 20

specific diseases, including various forms of cancer. The second question asked

whether “cigarettes that contain nicotine [are] addictive or dependence producing.”
9
 Case: 13-14590 Date Filed: 05/18/2017 Page: 10 of 284

The jury answered “yes.” The verdict form then contained nine questions about the

conduct of each tobacco company. One of the nine questions asked the jury to

decide whether each tobacco company was strictly liable. It asked if the tobacco

company “place[d] cigarettes on the market that were defective and unreasonably

dangerous.” Another question asked if each tobacco company was negligent. It

asked if the tobacco company “failed to exercise the degree of care which a

reasonable cigarette manufacturer would exercise under like circumstances.” The

jury answered “yes” to each of these nine questions for each tobacco company.

The last question on the verdict form asked the jury whether the actions of the

tobacco companies entitled the class to punitive damages, and the jury answered

“yes” for each tobacco company.

The trial court denied the tobacco companies’ motion for directed verdict.

Id. Regarding strict liability, the court ruled that the evidence supported a finding

that all of the tobacco companies’ cigarettes were defective even if some of the

cigarettes had brand-specific dangers:

There was more than sufficient evidence at trial to satisfy the legal
requirements of this Count and to support the jury verdict that
cigarettes manufactured and placed on the market by the defendants
were defective in many ways including the fact that the cigarettes
contained many carcinogens, nitrosamines, and other deleterious
compounds such as carbon monoxide. That levels of nicotine were
manipulated, sometime by utilization of ammonia to achieve a desired
“free basing effect” of pure nicotine to the brain, and sometime by
using a higher nicotine content tobacco called Y–1, and by other
means such as manipulation of the levels of tar and nicotine. The
10
 Case: 13-14590 Date Filed: 05/18/2017 Page: 11 of 284

evidence more than sufficiently proved that nicotine is an addictive
substance which when combined with other deleterious properties,
made the cigarette unreasonably dangerous. The evidence also
showed some cigarettes were manufactured with the breathing air
holes in the filter being too close to the lips so that they were covered
by the smoker thereby increasing the amount of the deleterious effect
of smoking the cigarette. There was also evidence at trial that some
filters being test marketed utilize glass fibers that could produce
disease and deleterious effects if inhaled by a smoker.

Engle v. R.J. Reynolds Tobacco, 2000 WL 33534572, at *2 (Fla. Cir. Ct. 2000).

Regarding negligence, the court ruled that the evidence supported a finding that the

tobacco companies were negligent in producing and selling all of their cigarettes:

The verdict of the jury on the issue of Negligence is well supported by
the evidence. . . . The defendants according to the testimony, well
knew from their own research, that cigarettes were harmful to health
and were carcinogenic and addictive. By allowing the sale and
distribution of said product under those circumstances without taking
reasonable measures to prevent injury, constitutes, in this Court[’]s
opinion, and in the opinion of the jury as it turns out, negligence.

Id. at *4.

In Phase II, the same jury determined that the tobacco companies were liable

to the three class representatives and awarded them compensatory damages

totaling $12.7 million. Walker, 734 F.3d at 1282. The jury awarded punitive

damages of $145 billion to the class. Id. The tobacco companies filed an

interlocutory appeal of the judgments in Phases I and II. Id.

The Florida Supreme Court approved in part and vacated in part the jury

verdicts. Engle III, 945 So. 2d at 1254. The Florida Supreme Court concluded that

11
 Case: 13-14590 Date Filed: 05/18/2017 Page: 12 of 284

the trial court did not abuse its discretion in certifying the class for purposes of

Phase I and II. Id. at 1267. But the court decertified the class for Phase III “because

individualized issues such as legal causation, comparative fault, and damages

predominate.” Id. at 1268. The Florida Supreme Court “retain[ed]” the findings of

liability by the jury from Phase I “other than those on the fraud and intentional

infliction of emotion[al] distress claims, which involved highly individualized

determinations, and the finding on entitlement to punitive damages questions,

which was premature.” Id. at 1269. The court explained, “Class members can

choose to initiate individual damages actions,” and those retained findings, which

include the findings that the companies acted negligently and that they sold

defective products, “will have res judicata effect in those trials.” Id. The court

affirmed the damages award in favor of two of the class representatives and

vacated the judgment in favor of the third class representative because the statute

of limitations barred his claims. Id. at 1276. The court vacated the award of

punitive damages. Id. at 1262–65.

After members of the Engle class filed thousands of individual actions in

state and federal courts, these courts had to determine the extent to which the

smokers could rely on the approved findings from Phase I to establish certain

elements of their claims. Walker, 734 F.3d at 1283. In Brown v. R.J. Reynolds

Tobacco Company, 611 F.3d 1324 (11th Cir. 2010), we stated that, under Florida

12
 Case: 13-14590 Date Filed: 05/18/2017 Page: 13 of 284

law, courts should give preclusive effect to the findings only to the extent that the

smoker can “show with a ‘reasonable degree of certainty’ that the specific factual

issue was determined in [his] favor.” Id. at 1335 (quoting Seaboard Coast Line R.

Co. v. Indus. Contracting Co., 260 So. 2d 860, 865 (Fla. Dist. Ct. App. 1972)). We

remanded to the district court to make that determination after considering the

“entire trial record.” Id. But several of the Florida district courts of appeal

disagreed with our decision that a member of the Engle class had to establish from

the trial record that an issue was actually decided. These district courts of appeal

all held that the Phase I findings established the duty and breach elements of the

smokers’ claims, though they disagreed about how the smokers would prove

causation in individual cases. See Philip Morris USA, Inc. v. Douglas, 83 So. 3d

1002, 1010 (Fla. Dist. Ct. App. 2012); R.J. Reynolds Tobacco Co. v. Brown, 70 So.

3d 707, 715–16 (Fla. Dist. Ct. App. 2011); R.J. Reynolds Tobacco Co. v. Martin,

53 So. 3d 1060, 1066–70 (Fla. Dist. Ct. App. 2010).

In Douglas, the Florida Supreme Court ruled that the approved findings

from Phase I established common elements of the claims of Engle class members.

110 So. 3d at 428–30. The court explained that, although the evidence submitted

during Phase I included both general and brand-specific defects, “the class action

jury was not asked to find brand-specific defects in the Engle defendants’

cigarettes.” Id. at 423. The jury was asked to determine “all common liability

13
 Case: 13-14590 Date Filed: 05/18/2017 Page: 14 of 284

issues,” and it heard evidence that the tobacco companies’ cigarettes were

“defective because they are addictive and cause disease.” Id. The court explained

that the approved findings concerned conduct that “is common to all class

members and will not change from case to case” and that “the approved Phase I

findings are specific enough” to establish some elements of the smokers’ claims.

Id. at 428. That is, the jury findings “conclusively establish” that the tobacco

companies manufactured defective products and that the companies failed to

exercise the degree of care of a reasonable person. Id. at 430. And the jury findings

establish general causation. Id. at 428. Going forward, “to prevail on either strict

liability or negligence Engle claims, individual plaintiffs must establish

(i) membership in the Engle class; (ii) individual causation, i.e., that addiction to

smoking the Engle defendants’ cigarettes containing nicotine was a legal cause of

the injuries alleged; and (iii) damages.” Id. at 430.

The Florida Supreme Court then held that giving preclusive effect to the

approved findings from Phase I did not violate the right to due process of the

tobacco companies. Id. The companies had argued that “accepting the Phase I

findings as res judicata violates their due process rights because it is not clear from

the Phase I verdict which theories of liability the Engle jury actually decided to

reach those findings.” Id. The Douglas court concluded that the tobacco companies

had notice and an opportunity to be heard and that the Engle proceedings did not

14
 Case: 13-14590 Date Filed: 05/18/2017 Page: 15 of 284

arbitrarily deprive them of property. Id. at 431. It explained that “the Phase I

verdict against the Engle defendants resolved all elements of the claims that had

anything to do with the Engle defendants’ cigarettes or their conduct.” Id. at 432.

The Douglas court stated, “[T]he defendants’ due process argument is an

attack on our decision in Engle to give the Phase I findings res judicata—as

opposed to issue preclusion—effect in class members’ individual damages

actions.” Id. The Douglas court explained that, when it gave “res judicata effect” to

the Phase I approved findings, Engle III, 945 So. 2d at 1269, it meant claim

preclusion, not issue preclusion. Douglas, 110 So. 3d at 432. The Douglas court

stated that claim preclusion prevents the same parties from relitigating the same

cause of action. Id. Issue preclusion prevents the parties from relitigating “the same

issues that were litigated and actually decided in a second suit involving a different

cause of action.” Id. at 433. The Douglas court ruled that the individual Engle

actions involved the same causes of action. Id. The Douglas court stated, “[T]o

decide here that we really meant issue preclusion even though we said res judicata

in Engle would effectively make the Phase I findings regarding the Engle

defendants’ conduct useless in individual actions.” Id. And the Douglas court

concluded that the tobacco companies “do not have the right to have issue

preclusion, as opposed to res judicata, apply to the Phase I findings.” Id. at 435.

15
 Case: 13-14590 Date Filed: 05/18/2017 Page: 16 of 284

In Walker, we held that giving res judicata effect to the findings of the jury

in Engle did not violate the rights of the tobacco companies to due process.

Walker, 734 F.3d at 1280–81. R.J. Reynolds had appealed the jury verdicts in favor

of two smokers after the district courts instructed the juries that R.J. Reynolds sold

defective cigarettes and was negligent. Id. at 1286. We explained that we were

obligated to give “full faith and credit to the decision in Engle, as interpreted in

Douglas,” unless it “would arbitrarily deprive R.J. Reynolds of its property without

due process of law.” Id. at 1287. We stated that no court “has ever held that due

process requires application of the federal common law of issue preclusion,” and

we did not decide whether it does. Id. at 1289. We concluded that, even if due

process requires that an issue be actually decided, the Florida Supreme Court ruled

in Douglas that the approved findings from Phase I concerned conduct that is

common to all class members and established negligence and defect elements of

the class members’ claims. Id. We concluded that the “actually decided”

requirement was satisfied and that it is “no concern of ours” what the Florida

Supreme Court calls the “relevant doctrine.” Id.

In this appeal, R.J. Reynolds and Philip Morris challenge a jury verdict in

favor of Earl Graham, as personal representative of the estate of his deceased wife,

Faye Graham, a member of the Engle class. Mr. Graham filed an individual Engle

action in the district court against R.J. Reynolds, Philip Morris, and other

16
 Case: 13-14590 Date Filed: 05/18/2017 Page: 17 of 284

defendants later dismissed. He alleged that his wife developed lung cancer and

died because of her addiction to cigarettes manufactured by R.J. Reynolds and

Philip Morris. He asserted claims of strict liability, breach of warranty, negligence,

fraudulent concealment, and conspiracy to fraudulently conceal.

Under the Engle framework articulated in Douglas, the jury was not asked to

find that the cigarettes Faye Graham smoked were defective or that the tobacco

companies were negligent. Graham v. R.J. Reynolds Tobacco Co., 782 F.3d 1261,

1273 (11th Cir. 2015), reh’g en banc granted, op. vacated, 811 F.3d 434 (11th Cir.

2016). The district court treated those findings as having already been established.

Id. For the claims of negligence and strict liability, the jury was asked to determine

only whether Faye Graham was a member of the Engle class and whether smoking

cigarettes manufactured by R.J. Reynolds or Philip Morris “was a legal cause” of

Faye Graham’s injuries. Id. The district court instructed the jury that, to find legal

causation, Graham’s addiction to cigarettes must have “directly and in natural and

continuous sequence produced or contributed substantially to producing” her

injuries.

The jury found for Graham on the claims of strict liability and negligence.

Id. The jury awarded Graham $2.75 million in damages and determined that Faye

Graham was 70 percent at fault, R.J. Reynolds was 20 percent at fault, and Philip

Morris was 10 percent at fault. Id. at 1273–74. The district court entered judgment

17
 Case: 13-14590 Date Filed: 05/18/2017 Page: 18 of 284

against R.J. Reynolds for $550,000 and against Philip Morris for $275,000. Id. at

1274. The district court denied the tobacco companies’ motion for judgment as a

matter of law. Id. Theresa Graham later replaced Earl Graham as personal

representative of the estate.

A panel of this Circuit reversed the judgment of the district court. Id. at

1285. The panel held that the Engle findings of strict liability and negligence are

preempted by federal law. Id. We later granted the petition for rehearing en banc

filed by Graham and vacated the panel opinion. Graham, 811 F.3d at 434–35. In

addition to briefing the preemption issue, we allowed the parties to brief whether

giving effect to the jury’s findings in Engle would “violate the tobacco companies’

rights under the Due Process Clause of the Fourteenth Amendment to the United

States Constitution notwithstanding the panel’s holding in Walker.” The Florida

Supreme Court has since ruled that federal law does not preempt “state tort”

actions against the tobacco companies and that, even if federal law preempted a

ban on the sale of cigarettes, the Engle Phase I findings do “not amount to . . . a

ban” that might conflict with federal law. R.J. Reynolds Tobacco Co. v. Marotta,

No. SC16-218, 2017 WL 1282111, at *9 (Fla. Apr. 6, 2017).

II. STANDARD OF REVIEW

We review de novo the denial of a motion for judgment as a matter of law.

Mendoza v. Borden, Inc., 195 F.3d 1238, 1244 (11th Cir. 1999) (en banc). We also

18
 Case: 13-14590 Date Filed: 05/18/2017 Page: 19 of 284

review de novo questions of constitutional law, Nichols v. Hopper, 173 F.3d 820,

822 (11th Cir. 1999), and whether federal law preempts a state law claim, Atwater

v. Nat’l Football League Players Ass’n, 626 F.3d 1170, 1179 (11th Cir. 2010).

III. DISCUSSION

We divide our discussion in two parts. First, we explain why giving full faith

and credit to the Engle jury findings of negligence and strict liability does not

deprive R.J. Reynolds and Philip Morris of property without due process of law.

Second, we conclude that the Engle jury findings of negligence and strict liability

are not preempted by federal law.

A. Giving Preclusive Effect to the Negligence and Strict Liability Findings
Does Not Violate Due Process.

The Full Faith and Credit Act, 28 U.S.C. § 1738, requires federal courts to

“give preclusive effect to a state court judgment to the same extent as would courts

of the state in which the judgment was entered,” Kahn v. Smith Barney Shearson

Inc., 115 F.3d 930, 933 (11th Cir. 1997) (quoting Battle v. Liberty Nat’l Life Ins.

Co., 877 F.2d 877, 882 (11th Cir. 1989)), subject to the requirements of the Due

Process Clause, see Kremer v. Chem. Const. Corp., 456 U.S. 461, 481 (1982). R.J.

Reynolds and Philip Morris argue that the Due Process Clause mandates that an

issue be actually decided in one case before it is given preclusive effect in another.

They argue that relying on the approved jury findings in individual actions by

Engle members is an application of issue preclusion and that the Florida courts did
19
 Case: 13-14590 Date Filed: 05/18/2017 Page: 20 of 284

not actually decide issues of strict liability and negligence for all class members.

They argue that by abandoning the “actually decided” requirement, the Florida

courts abrogated a fundamental protection against arbitrary deprivations of

property in violation of the Due Process Clause. See Honda Motor Co. v. Oberg,

512 U.S. 415, 430 (1994).

We need not determine whether the Due Process Clause requires that an

issue be actually decided in an earlier case before the judgment from that case is

given preclusive effect on that issue. We will assume, without deciding, that the

“actually decided” requirement is a fundamental requirement of due process under

Fayerweather v. Ritch, 195 U.S. 276 (1904). Even with that assumption, no

violation of due process occurred when the district court gave the Engle findings

preclusive effect. Based on our review of the Engle proceedings, we are satisfied

that the Engle jury actually decided common elements of the negligence and strict

liability of R.J. Reynolds and Philip Morris.

The Florida Supreme Court made clear in Douglas that the Engle jury

decided common elements of the negligence and strict liability of the tobacco

companies for all class members. And for that reason, the Florida Supreme Court

explained that the findings were binding in individual Engle actions. It stated,

“Because these findings go to the defendants’ underlying conduct, which is

common to all class members and will not change from case to case, we held that

20
 Case: 13-14590 Date Filed: 05/18/2017 Page: 21 of 284

these approved ‘Phase I common core findings . . . will have res judicata effect’ in

class members’ ‘individual damages actions.’” Douglas, 110 So. 3d at 428

(alteration in original) (quoting Engle III, 945 So. 2d at 1269).

The Florida Supreme Court rejected the same argument that R.J. Reynolds

and Philip Morris make here about what the Engle jury decided. R.J. Reynolds and

Philip Morris asserted that some of the evidence presented at the Engle trial

applied to specific brands of cigarettes. They argued that, although the Engle jury

found that the tobacco companies “place[d] cigarettes on the market that were

defective and unreasonably dangerous,” the jury did not necessarily find that all

cigarettes the defendants placed on the market were defective and unreasonably

dangerous. The Florida Supreme Court rejected this argument and stated that “this

Court in Engle necessarily decided that the approved Phase I findings” are

“specific enough to establish a causal link between their conduct and damages to

individual plaintiffs who prove injuries caused by addiction to smoking the Engle

defendants’ cigarettes.” Id. That is, the Phase I findings establish the causal link

between the tobacco companies’ conduct and the class members’ injuries because

the companies acted wrongfully toward all of the class members. Whether that

conduct was the legal cause of the individual class members’ injuries, and whether

the individual class members were entitled to damages, was left for later individual

trials.

21
 Case: 13-14590 Date Filed: 05/18/2017 Page: 22 of 284

After reviewing the Engle trial record, we are satisfied that the Florida

Supreme Court determined that the Engle jury found the common elements of

negligence and strict liability against Philip Morris and R.J. Reynolds. Both

companies admit that the smokers presented common “proof that the Engle

defendants’ cigarettes were defective because they are addictive and cause disease”

in addition to brand-specific evidence. Id. at 423. In two days of closing

arguments, the smokers’ attorneys recounted the ample body of evidence that

smoking cigarettes causes disease without focusing on the differences in the

designs of various brands. The trial court instructed the jury to “determine ‘all

common liability issues’ for the class concerning ‘the conduct of the tobacco

industry.’” Id. Moreover, the jury’s answers on the verdict form, when read

together with the entire record, were consistent with the general theories that the

tobacco companies’ cigarettes are defective and the sale of their cigarettes is

negligent because all of those cigarettes cause disease and are addictive.

The first two questions on the verdict form are most naturally read to apply

to all cigarettes manufactured by the tobacco companies. Question 1 asked whether

“smoking cigarettes cause one or more of the following diseases or medical

conditions.” The jury answered “yes” to 20 of 23 diseases. This question does not

admit of any limitation, nor did the accompanying jury instruction, and its natural

interpretation is that it was asking about all cigarettes manufactured by the tobacco

22
 Case: 13-14590 Date Filed: 05/18/2017 Page: 23 of 284

companies, not just some. Similarly, question 2 asked whether “cigarettes that

contain nicotine [are] addictive or dependence producing,” and the jury answered

“yes.” The evidence at trial was that nicotine, and not some other ingredient, made

cigarettes addictive. In closing arguments, the tobacco companies’ counsel told the

jury that the question should be understood to inquire whether “all cigarettes that

contain nicotine [are] addictive or dependence-producing,” not whether there is

“one cigarette or a brand of cigarettes or two brands of cigarettes” that are

addictive.

The strict liability and negligence questions presented to the jury used the

same unmodified noun—“cigarettes”—that was used to refer to all cigarettes

manufactured by the tobacco companies in questions 1 and 2. The strict liability

interrogatory asked whether “one or more of the defendant tobacco companies

place[d] cigarettes on the market that were defective and unreasonably dangerous,”

and the negligence interrogatory inquired whether the smokers had “proven that

one or more of the defendant tobacco companies failed to exercise the degree of

care which a reasonable cigarette manufacturer would exercise under like

circumstances.” The jury answered “yes” to both questions for R.J. Reynolds and

Philip Morris. When asked about strict liability, the jury found that R.J. Reynolds

and Philip Morris had sold defective cigarettes “both before and after July 1,

1974,” and, with respect to the negligence claim, that they had acted negligently by

23
 Case: 13-14590 Date Filed: 05/18/2017 Page: 24 of 284

selling, manufacturing, and distributing cigarettes “both before and after July 1,

1969.” That the jury found that these tobacco companies’ tortious conduct swept

across both time periods is consistent with a general theory of liability that applied

to all their cigarettes.

After the jury returned a verdict in favor of the class on all counts, the trial

court ruled that there was sufficient evidence to support those verdicts, including

negligence and strict liability, and cited evidence that applied to all of the

cigarettes made by the tobacco companies. For example, it stated, “The evidence

more than sufficiently proved that nicotine is an addictive substance which when

combined with other deleterious properties, made the cigarette unreasonably

dangerous.” Engle, 2000 WL 33534572, at *2. The only way to make sense of

these proceedings is that the Florida courts determined that the Engle jury actually

decided issues common to the class, and the district court did not abrogate a

protection against arbitrary deprivations of property in affording the Phase I jury’s

findings preclusive effect in Graham’s case.

R.J. Reynolds and Philip Morris argue that if the Florida Supreme Court had

determined that the Engle jury actually decided common elements of negligence

and strict liability for all class members, it would not have used the term “claim

preclusion” in Douglas to refer to the preclusive effect of the jury findings and

thereby evade the “actually decided” requirement, but we disagree. The Florida

24
 Case: 13-14590 Date Filed: 05/18/2017 Page: 25 of 284

Supreme Court explained that issue preclusion applies in actions involving

different causes of action and claim preclusion applies in actions involving the

same causes of action. Douglas, 110 So. 3d at 432–33. And in explaining the

differences between claim preclusion and issue preclusion, the Florida Supreme

Court reiterated that the Engle jury made findings about the tobacco companies’

conduct that applied to all class members. It said, “No matter the wording of the

findings on the Phase I verdict form, the jury considered and determined specific

matters related to the [Engle] defendants’ conduct. Because the findings are

common to all class members, [individual plaintiffs are] entitled to rely on them

. . . .” Id. at 433 (alterations in original) (quoting Martin, 53 So. 3d at 1067).

The terminology employed by the Florida Supreme Court was unorthodox,

but “[i]n determining what is due process of law, regard must be had to substance,

not to form.” Fayerweather, 195 U.S. at 297. The Supreme Court of the United

States has acknowledged that “[t]he preclusive effects of former adjudication are

discussed in varying and, at times, seemingly conflicting terminology, attributable

to the evolution of preclusion concepts over the years.” Migra v. Warren City

School Dist. Bd. of Educ., 465 U.S. 75, 77 n.1 (1984). As long as the state

proceedings “satisf[ied] the minimum procedural requirements” of due process,

Kremer, 456 U.S. at 481, what the Florida Supreme Court “calls the relevant

doctrine . . . is no concern of ours,” Walker, 734 F.3d at 1289.

25
 Case: 13-14590 Date Filed: 05/18/2017 Page: 26 of 284

Apart from their argument that the jury did not actually decide common

issues of negligence and strict liability, R.J. Reynolds and Philip Morris do not

deny that they were afforded due process. That is, they do not contend that they

were denied notice or an opportunity to be heard, the central features of due

process. See Fuentes v. Shevin, 407 U.S. 67, 80 (1972). The Florida courts

provided them notice that the jury findings would establish the “conduct elements

of the class’s claims.” Douglas, 110 So. 3d at 429. And the year-long trial provided

them “a full and fair opportunity to litigate the issues of common liability in Phase

I.” Walker, 734 F.3d at 1288. Both tobacco companies seized that opportunity,

presenting “testimony that cigarettes were not addictive and were not proven to

cause disease and that they had designed the safest cigarette possible.” Douglas,

110 So. 3d at 423. And they continue to contest liability in individual actions by

class members, in which new juries determine issues of individual causation,

apportionment of fault, and damages. Id. at 430; Engle III, 945 So. 2d at 1254.

The Due Process Clause does not require a state to follow the federal

common law of res judicata and collateral estoppel. “State courts are generally free

to develop their own rules for protecting against the relitigation of common issues

or the piecemeal resolution of disputes.” Richards v. Jefferson Cty., Ala., 517 U.S.

793, 797 (1996). For example, a state might allow offensive, non-mutual collateral

estoppel. E.g., In re Owens, 532 N.E.2d 248, 252 (Ill. 1988). And courts, both state

26
 Case: 13-14590 Date Filed: 05/18/2017 Page: 27 of 284

and federal, frequently manage class actions by splitting them into separate phases.

See generally William B. Rubenstein, Newberg on Class Actions §§ 10.6, 11.3 (5th

ed.). Engle is not the first time that “a defendant’s common liability [was]

established through a class action and given binding effect in subsequent individual

damages actions.” Douglas, 110 So. 3d at 429 (collecting cases); see also Brown v.

Electrolux Home Prods., Inc., 817 F.3d 1225, 1239 (11th Cir. 2016) (discussing

several “tools to decide individual damages” in a class action, including “(1)

bifurcating liability and damage trials with the same or different juries; (2)

appointing a magistrate judge or special master to preside over individual damages

proceedings; [and] (3) decertifying the class after the liability trial and providing

notice to class members concerning how they may proceed to prove damages”

(quoting In re Visa Check/MasterMoney Antitrust Litig., 280 F.3d 124, 141 (2d

Cir. 2001))). The Due Process Clause requires only that the application of

principles of res judicata by a state affords the parties notice and an opportunity to

be heard so as to avoid an arbitrary deprivation of property. Fuentes, 407 U.S. at

80.

We recognize that the Engle Court defined a novel notion of res judicata, but

we cannot say that the substance of that doctrine or its application in these trials

was so unfair as to violate the constitutional guarantee of due process. “The very

nature of due process negates any concept of inflexible procedures universally

27
 Case: 13-14590 Date Filed: 05/18/2017 Page: 28 of 284

applicable to every imaginable situation,” Cafeteria & Rest. Workers Union, Local

473, AFL-CIO v. McElroy, 367 U.S. 886, 895 (1961), and our review of the record

establishes that the tobacco companies had notice that the Engle trial involved

common evidence and theories of negligence and strict liability that applied to all

cigarettes manufactured by all tobacco companies and sold to all members of the

class during the relevant periods. The tobacco companies were given an

opportunity to be heard on the common theories in a year-long trial followed by an

appeal to the Florida Supreme Court and later individual trials and appeals on the

remaining issues of proximate causation, comparative fault, and damages. See

Engle III, 945 So. 2d at 1254–56.

Contrary to the dissent’s view, see Dissenting Op. of Tjoflat, J., at 90–91, no

tobacco company can be held liable to any smoker without proof at trial that the

smoker belongs to the Engle class, that she smoked cigarettes manufactured by the

company during the relevant class period, and that smoking was the proximate

cause of her injury. Every tobacco company must also be afforded the opportunity

to contest the smokers’ pleadings and evidence and to plead and prove the

smokers’ comparative fault. Indeed, in this appeal, after the district court instructed

it, the jury reduced Graham’s damages award for his deceased spouse’s

comparative fault. And in other Engle progeny litigation, tobacco companies have

won defense verdicts. E.g., Suarez v. R.J. Reynolds Tobacco Co., No. 09-79584-

28
 Case: 13-14590 Date Filed: 05/18/2017 Page: 29 of 284

CA-01 (11th Fla. Cir. Ct., Nov. 25, 2015) (final judgment). “[S]tate proceedings

need do no more than satisfy the minimum procedural requirements” of due

process to receive full faith and credit. Kremer, 456 U.S. at 481. The record in this

appeal establishes that R.J. Reynolds and Philip Morris were afforded the

protections mandated by the Due Process Clause.

“Under the Full Faith and Credit Act, federal courts generally should respect

state court judgments, even where erroneous.” Lops v. Lops, 140 F.3d 927, 938

(11th Cir. 1998); see also Hickerson v. City of New York, 146 F.3d 99, 107 (2d Cir.

1998) (“[T]o second-guess that court’s determination of this issue would violate

the full faith and credit statute.”). We decide only whether applying Florida law in

this case violates due process. We do not endorse or condemn the use of a class

action in Phase I of the Engle litigation. Nor do we endorse or condemn the

explication of res judicata by the Supreme Court of Florida. We say only that

applying Florida law in this trial did not violate the tobacco companies’ rights to

due process of law.

R.J. Reynolds and Philip Morris argue that we are not compelled to give full

faith and credit to Douglas because Graham was not a party in Douglas and

Florida law does not allow non-mutual issue preclusion. Because state courts

would not be bound by the Douglas decision in this circumstance, they argue, we

are also not bound. But this argument is a straw man.

29
 Case: 13-14590 Date Filed: 05/18/2017 Page: 30 of 284

We do not give full faith and credit to the decision in Douglas; we instead

give full faith and credit to the jury findings in Engle. The Florida Supreme Court

in Engle interpreted those findings to determine what the jury actually decided, and

the Florida Supreme Court in Douglas decided a matter of state law when it

explained the preclusive effect of the Engle jury’s Phase I findings. We are bound

by the decisions of state supreme courts on matters of state law when we exercise

diversity jurisdiction, subject to the constraints of due process. See Walker, 734

F.3d at 1284. We conclude that giving preclusive effect to the findings of

negligence and strict liability by the Engle jury in individual actions by Engle class

members against R.J. Reynolds and Philip Morris does not deprive these tobacco

companies of property without due process of law.

B. Federal Law Does Not Preempt the Jury Findings of Negligence and Strict
Liability.

“The Supremacy Clause of Art. VI of the Constitution provides Congress

with the power to pre-empt state law.” La. Pub. Serv. Comm’n v. FCC, 476 U.S.

355, 368 (1986). “State action may be foreclosed by express language in a

congressional enactment, by implication from the depth and breadth of a

congressional scheme that occupies the legislative field, or by implication because

of a conflict with a congressional enactment.” Lorillard Tobacco Co. v. Reilly, 533

U.S. 525, 541 (2001) (citations omitted). Conflicts arise in two ways: “when

compliance with both federal and state regulations is impossible or when the state
30
 Case: 13-14590 Date Filed: 05/18/2017 Page: 31 of 284

law ‘stands as an obstacle to the accomplishment and execution of the full

purposes and objectives of Congress,’” Hillman v. Maretta, 133 S. Ct. 1943, 1950

(2013) (citation omitted) (quoting Hines v. Davidowitz, 312 U.S. 52, 67 (1941)).

“‘[T]he purpose of Congress is the ultimate touchstone’ in every pre-emption

case.” Medtronic, Inc. v. Lohr, 518 U.S. 470, 485 (1996) (quoting Cipollone v.

Liggett Grp., Inc., 505 U.S. 504, 516 (1992)). “Congress’ intent, of course,

primarily is discerned from the language of the pre-emption statute and the

‘statutory framework’ surrounding it.” Id. at 486 (quoting Gade v. Nat’l Solid

Wastes Mgmt. Ass’n, 505 U.S. 88, 111 (1992) (Kennedy, J., concurring in part and

concurring in the judgment)).

This appeal presents an issue of conflict preemption. A party asserting

conflict preemption faces a high bar:

[I]n all pre-emption cases, and particularly in those in which Congress
has “legislated . . . in a field which the States have traditionally
occupied,” . . . we “start with the assumption that the historic police
powers of the States were not to be superseded by the Federal Act
unless that was the clear and manifest purpose of Congress.”

Wyeth v. Levine, 555 U.S. 555, 565 (2009) (second and third alteration in original)

(quoting Lohr, 518 U.S. at 485).

R.J. Reynolds and Philip Morris argue that the obstacle form of conflict

preemption defeats the findings of negligence and strict liability in Engle. They

argue that this Circuit avoided finding a violation of due process in Walker by

31
 Case: 13-14590 Date Filed: 05/18/2017 Page: 32 of 284

construing the Engle findings as embracing a theory that all cigarettes

manufactured by the tobacco companies are defective and the sale of all of those

cigarettes is negligent because all of those cigarettes are dangerous—that is, that all

of those cigarettes are addictive and cause disease. Federal law, they contend,

preempts state law claims premised on the theory that all of the cigarettes

manufactured by the tobacco companies are inherently dangerous.

We disagree. We conclude that federal tobacco laws do not preempt state

tort claims based on the dangerousness of all the cigarettes manufactured by the

tobacco companies. In other words, federal law does not preempt the Engle jury

findings.

Congress has enacted six tobacco-specific laws that are relevant to this

appeal. In 1965, Congress passed the Federal Cigarette Labeling and Advertising

Act, Pub. L. No. 89-92, 79 Stat. 282, which made it unlawful to sell cigarettes

without the following warning label: “Caution: Cigarette Smoking May Be

Hazardous to Your Health.” Id. § 4, 79 Stat. at 283. And the Act prohibited

requiring any additional “statement relating to smoking and health” on cigarette

packages or in cigarette advertising. Id. § 5, 79 Stat. at 283. Congress then passed

the Public Health Cigarette Smoking Act of 1969, Pub. L. No. 91-222, 84 Stat. 87,

which changed the language of the warning label to “Warning: The Surgeon

General Has Determined That Smoking Is Dangerous to Your Health.” Id. § 2, 84

32
 Case: 13-14590 Date Filed: 05/18/2017 Page: 33 of 284

Stat. at 88. The Act made it “unlawful to advertise cigarettes on any medium of

electronic communication subject to the jurisdiction of the Federal

Communications Commission.” Id., 84 Stat. at 89. And it amended the preemption

provision in the 1965 Act by adding the following statement: “No requirement or

prohibition based on smoking and health shall be imposed under State law with

respect to the advertising or promotion of any cigarettes the packages of which are

labeled in conformity with the provisions of this Act.” Id., 84 Stat. at 88 (codified

as amended at 15 U.S.C. § 1334(b)). Congress again amended the Labeling Act by

passing the Comprehensive Smoking Education Act, Pub. L. No. 98-474, 98 Stat.

2200 (1984). The Act replaced the warning with a series of warnings that must

appear on cigarette packages and advertisements on a rotating basis. Id. § 4, 98

Stat. at 2201–03. The Act also required the Secretary of Health and Human

Services to “establish and carry out a program to inform the public of any dangers

to human health presented by cigarette smoking.” Id. § 3, 98 Stat. at 2200. The

Alcohol and Drug Abuse Amendments of 1983, Pub. L. No. 98-24, 97 Stat. 175,

required the Secretary of Health and Human Services to issue a report to Congress

every three years on, among other things, “the addictive property of tobacco.” Id.

§ 2, 97 Stat. at 178. The Comprehensive Smokeless Tobacco Health Education Act

of 1986, Pub. L. No. 99-252, 100 Stat. 30, regulates smokeless tobacco products.

The Act requires that a warning appear on smokeless tobacco products, id. § 3, 100

33
 Case: 13-14590 Date Filed: 05/18/2017 Page: 34 of 284

Stat. at 30–32, prohibits the advertising of smokeless tobacco products “on any

medium of electronic communications subject to the jurisdiction of the Federal

Communications Commission,” id. § 3(f), 100 Stat. at 32, and requires the

Secretary of Health and Human Services to create a program to inform the public

about the health effects of using smokeless tobacco products, id. § 2, 100 Stat. at

30. Last, the ADAMHA Reorganization Act, Pub. L. No. 102-321, 106 Stat. 323

(1992), conditions certain block grants on states making it unlawful “for any

manufacturer, retailer, or distributor of tobacco products to sell or distribute any

such product to any individual under the age of 18.” Id. § 202, 106 Stat. at 394

(codified at 42 U.S.C. § 300x-26(a)(1)). We do not consider the Family Smoking

Prevention and Tobacco Control Act, Pub. L. No. 111-31, 123 Stat. 1776 (2009),

because the Act does not affect actions, like the individual Engle actions, that were

pending in federal or state court during its passage, id. § 4, 123 Stat. at 1782.

Affording preclusive effect to the Engle jury findings does not frustrate the

objectives of these federal laws on tobacco. The only significant requirement

imposed on cigarette manufacturers by the six federal laws in question is the

warning label requirement for cigarette packages and advertising. Three of the six

statutes—the Federal Cigarette Labeling and Advertising Act, the Public Health

Cigarette Smoking Act of 1969, and the Comprehensive Smoking Education Act—

concern this warning label. Fittingly, the Labeling Act expressly preempts state

34
 Case: 13-14590 Date Filed: 05/18/2017 Page: 35 of 284

laws that would impose labeling requirements. See 15 U.S.C. § 1334; Altria Grp.,

Inc. v. Good, 555 U.S. 70, 79 (2008). The other three statutes impose no significant

requirements on cigarette manufacturers: the Comprehensive Smokeless Tobacco

Health Education Act of 1986 concerns smokeless products, not cigarettes; the

Alcohol and Drug Abuse Amendments imposed a requirement on the Secretary of

Health and Human Services to submit reports about cigarettes; and the ADAMHA

Reorganization Act conditions block grants to states.

Contrary to R.J. Reynolds and Philip Morris’s argument, the statement of

purpose in the Labeling Act, 15 U.S.C. § 1331, does not preserve cigarette sales.

The second listed purpose of establishing a program to “deal with cigarette

labeling and advertising” states, “[C]ommerce and the national economy may be

(A) protected to the maximum extent consistent with this declared policy and (B)

not impeded by diverse, nonuniform, and confusing labeling and advertising

regulations.” Id. Congress sought to protect “commerce and the national economy”

specifically from the effect of “diverse, nonuniform and confusing cigarette

labeling and advertising” rules, id., not from more stringent regulation generally.

See Altria Grp., 555 U.S. at 78–79 (explaining that the “Act’s pre-emption

provisions promote its second purpose” by preventing States from “enforcing rules

that are based on an assumption that the federal warnings are inadequate”); Reilly,

533 U.S. at 542–43 (paraphrasing the second purpose as “to protect the national

35
 Case: 13-14590 Date Filed: 05/18/2017 Page: 36 of 284

economy from interference due to diverse, nonuniform, and confusing cigarette

labeling and advertising regulations”); Marotta, 2017 WL 1282111, at *7 (“Thus,

Congress clearly intended to ‘protect the national economy from the burden

imposed by diverse, nonuniform, and confusing cigarette labeling and advertising

regulations,’ but did not clearly intend to extend broad immunity from common

law liability to cigarette manufacturers.” (citation omitted)).

Nothing in these six statutes reflects a federal objective to permit the sale or

manufacture of cigarettes. As a result, we cannot say that Congress created a

regulatory scheme that does not tolerate tort liability based on the dangerousness of

all cigarettes manufactured by the tobacco companies but tolerates tort actions

based on theories with a more limited scope. Cf. Altria Grp., 555 U.S. at 551

(holding that federal law did not preempt common-law fraud claim against

cigarette manufacturer based on advertising of light cigarettes); Boerner v. Brown

& Williamson Tobacco Co., 394 F.3d 594, 600 (8th Cir. 2005) (holding that the

Labeling Act did not preempt design defect claim against cigarette manufacturer);

Spain v. Brown & Williamson Tobacco Corp., 363 F.3d 1183, 1197 (11th Cir.

2004) (holding that the Labeling Act did not preempt negligent and wanton design

and manufacture claims against cigarette manufacturer). Federal law is silent both

by its terms and by its operation.

36
 Case: 13-14590 Date Filed: 05/18/2017 Page: 37 of 284

Determinations of strict liability and negligence based on the Engle findings

create no conflict with a federal objective. R.J. Reynolds and Philip Morris do not

contend that the Engle jury based its findings of liability on a determination that

the warnings on cigarette packages and advertisements were inadequate such that

the jury’s findings imposed labeling requirements preempted by federal law. Rules

governing the design of cigarettes or even banning the sale of cigarettes do not

frustrate accomplishing a rule that requires a certain label when and if cigarettes

are sold. See Hunter v. Philip Morris USA, 582 F.3d 1039, 1048 (9th Cir. 2009)

(explaining that product-liability claim against cigarette manufacturer “does not

present an obstacle to the congressional policy concerning the regulation of

tobacco” because the federal laws “concern labeling, research and education and

do not provide strong evidence of a federal policy against more stringent state

regulation”); Marotta, 2017 WL 1282111, at *9 (“Strict liability and negligence

claims, such as those brought . . . under Engle, do not interfere with the regulation

of advertising and promotion of cigarettes and, therefore, do not clearly conflict

with congressional objectives.”).

That the express-preemption provision in the Labeling Act does not cover

the negligence and strict liability findings in Engle supports an inference that there

is no implied preemption of those findings. See Wyeth, 555 U.S. at 574–75; Riegel

v. Medtronic, Inc., 552 U.S. 312, 327 (2008). Granted, “[i]f a federal law contains

37
 Case: 13-14590 Date Filed: 05/18/2017 Page: 38 of 284

an express pre-emption clause, it does not immediately end the inquiry because the

question of the substance and scope of Congress’ displacement of state law still

remains.” Altria Grp., 555 U.S. at 76; see also Geier v. Am. Honda Motor Co., 529

U.S. 861, 874 (2000). But, with the Federal Cigarette Labeling and Advertising

Act and the Public Health Cigarette Smoking Act of 1969, in Cipollone the

Supreme Court interpreted the express-preemption provision as exclusively

defining the preemptive scope of the Acts:

In our opinion, the pre-emptive scope of the 1965 Act and the 1969
Act is governed entirely by the express language in § 5 of each Act.
When Congress has considered the issue of pre-emption and has
included in the enacted legislation a provision explicitly addressing
that issue, and when that provision provides a “reliable indicium of
congressional intent with respect to state authority,” “there is no need
to infer congressional intent to pre-empt state laws from the
substantive provisions” of the legislation. Such reasoning is a variant
of the familiar principle of expression unius est exclusio alterius:
Congress’ enactment of a provision defining the pre-emptive reach of
a statute implies that matters beyond that reach are not pre-empted. In
this case, the other provisions of the 1965 and 1969 Acts offer no
cause to look beyond § 5 of each Act. Therefore, we need only
identify the domain expressly pre-empted by each of those sections.

Cipollone, 505 U.S. at 517 (citations omitted) (quoting Malone v. White

Motor Corp., 435 U.S. 497, 505 (1978); Cal. Fed. Sav. & Loan Ass’n v.

Guerra, 479 U.S. 272, 282 (1987) (plurality opinion)).

The Supreme Court has explained that “in Cipollone, we engaged in a

conflict pre-emption analysis of the Federal Cigarette Labeling and Advertising

Act, and found ‘no general, inherent conflict between federal preemption of state
38
 Case: 13-14590 Date Filed: 05/18/2017 Page: 39 of 284

warning requirements and the continued vitality of state common-law damages

actions.’” Freightliner Corp. v. Myrick, 514 U.S. 280, 288–89 (1995) (citation

omitted) (quoting Cipollone, 505 U.S. at 518). Although the Supreme Court

considered only the 1965 and 1969 statutes in Cipollone, “[s]ince the Labeling

Act’s passage, Congress’s basic goals have remained largely unchanged.” Graham,

782 F.3d at 1277. We find nothing in the four statutes passed later that alters the

preemptive scope of federal law on tobacco in a way that is relevant to this appeal.

R.J. Reynolds and Philip Morris argue that, by passing legislation that does

not ban cigarettes, Congress has established a policy of allowing the sale of

tobacco products, but this argument is contrary to settled law that inaction by

Congress cannot serve as justification for finding federal preemption of state law.

See Wyeth, 555 U.S. at 602–03 (Thomas, J., concurring in the judgment)

(collecting cases); Sprietsma v. Mercury Marine, 537 U.S. 51, 65 (2002) (“History

teaches us that a Coast Guard decision not to regulate a particular aspect of boating

safety is fully consistent with an intent to preserve state regulatory

authority . . . .”); Schneidewind v. ANR Pipeline Co., 485 U.S. 293, 306 (1988)

(“This Court generally is reluctant to draw inferences from Congress’ failure to

act.”). “[O]therwise, deliberate federal inaction could always imply pre-emption,

which cannot be. There is no federal pre-emption in vacuo, without a constitutional

39
 Case: 13-14590 Date Filed: 05/18/2017 Page: 40 of 284

text or a federal statute to assert it.” P.R. Dep’t of Consumer Affairs v. Isla

Petroleum Corp., 485 U.S. 495, 503 (1988).

R.J. Reynolds and Philip Morris also rely on the discussion of federal law

regulating cigarettes in FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120

(2000), but that decision does not support their argument for preemption. In Brown

& Williamson, the Supreme Court considered whether the Food and Drug

Administration had jurisdiction over tobacco products. Id. at 125–26. The Court

held that it did not. Id. at 126. The Supreme Court reasoned that, if the

Administration had jurisdiction, the Food, Drug, and Cosmetic Act would require

the administration to remove cigarettes from the market. Id. at 135. The Supreme

Court considered the six federal statutes that regulate cigarette labeling and

concluded that Congress would not have enacted these laws if it intended the

Administration to ban cigarettes. See id. at 137–38. “[T]he collective premise of

these statutes is that cigarettes and smokeless tobacco will continue to be sold in

the United States.” Id. at 139. The Supreme Court stated that Congress has

“foreclosed the removal of tobacco products from the market” in this context, id. at

137—surmising that Congress would not have bothered to regulate a product that it

intended to have removed from the market nationwide by a federal agency.

Although federal agencies have only the authority granted to them by

Congress, states are sovereign. Brown & Williamson does not address state

40
 Case: 13-14590 Date Filed: 05/18/2017 Page: 41 of 284

sovereignty, and it does not consider the preemptive reach of federal legislation on

tobacco. Marotta, 2017 WL 1282111 at *6 (“[W]hile Brown & Williamson held

that the FDA did not have the authority to regulate tobacco products, it said

nothing about the states’ power to do the same.”). Cipollone does.

State governments retain their historic police powers to protect public health.

See U.S. Const. Amend. X. “It is one of the happy incidents of the federal system

that a single courageous state may, if its citizens choose, serve as a laboratory; and

try novel social and economic experiments without risk to the rest of the country.”

New State Ice Co. v. Liebmann, 285 U.S. 262, 386–87 (1932) (Brandeis, J.,

dissenting). Over a hundred years ago, Tennessee, like some other states, passed a

law making it a crime to sell cigarettes. 6 Clark Bell, Medico-Legal Studies 50–65

(1902). Although that experiment in prohibition, like so many others, failed,

Tennessee did not violate the federal Constitution. In upholding the law as not

infringing the power of Congress under the Commerce Clause, the Supreme Court

described the cigarette ban as the type of legislation that states may enact “for the

preservation of the public health or safety” under their police powers. Austin v.

Tennessee, 179 U.S. 343, 349 (1900). Today, state and local governments continue

to enact public health measures to respond to the dangers associated with smoking,

see, e.g., Paul A. Diller, Why Do Cities Innovate in Public Health? Implications of

Scale and Structure, 91 Wash. U. L. Rev 1219, 1234–35 (2014) (discussing state

41
 Case: 13-14590 Date Filed: 05/18/2017 Page: 42 of 284

and local bans of flavored cigarettes passed before the Tobacco Control Act

banned cigarette flavorings); Patrick Kabat, Note, “Till Naught but Ash is Left to

See”: Statewide Smoking Bans, Ballot Initiatives, and the Public Sphere, 9 Yale J.

Health Pol’y L. & Ethics 128, 138–45 (2009) (surveying state prohibitions of

smoking in public places), and to combat other public health risks, see, e.g., Cal.

Health & Safety Code § 114377 (banning certain trans fats); N.Y. State Rest. Ass’n

v. N.Y.C. Bd. of Health, 556 F.3d 114 (2d Cir. 2009) (upholding New York City

law requiring caloric disclosure on chain restaurant menus against preemption

challenge); Trans Fat and Menu Labeling Legislation, Nat’l Conference of State

Legislatures (Jan. 2013), http://www.ncsl.org/research/ health/trans-fat-and-menu-

labeling-legislation.aspx (all Internet materials as visited July 9, 2016 and available

in Clerk of Court’s case file) (listing six states that had enacted menu labeling

legislation as of 2010).

Florida may employ its police power to regulate cigarette sales and to

impose tort liability on cigarette manufacturers. We may not supersede the

“historic police powers of the States” unless it is the “clear and manifest purpose of

Congress.” Wyeth, 555 U.S. at 565 (quoting Lohr, 518 U.S. at 485). And “[t]hat

assumption applies with particular force when Congress has legislated in a field

traditionally occupied by the States,” Altria Grp., 555 U.S. at 77, like public

health, Lohr, 518 U.S. at 475.

42
 Case: 13-14590 Date Filed: 05/18/2017 Page: 43 of 284

R.J. Reynolds and Philip Morris would have us presume that Congress

established a right to sell cigarettes based on a handful of federal labeling

requirements. We decline to do so. We discern no “clear and manifest purpose” to

displace tort liability based on the dangerousness of all cigarettes manufactured by

the tobacco companies.

IV. CONCLUSION

We AFFIRM the judgments against R.J. Reynolds and Philip Morris.

43
 Case: 13-14590 Date Filed: 05/18/2017 Page: 44 of 284

JULIE CARNES, Circuit Judge, concurring in part and dissenting in part:

I concur in the majority opinion’s decision that federal law does not preempt

the jury findings in the underlying Engle litigation. As to defendants’ Due Process

Clause challenge, the latter presents a close question on which reasonable minds

can differ. I do not disagree that the majority opinion articulates reasonable

arguments in explaining why it rejects defendants’ challenge. On balance,

however, I agree with Judges Tjoflat and Wilson that on the particular and unusual

facts of the underlying Engle litigation, its jury findings are too non-specific to

warrant them being given preclusive effect in subsequent trials. Concluding that

defendants’ due process rights were therefore violated, I respectfully dissent as to

the Majority’s contrary holding.

44
 Case: 13-14590 Date Filed: 05/18/2017 Page: 45 of 284

TJOFLAT, Circuit Judge, dissenting:

In 1998, the Circuit Court of the Eleventh Judicial Circuit in and for Dade

County, Florida held a trial—Phase I of the Engle class action—to determine

whether the largest domestic tobacco companies (the “Engle defendants”) engaged

in conduct proscribed by Florida tort law.1 The Phase I jury found that each of the

Engle defendants engaged in nine different kinds of proscribed conduct.2

Eight years later, the Florida Supreme Court “retain[ed] the jury’s Phase I

findings other than those on the fraud and intentional infliction of emotion distress

claims.” Engle v. Liggett Group, Inc. (Engle III), 945 So. 2d 1246, 1269 (Fla.

2006). It then instructed progeny courts tasked with adjudicating causation and

1
Phase I also involved two contract claims, breach of implied warranty and breach of
express warranty. Nevertheless, for convenience, I will refer to all Phase I claims as tort claims.
2
Specifically, the Phase I jury found that each of the Engle defendants (1) “place[d]
cigarettes on the market that were defective and unreasonably dangerous”; (2) “[made] a false
statement of material fact, either knowing the statement was false or misleading, or being
without knowledge as to its truth or falsity, with the intention of misleading smokers”; (3)
“conceal[ed] or omit[ted] material information, not otherwise known or available, knowing the
material was false and misleading, or failed to disclose a material fact concerning or proving the
health effects and/or addictive nature of smoking cigarettes”; (4) “enter[ed] into an agreement to
misrepresent information relating to the health effects of cigarette smoking, or the addictive
nature of smoking cigarettes, with the intention that smokers and members of the public rely to
their detriment”; (5) “enter[ed] into an agreement to conceal or omit information regarding the
health effects of cigarette smoking, or the addictive nature of smoking cigarettes with the
intention that smokers and members of the public rely to their detriment”; (6) “[sold] or
[supplied] cigarettes that were defective in that they were not reasonably fit for the uses
intended”; (7) “[sold] or [supplied] cigarettes that, at the time of sale or supply, did not conform
to representations of fact . . . either orally or in writing”; (8) “failed to exercise the degree of care
which a reasonable cigarette manufacturer would exercise under like circumstances”; (9)
“engaged in extreme and outrageous conduct or with reckless disregard relating to cigarettes sold
or supplied to Florida smokers with the intent to inflict severe emotional distress.”
 Case: 13-14590 Date Filed: 05/18/2017 Page: 46 of 284

damages in individual class-member tort actions to give “res judicata effect” to the

retained findings. Id. at 1254.

But progeny courts had trouble understanding Engle III’s res judicata

instruction. For starters, issuing such a mandate was strange because courts that

render a judgment ordinarily do not attempt to predetermine the res judicata effect

of that judgment.3 Courts tasked with determining whether to enforce a rendering

court’s judgment make those determinations themselves.4 On top of that, the Phase

I findings only established that the Engle defendants engaged in proscribed

conduct; the findings revealed nothing about what the defendants actually did.

Thus, the findings were useless in helping plaintiffs prove, as Florida law requires,5

that their injuries were caused by a defendant’s proscribed conduct. Judge May of

Florida’s Fourth District Court of Appeal lamented that “parties to the tobacco

litigation [were left to] . . . play legal poker, placing their bets on questions left

unresolved by Engle.” Brown R.J. Reynolds Tobacco Co. v. Brown (Jimmie Lee

Brown II), 70 So. 3d 707, 720 (Fla. 4th Dist. Ct. App. 2011) (May, J., concurring).

Seven years after it had issued its res judicata mandate, the Florida Supreme

Court finally stepped in to explain it. The Court conceded that the Phase I findings

3
“A court conducting an action cannot predetermine the res judicata effect of the
judgment; that effect can be tested only in a subsequent action.” Matsushita Elec. Indus. Co. v.
Epstein, 516 U.S. 367, 396, 116 S. Ct. 873, 888, 134 L. Ed. 2d 6 (1996) (Ginsburg, J., concurring
in part and dissenting in part) (citation omitted).
4
See infra note 77 and accompanying text.
5
See, e.g., infra notes 57–58 and accompanying text.
46
 Case: 13-14590 Date Filed: 05/18/2017 Page: 47 of 284

were “useless in individual actions.” Philip Morris USA, Inc. v. Douglas (Douglas

III), 110 So. 3d 419, 433 (Fla. 2013) (emphasis added). Nevertheless, by retaining

certain Phase I findings and instructing progeny courts to give those findings res

judicata effect in Engle III, the Court claimed it had intended to allow class

members to simply “assume[]” “injury as a result of the Engle defendants’

conduct.” Id. at 430 (emphasis added). Thus, regardless of the tort a class member

alleged, she only needed to prove that she was injured as a result of “‘smoking

cigarettes’ manufactured by [a defendant]” to recover. Id. at 426. In effect, then,

the Douglas III Court proscribed the very act of selling cigarettes, albeit under

color of traditional tort law. So long as a defendant’s sale of cigarettes caused a

plaintiff’s injury—that is, so long as a plaintiff was injured by smoking

cigarettes—the plaintiff had no need to identify, for example, the defendant’s

negligent conduct or unreasonably dangerous product defect.

This case was litigated pursuant to the state law set forth in Douglas III.

Earl Graham, as personal representative of the estate of Faye Graham, alleged in

his complaint all of the torts for which Engle III had retained findings. Yet, he was

never required to identify any proscribed conduct other than the sale of cigarettes.

With respect to both negligence and strict liability, the District Court instructed the

jury to determine only “whether smoking cigarettes manufactured by [the]

Defendant was a legal cause of Faye Graham’s death.”

47
 Case: 13-14590 Date Filed: 05/18/2017 Page: 48 of 284

The Majority purport to give effect to the “state law” created in Douglas III.

Ante at 30. They recognize that it is “unorthodox,” “novel,” and amounts to an

irrebuttable presumption of liability. Id. at 25, 27, 28. Yet, they believe that due

process is flexible enough to accommodate such a law. Id. at 29. It is not.6

This is not to say, as the Majority imply, that I would “require a state to

follow the federal common law of res judicata and collateral estoppel.” Id. at 26. I

agree that states are free to fashion “novel” and even “unorthodox” laws. Id. at 25,

27. I do not agree, however, that federal courts must apply such laws when doing

so deprives litigants of an opportunity to be heard on essential elements of their

case.7 Id. at 29–30. I also disagree with the Majority’s conclusion that Florida’s

proscribing the sale of cigarettes is not preempted by federal law.

6
In W. & A.R.R. v. Henderson, 279 U.S. 639, 643, 49 S. Ct. 445, 447, 73 L. Ed. 884
(1929), the Supreme Court held that a defendant railroad company’s due process rights were
violated where it was held liable even though the plaintiff offered no evidence of a connection
between tortious conduct and the injury at issue. Id. at 640–44, 49 S. Ct. 445–48. Instead of
presenting such evidence, the plaintiff relied on a state-law presumption that “[t]he mere fact of
collision between a railway train and a vehicle . . . was caused by negligence of the railway
company.” Id. at 642–43, 49 S. Ct. 445, 447. Because, as a factual matter, a collision could
result from “negligence of the railway, or of the traveler on the highway, or of both, or without
fault of any one,” the Supreme Court struck down the presumption as “unreasonable and
arbitrary.” Id. at 644, 49 S. Ct. 445, 447.
Here, Mr. Graham neither alleged nor proved that Ms. Graham’s death was caused by the
defendants’ tortious conduct. Instead, he was allowed to “assume[]” “injury as a result of the
Engle defendants’ conduct” on the basis of a smoking-related injury. Douglas III, 110 So. 3d at
430. This presumption is just as unreasonable and arbitrary as one that allows plaintiffs to
assume injury as a result of a defendant’s conduct on the basis of a collision-related injury.
7
Litigants enjoy a “due process right to fully and fairly litigate each issue in their case.”
DuPont v. Southern, 771 F.2d 874, 880 (5th Cir. 1985); see also Bell v. Burson, 402 U.S. 535,
542, 91 S. Ct. 1586, 1591, 29 L. Ed 90 (1971) (“It is a proposition which hardly seems to need
48
 Case: 13-14590 Date Filed: 05/18/2017 Page: 49 of 284

To navigate the surprising evolution from Phase I’s nonprobative findings of

fact to Douglas III’s sweeping new tort law, I start from the beginning of the Engle

litigation and proceed painstakingly to the end. As the Table of Contents indicates,

I begin with Phase I of Engle and proceed through Walker v. R.J. Reynolds

Tobacco Co. (Walker II), 734 F.3d 1278 (11th Cir. 2013), and to the opinion the

Court issues today. Along the way, I comment on the decisions in light of relevant

legal principles. My commentaries are set aside by conspicuous section breaks or

headings, and my preemption discussion is set within its own part as it is more

legally complex than the basic principles of procedural fairness that animate the

rest of the opinion.

As I detail below, Engle-progeny opinions examining the same basic legal

issues vary drastically in both their analysis and recitation of the facts. The

Majority, for example, portray Engle III differently from the way all other courts,

including the Florida Supreme Court, see that case.8 Unfortunately, the one theme

that remains constant throughout—with a few exceptions—is that Engle-progeny

courts have rested their thumbs on the scales to the detriment of the unpopular

Engle defendants.

explication that a hearing which excludes consideration of an element essential to the decision . .
. does not meet [the requirements of the Due Process Clause].”).
8
See infra Part VI.
49
 Case: 13-14590 Date Filed: 05/18/2017 Page: 50 of 284

I dissent for eight reasons. First, I reject the Majority’s false narrative of

Engle III. Second, in injecting their false narrative into the case, the Majority

improperly act as advocates and relieve the plaintiff of his burden of proving

preclusion. Third, the Majority fail to provide the defendants with an opportunity

to be heard on the accuracy and applicability of their narrative. Fourth, even if that

narrative were not false, Engle III, as portrayed by the Majority, would not be

entitled to full faith and credit because its key holdings were rendered without

affording the Engle defendants notice or opportunity to be heard. Fifth, and most

importantly, we cannot deprive R.J. Reynolds (“RJR”) and Philip Morris of their

property because they have never been afforded an opportunity to be heard on

whether their unreasonably dangerous product defect(s) or negligent conduct

caused Ms. Graham’s death. Sixth, we cannot give effect to a state law that

amounts to an unreasonable and arbitrary presumption of liability.9 Seventh, we

cannot give effect to a state law that operates to deprive the defendants of their

Seventh Amendment right to a jury trial on contested and material elements of the

claims against them. Eighth, the way in which the Engle-progeny litigation has

been carried out has resulted in a functional ban on cigarettes, which is preempted

by federal regulation premised on consumer choice.

9
Not only does the presumption itself raise due process concerns, the fact that it applies
only to the detriment of a small group of unpopular defendants also raises serious equal
protection concerns.
50
 Case: 13-14590 Date Filed: 05/18/2017 Page: 51 of 284

TABLE OF CONTENTS

I. Procedural History of Engle…………………….…….……………………54

A. Certifying the Engle Class…………………………..………………54

B. Engle Trial to Proceed in Three Phases……………………..………57

1. Phase I…………………………………………………..…….58

2. Phase II……………………………………………………….67

3. Posttrial Motions……………………...………………………72

C. Appeal to the Third District Court of Appeal in Engle II…………...77

D. Petition for Review to the Florida Supreme Court in Engle III……..81

II. What “Res Judicata” Traditionally Means…………………….... . . . . . . . .89

A. Res Judicata 101: The Elements of Issue and Claim Preclusion……92

B. Res Judicata 102: Procedures to Invoke Issue and Claim
Preclusion……………………………………………………………99

III. Engle III Instructed Courts to Disregard Traditional Res Judicata Law so as
to Hold the Defendants Liable without Regard to the Phase I
Findings.......................................................................................................107

A. The U.S. District Court for the Middle District of Florida in Brown I
Rejected the Florida Supreme Court’s Interference with Its Duties as a
Recognizing Court..………...……………………………..……….109

B. In Brown II, We Upheld the District Court’s Decision as a
Recognizing Court to Apply Florida’s Traditional Issue-Preclusion
Doctrine to the Phase I Findings…..……………………………….131

C. The Florida District Courts of Appeal Rejected Brown II on the Basis
of Engle III’s Instruction……………………………..…………...136

51
 Case: 13-14590 Date Filed: 05/18/2017 Page: 52 of 284

1. The Martin I Circuit Court Concluded That Engle III’s
Instruction Required It to Hold the Defendants Liable If the
Plaintiff Simply Proved Class Membership Irrespective of the
Phase I Findings……….…….………..……………………..137

2. The First District Court of Appeal in Martin II Agreed That
Engle III’s Instruction Required It to Hold the Defendants Liable
to all Class Members Irrespective of the Phase I Findings...…141

3. The Fourth District Court of Appeal in Jimmie Lee Brown II
Held That Engle III’s Instruction Meant Issue Preclusion but
That the Plaintiff Did Not Need to Identify a Specific Defect or
Negligent Conduct…………………………………..……….153

D. In Light of Martin II and Jimmie Lee Brown II, the Middle District of
Florida in Waggoner Ruled That the Preclusive Application of the
Phase I Findings to Hold the Defendants Liable Would Not Violate
Due Process………….……………………………………………..160

E. The Second District Court of Appeal in Douglas II Accepted Martin
II’s Reasoning, But Certified the Due Process Question to the Florida
Supreme Court…………………………………………………..…170

IV. The Florida Supreme Court in Douglas III Held That the Engle III Court
Had (1) Implicitly Determined That the Phase I Findings Were Full-Blown
Liability Determinations and (2) Implicitly Entered Judgment Against All
Defendants on Behalf of All Class Plaintiffs ……………..………….…..176

V. The Walker Panel Effectively Rewrote and then Gave Full Faith and Credit
to Douglas III Before Issuing a New Opinion That Gave Full Faith and
Credit to Engle III, Yet Left the Original Opinion’s Inapposite Reasoning
Intact... …………………………………………………………………....211

VI. The Majority Repeat and Add to the Walker Panel’s Errors………..…….234

VII. The Functional Ban on Cigarettes is Preempted by Federal Law………...247

A. Obstacle Preemption…………………………………………….....250

52
 Case: 13-14590 Date Filed: 05/18/2017 Page: 53 of 284

B. Federal Regulation of Tobacco Consumers’ Ability to Choose…...253

C. Florida Has Imposed a Duty Not to Sell Cigarettes Contrary to
Federal Law……………………………………………………...…258

D. The Majority Misinterpret the Statutory Framework of Tobacco
Regulation………………………………………………………….264

Conclusion........………………………………………...……………………......265

53
 Case: 13-14590 Date Filed: 05/18/2017 Page: 54 of 284

I.
PROCEDURAL HISTORY OF ENGLE

The Engle litigation epic began in 1994 when six plaintiffs filed a putative

class action in the Circuit Court for Miami-Dade County, Florida against the Engle

defendants seeking over $100 billion in both compensatory and punitive damages

for injuries allegedly caused by smoking cigarettes. Walker II, 734 F.3d at 1278 .

The plaintiffs asserted an array of claims, including “strict liability, negligence,

breach of express warranty, breach of implied warranty, fraud, conspiracy to

commit fraud, and intentional infliction of emotional distress.” Liggett Grp. Inc. v.

Engle (Engle II), 853 So. 2d 434, 441 (Fla. 3d Dist. Ct. App. 2003).

A. Certifying the Engle Class

On May 5, 1994, the plaintiffs moved the Circuit Court pursuant to Florida

Rule of Civil Procedure 1.220(b)(3)10 to certify a class consisting of all smokers in

the United States and their survivors. They estimated that the class would include

“in excess of one million addicted smokers.” R.J. Reynolds Tobacco Co. v. Engle,

672 So. 2d 39, 41 (Fla. 3d Dist. Ct. App. 1996). The defendants opposed the

motion, arguing that it failed to establish the “predominance” and “superiority”

10
This rule allows certification of a class action when common issues “predominate over
any question of law or fact affecting only individual members of the class, and class
representation is superior to other available methods for the fair and efficient adjudication of the
controversy.” Fla. R. Civ. P. 1.220(b)(3).
54
 Case: 13-14590 Date Filed: 05/18/2017 Page: 55 of 284

requirements imposed by Rule 1.220(b)(3). Id. at 39. They further argued that a

nationwide class would be unmanageable and would unduly burden Florida’s

courts and taxpayers. Id. at 41–42. The Circuit Court disagreed. It granted the

plaintiffs’ motion and certified a nationwide class.

The defendants appealed the decision to the District Court of Appeal, Third

District.11 Id. at 39. The Third District found that the plaintiffs’ motion satisfied

the Rule 1.220(b)(3) “predominance” requirement but agreed with the defendants

that a nationwide class was too large in that it “would unduly burden Florida courts

and taxpayers,” and would “require the sustained attention of all . . . circuit judges

in Dade County, if not the entire state.” Id. at 40, 41. After the nationwide class

had been rejected, the plaintiffs responded with their fallback position—a

statewide class, which, they later represented, would consist of roughly 40,000

members.

Appeased, the Court affirmed the certification order on January 31, 1996,

but limited the class to “[a]ll Florida citizens and residents,” “and their survivors,

who have suffered, presently suffer or have died from diseases and medical

conditions caused by their addiction to cigarettes that contain nicotine.” Id. at 40–

42. In their motion for rehearing, the defendants rejected a 40,000-claimants

estimate, insisting that a statewide class would consist of an unmanageable host of

11
The Court had jurisdiction to entertain the appeal, as an interlocutory appeal, pursuant
to Fla. R. App. P. 9.130(a)(3)(C)(ii).
55
 Case: 13-14590 Date Filed: 05/18/2017 Page: 56 of 284

hundreds of thousands of class members. Nevertheless, the Court denied their

motion on May 10, 1996, and the Florida Supreme Court denied review on October

2, 1996. R.J. Reynolds Tobacco Co. v. Engle, 682 So. 2d 1100 (Fla. 1996). Three

months later, plaintiffs’ counsel wrote thousands of Florida physicians informing

them of the class action and stating that the class included “well over one-half

million” people.

By the end of 1997, as the case proceeded through its pretrial stages, the

class had indeed grown to hundreds of thousands of claimants.12 In light of the

class size and plaintiffs’ counsel’s concession that addiction to nicotine was an

individual issue, the Engle defendants moved to decertify the class. The Court

heard the motion on January 15, 1998.13 It denied the motion with this comment:

12
On May 13, 1997, the plaintiffs represented to the court that the class included a half-
million members.
13
The defendants’ motion to decertify the class was akin to the motions tobacco
companies had been asserting in the scores of smoker class actions that had been filed in state
and federal courts across the country. The courts in most of the cases had declined to certify a
class of Engle’s magnitude, concluding that the claims were too individualized to make class-
wide adjudication viable. See, e.g., Barnes v. Am. Tobacco Co., 161 F.3d 127, 143 (3d Cir.
1998) (affirming the denial of class certification because cigarette litigation and addiction claims
involved too many disparate, individual issues to make class treatment appropriate); Castano v.
Am. Tobacco Co., 84 F.3d 734, 746–48 (5th Cir. 1996) (noting that “historically, certification of
mass tort litigation cases has been disfavored” and reversing district court’s grant of class
certification because of both severe manageability problems and the fact that “the most
compelling rationale for finding superiority in a class action—the existence of a negative value
suit—is missing in this case”); Arch v. Am. Tobacco Co., 175 F.R.D. 469, 492 (E.D.PA. 1997)
(refusing to certify the class because “there are simply too many individual issues and class
members to try this class efficiently. The manageability problems . . . are staggering”); Smith v.
Brown & Williamson Tobacco Corp., 174 F.R.D. 90, 98–99 (W.D. Mo. 1997) (finding that the
individualized nature of claims defeated class certification and expressing concern “that forcing
the plethora of individual issues into a class action constitutes a disservice to both potential class
56
 Case: 13-14590 Date Filed: 05/18/2017 Page: 57 of 284

I believe changes have occurred. I also believe that the case may be
unmanageable. I do have substantial reservations regarding the class
action. However, I’m going to deny [defendants’] motion. I do
implore the Third District Court of Appeal to accept review on an
expedited basis and to take into consideration a lot of what
[defendants] have covered.

The defendants appealed the Court’s ruling to the Third District. That Court

dismissed the appeal for lack of jurisdiction, but stated that the defendants had a

right to obtain review of “the propriety of the order by plenary appeal from any

adverse final judgment.” Engle II, 853 So. 2d at 443.

B. Engle Trial to Proceed in Three Phases

In February 1998, the Circuit Court announced that it had developed a

tentative three-phase trial plan to manage the litigation.14 Engle III, 945 So. 2d

1246, 1256 (Fla. 2006). In Phase I, the Court would conduct a jury trial of the

issues common to the entire statewide class. The plan defined the issues as those

“which form integral elements of the claims” the named plaintiffs were asserting

members and the Defendant”); Small v. Lorillard Tobacco Co., 252 A.2d 1, 12 (N.Y. App. Div.
1998) (finding the proposed class action “unmanageable because of the individual issues of
reliance, causation and damages with respect to each of the [class members]”); Reed v. Philip
Morris, Inc., No. 96-5070 (D.C. Super. Aug. 18, 1997), 1997 WL 538921, at *9 (denying class
certification in light of the fact that “the individual issues raised not only predominate over the
common issues raised but overwhelm [them]” (quotation marks and citation omitted)).
14
The Court’s plan was tentative in that it made changes in the plan before and after the
trial began.
57
 Case: 13-14590 Date Filed: 05/18/2017 Page: 58 of 284

for themselves and the members of their class.15 The jury would evaluate evidence

exclusively related to the defendants’ conduct and would determine whether such

conduct rendered the defendants answerable in punitive damages.

If the jury found that the defendants had engaged in the tortious conduct

alleged, the litigation would proceed to Phase II-A to determine whether that

conduct caused the class representatives’ injuries. In Phase II-B, the same jury

would also decide whether the entire class was entitled to punitive damages, and, if

so, make a “lump-sum” award. Engle III, 945 So. 2d at 1257. Finally, in Phase

III, new juries would try the individual class members’ claims—that the Engle

defendants’ tortious conduct caused their harm. Id. at 1268. The punitive

damages, if any, awarded in Phase II-B would be divided among the class

members who prevailed.

1. Phase I

The Phase I trial commenced on July 6, 1998. In accordance with the plan,

the Phase I jury considered evidence pertaining to the defendants’ conduct between

1953 and 1994 and to whether cigarettes manufactured during that time were

addictive and caused diseases. Over the course of the yearlong trial, the plaintiffs

15
The trial plan did not shape the issues more concretely than this. Indeed, plaintiffs’
counsel argued that the trial plan represented a judgment that the Phase I jury need not “get
involved in unnecessary complexity and fragmentation by asking a zillion specific questions, but
rather to have the jury take the common sense approach.”
58
 Case: 13-14590 Date Filed: 05/18/2017 Page: 59 of 284

presented evidence that was sweeping in its scope, spanning decades of tobacco-

industry history. Ante at 5–8. Witnesses testified that cigarettes were addictive

and could cause a variety of diseases, including lung cancer. Douglas III, 110 So.

3d at 423 (Fla. 2013). Witnesses also described differences among cigarette

brands, filtered and nonfiltered, in terms of their tar and nicotine levels and the

way in which they were designed, tested, manufactured, advertised, and sold. Id.

at 423–24.

With such wide-ranging evidence and disparity among cigarettes, the

defendants registered early on their concerns that the jury would have a hard time

sorting through the evidence and connecting it to particular defendants and

particular assertions of wrongdoing. They repeatedly argued, for example, “that

[the] wide spectrum of views . . . represented by counsel . . . [make it] hard [to]

figur[e] out where we’re going as a common question.” The defendants later

summarized their concerns:

The Court subjected defendants to an artificial proceeding, not a real
trial, in which the jury was inundated with evidence of abstract
“misconduct” unconnected to any real person’s knowledge, choices,
or other circumstances—thus setting the stage for an enormous
punitive award in Phase II-B. Plaintiffs were allowed to “mix and
match” their evidence, creating a hypothetical plaintiff who was
exposed to and relied on every alleged misstatement over the course
of nearly 50 years, smoked every band of cigarette, and suffered every
asserted disease plus “emotional distress.”

59
 Case: 13-14590 Date Filed: 05/18/2017 Page: 60 of 284

Undeterred, the Court responded that it would make sense of the scattershot

theories and evidence by means of jury instructions at the end of Phase I.

In March 1999, the plaintiffs rested, and the defendants moved the Court for

decertification of the class and a directed verdict on all counts. After eight months

of trial, the defendants pressed the Court to address the manageability problems

that had been looming since the beginning. Although the plaintiffs had, to that

point, successfully urged the Court to postpone such issues until “later,” the

defendants insisted that “later is here. Later is now.”

Given the jumble of evidence and theories that had been put forward, the

defendants argued, the jury would be unable to match theories with evidence as

required unless it was instructed with precision:

If we asked the question, Judge, can smoking cause heart disease? and
they answer that yes, so what? So what? The question is going to be,
did it cause this class member’s heart disease? That’s got to be the
only significant question. It’s a “so what?”
Take the easy one, the one that you could apply most
meaningfully: Product defect. There is one, and we ought to be able
to get a jury to give us something on product defect. If they decide in
favor of the plaintiff, we can take that and we can transport it into
Phase II and Phase III.
Well, when you think about that, how are you going to do that,
because we have no actual plaintiff in the common issue part of this
trial, all kinds of evidence has been introduced from which a jury
could conceivably find that there’s a defect in the product?
They might find that it has something to do with a particular
filter construction; they might find that it’s products with a certain
amount of nicotine; they could say that it’s additives, that when
certain additives were put into the cigarette; they could say that it has
to do with low tar, the fact that people who smoke low-tar cigarettes
60
 Case: 13-14590 Date Filed: 05/18/2017 Page: 61 of 284

get a different tar level than the FTC machines, and that that’s a
defect.
But how are we ever going to know? And this is the easy one.
Forget the fraud, misrepresentations. But how are we ever going to
know on what basis the jury found the defect? Are they going to tell
us on what basis they found the defect? And if we don’t know on
what basis they found the defect, how are we going to apply that to
people in subsequent phases?
If the defect is in connection with low tar, then people who
smoked high-tar cigarettes their entire lives . . . wouldn’t have a claim
[because] there would be no proximate cause with regard to their
particular allegations. But we won’t know that.
And it’s uniquely caused by the nature of this trial. If this was a
single smoker, we’d know the particular circumstances of that
smoker. We'd know what evidence was relevant, what evidence
wasn’t relevant, and we would be able to look at and apply it.

Such reasoning undergirded the defendants’ motions for directed verdict as

well. In those motions, the defendants argued that the plaintiffs had spread

themselves too thin by sporadically referencing, while never fully substantiating,

numerous theories of liability. The defendants worried that these shotgun-style

allegations would unfairly disadvantage them if their motions were denied:

You take all the stuff that you think sounds bad. You say it all real
fast. You say: We had 57 witnesses, and all this. And then you say:
Therefore, we have a case. We have law that requires certain
evidence. We have to know what to defend against, and we all have
to know what to put on that verdict form.

The plaintiffs did not confront the merits of such arguments directly,

countering instead with two process-oriented arguments. First, they argued that the

defendants failed to satisfy the directed verdict standard because “the burden of the

defendants is an almost impossible burden. In most instances because the
61
 Case: 13-14590 Date Filed: 05/18/2017 Page: 62 of 284

defendants have to convince the Court that there is not minimal but zero, zero

evidence and zero inferences from the evidence that would support our claims.”

Second, the plaintiffs argued that the Court should defer its ruling because the law

demands “that in those rare instances where the Court really doesn’t feel there’s

enough to go to a jury, the Court should wait,” let the jury render a verdict, and

then rule, so the appellate court can reinstate the jury verdict if it disagrees with the

trial judge.

Persuaded by the plaintiffs, the Court reserved ruling on the motion,16 and

the defendants went on to present their case.17 On June 9, 1999, the parties rested,

and a charge conference with counsel followed. The plaintiffs conceded that there

were “many hundreds and hundreds” of things at issue for each claim. To account

for the many theories presented, and, concomitantly, to provide the jury with the

widest possible range of bases upon which to premise tortious-conduct findings,

the plaintiffs proposed that the jury be instructed on eight different theories of

negligence and five theories of strict liability. The defendants did not take issue

with instructing the jury on an array of tortious-conduct theories. They did object,

however, to the plaintiffs’ proposed verdict-form questions, which were generic

rather than disaggregated and specific. They warned that a verdict form that failed

16
See Fla. R. Civ. P. 1.480(a)–(b).
17
With one exception: the Court did grant the defendants’ motions for directed verdict
with respect to certain diseases and medical conditions.
62
 Case: 13-14590 Date Filed: 05/18/2017 Page: 63 of 284

to specify the particular theories on which the jury based its findings could not be

“meaningful[ly] imported into Phase II and Phase III”:

If the jury in this case were to simply answer the question, “Have one
or more of the defendants, during whatever time period, manufactured
a cigarette that is defective and unreasonably dangerous?” and the
answer to that is “Yes,” what in the world are we going to do with that
in an individual case? We won’t know what the defect was. We
won’t know when or during what period of time, what brand or brand
style. What in the world are we going to do with that finding?18

Because, the defendants argued, a generic verdict form would make it “completely

impossible to import intelligently and rationally the findings from the verdict form

in Phase I to any particular plaintiff in Phase II and III,” relying on such a verdict

18
The defendants were pointing out that the class plaintiffs would be unable to prove, as
relevant tort law requires, that a particular defect caused harm if they were unable to even
identify the product features the jury deemed defective and unreasonably dangerous. Similarly,
plaintiffs in Phase III trials would be unable to prove that the defendants’ negligent conduct
caused harm if they could not identify the conduct the Phase I jury deemed negligent. In other
words, the Phase I findings would be utterly useless to plaintiffs if they could not rely on those
findings to identify the defendants’ product defect(s) and tortious conduct in the Phase III
causation trials.
The problems associated with generic findings extended beyond the negligence and
strict-liability claims. For example, regarding the claims of fraud, the defendants argued,
If you merely ask this jury whether the defendants made a misstatement of a
material fact, and they are not required to identify what it is, when you go into the
Phase II and Phase III trials of the individual smokers’ claims, that finding will
have no meaning. So we believe that, for it to have meaning going forward, it
needs specificity.
The trial judge did not heed the defendants’ warnings. As a result, the Phase I
findings were, as the Florida Supreme Court conceded seven years later, “useless.”
Douglas III, 110 So. 3d at 433. To remedy this problem, the Florida Supreme Court
sanctioned a conclusive presumption that eliminated the class members’ burden of
proving that a defendant’s unreasonably dangerous product defect or tortious conduct
caused their harm. See supra note 6 and accompanying text.

63
 Case: 13-14590 Date Filed: 05/18/2017 Page: 64 of 284

form to preclude defendants’ defenses in later phases would result in a “due

process violation under the U.S. Constitution as well as the Florida Constitution.”

The defendants accordingly requested a verdict form that would elicit

specific findings that class members could later allege, in a meaningful way and in

accordance with due process, in their Phase III complaints. See Walker II, 734

F.3d at 1282 (The defendants “requested that the trial court submit to the jury a . . .

detailed verdict form that would . . . ask[] the jury [among other things] to identify

the brands of cigarettes that were defective.”). Plaintiffs repeatedly opposed such

requests, arguing that specificity burnished a slippery slope to complexity and

delay: “[O]nce you start [being more specific], then you’ve got to include a lot

more . . . . And that becomes a 20, 25-page verdict form for the jury to complete,

yes, no, and be here for a long time.” The Court sided with the plaintiffs.

Hence, the first two questions on the finalized verdict form made no

distinction between cigarette brands and did not even refer to the defendants’

conduct.19 Instead, the questions asked the jury to determine whether cigarettes

could cause certain diseases and addiction. The remaining verdict-form questions

charged the jury to determine whether the defendants had engaged in tortious

conduct, but did not require the jury to reveal the theory or theories on which it

19
The verdict form is Appendix A to this dissent. The answers to Questions No. 3 (strict
liability) and No. 8 (negligence) are the Phase I findings underpinning the judgment in this case
and are directly at issue in this appeal.
64
 Case: 13-14590 Date Filed: 05/18/2017 Page: 65 of 284

premised its tortious-conduct findings.20 Thus, as the defendants had feared, the

verdict form did not prompt the jury to indicate whether it had accepted, for

example, just one or all eight instructed theories of negligence.21 Nor did it prompt

the jury to reveal which of the five instructed theories of strict liability22 it accepted

or which particular brands of cigarettes or cigarette features it identified as

defective and unreasonably dangerous. With respect to strict liability and

negligence—the two claims at issue in this appeal—the form simply asked the jury

20
Questions No. 6 (breach of implied warranty) and No. 7 (breach of express warranty)
deal with contract law rather than tort law. Nevertheless, for convenience, I refer to the Phase I
findings as “tortious-conduct findings” throughout this opinion.
21
This appeal involves negligence and strict-liability claims. With respect to negligent
conduct, the jury was asked to determine
whether one or more of the defendants were negligent in manufacturing,
designing, marketing, selling and distributing cigarettes which defendants knew
or should have known would cause serious and fatal diseases, including lung
cancer, or dependence-producing substances; in negligently not testing tobacco
and commercial cigarettes to confirm that smoking causes human disease; in
failing to design and produce a reasonably safe cigarette with lower nicotine
levels; in negligently measuring and . . . understating nicotine and tar levels in
low-tar cigarettes; and in failing to warn smokers of the dangers of smoking and
the addictiveness or dependence-producing effects of cigarettes prior to July 1 of
1969.
The jury was also instructed as follows:
The issue for determination on the negligence claims of the plaintiffs against each
of the tobacco companies is whether one or more of the tobacco companies were
negligent in designing, manufacturing, testing, or marketing of cigarettes.
Another issue for your determination is whether one or more of the defendant
tobacco companies were negligent prior to July l, 1969 in failing to warn smokers
of the health risks of smoking or the addictiveness of smoking.
22
As for strict liability, the Court initially instructed the jury that “the issues are whether
one or more of the defendants designed, manufactured and marketed cigarettes which were
defective and unreasonably dangerous to smokers.” Later on, it instructed the jury that the “issue
for [its] consideration is whether cigarettes sold by these tobacco companies were defective
when they left the possession of the companies.”
65
 Case: 13-14590 Date Filed: 05/18/2017 Page: 66 of 284

to respond “yes” or “no” to whether “one or more of the Defendant Tobacco

Companies” (1) “place[d] cigarettes on the market that were defective and

unreasonably dangerous”23 and (2) “failed to exercise the degree of care which a

reasonable cigarette manufacturer would exercise under like circumstances.”24

The defendants objected to both questions, arguing that “[t]he [defect]

question does not require specificity as to the product (brand or brand style), the

defect, or the time of occurrence, which renders it useless for application to

individual plaintiffs in other Phases of this case,” and “the [negligence] question

does not require specificity as to the product (brand and brand style), the alleged

negligent act, and the date of the act, which renders it useless for application to

individual plaintiffs in other Phases of this case.” They objected to verdict-form

questions related to other tortious-conduct claims as well, insisting that the answers

to such questions would be useless in Phase III because a different jury would be

unable to discern what conduct the Phase I jury deemed tortious, making it

impossible to prove that such conduct caused harm.

The Court overruled the defendants’ objections, and the jury, in the verdicts

they returned on July 7, 1999, answered “yes” to every question. 25 Walker II, 734

23
The time periods for strict liability were “before July 1, 1974,” “after July 1, 1974,” or
both.
24
The time periods for negligence were “before July 1, 1969,” “after July 1, 1968,” or
both.
25
On the whole, the jury found:

66
 Case: 13-14590 Date Filed: 05/18/2017 Page: 67 of 284

F.3d at 1282. The defendants moved the Court to set aside the verdicts in

accordance with their motions for directed verdict and alternatively for a new

(1) that smoking cigarettes caused twenty of twenty-three listed diseases or
medical conditions; (2) that cigarettes containing nicotine were addictive or
dependence producing; (3) that the defendants placed cigarettes on the market
that were defective and unreasonably dangerous both before and after July 1,
1974; (4) that the defendants made a false statement of a material fact, either
knowing the statement was false or misleading, or being without knowledge
as to its truth or falsity, with the intention of misleading smokers both before
and after May 5, 1982; (4a) that the defendants concealed or omitted material
information, not otherwise known or available, knowing the material was false
and misleading, or failed to disclose a material fact concerning or proving the
health effects and/or addictive nature of smoking cigarettes both before and
after May 5, 1982; (5) that the defendants entered into an agreement to
misrepresent information relating to the health effects of cigarette smoking, or
the addictive nature of smoking cigarettes, with the intention that smokers and
members of the public rely to their detriment; (5a) that the defendants entered
into an agreement to conceal or omit information regarding the health effects
of cigarette smoking, or the addictive nature of smoking cigarettes, with the
intention that smokers and members of the public rely to their detriment; (6)
that the defendants sold or supplied cigarettes that were defective in that they
were not reasonably fit for the uses intended before July 1, 1969 and up to and
after July 1, 1974; (7) that the defendants sold or supplied cigarettes that, at
the time of sale or supply, did not conform to representations of fact made by
the defendants either orally or in writing both before and after July 1, 1974;
(8) that the defendants failed to exercise the degree of care that a reasonable
cigarette manufacturer would exercise under like circumstances both before
and after July 1, 1969; (9) that the defendants engaged in extreme and
outrageous conduct or with reckless disregard relating to cigarettes sold to
Florida smokers with the intent to inflict severe emotional distress; and (10)
that the defendants' conduct rose to a level that would permit a potential award
or entitlement to punitive damages.

Brown v. R.J. Reynolds Tobacco Co. (Brown II), 611 F.3d 1324, 1327 (11th Cir. 2010) (citation
omitted).

67
 Case: 13-14590 Date Filed: 05/18/2017 Page: 68 of 284

trial.26 They also moved the Court to decertify the class. On July 29th, the Court

summarily denied these motions.

2. Phase II

The trial of Phase II-A—the cases of three class representatives, Mary

Farnan, Frank Amodeo, and Angie Della Vecchia,27 against six tobacco

companies28—began on November 1, 1999. All alleged that they were addicted to

cigarettes, smoked a variety of the companies’ brands, both filtered and

nonfiltered,29 and in time contracted cancer. All sought damages against each

defendant on theories of strict liability and negligence.30 Because the Phase I trial

did not involve the class representatives’ claims (or those of any class members),

the Phase I jury was not instructed to determine whether any of the brands these

plaintiffs actually smoked were defective, unreasonably dangerous, or negligently

26
See Fla. R. Civ. P. 1.480(c).
27
Angie Della Vecchia was deceased. Her claims were brought by her personal
representative, Ralph Della Vecchia. For convenience, I refer to Ms. Della Vecchia as the class
representative plaintiff rather than Mr. Della Vecchia.
28
The companies were Philip Morris, R.J. Reynolds, Brown & Williamson, Lorillard,
Liggett Group and Brook Group Holding.
29
The cigarettes included the following brands: Camels, Salem, Winston, Winston
Lights, Marlboro, Viceroy, Raleighs, Tareyton, Carlton, Pall Mall, Kent, Lucky Strike, Virginia
Slims, Benson & Hedges, Cambridge Lights, and Parliament.
30
In addition, the plaintiffs sought damages based on fraud and misrepresentation,
conspiracy to misrepresent and commit fraud, breach of implied warranty, intentional infliction
of emotional distress, and breach of express warranty. I limit my discussion, for the most part, to
the claims of strict liability and negligence because those are the claims pertinent to this appeal.
68
 Case: 13-14590 Date Filed: 05/18/2017 Page: 69 of 284

produced.31 In Phase I, the jury had determined “issues . . . concerning the conduct

of the tobacco industry.” In Phase II-A, the same jury32 was tasked with deciding

inter alia whether the tortious conduct it identified in Phase I caused the class

representatives’ injuries.

The jury had the Phase I trial record before it, and the three plaintiffs

augmented that record by alleging the various brands of cigarettes they smoked,

their inability to stop smoking, and that cigarette smoking caused the cancer they

contracted.33 After they rested their cases, the defendants moved the Court for

directed verdicts on the ground that the plaintiffs failed to prove all elements of

their claims, including whether the cigarettes the plaintiffs smoked were defective,

unreasonably dangerous, or negligently produced. Evidence that cigarettes could

31
The three class representatives were not class representatives when the Phase I trial
began. Along the way, they were substituted for the original class representatives, but as far as I
can tell from the record, the complaint was not amended to allege, among other things, the
brands of cigarettes they smoked. The brands they smoked were disclosed during the
presentation of the evidence in Phase II-A.
32
Phase II-A was different from other progeny cases in this regard. The Phase II-A jury
was the same as the Phase I jury. The Phase II-A jury, therefore, knew the particular defects and
tortious conduct it had in mind when it answered “yes” to the Phase I verdict-form questions.
33
Fifteen witnesses testified on behalf of Mary Farnan, eighteen testified on behalf of
Frank Amodeo, and sixteen testified on behalf of Angie Della Vecchia. The testimony and
evidence focused on their smoking and medical histories; awareness of the health risks of
smoking; exposure to and purported reliance on statements the tobacco companies made; ability
to quit smoking; the cause of the cancer; and various other individual-specific issues, including
comparative negligence.

69
 Case: 13-14590 Date Filed: 05/18/2017 Page: 70 of 284

cause disease, the defendants argued, did not establish that their tortious conduct

caused the plaintiffs’ diseases.

The Court deferred its ruling on the motion until after the jury rendered its

verdicts on the plaintiffs’ claims. In the Court’s view, the jury’s answers to the

Phase I verdict-form questions, coupled with the plaintiffs’ testimony that they

could not stop smoking and their experts’ testimony that their smoking caused their

cancer, were all the plaintiffs needed to make out a case for the jury under the

theories of strict liability and negligence they were advancing.

The Court’s instructions to the jury reflected this view.34 The Court began

by explaining that the issues the jury decided in Phase I were not being litigated

anew. What it had to decide now was whether the defendants’ “conduct” on which

it based its Phase I verdict was the “legal cause of injury to Mary Farnan, Frank

Amodeo and Angie Della Vecchia.”35 Turning to the verdict form it would be

submitting to the jury, the Court informed the jury of the issues it had to decide by

answering “yes” or “no” to a series of questions, each prefaced with a finding the

jury made in Phase I.

34
Before charging the jury, the Court explained to counsel that “[t]his is not two separate
trials, although a lot of people thought it was. It’s really a continuation of one trial, a bifurcated
trial. And although we had different issue in trial one and different exhibits in evidence, it is part
of the same trial. And one relates to the other.”
35
The Court stated that a “defective and unreasonably dangerous product” or negligence
“is a legal cause of loss, injury or damage if it directly and in natural and continuous sequence
produces or contributes substantially to producing such loss, injury or damage, so that it can
reasonably be said that but for the [defective product or negligence], the loss injury or damage
would not have occurred.”
70
 Case: 13-14590 Date Filed: 05/18/2017 Page: 71 of 284

The first question was prefaced with this statement: “In your [Phase I

verdict], you found that smoking cigarettes causes . . . lung cancer and laryngeal

(throat) cancer.” The question that followed asked, “[W]as smoking cigarettes a

legal cause” of the plaintiff’s cancer? If the jury answered “yes,” it would proceed

to the question pertaining to the claims of strict liability. The preface read, “You

found in your [Phase I verdict] that each of the Defendant Tobacco Companies

placed cigarettes on the market that were defective and unreasonably dangerous,

both before and after July 1 of 1974 (except for Brooke, whose liability is limited

to after July, 1974).” That preface was followed by a question: “Were defective

and unreasonably dangerous cigarettes placed on the market by one or more of the

Defendant tobacco companies a legal cause of [the plaintiff’s cancer]”?

In addition to answering this question regarding strict liability, the jury had

to answer the question pertaining to the claims of negligence. The preface to the

question was, “[I]n your [Phase I verdict], you found that all of the Defendant

Tobacco Companies failed to exercise the degree of care which a reasonable

cigarette manufacturer would exercise under like circumstances, both before and

after July 1 of 1969 (except for Brooke whose liability is limited to after July 1,

1969).” The corresponding question was, “As to each of the Defendants . . . please

state whether that Defendant’s negligence was a legal cause of [plaintiff’s cancer].”

71
 Case: 13-14590 Date Filed: 05/18/2017 Page: 72 of 284

The Court sent the case to the jury on April 5, 2000. The jury returned its

verdicts on April 7, 2000, responding “yes” to each of the questions and therefore,

pursuant to the Court’s instructions, proceeded to determine the amount of the

plaintiffs’ compensatory damages, which were offset by comparative fault. The

total award was $12.7 million. Engle II, 853 So. 2d at 441.

The trial of Phase II-B began on May 22, 2000. In Phase I, the jury

determined that the defendants’ conduct warranted the imposition of punitive

damages,36 so the Phase II-B trial focused on the monetary sum that should be

imposed. On July 14, 2000, the jury fixed that sum at $145 billion to be awarded

incrementally to class members who prevailed in subsequent Phase III lawsuits.

Id. at 1257.

3. Posttrial Motions

At the conclusion of Phase II-B, the defendants moved the Court for the

entry of judgment (as to Phases II-A and II-B) in accordance with their motion for

directed verdict.37 They also moved the Court to decertify the plaintiffs’ class.38

36
In Engle II, the Court stated that the Phase I “jury . . . made a general finding that the
defendants had engaged in unspecified conduct that ‘rose to a level that would permit a potential
award or entitlement to punitive damages.’” 853 So. 2d at 443.
37
The motion adopted and incorporated by reference “all of the grounds set forth” in the
defendants’ July 19, 1999, motion made at the conclusion of Phase I. This motion, like most of
the defendants’ motions, was filed jointly. I therefore treat each of the defendants’ motions as a
joint motion unless otherwise indicated.
72
 Case: 13-14590 Date Filed: 05/18/2017 Page: 73 of 284

The plaintiffs filed no opposition to the defendants’ motions. The Court took the

motions under advisement and, without entertaining oral argument, denied them on

November 6, 2000, in an Amended Final Judgment and Amended Omnibus Order

(“Omnibus Order”).39

In denying the defendants’ motion for the entry of judgment in accordance

with their motion for directed verdict, the Omnibus Order addressed the plaintiffs’

claims separately40 and concluded that each claim had substantial evidentiary

support.41

38
The defendants moved the Court to decertify the plaintiff class following the jury’s
verdicts in Phase II-A and following its verdict in Phase II-B. In these motions, they argued that
“the [Phase I] verdict . . . does not advance the claim of any individual class member. The Court
is now left with potentially hundreds of thousands of class member trials involving all the
individual issues that courts around the country have held preclude class treatment.”
39
The Omnibus Order disposed of forty-six motions the defendants had filed during the
course of the litigation. Forty-four motions were denied and two were granted. “First, the trial
court granted Tobacco’s motion for directed verdict on a statute of limitations basis with regard
to named plaintiff Frank Amodeo on the counts based on strict liability, implied warranty,
express warranty, negligence, and intentional infliction of emotional distress.” Engle III, 945 So.
2d at 1257. “Second, the court granted Tobacco’s motion for directed verdict with regard to
count seven of the complaint, in which the Engle Class sought equitable relief, upon the basis
that the count had previously been dismissed by the court.” Id.
The Court entered the Omnibus Order after withdrawing a Final Judgement and Omnibus
Order it entered on November 3, 2000. Between November 3rd and 7th, the Court made several
minor alterations to November 3rd order. The Court “reserve[d] jurisdiction . . . to enter any
further Orders and conduct further proceedings to the Mandate of the Third District Court of
Appeal of Florida.”
40
The Omnibus Order referred to the claims that were pursued during the trial of Phase
I: Count I, Strict Liability; Count II, Fraud and Misrepresentation; Count III, Conspiracy to
Misrepresent and Commit Fraud; Count IV, Breach of Implied Warranty of Merchantability and
Fitness; Count V, Intentional Infliction of Emotional Distress; Count VI, Negligence; Count VII,
Equitable Relief; and Count VII, Breach of Express Warranty. At some point prior to the
73
 Case: 13-14590 Date Filed: 05/18/2017 Page: 74 of 284

The evidence introduced during the trial of Phase II-A was sufficient to

prove that the plaintiffs had become addicted to the defendants’ cigarettes and that

smoking those cigarettes caused the plaintiffs’ disease, cancer. That was all the

plaintiffs had to show to prevail on their claims of strict liability, the Omnibus

Order indicated, because the evidence introduced during the trial of Phase I

established that both before and after July 1, 1974, the defendants had “placed

cigarettes on the market that were defective and unreasonably dangerous.”42 The

conclusion of the Phase II proceedings, the Court dismissed Count VII “under the heading
Medical Monitoring.”
41
It was not until it entered the Omnibus Order that the Court passed on the question of
whether the Phase I findings had sufficient evidentiary support to withstand a motion for directed
verdict.
42
Addressing Count I, strict liability, the Court said,

There was more than sufficient evidence at trial to satisfy the legal requirements
of this Count and to support the jury verdict that cigarettes manufactured and
placed on the market by the defendants were defective in many ways including
the fact that the cigarettes contained many carcinogens, nitrosamines, and other
deleterious compounds such as carbon monoxide. . . . The evidence more than
sufficiently proved that nicotine is an addictive substance which when combined
with other deleterious properties, made the cigarette unreasonably dangerous.
The evidence also showed some cigarettes were manufactured with the breathing
air holes in the filter being too close to the lips so that they were covered by the
smoker thereby increasing the amount of the deleterious effect of smoking the
cigarette. There was also evidence at trial that some filters being test marketed
utilize glass fibers that could produce disease and deleterious effects if inhaled by
a smoker. In addition, there was adequate evidence that all three of the class
members whose claims were tried in Phase II–A smoked one or more brands
manufactured by one of more of the defendants.

74
 Case: 13-14590 Date Filed: 05/18/2017 Page: 75 of 284

plaintiffs also prevailed on their claims of negligence because, the Omnibus Order

indicated, the defendants “failed to exercise the degree of care which a reasonable

cigarette manufacturer would exercise under like circumstances.”43 The plaintiffs

were relieved of the burden of proving that specific defects in the defendants’

cigarettes or specific tortious conduct caused their injuries. Instead, plaintiffs were

required to prove only that smoking the defendants’ cigarettes caused their

injuries.

In sum, the Court held that the evidence was sufficient to support the jury’s finding that
“the Defendant Tobacco Companies place[d] cigarettes on the market that were defective and
unreasonably dangerous” during certain date ranges.
43
Addressing Count VI, negligence, the Court said,
The verdict of the jury on the issue of Negligence is well supported by the
evidence. The elements of negligence have certainly been sufficiently proven by
the testimony in this case in that any reasonable person or entity, armed with the
information the defendants had, should have done that which a reasonable person
would have done under like circumstances, or should not have done what a
reasonable person would not do under like circumstances[]. It is obvious that a
reasonable person or entity would not have allowed a condition to exist that he or
it knew would injure someone, without taking appropriate measures to prevent it.
The defendants according to the testimony, well knew from their own research,
that cigarettes were harmful to health and were carcinogenic and addictive. By
allowing the sale and distribution of said product under those circumstances
without taking reasonable measures to prevent injury, constitutes, in this Courts
[sic] opinion, and in the opinion of the jury as it turns out, [] negligence.

In other words, the Court held that the evidence was sufficient to support the jury’s
finding that “the Defendant Tobacco Companies failed to exercise the degree of care which a
reasonable cigarette manufactuer would exercise under like circumstances” during certain date
ranges.

75
 Case: 13-14590 Date Filed: 05/18/2017 Page: 76 of 284

The Court had previously forecast that it would ease the plaintiffs’ burden of

proof in this way in a colloquy with Philip Morris’ counsel during closing

arguments in Phase II-A. The Court said,

Okay. Number One, cigarettes cause a disease. We know there is a
causal effect between cigarettes.
If you put the product out and people smoke it, and they get
disease, that is a causal effect. The jury has already made that
determination.
The question is whether you did it. You did. The jury found
you put these things on the market, somebody smoked it, and they got
sick. That is strict liability. You are liable. That’s what the jury
indicates from Phase I.

The Court upheld the jury’s punitive-damages award because “[i]n Phase I

of the trial, the jury, having heard the testimony concerning the behavior and

conduct of the defendants, decided that punitive damages were indeed appropriate

in this case.”44 The Court found that the $145 billion award was not unreasonable

because “the amount of the jury verdict is within the parameters of the evidence at

trial—within the limits of the highs and lows, albeit on the high side, but when the

enormity of the facts and issues of this case are considered, the award cannot be

said to be unreasonable.”

The Court made one further reference to Phase I. “[I]t should be noted that

the jury in . . . Phase I . . . found each of the defendants Guilty as to all counts with

44
The Court denied each of the defendants’ challenges to the punitive-damages award in
a lengthy discussion. It also denied their motion to decertify the plaintiff class.
76
 Case: 13-14590 Date Filed: 05/18/2017 Page: 77 of 284

the exception of count 7 for Equitable relief which the court dismissed previously

under the plaintiffs request for Medical Monitoring.”

C. Appeal to the Third District Court of Appeal in Engle II

The defendants appealed the Omnibus Order to the Third District Court of

Appeal. They argued that plaintiffs’ counsel’s race-based incendiary remarks

throughout trial merited the judgment’s complete reversal. They argued

alternatively that the punitive damages should be set aside as foreclosed by Florida

precedent and that the class should be decertified because the Phase I findings were

useless. The findings of tobacco-company misconduct were “generalized”; hence,

the defendants contended, the “Phase III juries [would be] unable to determine

whether the conduct found to be wrongful in Phase I was the legal cause of any

Phase III claimant’s injury.”45 Thus, the defendants warned, they would be faced

with “an infinite re-examination of issues by different juries and the consequent

risk of inconsistent verdicts, in violation of [their] constitutional right to have

one—and only one—jury decide the same or interrelated issues.”

On May 21, 2003, the Third District, persuaded by the defendants’

arguments, held that “the entire judgment must be reversed and the class

decertified.” Engle II, 853 So. 2d at 470. The Court began its opinion by noting

45
Although the Phase II-A jury was the same as the Phase I jury, the trial plan dictated
that the Phase III juries, in individual class-member cases, would decide the factual issues.
77
 Case: 13-14590 Date Filed: 05/18/2017 Page: 78 of 284

that “[a]lthough the emotional appeal of the class representatives’ claims is

compelling, our job as appellate judges is not to be swayed by emotion where to do

so results in violating established legal principles.” Id. at 442. The Court found

that the plaintiffs had “incit[ed] juror prejudice against an unpopular industry,”

concocted ostensibly “common” issues only by “creat[ing] a composite plaintiff

who smoked every single brand of cigarettes, saw every single advertisement, read

every single piece of paper that the tobacco industries ever created or distributed,

and knew about every single allegedly fraudulent act.” Id. at 467 n.48. Doing so

enabled the class “to try fifty years of alleged misconduct that they never would

have been able to introduce in an individual trial, which was untethered to any

individual plaintiff.” Id. Making matters worse moving forward, “there were no

specific findings as to any act by any defendant at any period of time.” Id. The

Court acknowledged what the defendants had been arguing—the Phase I findings

were useless.

The Court concluded that “Florida’s class action rules, substantive tort law,

and state and federal guarantees of due process and a fair trial, [all] require[d] class

decertification.” Id. at 450. In reaching this conclusion, the Court noted that

“virtually all courts that have addressed the issue have concluded that certification

of smokers’ cases is unworkable and improper.” Id. at 444 (collecting cases). This

is in large part because “issues of liability, affirmative defenses, and damages,

78
 Case: 13-14590 Date Filed: 05/18/2017 Page: 79 of 284

outweigh[] any ‘common issues’ in th[e] case.” Id. at 445. The impropriety of

class certification was especially clear in this particular case, the Court explained,

because “the jury did not determine whether defendants were liable to anyone.” Id.

at 450 (emphasis in original). And, “[a]s evidenced by the proceedings in Phase 2,

each claimant will have to prove that his or her illness not only was caused by

smoking, but was also proximately caused by defendants’ alleged misconduct.” Id.

at 446 (emphasis added). Because “each class member had unique and different

experiences that will require the litigation of substantially separate issues, class

representation is not ‘superior’ to individual suits.” Id. at 446–47.

In addition to decertifying the class, the Court vacated the punitive-damages

award on a host of independent grounds. First, the award violated “well-

established Florida precedent” by

a) improperly requiring the defendants to pay punitive damages for
theoretical injuries to hundreds of thousands of class members,
without a determination that defendants are liable for such injuries;
b) precluding the constitutionally required comparison of punitive
damages and compensatory damages; and c) eliminating the jury’s
discretion to assess punitive damages based upon the individual
class members' varying circumstances.

Id. at 450. Second, the size of the punitive-damages award was excessive under

state and federal law, noting that “the $145 billion verdict is roughly 18 times the

defendants’ proven net worth.” Id.at 457. Third, as explained in Young v. Miami

Beach Improvement Co., 46 So. 2d 26 (Fla. 1950), the punitive award was

79
 Case: 13-14590 Date Filed: 05/18/2017 Page: 80 of 284

precluded by settlement agreements between the tobacco companies and the states,

“which expressly included claims for punitive damages.” Engle II, 845 So. 2d at

467–70.

Lastly, the Court held that “Plaintiffs’ counsel’s improper race-based

appeals for nullification caused irreparable prejudice and require reversal.” Id. at

458. “The trial was book-ended with prejudicial attorney misconduct which

incited the jury to disregard the law because the defendants are tobacco

companies.” Id. The Court explained that “Plaintiffs’ counsel began making

racially-charged arguments on the first day of trial,” and perpetuated through

closing. Id. Specifically,

Plaintiffs’ counsel . . . explicitly tied . . . racial references to appeals
for jury nullification of the law during closing argument. He set the
stage by telling the jury, “And let’s tell the truth about the law, before
we all get teary-eyed about the law. Historically, the law has been
used as an instrument of oppression and exploitation.” Plaintiff’s
counsel then juxtaposed defendants’ conduct with genocide and
slavery. Although the trial court sustained a defense objection,
plaintiffs’ counsel proceeded to tell the jury that, like slavery and the
Holocaust, there was just one “side” to whether the defendants should
continue to sell cigarettes . . . . [C]ounsel repeatedly urged the jury to
fight what he called “unjust laws” citing the civil disobedience of
Martin Luther King and Rosa Parks.

Id. at 459–60. After citing many further examples of prejudicial conduct, the

Court explained that “the improper comments of plaintiffs’ counsel further

deprived the defendants of due process and a fair trial, thus additionally requiring

reversal.” Id. at 466.
80
 Case: 13-14590 Date Filed: 05/18/2017 Page: 81 of 284

The Court ultimately summarized its holding thus: “The fate of an entire

industry and of close to a million Florida residents, cannot rest upon such a

fundamentally unfair proceeding.” Id. at 470.

D. Petition for Review to the Florida Supreme Court in Engle III

The plaintiffs petitioned the Florida Supreme Court for review under Article

V, Section 3(b)(3) of the Florida Constitution, which grants the Court jurisdiction

to “review any decision of a district court of appeal that . . . expressly and directly

conflicts with a decision . . . of the supreme court on the same question of law.”

The defendants opposed the Court’s exercise of such jurisdiction by arguing that

Engle II did not in fact come in “express and direct conflict” with any such

precedent. The Supreme Court rejected the defendants’ arguments and accepted

jurisdiction based on a conflict between Engle II and its decision in Young v.

Miami Beach Improvement Co. as to whether a settlement agreement between the

state and the defendants would bind private citizens in their punitive-damages

claims. Engle III, 945 So. 2d at 1254, 1260. Following the Supreme Court’s

acceptance of jurisdiction, the parties jointly briefed the issues the Third District

resolved in reaching its decision in Engle II.46

46
These issues were whether it was error for the Third District to reverse the Circuit
Court’s final judgment based on the plaintiffs’ counsel’s conduct or, alternatively, reverse the
judgments for the three class representatives and the punitive-damages award, and decertify the
class.
81
 Case: 13-14590 Date Filed: 05/18/2017 Page: 82 of 284

On July 6, 2006, a divided Supreme Court issued its decision.47 Resolving

the issues the parties had briefed, the Court quashed the Third District’s judgment

as to each of its holdings except for its rejection of the punitive-damages award.48

Engle III, 945 So. 2d at 1254–56. However, in its rejection of the holdings, the

Court actually echoed many of the Third District’s criticisms regarding the

unmanageably expansive class action litigation. Id. at 1267–71. Most notably, the

Court “agree[d] with the Third District that problems with the three-phase trial

plan” required the class to be decertified. Id. at 1267–68.

To the Florida Supreme Court, however, decertification would not serve as

an acceptable outcome for the class members who had been standing idly by while

their attorneys tried Phases I and II of their case. To thus accommodate such class

members, the Supreme Court sua sponte fashioned a “pragmatic solution” in which

it preserved some of the Phase I findings for use in the class members’ cases to

establish tobacco-company liability. Id. at 1269.

In addition to these issues, the defendants’ briefs argued issues the Third District had not
addressed: whether the Circuit Court erred in allowing the plaintiffs to prosecute claims
preempted by federal law and abused its discretion in failing to instruct the jury that it could not
punish lawful conduct.
47
This decision, Engle v. Liggett Grp., No. SC03-1856, 2006 WL 1843363 (Fla. 2006),
was later withdrawn and replaced with a new, but mostly identical, decision, Engle III, 945 So.
2d 1246.
48
With Justice Cantero recused, the Court was divided 4-2 on all issues except the
reversal of punitive damages. On that issue, two of the justices from the majority joined the two
dissenters. Engle III, 945 So. 2d at 1254–56.
82
 Case: 13-14590 Date Filed: 05/18/2017 Page: 83 of 284

The Court implemented its pragmatic solution in two steps. First, it

certified, pursuant to Florida Rule of Civil Procedure 1.220(d)(4)(A),49 a class

limited to liability issues; that is, limited to eight of the ten Phase I findings,50 the

49
The rule provides that “a claim or defense may be brought or maintained on behalf of a
class concerning particular issues.” Fla. R. Civ. P. 1.220(d)(4)(A).
50
According to the Supreme Court,
The Phase I findings were: (1) that cigarettes cause some of the diseases at issue;
(2) that nicotine is addictive; (3) that the defendants placed cigarettes on the
market that were defective and unreasonably dangerous; (4) that the defendants
made a false or misleading statement of material fact with the intention of
misleading smokers; (4)(a) that the defendants concealed or omitted material
information not otherwise known or available knowing that the material was false
or misleading or failed to disclose a material fact concerning the health effects or
addictive nature of smoking cigarettes or both; (5) that all of the defendants
agreed to misrepresent information relating to the health effects of cigarettes or
the addictive nature of cigarettes with the intention that smokers and the public
would rely on this information to their detriment; (5)(a) that the defendants agreed
to conceal or omit information regarding the health effects of cigarettes or their
addictive nature with the intention that smokers and the public would rely on this
information to their detriment; (6) that all of the defendants sold or supplied
cigarettes that were defective; (7) that all of the defendants sold or supplied
cigarettes that at the time of the sale or supply did not conform to representations
of fact made by the defendants; (8) that all of the defendants were negligent; (9)
that all of the defendants engaged in extreme and outrageous conduct or with
reckless disregard relating to cigarettes sold or supplied to Florida smokers with
the intent to inflict severe emotional distress; and (10) that all of the defendants'
conduct rose to a level that would permit an award of punitive damages.

Engle III, 945 So. 2d at 1257 n.4. “A majority of Court” held that findings (1), (2), (3),
4(a), 5(a), (6), (7), and (8) “in favor of the Engle class can stand.” Id. at 1254–55.
In listing the findings, the Court omitted to state when, according to some of the findings,
the acts they depict occurred. The acts in findings (3), (6) and (7) occurred both before and after
July 1, 1974; those in findings (4) and (4)(a) before and after May 5, 1982, and those in (8)
before and after July 1, 1969. In applying findings (3), (4)(a), and (6) through (8), the courts in
the cases the class members brought cited and relied on the findings as listed in this footnote,
without regard to the designated time frames. That is, they applied those findings as if the

83
 Case: 13-14590 Date Filed: 05/18/2017 Page: 84 of 284

findings on “Questions 1 (general causation), 2 (addiction of cigarettes), 3 (strict

liability), 4(a) (fraud by concealment), 5(a) (civil-conspiracy-concealment), 6

(breach of implied warranty), 7 (breach of express warranty), and 8

(negligence).”51 Engle III, 945 So. 2d at 1255, 1268. It labeled these findings

“common core findings.” Id. at 1269. The Court did not certify the issue of legal

causation—whether the Engle defendants’ tortious conduct caused the class

members’ harm—noting that it was “highly individualized and [did] not lend

[itself] to class action treatment.” Id. at 1254. The Court acknowledged that “no

Florida cases address whether it is appropriate under rule 1.220(d)(4)(A) to certify

class treatment for only limited liability issues.” Id. at 1268. It nevertheless

certified the issues class because “several decisions by federal appellate courts

applying a similar provision in the Federal Rules of Civil Procedure provide

persuasive authority for this approach.” Id.

Under step two of its pragmatic solution, the Court declared that these

“common core findings . . . will have res judicata effect” in the subsequent

“damages actions” the class members would bring. Id. at 1269. The Phase I

defendants committed the designated acts at all times after July 1, 1969, (8), July 1, 1974, (3),
(6) and (7), or May 5,1982 (4), and (4)(a).
51
It declined to certify the issues “on the fraud and intentional infliction of emotion
distress claims, which involved highly individualized determinations, and the finding on
entitlement to punitive damages questions, which was premature.” Engle III, 945 So. 2d at 1269.
The new issues class served as a substitute for the class the Court decertified.
84
 Case: 13-14590 Date Filed: 05/18/2017 Page: 85 of 284

findings, which, as the Third District observed, were decided with reference to a

“composite plaintiff who smoked every single brand of cigarettes, saw every single

advertisement, read every single piece of paper that the tobacco industry ever

created or distributed, and knew about every single allegedly fraudulent act,”

Engle II, 853 So. 2d at 467 n.48, would now have the legal effect of a partial final

judgment resolving issues for individual class members.52 The Court’s pragmatic

solution was therefore intended to enable class members—in suing an Engle

defendant on claims of strict liability, negligence, breach of express warranty,

breach of implied warranty, fraud, and conspiracy to commit fraud—to plead the

Phase I findings to conclusively establish elements of their claims and thereby

foreclose the defendant from denying such elements.53 Although the Phase I jury

52
In a dissent joined by Justice Bell, Justice Wells recognized that the decision to grant
“res judicata effect” to these Phase I findings was, to put it mildly, problematic. As Justice Wells
presciently observed,
In what I conclude will be harmful and confusing precedent, the majority saves
some of the jury findings in Phase I of the class action before decertifying the class.
I do not join in doing that; rather, I would follow the overwhelming majority of
courts and hold that this was not a proper class action. The result of the majority
“retaining the jury’s Phase I findings” is not, as the majority asserts, “pragmatic”;
rather, it is problematic. Under the majority’s holding, the class closed a decade
ago. Who are the individuals that are to get the use of these “findings”? How will
a trial court make that determination? . . . How are these findings to be used in
cases in which the findings are used? . . . These are only a few of the issues which
arise in application of the majority’s holding.
Engle III, 945 So. 2d at 1284 (Wells, J., dissenting) (citation omitted).
53
The class members understood that this was the Florida Supreme Court’s intent. Their
complaints cited the Engle III opinion and its holdings and pled the Phase I findings verbatim, as
85
 Case: 13-14590 Date Filed: 05/18/2017 Page: 86 of 284

found none of the defendants foreclosed of their defenses,54 id. at 450, the Florida

Supreme Court did. All that remained for progeny plaintiffs to prove, and for

progeny juries to consider, was “individualized issues such as legal causation,

comparative fault, and damages.” Engle III, 945 So. 2d at 1268.

After ruling on these two matters without providing the parties notice or

opportunity to be heard on them,55 the Supreme Court remanded the case to the

if they were filing suit to domesticate a foreign judgment. E.g., Amended Complaint, Brown v.
R.J. Reynolds Tobacco Co. (Brown I), 576 F. Supp. 2d 1328 (M.D. Fla. 2008) (No. 3:07-cv-
00761).
54
Indeed, the Phase I jury was instructed not to “determine any issues regarding the
conduct of individual class members of the Florida class, including any issues as to
compensatory damages for individual class members.”
55
The Florida Supreme Court has the power to request supplemental briefs on any issue
“where confusion or doubt remains.” In re Order of First Dist. Ct. App. Regarding Brief Filed in
Forrester v. State, 556 So. 2d 1114, 1116 (Fla. 1990). The Court regularly employs this tool,
and it should have been especially inclined to do so in a case involving hundreds of thousands of
plaintiffs, an entire industry, and the potential for billions of dollars in compensatory and
punitive damages. The question, then, is why did the Engle III Court not do so?
Consider what would have happened if the Court had requested briefing. Imagine what
the parties’ responses would have been if they were to asked to comment on whether the Court
should (1) certify a class of “limited liability issues” pursuant to Florida Rule of Civil Procedure
1.220(d)(4)(A); (2) retain the Phase I findings (with the exception of findings 4, 5 and 9); and (3)
order the courts that would be handling the progeny cases to give the Phase I findings res
judicata effect. The plaintiffs, sensing that a majority of the justices were seeking a way to
enhance the class members’ chances of recovery against the Engle defendants, would have
responded affirmatively. The defendants, realizing that the questions had no bearing on the
matters before the Court and sensing that a majority of the justices wanted to stack the deck
against them, would have responded in the negative.
Surely, before briefing the first question, the defendants would be wondering why the
Court, acting as though it were a trial court, was contemplating the certification of an issues class
after the litigation ended. The defendants would contemplate asking the Court for clarification.
Why certify the class? What issues might be certified?
The second question would leave the defendants bewildered, since the Phase I findings
were not before the Court at all. The defendants had not challenged the findings in appealing the
86
 Case: 13-14590 Date Filed: 05/18/2017 Page: 87 of 284

Third District “with directions that the class should be decertified without

prejudice to the class members filing individual claims within one year of the

issuance of our mandate in this case with res judicata effect given to certain Phase I

findings,” and “for further proceedings consistent with [its] opinion.” Engle III,

945 So. 2d at 1254, 1277.

On August 7, 2006, the tobacco companies moved the Supreme Court for

rehearing. Their motion contended that the Court’s certification of an issues class

under Rule 1.220(d)(4)(A) and its pronouncement that the Phase I jury findings

would “have res judicata effect” in the cases brought by class members denied

trial judge’s Omnibus Order to the Third District. And the Third District had not passed sua
sponte on the legal status of the findings in deciding Engle II. That aside, why would the Court
“retain” useless jury findings? Would the retention of the findings, vague or irrelevant facts,
amount to an affirmance of the findings on appeal, following a review of the Phase I jury
instructions, the jury’s answers to the special interrogatories, and the jury’s verdict? The
defendants would likely ask the Court for clarification.
The third question would have informed the defendants that in declaring that the findings
“will have res judicata effect” in future progeny cases, the Court’s majority were so intent on
stacking the deck that they were willing, in this case only, (1) to disregard the Court’s well-
established precedent that bars a rendering court from determining the res judicata effect of its
own decisions and (2) to enjoin the progeny courts, in case after case, from obeying their federal
constitutional duty to examine the Engle litigation to determine whether the defendants were
afforded basic common-law protections against the arbitrary deprivation of property. In taking
these steps, the majority would be risking the Court’s integrity and, worse yet, they would be
inducing the lower courts to risk their integrity as well. Why would the majority do all of that?
The defendants would very carefully weigh their response to the third question. The
Florida precedent that bars a rendering court from declaring the preclusive effect of its own
decisions is so strong and time-honored that calling that precedent to the Court’s attention,
reminding the Court of the policies underpinning the precedent, and, in particular, the
recognizing court’s constitutional duty, would require the wisdom of Solomon and then some.
In sum, had the Court abided by its procedures, and provided the parties with the
requisite notice and opportunity to be heard, the Engle III opinion would have never been
written.
87
 Case: 13-14590 Date Filed: 05/18/2017 Page: 88 of 284

them due process in that the Court provided them with no notice that it was

contemplating such action and no opportunity to be heard. The denial of due

process aside, the tobacco companies contended that the Court erred in certifying

the issues class. The “basic principle of class-action law throughout the country . .

. [is] that certification—under any subdivision of the rules—must be addressed and

determined before there is a trial on the merits.” The companies’ final contention

was that the Phase I jury findings relating to the claims of strict liability and

negligence, among others, could not be given “res judicata effect” because the

findings were too generalized to provide a basis for individual causation consistent

with due process.

The Florida Supreme Court withdrew its July 6, 2006, opinion, Engle v.

Liggett Grp., No. SC03-1856, 2006 WL 1843363 (Fla. 2006), and on December

21, 2006, published Engle III as a substitute. Engle III made minor modifications

to the withdrawn opinion, but none are pertinent here. That same day, the Court

summarily denied the tobacco companies’ motion for rehearing in an order it chose

not to publish. The order instructed the companies not to file another motion for

rehearing.56 The Engle defendants petitioned the U.S. Supreme Court for certiorari

relief, but their petition was denied. R.J. Reynolds Tobacco Co. v. Engle, 552 U.S.

941, 128 S. Ct. 96, 169 L. Ed. 2d 244 (2007).

56
“No further motions for rehearing will be entertained by this Court.”
88
 Case: 13-14590 Date Filed: 05/18/2017 Page: 89 of 284

II.
WHAT “RES JUDICATA” TRADITIONALLY MEANS

In this appeal, RJR and Philip Morris challenge a judgment in favor of Earl

Graham, as personal representative of the estate of his deceased wife, Faye

Graham, on claims of strict liability and negligence. Ante at 17. Under traditional

Florida tort law, a plaintiff alleging strict liability in the products-liability context

must prove inter alia (a) that the product in question was defective57 and (b) that

the “defect caused the injury or harm alleged.” Aubin v. Union Carbide Corp., 177

So. 3d 489, 513 (Fla. 2015). Similarly, under traditional Florida tort law, a

plaintiff alleging negligence must prove inter alia (a) that the defendant breached a

duty of care owed to her and (b) that the defendant’s breach caused her harm.

Williams v. Davis, 974 So. 2d 1052, 1056 (Fla. 2007). In the wake of Engle III’s

res judicata dicta,58 these traditional requirements have gone by the wayside in

Engle-progeny cases.

In this case, for example, the District Court held the defendants liable even

though Mr. Graham never proved that his late wife’s injury was caused by the

57
Under traditional Florida tort law, “a product may be defective by virtue of a design
defect, a manufacturing defect, or an inadequate warning.” Jennings v. BIC Corp., 181 F.3d
1250, 1255 (11th Cir. 1999) (citing West v. Caterpillar Tractor Co., 336 So. 2d 80, 86 (Fla.
1976). A “design defect . . . [is] a defect which renders the product unreasonably dangerous.”
Liggett Grp., Inc. v. Davis, 973 So. 2d 467, 475 (Fla. 4th Dist. Ct. App. 2007).
58
See infra notes 77, 89, and accompanying text.
89
 Case: 13-14590 Date Filed: 05/18/2017 Page: 90 of 284

defendants’ product defect(s) or negligent conduct. Instead, the Court allowed Mr.

Graham to take advantage of state-law conclusive presumptions—which did not

exist when the parties litigated Phase I and apply only in Engle-progeny cases—

under which “injury as a result of the Engle defendants’ conduct is assumed.”

Douglas III, 110 So. 3d at 429.

The conclusive presumptions on which Engle-progeny plaintiffs rely

effectively transform the Phase I findings from “useless,” Douglas III, 110 So. 3d

at 433, to dispositive. For example, the Phase I finding that each defendant

“place[d] cigarettes on the market that were defective and unreasonably

dangerous” now establishes as a matter of law that (a) every cigarette smoked by

every class plaintiff was defective and unreasonably dangerous59 and (b) such

(unidentified) unreasonably dangerous defect(s) caused every class member’s

injury, including Ms. Graham’s.60 Similarly, the Phase I finding that each

defendant “failed to exercise the degree of care which a reasonable manufacturer

would exercise under like circumstances” now establishes under state law that (a)

59
This finding only necessarily establishes that each defendant produced at least one
defective and unreasonably dangerous cigarette. That the Engle jury also answered “yes” with
respect to the conduct element of fraud and intentional infliction of emotional distress—claims
“which involve[] highly individualized determinations,” Engle III, 945 So. 2d at 1269—shows
that the jury felt empowered to make findings that would be narrowly applicable to only some
class members. At the Phase I trial, plaintiffs’ counsel acknowledged that “[i]t’s a fallacy that
every common issue has to apply to one hundred percent of the class members.”
60
Phase I jurors were explicitly instructed not to “determine any issues regarding the
conduct of individual members of the Florida class.” Moreover, under the original Engle trial
plan, “individual causation” was to be determined by new juries in Phase III because it was
“highly individualized.” Engle III, 945 So. 2d at 1254.
90
 Case: 13-14590 Date Filed: 05/18/2017 Page: 91 of 284

the Engle defendants breached their duty of care to every class plaintiff61 and (b)

their (unidentified) breach(es) caused every class member’s injury, including Ms.

Graham’s.

That Engle III’s dicta62 regarding the res judicata effect of the Phase I

findings could so drastically alter the Phase I findings and Florida’s preclusion

doctrines and tort law is startling. Even more alarming is that progeny courts,

including the Majority today, have consistently failed to address the resulting

constitutional violations.63 In this dissent, I lay bare these violations, which have

been carried forward and incrementally exacerbated for twenty years.

So far, I have traced the relevant procedural history preceding this case

through Engle III. Below, I continue the narrative by detailing layer upon layer of

judicial error committed by numerous state and federal courts, culminating finally

with the Majority’s errors today. To illuminate that narrative, I pause to explain

some fundamental principles of common and constitutional law that progeny

61
The finding only necessarily establishes that each defendant breached its duty to some
class members at least once over a fifty-year period. See supra note 59. Some of the theories of
breach on which jurors were instructed pertained only to class members who smoked low-tar
cigarettes. See supra note 21 and accompanying text. Another theory related to “youth
marketing,” which clearly cannot serve as the basis of a breach of duty owed to adults. Another
rested on “minority marketing,” which clearly cannot serve as the basis of a breach of duty owed
to those in the majority. Finally, yet another theory of breach rested on marketing aimed at
women and other discrete portions of the population, which clearly cannot serve as the basis of a
breach of duty owed to males.
62
See infra notes 77, 89, and accompanying text.
63
See infra Part VII.
91
 Case: 13-14590 Date Filed: 05/18/2017 Page: 92 of 284

courts have either failed to understand or chosen to ignore. Specifically, I provide

an overview of preclusion law and explain the U.S. Constitution’s role in its

effective operation. I then explain how progeny courts have interpreted Engle III’s

“res judicata” dicta as a mandate to disregard traditional preclusion law, tort law,

and the Constitution; an invitation that many progeny courts have accepted.

A. Res Judicata 101: The Elements of Issue and Claim Preclusion

The term “res judicata” refers to all the ways in which the judgment of one

court will have a binding effect in a subsequent case. Res judicata, Black’s Law

Dictionary 1425 (9th ed. 2009). This definition is the most common, but “lumps

under a single name two quite different effects of judgments.”64 Id. The first—

“issue preclusion” or “collateral estoppel”—is the effect of foreclosing relitigation

of matters that have been litigated and decided. Id. The second—“claim

preclusion,” “merger,” or “bar”—is the effect of foreclosing any litigation of

64
A court’s “limit[ing] the res judicata phrase so as to exclude the doctrines of issue
preclusion or collateral estoppel. . . . is potentially confusing, and it is better to use res judicata in
its broader sense to encompass both sets of doctrine.” 18 Charles Alan Wright & Arthur R.
Miller, Federal Practice and Procedure § 4402 (3d ed. 2016).
92
 Case: 13-14590 Date Filed: 05/18/2017 Page: 93 of 284

matters that have never been litigated because they should have been advanced in

an earlier suit.65 Id.

Both issue preclusion and claim preclusion operate across a two-lawsuit

continuum.66 First, parties litigate a dispute to a final judgment on the merits.

Second, in a later, separate suit between the parties, one party brings to court

evidence of an earlier judgment and contends that issue or claim preclusion should

apply to prevent her opponent from litigating a previously decided issue67 or cause

of action.68 In this two-lawsuit scheme, the first court is the “rendering” court and

the second is the “recognizing” court. In this subpart, I elaborate on the elements

of each doctrine.

Issue preclusion, as developed in the common law, “bars relitigation of an

issue of fact or law that has been decided in a prior suit.” Baloco v. Drummond

65
Considering that “[t]he preclusive effects of former adjudication are discussed in
varying and, at times, seemingly conflicting terminology,” Migra v. Warren City Sch. Dist. Bd.
of Ed., 465 U.S. 75, 77 n.1, 104 S. Ct. 892, 894 n.1, 79 L .Ed. 2d 56 (1984), to make matters
easier, I will refer to effect one exclusively as “issue preclusion,” and effect two as “claim
preclusion.”
66
The two-lawsuit nature of claim and issue preclusion distinguishes these doctrines from
the law-of-the-case doctrine: whereas the former apply only when a “new and different” suit is
involved, the latter operates within a single proceeding. See, e.g., Florida Dep’t of Transp. v.
Juliano, 801 So. 2d 101, 105 (Fla. 2001) (providing a typical explanation of the distinction).
67
“Issue” is defined as “a single, certain, and material point arising out of the allegations
and contentions of the parties; it is matter affirmed on one side and denied on the other.” Issue,
Black’s Law Dictionary 907 (9th ed. 2009) (citation omitted).
68
A cause of action is defined as “a situation or state of facts that entitles a party to
maintain an action in a judicial tribunal.” Cause of action, Black’s Law Dictionary 251 (9th ed.
2009) (citing Edwin E. Bryant, The Law of Pleading Under the Codes of Civil Procedure 170
(2d ed. 1899)).
93
 Case: 13-14590 Date Filed: 05/18/2017 Page: 94 of 284

Co., 767 F.3d 1229, 1251 (11th Cir. 2014). Drawing from its common-law roots,

the doctrine only applies when

(1) the issue at stake is identical to the one involved in the prior
litigation; (2) the issue was actually litigated in the prior suit; (3) the
determination of the issue in the prior suit was a necessary part of the
judgment in that action; and (4) the parties are the same or in privity
with each other and the party against whom the earlier decision is
asserted had a full and fair opportunity to litigate the issue in the
earlier proceeding.

Id. Although some states articulate these elements differently, the core

requirements are largely the same across all jurisdictions.

In Florida, the elements are set forth in a five-prong test. For issue

preclusion to apply there must be (1) identical parties,69 (2) identical issue(s), (3)

full litigation of the particular matter, (4) determination of the particular matter,

and (5) a “final decision” in the prior proceeding by a court of competent

jurisdiction. Dadeland Depot, Inc. v. St. Paul Fire and Marine Ins. Co., 945 So.

2d 1216, 1235 (Fla. 2006) (quoting Dep’t of Health & Rehab. Servs. v. B.J.M., 656

So. 2d 906, 910 (Fla. 1995)).

Elements (2), (3), and (4) of the Florida doctrine culminate in an “actually

decided” requirement, which is fundamental to issue preclusion. The requirement

69
With respect to this element at least, the Florida Supreme Court has shown a greater-
than-normal reticence to depart from traditional common law. Florida Bar v. Clement, 662 So.
2d 690, 697–98 (Fla. 1995) (per curiam) (citation omitted) (“Although federal courts and some
other jurisdictions no longer require mutuality of parties . . . Florida courts have held that [issue
preclusion] can be asserted only when the identical issue has been litigated between the same
parties.”).
94
 Case: 13-14590 Date Filed: 05/18/2017 Page: 95 of 284

originated with early English authorities, which explained that preclusion requires

a determination “directly upon point”; recognizing courts could not preclude

parties from litigating issues on the basis that such issues might have been or

probably were decided. The Duchess of Kingston’s Case, 20 Howell’s State Trials

538 (House of Lords 1776). Rather, courts could estop litigation only when the

“estoppell” was “certaine to every intent, and not . . . taken by argument or

inference.” 2 Coke, The First Part of the Institutes of the Laws of England; Or, A

Commentary on Littleton ¶352a (1817).

This early English common-law requirement is now deeply ingrained in the

American judicial system. Federal and state issue-preclusion doctrines have

included the requirement for well over a century. See, e.g., Cromwell v. County of

Sacramento, 94 U.S. 351, 353, 24 L. Ed. 195 (1876) (“[T]he inquiry must always

be as to the point or question actually litigated and determined in the original

action, not what might have been thus litigated and determined.” (emphasis

added)); Burlen v. Shannon, 99 Mass. 200, 203 (1868) (noting that “according to

all the well considered authorities, ancient and modern,” the inference that an issue

was decided by prior litigation had to “be inevitable, or it [could not] be drawn”).

And, to this day, federal and state courts uniformly adhere to it.70 Florida is no

70
E.g., SEC v. Monarch Funding Corp., 192 F.3d 295, 309 (2d Cir.1999); United States
v. Rigas, 605 F.3d 194, 217–19 (3d Cir. 2010); Haywood v. Ball, 634 F.2d 740, 743 (4th Cir.
1980); United States v. Patterson, 827 F.2d 184, 187–90 (7th Cir. 1987); Kelly v. Armstrong, 141
95
 Case: 13-14590 Date Filed: 05/18/2017 Page: 96 of 284

exception. See Brown v. R.J. Reynolds Tobacco Co. (Brown II), 611 F.3d 1324,

1334 (“Florida courts have enforced the ‘actually adjudicated’ requirement with

rigor.” (citation omitted)).

The universality of the actually decided requirement is no accident; the

requirement helps facilitate due process. When a rendering court decides an issue

and a recognizing court later accords that issue preclusive effect, two consequences

result: First, the precluded party is gagged from litigating that issue.

Fayerweather v. Ritch, 195 U.S. 276, 307, 25 S. Ct. 58, 68, 49 L. Ed. 193 (1904).

Second, the parties are bound to the rendering court’s decision with respect to that

issue. Id. at 299, 25 S. Ct. at 64. A litigant is therefore susceptible to being denied

her due process right of having an opportunity to be heard on each issue of her

case, duPont v. Southern, 771 F.2d 874, 880 (5th Cir. 1985), unless the recognizing

court, before giving preclusive effect to an issue determination, first identifies with

F.3d 799, 801–02 (8th Cir. 1998); Chew v. Gates, 27 F.3d 1432, 1438 (9th Cir. 1994); Dodge v.
Cotter Corp., 203 F.3d 1190, 1198–99 (10th Cir. 2000); Lary v. Ansari, 817 F.2d 1521, 1524–25
(11th Cir. 1987); Moody v. Rambo, 727 So. 2d 116, 118 (Ala. Civ. App. 1998); JeToCo Corp. v.
Hailey Sales Co., 596 S.W.2d 703, 706–07 (Ark. 1980); Brake v. Beech Aircraft Corp., 229 Cal.
Rptr. 336, 343 (Cal. Ct. App. 1986); Dowling v. Finley Assocs., 727 A.2d 1245, 1251–53 (Conn.
1999); Major v. Inner City Prop. Mgmt., Inc., 653 A.2d 379, 382–83 (D.C. 1995); Herzog v.
Lexington Twp., 657 N.E.2d 926, 931 (Ill. 1995); Conn. Indem. Co. v. Bowman, 652 N.E.2d 880,
883 (Ind. Ct. App. 1995); Day v. Crowley, 172 N.E.2d 251, 254 (Mass. 1961); People v. Gates,
452 N.W.2d 627, 631–32 (Mich. 1990); Parker v. MVBA Harvestore Sys., 491 N.W.2d 904, 906
(Minn. Ct. App. 1992); In re Breuer’s Income Tax, 190 S.W.2d 248, 250 (Mo. 1945); Manard v.
Hardware Mut. Cas. Co., 207 N.Y.S.2d 807, 809 (App. Div. 1960); Buckeye Union Ins. Co. v.
New England Ins. Co., 720 N.E.2d 495, 501 (Ohio 1999); Nealis v. Baird, 996 P.2d 438, 458–59
(Okla. 1999); Lee v. U.S. Fid. & Guar. Co., 538 P.2d 359, 361 (Or. 1975).
96
 Case: 13-14590 Date Filed: 05/18/2017 Page: 97 of 284

specificity what the rendering court allegedly decided and determines it was,

indeed, actually decided.

Though similar to issue preclusion in some respects, claim preclusion is a

distinct doctrine carrying its own elements. Unlike issue preclusion, which can be

asserted offensively or defensively, claim preclusion is an affirmative defense.71

Fed. R. Civ. P. 8(c)(1); Fla. R. Civ. P. 1.110. To invoke claim preclusion, a

defendant must prove in a recognizing court that the plaintiff’s cause of action was

adjudicated on the merits in a previous case involving the same parties.72 Fla.

Dep’t of Transp. v. Juliano, 801 So. 2d 101, 105 (Fla. 2001). Thus, under both

Florida and federal law, claim preclusion carries four elements: (1) “a final

judgment on the merits”; (2) a “decision . . . rendered by a court of competent

jurisdiction”; (3) “the same cause of action . . . involved in both cases”; and (4)

“the parties, or those in privity with them, are identical in both suits.”73 Baloco v.

Drummond Co., 767 F.3d 1229, 1246 (11th Cir. 2014).

71
“[T]he obvious purpose” of claim preclusion is to force the plaintiff to present all of his
grounds of recovery in a single action. Manning v. Grimsley, 643 F.2d 20, 24 (1st Cir. 1981).
72
Florida’s common-law doctrine of claim preclusion remained remarkably stable for
much of the state’s history prior to Engle III and Douglas III. Compare Yulee v. Canova, 11 Fla.
9, 29 (Fla. 1864) with Florida Bar v. Rodriguez, 959 So. 2d 150, 158 (Fla. 2007) (showing that
core elements of claim preclusion remained unchanged in Florida across a period of nearly 150
years).
73
Under Florida law, these elements have not changed in non-Engle-progeny cases in the
years following Engle III and Douglas III. See Seminole Tribe of Fla. v. State, 202 So. 3d 971,
973 (Fla. 1st Dist. Ct. App. 2016) (setting forth the same four claim-preclusion elements).
97
 Case: 13-14590 Date Filed: 05/18/2017 Page: 98 of 284

Like issue preclusion’s actually decided requirement, elements (1) and (3) of

claim preclusion are ubiquitous and deeply ingrained because they help protect

parties’ due process rights.74 Element (1), the final-judgment requirement,75 has

long been a “cardinal rule” in Florida and all other traditional common-law

jurisdictions. Douglas III, 110 So. 3d at 438 (Canady, J., dissenting) (quoting

Juliano, 801 So. 2d at 105) (citing Kimbrell v. Paige, 448 So. 2d 1009, 1012 (Fla.

1984)). The requirement is important because a defendant who successfully

invokes claim preclusion bars a plaintiff from litigating a previously adjudicated

cause of action, both as to “issues that were raised . . . [and] issues that could have

been raised but were not raised in the first case.”76 Juliano, 801 So. 2d at 105.

Barring a cause of action that was never fully litigated to a final judgment unjustly

“blockades [an] unexplored path[] that may lead to the truth.” Brown v. Felsen,

442 U.S. 127, 132, 99 S. Ct. 2205, 2210, 60 L. Ed. 2d 767 (1979).

74
The other elements have due process implications as well, but those elements are not as
relevant in this case.
75
In the context of claim preclusion, the Florida Supreme Court has endorsed as “most
comprehensive,” a definition of “[a] judgment on the merits” as “one based on the legal rights
and liabilities of the parties.” Allie v. Ionata, 503 So. 2d 1237, 1241 (Fla. 1987) (emphasis
added) (citing 46 Am. Jur. 2d Judgments § 74 (1964)). Further, for a judgment to be final it must
“leav[e] nothing more to be done in the cause except execution.” Id. at 1240.
76
For example, a defendant might successfully assert claim preclusion to prevent a
plaintiff from relitigating a car collision on an intentional-torts theory when the plaintiff
previously won a negligence suit arising from that collision. See 18 Wright, supra, § 4408 (“A
single injury gives a single cause of action.” (citation omitted)).
98
 Case: 13-14590 Date Filed: 05/18/2017 Page: 99 of 284

Element (3) of claim preclusion, the same-cause-of-action requirement, has

similar constitutional significance. Litigants enjoy a “due process right to fully and

fairly litigate each issue in their case.” duPont, 771 F.2d at 874; see also Bell v.

Burson, 402 U.S. 535, 542, 91 S. Ct. 1586, 1591, 29 L. Ed 90 (1971) (“It is a

proposition which hardly seems to need explication that a hearing which excludes

consideration of an element essential to the decision . . . does not meet [the

requirements of the Due Process Clause].”). Claim preclusion—which bars

litigation both as to issues that were and were not litigated in a prior case, Juliano,

801 So. 2d at 105—stands in tension with this due process right. The doctrine is

reconciled with due process by means of the same-cause-of-action requirement,

which functions to “bar[] only those claims that could have been raised in the prior

litigation.” Griswold v. City of Hillsborough, 598 F.3d 1289, 1293 (11th Cir.

2010) (emphasis added); see also Dennard v. State, No. SC15-300, 2016 WL

1252516, at *2 (Fla. Mar. 30, 2016) (explaining that res judicata only extends to

“claims that could have been raised in the prior action” (emphasis in original)

(quotation marks and citation omitted)).

B. Res Judicata 102: Procedures to Invoke Issue and Claim Preclusion

When applied properly, issue and claim preclusion facilitate the worthy aim

of efficiency: “By ‘preclud[ing] parties from contesting matters that they have had

99
 Case: 13-14590 Date Filed: 05/18/2017 Page: 100 of 284

a full and fair opportunity to litigate,’ these two doctrines protect against ‘the

expense and vexation attending multiple lawsuits, conserv[e] judicial resources,

and foste[r] reliance on judicial action by minimizing the possibility of inconsistent

decisions.’” Taylor v. Sturgell, 553 U.S. 880, 892, 128 S. Ct. 2161, 2171, 171 L.

Ed. 2d 155 (2008) (alterations in original) (quoting Montana v. United States, 440

U.S. 147, 153–154, 99 S. Ct. 970, 59 L. Ed. 2d 210 (1979)). The doctrines,

however, carry the risk of depriving litigants of their property without ever

affording them an opportunity to be heard on a central element of their case.

Hence, recognizing courts should apply the doctrines “only after careful inquiry.”

Felsen, 442 U.S. at 132, 99 S. Ct. at 2210. “[I]n properly seeking to deny a litigant

two days in court, [recognizing] courts must be careful not to deprive him of one.”

Criales v. Am. Airlines, Inc., 105 F.3d 93, 97 (2d Cir. 1997).

Recognizing courts therefore strictly abide by certain common-law

procedures designed to help protect the integrity of their proceedings and litigants’

due process rights. Such procedures are so ubiquitous and rudimentary that

litigants and courts have had little, if any, reason to test their boundaries. See

Honda Motor Co., Ltd. v. Oberg, 512 U.S. 415, 430, 114 S. Ct. 2331, 2340, 129 L.

Ed. 2d 336 (1994) (“Because the basic procedural protections of the common law

have been regarded as so fundamental, very few cases have arisen in which a party

has complained of their denial.”). The rare court that does deviate from, or

100
 Case: 13-14590 Date Filed: 05/18/2017 Page: 101 of 284

abrogate, such procedures risks violating litigants’ due process rights. See Douglas

III, 110 So. 3d at 430–31 (“[E]liminating the basic common law protections

against an arbitrary deprivation of property violates due process.” (citing Oberg,

512 U.S at 432, 114 S. Ct. at 2341)). I detail some of these procedures in a

hypothetical.

A lawsuit is tried to a jury in a rendering court on claims and defenses

framed by the plaintiff’s complaint and the defendant’s answer. After receiving

the jury’s verdict, the court enters a final judgment for the plaintiff. In doing so,

the rendering court does not declare or predict whether, and if so to what extent, a

recognizing court will give preclusive effect to its judgment, that is, to any of the

claims or defenses or to any of the issues that were litigated. To do so would result

in mere dicta, because those determinations are within the recognizing court’s sole

purview.77

77
Prior to Engle III and Douglas III, the Florida Supreme Court had, for more than a
century, consistently implemented the common-law principle that the recognizing court decides
for itself whether claim or issue preclusion should apply. In its 1896 decision of Little v. Barlow,
for example, the Florida Supreme Court held that a defendant asserting a “res judicata” defense
needed to produce “the complete record [of] the former suit” to allow the recognizing court to
evaluate the defense’s merits. 20 So. 240, 241 (Fla. 1896) (emphasis added). Since then, the
Court has repeatedly reiterated this recognizing-court-decides principle. See, e.g., Prall v. Prall,
50 So. 867, 870 (Fla. 1909) (requiring the party asserting issue preclusion to establish, “with
sufficient certainty,” its elements in the second lawsuit); Rodriguez, 959 So. 2d at 159 (finding
that res judicata does not apply because it—acting as the recognizing court—determined that
“the current case . . . is based on a different cause of action” from the first case); Gordon v.
Gordon, 59 So. 2d 40, 44–45 (Fla. Div. A 1952) (explaining that the recognizing court evaluates
whether claim or issue preclusion should apply); Bagwell v. Bagwell, 14 So. 2d 841, 843 (Fla.
Div. A 1943) (requiring the party “claim[ing] the benefit of the former judgment” to produce
evidence in the second lawsuit of “the matter formerly adjudicated”). In fact, so basic is this
101
 Case: 13-14590 Date Filed: 05/18/2017 Page: 102 of 284

Later, the plaintiff sues the defendant78 in a Title VII action in a different

court, a recognizing court. Her complaint alleges several discrete acts of conduct

severe or pervasive enough to create a hostile work environment. The defendant

denies each allegation. The plaintiff, invoking issue preclusion, then moves the

court to strike the defendant’s denial of two of the acts on the ground that they

were adjudicated in her favor in the previous lawsuit. The defendant opposes the

motion, so the court requires the plaintiff—the party with the burden of proof79 —

principle that it extends beyond Florida and is viewed as a general rule of common law in the
United States. E.g., Matsushita Elec. Indus. Co. v. Epstein, 516 U.S. 367, 396, 116 S. Ct. 873,
888, 134 L. Ed. 2d 6 (1996) (Ginsburg, J., concurring in part and dissenting in part) (citation
omitted) (“A court conducting an action cannot predetermine the res judicata effect of the
judgment; that effect can be tested only in a subsequent action.”); 18 Wright, supra, § 4405
(“The first court does not get to dictate to other courts the preclusion consequences of its own
judgment.”); Herbert Newberg & Alba Conte, Newberg on Class Actions, § 16:24 (4th ed. 2002)
(“[T]he potential impact of a class court judgment is not a matter for determination by the
deciding court. The res judicata effect of a class judgment can only be determined by a later
court in light of a specific controversy.”).
78
Recall that under Florida’s preclusion doctrines, a recognizing-court lawsuit
necessarily involves litigants who are the same as, or privy to, those who litigated in the
rendering court. See Dadeland Depot, 945 So. 2d at 1235 (listing identical parties as one of the
elements of issue preclusion); Juliano, 801 So. 2d at 105 (listing identical parties as one of the
elements of claim preclusion).
79
The party claiming preclusion bears the burden of proving its elements. 18 Wright,
supra, § 4405 (“[T]he burden of establishing preclusion is placed on the party claiming it.”).
The Florida Supreme Court made this clear more than a hundred years ago in Prall: “If there is
any uncertainty as to the matter formerly adjudicated, the burden of showing it with sufficient
certainty . . . is upon the party who claims the benefit of the former judgment.” 50 So. at 870;
see also Bagwell, 14 So. 2d at 843 (“The burden of proof to establish a former adjudication, by
law, was cast on the defendant below.”). At least until Douglas, modern Florida courts were still
consistently hearkening to this common-sense principle. See, e.g., Campbell v. State, 906 So. 2d
293, 295 (Fla. 2d Dist. Ct. App. 2004) (“The party claiming the benefit of res judicata has the
burden of establishing with sufficient certainty, by the record or by extrinsic evidence, that the
matter was formerly adjudicated”); State St. Bank & Trust Co. v. Badra, 765 So. 2d 251, 253
(Fla. 4th Dist. Ct. App. 2000) (“[T]o establish res judicata . . . the party claiming the benefit of
102
 Case: 13-14590 Date Filed: 05/18/2017 Page: 103 of 284

to present the portions of the previous lawsuit’s record that establish the

adjudication of the issues. The plaintiff responds by introducing from that record

the complaint and answer, the jury instructions, the jury’s verdict, and the final

judgment.

Upon receiving the plaintiff’s evidence, the court decides whether to grant

her motion to strike. First, the court determines whether the plaintiff has

established the elements of issue preclusion under the rendering state’s laws.80

Because every state has a presumption against preclusion, recognizing courts must

not apply preclusion if any doubt exists that the elements of preclusion have been

satisfied. Issue preclusion’s actually decided requirement, for example, is

stringent: If a rendering court’s jury instructions leave “it open to the jury to find

for the defendant upon either of . . . two [or more] propositions, and the verdict

does not specify upon which the jury acted, there can be no certainty that they

the former adjudication has the burden of establishing, with sufficient certainty by the record or
by extrinsic evidence, that the matter was formerly adjudicated.”); Meyers v. Shore Inds. Inc.,
597 So. 2d 345, 346 (Fla. 2d Dist. Ct. App. 1992) (“The party asserting the defense of estoppel
by judgment has the burden of demonstrating with sufficient certainty through the record or
extrinsic evidence that the issue was adjudicated fully.”).
80
“[The full faith and credit statute] requires a federal court to look first to state
preclusion law in determining the preclusive effects of a state court judgment.” Marrese v. Am.
Acad. of Orthopaedic Surgeons, 470 U.S. 373, 381, 105 S. Ct. 1327, 1332, 84 L. Ed. 2d 274
(1985) (citations omitted). Similarly, the Full Faith and Credit Clause requires recognizing state
courts to determine the preclusive effect of a judgment in accordance with the rendering state’s
preclusion law. Kremer v. Chem. Const. Corp., 456 U.S. 461, 466, 102 S. Ct. 1883, 1889, 72 L.
Ed. 2d 262 (1982).
103
 Case: 13-14590 Date Filed: 05/18/2017 Page: 104 of 284

found upon one rather than the other,” and preclusion is inappropriate.81 De Sollar

v. Hanscome 158 U.S. 216, 222, 15 S. Ct. 816, 818, 39 L. Ed. 956 (1895). In other

words, if the jury in the previous case could have returned a verdict for the plaintiff

without deciding whether the two acts at issue actually occurred, the recognizing

court could not grant the motion to strike.82

If, on the other hand, the recognizing court concludes that the plaintiff has

met her burden, and preclusion is appropriate under the rendering state’s laws, the

81
We have observed that Florida courts abide a similarly stringent actually
decided requirement:
[P]reclusive effect is not given to issues which could have, but may not have,
been decided in an earlier lawsuit between the parties. See, e.g., Acadia Partners,
L.P. v. Tompkins, 673 So. 2d 487, 488–89 (Fla. 5th DCA 1996) (holding that
jury’s verdict “for [the defendant]” in a breach of contract action did not establish
the absence of breach because the jury was instructed that it could find for the
defendant if it concluded that the defendant had not breached the contract or if the
defendant proved an affirmative defense); Allstate Ins. Co. v. A.D.H., Inc., 397
So. 2d 928, 929–30 (Fla. 3d DCA 1981) (concluding that subcontractor could not
show that general contractor was at fault and therefore not entitled to
indemnification based on jury’s “undifferentiated general verdict finding [the
general contractor] ‘negligent’” in an earlier lawsuit; the jury could have
determined that the general contractor was at fault or vicariously liable);
Seaboard, 260 So. 2d at 864–65 (finding that general verdict “in favor of the
defendant” could have been based on jury’s conclusion that the defendant was not
negligent or that the plaintiff was contributorily negligent); see id. at 865 (“[I]t is
impossible to ascertain with any reasonable degree of certainty as to what issue
was adjudicated in the former suit except to say that the jury found in favor of [the
defendant]. Such uncertainty as to the effect of the prior adjudication renders the
doctrine of collateral estoppel inapplicable.”).
Brown II, 611 F.3d at 1334.
82
So stringent is the actually decided requirement that a recognizing court cannot apply
preclusion where the record does not reveal the theory on which the jury rendered its decision,
even if the plaintiff produces sworn affidavits from all the jury members to establish that they
based their determination on a particular theory. See Washington, A. & G. Steam Packet Co. v.
Sickles, 72 U.S. 580, 593, 18 L. Ed. 550 (1866) (“[T]he secret deliberations of the jury, or
grounds of their proceedings while engaged in making up their verdict, are not competent or
admissible evidence of the issues or finding.”).
104
 Case: 13-14590 Date Filed: 05/18/2017 Page: 105 of 284

court will grant the plaintiff’s motion unless the defendant objects further. If the

defendant objects on due process grounds, the recognizing court must ensure that

applying the rendering state’s preclusion law will not violate the defendant’s due

process rights. 83 See Hansberry v. Lee, 311 U.S. 32, 40, 61 S. Ct. 115, 117, 85 L.

Ed. 22 (1940) (“[When a due process objection is raised] it becomes the duty of

[the recognizing court] to examine the course of procedures in both litigations to

ascertain whether the litigant whose rights have thus been adjudicated has been

afforded . . . due process.”); Douglas III, 110 So. 3d at 430–31 (expressing the

same principle); Adams v. State Farm Bureau Life Ins. Co., 493 F.3d 1276, 1285

(11th Cir. 2007) (“[W]e have stated that res judicata can only be applied to an

action if it is first shown that doing so would be consistent with due process.”

(citing Twigg v. Sears & Roebuck & Co., 153 F.3d 1222, 1226 (11th Cir.1998))).

In conducting its due process inquiry, the recognizing court must determine (a)

whether the determination in the rendering court was made with adequate notice

and opportunity to be heard, (b) whether state preclusion law contains adequate

83
Many cases make this point plain. See, e.g., Kremer, 456 U.S. at 482–83, 102 S. Ct. at
1898 (“A State may not grant preclusive effect in its own courts to a constitutionally infirm
judgment, and other state and federal courts are not required to accord full faith and credit to
such a judgment. Section 1738 does not suggest otherwise.”). If this point seems obvious, it is
because it is. The duty of courts to refrain from applying laws so as to violate the Constitution
has long been understood. See Marbury v. Madison, 5 U.S. 137, 177–80 (1803) (“If two laws
conflict with each other, the courts must decide on the operation of each. . . . [A] law repugnant
to the constitution is void.”).
105
 Case: 13-14590 Date Filed: 05/18/2017 Page: 106 of 284

safeguards to ensure that courts do not arbitrarily deprive litigants of property,84

and (c) whether such safeguards were, in fact, applied.

To conduct its inquiry appropriately, the recognizing court must “look past

the linguistic label[s] employed by the [rendering court]” and conduct a

meaningful review.85 Davila v. Delta Air Lines, Inc., 326 F.3d 1183, 1189 (11th

Cir. 2003); see also Gooch v. Life Inv’rs Ins. Co. of Am., 672 F.3d 402, 420–21

(6th Cir. 2012) (“[In conducting this inquiry] it is incumbent upon us to apply the

same scrutiny to state-court judgments that the Supreme Court would apply.”);

Criales v. Am. Airlines, Inc., 105 F.3d 93, 97 (2d Cir. 1997) (“[W]e would not

permit the choice of labels to distort substance, especially where the consequence

would be so drastic as to deprive a party of the opportunity to be heard.”). If its

due process inquiry so warrants,86 the recognizing court then grants the plaintiff’s

motion to strike.

84
As the Florida Supreme Court has correctly noted, “eliminating the basic common law
protections against an arbitrary deprivation of property violates due process.” Douglas III, 110
So. 3d at 430–31 (citing Oberg, 512 U.S. at 432, 114 S. Ct. at 2341).
85
On appeal, an appellate court reviews “‘de novo a district court’s determination of res
judicata or collateral estoppel.’” Vasquez v. YII Shipping Co., 692 F.3d 1192, 1196 (11th Cir.
2012) (quoting EEOC v. Pemco Aeroplex, Inc., 383 F.3d 1280, 1285 (11th Cir.2004)).
86
The Sixth Circuit articulated well the importance of recognizing courts’ due process
inquiry: “Even though reconsidering whether the class judgment complied with the due process
clause may not promote judicial ‘efficiency’ or protect the ‘finality’ of the original judgment, it
is a due-process imperative that we are not free to ignore.” Gooch v. Life Inv’rs Ins. Co. of Am.,
672 F.3d 402, 420 (6th Cir. 2012) (citation omitted).
106
 Case: 13-14590 Date Filed: 05/18/2017 Page: 107 of 284

Given the essential inquiries for which a recognizing court is responsible, a

rendering court cannot “predetermine the res judicata effect of [its] judgment.”87

Matsushita Elec. Indus. Co. v. Epstein, 516 U.S. 367, 396, 116 S. Ct. 873, 888, 134

L. Ed. 2d 6 (1996) (Ginsburg, J., concurring in part and dissenting in part) (citation

omitted). This is so even if the rendering court, like the Supreme Court in Engle

III, is convinced that its proceedings were constitutionally sound.

III.
ENGLE III INSTRUCTED PROGENY COURTS TO DISREGARD
TRADITIONAL RES JUDICATA LAW SO AS TO HOLD THE DEFENDANTS
LIABLE WITHOUT REGARD TO THE PHASE I FINDINGS

When Engle III accepted jurisdiction under Article V, Section 3(b)(3) of the

Florida Constitution, it assumed the role of an appellate rendering court, reviewing

the proceedings in Engle I and Engle II for certain issues that were “properly

briefed[,]argued and [ ] dispositive of the case.” Murray v. Regier, 872 So. 2d 217,

225 n.5 (Fla. 2002) (citing Savona v. Prudential Ins. Co. of America, 648 So. 2d

705, 707 (Fla.1995)). The Florida Supreme Court fulfilled this role by considering

the briefed issues and quashing much of the Third District’s judgment.

When Engle III retroactively certified an issues class limited to eight of the

ten of Phase I findings, and declared that those “findings . . . will have res judicata

effect” in future “damages actions” to be brought by individual members of the

87
See supra note 77 and accompanying text.
107
 Case: 13-14590 Date Filed: 05/18/2017 Page: 108 of 284

decertified class, it usurped the role of a recognizing court.88 Engle III, 945 So. 2d

at 1269. Recognizing progeny courts could have disregarded Engle III’s res

judicata instruction as mere dicta,89 and some did.90 Many others, however, in

deference to the state’s highest court, interpreted the instruction as a binding

mandate. See, e.g., R.J. Reynolds Tobacco Co. v. Martin (Martin II), 53 So. 3d

1060, 1066–67 (Fla 1st Dist. Ct. App. 2010) (interpreting the Florida Supreme

Court’s res judicata instruction as a mandate that “district courts of appeal do not

have the prerogative to overrule”); Jimmie Lee Brown II, 70 So. 3d at 715 (“We are

constrained by the Florida Supreme Court’s decision in Engle III.”). As shown

below, recognizing courts that interpreted the instruction as a mandate treated

preclusion as a foregone conclusion, thereby abandoning their recognizing-court

duties and putting their integrity at risk while sparing progeny plaintiffs their

burden of proving the elements of preclusion.

88
In retroactively certifying an issues class under Florida Rules of Civil Procedure
1.220(d)(4)(A) pertaining to Phase I findings that were not before the Court, Engle III also
relieved the plaintiffs of their burden of proving that the prerequisites of class certification had
been met, disregarded its procedural rules which require certification to be left to the discretion
of the trial court, and exceeded the scope of its lawful jurisdiction under the state constitution.
89
The Brown I and Brown II Courts performed their recognizing-court duties
notwithstanding Engle III’s res judicata instruction. See infra note 77 and accompanying text.
90
See supra note 117 and accompanying text.
108
 Case: 13-14590 Date Filed: 05/18/2017 Page: 109 of 284

In accordance with mutuality requirements under Florida preclusion law,91

we are not bound by previous recognizing-court determinations.92 Nevertheless, I

review such cases to demonstrate how progeny courts have incrementally grown

ever-more absurd in their reasoning, ever-more disingenuous in their portrayal of

facts, and ever-more cavalier in their abrogation of due process. The Majority’s

opinion is best understood in the context of the steady downslide that preceded it.

A. The U.S. District Court for the Middle District of Florida in Brown I
Rejected the Florida Supreme Court’s Interference with Its Duties as a
Recognizing Court

Within the one-year limitations period Engle III provided, 9,000 class

members—smokers and personal representatives of deceased smokers— filed suit

against the Engle defendants in state and federal court, the “Engle-progeny

cases.”93 Approximately 4,000 members brought suit in the Circuit Court of Duval

91
As explained above in Part II.A, Florida preclusion law—both issue and claim
preclusion—does not allow parties to successfully assert preclusion unless the recognizing-court
litigants are identical to the rendering-court litigants. In accordance with Engle III, due to the
“highly individualized” issues being litigated, the plaintiffs differ in each progeny case. 945 So.
2d at 1263.
92
That is, of course, unless such determinations represent a change in Florida preclusion
law. But Florida courts have not explicitly indicated a change in preclusion or tort law that
applies only to Engle-progeny litigants. If they had, this appeal would entail different, but
equally serious, constitutional questions.
93
In most situations, several members joined together as plaintiffs.
109
 Case: 13-14590 Date Filed: 05/18/2017 Page: 110 of 284

County, Florida.94 The tobacco companies, invoking the Class Action Fairness Act

of 2005, Pub. L. No. 109-2,119 Stat. 4 (codified in scattered sections of 28 U.S.C.),

successfully removed the cases to the U.S. District Court for the Middle District of

Florida.95 Graham v. R.J. Reynolds Tobacco Co. (Graham I), No. 3:09-cv-13603-

MMH-JBT (M.D. Fla. May 28, 2013) was one of them.

After the cases removed to the Middle District of Florida were assembled,96

the tobacco companies moved the District Court in one of the cases, Brown v. R.J.

Reynolds Tobacco Co. (Brown I), 576 F. Supp. 2d 1328 (M.D. Fla. 2008),97 to

decide the preclusive effect, if any, of the Phase I findings based on Engle III’s

declaration that “the Phase I common core findings [it] approved will have res

judicata effect” in the progeny cases. 98 Engle III, 945 So. 2d at 1269. The District

Court granted the motion.

94
Brown I, 576 F. Supp. 2d at 1334. In January 2008, the parties moved the Judicial
Panel on Multi-District Litigation to consolidate and transfer all of the Engle-progeny cases to
the Middle District of Florida. The motion was denied. All of the progeny complaints asserted
the claims Engle III approved.
95
The plaintiffs contested the removals, but were unsuccessful. Cooper v. R.J. Reynolds
Tobacco Co., 586 F. Supp. 2d 1312, 1315 (M.D. Fla. 2008).
96
Many of the cases had been filed in the District Court, instead of state court.
97
Brown I had been brought by several class members.
98
Throughout the litigation of the progeny cases, the parties and the courts, focusing on
Engle III’s use of res judicata, have discussed the Engle III “res judicata effect” declaration in
preclusion language. In doing so, they have honored form over substance. If the Florida
Supreme Court had issued an opinion expressly holding that the Phase I findings were such that a
class plaintiff could recover damages against an Engle defendant merely by alleging and proving
addiction to the defendant’s cigarettes, the Court would not have used words res judicata.
Rather, the Court would have entered a judgment for the class plaintiffs on their Engle III-
110
 Case: 13-14590 Date Filed: 05/18/2017 Page: 111 of 284

The preclusion issue was framed by Brown I’s amended complaint99 and the

defendants’ answers.100 I quote parts of these pleadings because they set the stage

for, and were integral to, the District Court’s decision.

The amended complaint was materially identical to the complaints filed in

the other Engle-progeny cases in that all asserted the same Engle III-approved tort

claims and sought compensatory and punitive damages. None of the complaints

specified the brand(s) of the defendants’ cigarettes the plaintiff smoked, how the

defendants’ tortious conduct caused the plaintiff’s injuries, or even what the

tortious conduct was in the first place. The facts on which a specific tort claim

rested consisted of a citation to the Engle III decision and the Phase I findings.

Amended Complaint at 1, 5, 12–14, Brown I, 576 F. Supp. 2d 1328 (No. 3:07-cv-

00761).

I begin with the pertinent allegations of the complaint and then move to the

defendants’ answers.

approved tort claims, provided that a plaintiff would have to prove addiction to the defendant’s
cigarettes in order to prevail. Nevertheless, courts have treated Engle III as doing the former.
99
The complaint was filed by twenty plaintiffs, all personal representatives of deceased
smokers. Amended Complaint at 1, Brown I, 576 F. Supp. 2d 1328 (No. 3:07-cv-00761).
100
The defendants were the tobacco companies sued in Engle. Answer, Defenses and
Jury Demand at 1–2, Brown I, 576 F. Supp. 2d 1328 (No. 3:07-cv-00761).
111
 Case: 13-14590 Date Filed: 05/18/2017 Page: 112 of 284

AMENDED COMPLAINT

Plaintiffs, as Personal Representatives of the Estates of
Decedents, hereby sue the Defendants as follows:

INTRODUCTION AND GENERAL ALLEGATIONS

1. This is a complaint against the Defendants seeking compensatory
and punitive damages in accordance with the Florida Supreme Court’s
class action decision and mandate in Engle v. Liggett Group, Inc., 945
So.2d 1246 (Fla. 2006). In approving the Engle Phase I class
certification and trial, but ordering post Phase I class decertification,
the Florida Supreme Court provided this opportunity to complete
unresolved individual damages claims. The Court held: “that it was
proper to allow the jury to make findings in Phase I on Questions 1
(general causation), 2 (addiction of cigarettes), 3 (strict liability), 4(a)
(fraud by concealment), 5(a) (civil-conspiracy-concealment), 6
(breach of implied warranty), 7 (breach of express warranty), and 8
(negligence). Therefore, these findings in favor of the Engle class can
stand.” The Court further held that specified liability and general
causation findings by the Engle jury did not need to be proved again
as they shall be given res judicata effect. Consequently, Plaintiffs
bring this action upon the limited remaining issues in dispute, to-wit:
specific causation, apportionment of damages, comparative fault,
compensatory damages, entitlement to punitive damages, and punitive
damages.

2. The Florida Supreme Court expressly reserved to class members,
including Plaintiffs and their Decedents, the right to bring individual
actions against Defendants for smoking-related injuries and damages,
including punitive damages. This action is timely because it is brought
within one (1) year of the Florida Supreme Court’s mandate in Engle.

3. Plaintiffs are the Personal Representative for the Estate of the
Decedents. Letters of Administration will be forthcoming and filed
with the Clerk of this Court. This action is brought on behalf of the
Decedent’s survivors and Estate. The potential beneficiaries of a
recovery in this action and the relationship to the Decedents follow
Fla. Stat. § 768, et seq.

112
 Case: 13-14590 Date Filed: 05/18/2017 Page: 113 of 284

...

5. The Defendants are manufacturers of cigarettes, or their
successors/predecessors are manufacturers of cigarettes, and they are
foreign corporations doing business in Florida who, at times material
to this action, designed, manufactured, advertised, marketed, and sold
tobacco products for human consumption which proximately caused
injury to Decedents.

...

12. Cigarette Products. Decedents purchased, smoked, and were
addicted to cigarette products manufactured and sold by Defendants
which were the subject of Engle. They were designed, manufactured,
advertised, marketed, and sold by the Defendants at all times material
to these claims.

13. Common Liability Findings. Plaintiffs assert the jury findings in
the Phase I Engle trial which were given res judicata effect by the
Florida Supreme Court, including but not limited to the following:

a. Smoking cigarettes causes aortic aneurysm, bladder cancer, cerebral
vascular disease, cervical cancer, chronic obstructive pulmonary
disease, coronary heart disease, esophageal cancer, kidney cancer,
laryngeal cancer, lung cancer (specifically, adenocarcinoma, large cell
carcinoma, small cell carcinoma, and squamous cell carcinoma),
complications of pregnancy, oral cavity/tongue cancer, pancreatic
cancer, peripheral vascular disease, pharyngeal cancer, and stomach
cancer.
b. Nicotine is addictive.
c. All of the Defendants placed cigarettes on the market that were
defective and unreasonably dangerous.
d. All of the Defendants concealed or omitted material information
not otherwise known or available, knowing that the material was false
or misleading, or failed to disclose a material fact concerning the
health effects or addictive nature of smoking cigarettes or both.
e. All of the Defendants agreed to conceal or omit information
regarding the health effects of cigarettes or their addictive nature with
the intention that smokers and the public would rely on this
information to their detriment.
113
 Case: 13-14590 Date Filed: 05/18/2017 Page: 114 of 284

f. All of the Defendants sold or supplied cigarettes that were efective.
g. All of the Defendants were negligent.
h. All Defendants sold or supplied cigarettes that, at the time of sale or
supply, did not conform to representations of fact made by
Defendants.

14. As a direct and proximate result of Decedents’ smoking of
Defendants’ cigarettes, Decedents suffered bodily injury and died.
Defendants’ cigarettes caused Decedents to develop one or more
cigarette-related diseases or medical conditions and one or more of
them resulted in or substantially contributed to Decedents’ death.

...

17. The threshold requirement for pleading punitive damages has been
previously met in the Engle Phase I proceeding.

COUNT I – STRICT LIABILITY

18. The Introduction and General Allegations above are re-alleged and
incorporated herein by reference.

19. As a direct and proximate result of Defendants’ defective and
unreasonably dangerous cigarettes, Decedents were injured and died.

...

COUNT II – BREACH OF EXPRESS WARRANTY

20. The Introduction and General Allegations above are re-alleged and
incorporated herein by reference.

21. As a direct and proximate result of Defendants’ breach of express
warranty, Decedents were injured and died.

...

COUNT III – BREACH OF IMPLIED WARRANTY

114
 Case: 13-14590 Date Filed: 05/18/2017 Page: 115 of 284

22. The Introduction and General Allegations above are re-alleged and
incorporated herein by reference.
23. As a direct and proximate result of Defendants’ breach of implied
warranty, Decedents were injured and died.
...

COUNT IV – CIVIL CONSPIRACY TO FRAUDULENTLY
CONCEAL

24. The Introduction and General Allegations above are re-alleged and
incorporated herein by reference.

25. As a direct and proximate result of Defendants’ conspiracy to
fraudulently deceive, Decedents were injured and died.
...

COUNT V – FRAUDULENT CONCEALMENT

26. The Introduction and General Allegations above are re-alleged and
incorporated herein by reference.

...

27. As a direct and proximate result of Defendants’ fraudulent
concealment, Decedents were injured and died.

...

COUNT VI – NEGLIGENCE

28. The Introduction and General Allegations above are re-alleged and
incorporated herein by reference.

29. As a direct and proximate result of Defendants’ negligence,
Decedents were injured and died.

Id. at 2–7.

115
 Case: 13-14590 Date Filed: 05/18/2017 Page: 116 of 284

The complaints in Brown I and the other progeny cases were pleaded

pursuant to Rule 8(a) of the Federal Rules of Civil Procedure101 and the Supreme

Court’s instructions in Ashcroft v. Iqbal, 556 U.S. 662, 129 S. Ct. 1937, 173 L. Ed.

2d. 868 (2009). Iqbal requires that

[t]o survive a motion to dismiss, a complaint must contain sufficient
factual matter . . . to “state a claim to relief that is plausible on its
face.” [Bell Atlantic Corp. v. Twombly, 550 U.S. 544,] 570, 127 S. Ct.
1955, [167 L. Ed. 2d. 929 (2007)]. A claim has facial plausibility
when the plaintiff pleads factual content that allows the court to draw
the reasonable inference that the defendant is liable for the misconduct
alleged. Id. at 556, 127 S. Ct. 1955. The plausibility standard is not
akin to a “probability requirement,” but it asks for more than a sheer
possibility that a defendant has acted unlawfully. Ibid. . . .
Threadbare recitals of the elements of a cause of action,
supported by mere conclusory statements, do not suffice. . . .
Determining whether a complaint states a plausible claim for relief
will . . . be a context-specific task that requires the reviewing court to
draw on its judicial experience and common sense. [Iqbal v. Hasty,
490 F.3d 143, 157–58 (2d Cir. 2007) rev’d, Iqbal, 556 U.S. 662, 129
S. Ct. 1937]. But where the well-pleaded facts do not permit the court
to infer more than the mere possibility of misconduct, the complaint
has alleged—but it has not “show[n]”—“that the pleader is entitled to
relief.” Fed. Rule Civ. Proc. 8(a)(2).

Id. at 678–79, 120 S. Ct. at 1949–50.

The amended complaint did not satisfy Rule 8(a) and Iqbal’s pleading

standards because it merely recited conclusory statements from the Engle III

opinion. Moreover, it failed altogether to identify the tortious conduct that caused

the plaintiffs’ injuries.

101
“A pleading that states a claim for relief must contain . . . a short and plain statement
of the claim showing that the pleader is entitled to relief.” Fed. R. Civ. P. 8(a).
116
 Case: 13-14590 Date Filed: 05/18/2017 Page: 117 of 284

In drafting their amended complaints in Brown I and other progeny cases,

progeny plaintiffs simply lifted language from the Engle III opinion as a means of

pleading res judicata offensively, using the doctrine as a substitute for alleging the

facts necessary to establish the elements of their causes of action. E.g., Amended

Complaint at 4, Brown I, 576 F. Supp. 2d 1328 (No. 3:07-cv-00761). This method

of pleading is foreign to what Rule 8 prescribes102 and helps explain the difficulty

federal district courts have encountered in adjudicating progeny cases.

Such difficulty could have been mitigated if the following procedures had

been followed: To satisfy the requirements of Rule 8 and Iqbal, plaintiffs’ counsel

should have drafted a condensed version of the Phase I complaint, one tailored to

the individual plaintiff’s claims. With respect to negligence, for example, the

plaintiffs’ complaints should have identified each defendant’s negligent conduct,

noting when it took place and explaining how it caused the plaintiff’s injury. Each

defendant, in turn, would answer the complaint and might choose to deny inter alia

(1) that it engaged in the alleged conduct, (2) that such conduct was negligent, and

(3) that such conduct caused the plaintiff’s injuries. Next, the plaintiff, invoking

res judicata, would likely respond with a motion to strike denials (1) and (2) as

being foreclosed by the Phase I findings and/or by Engle III. If a defendant then

102
Rule 8(a) provides that “a claim for relief must contain . . . a short and plain statement
of the claim showing that the pleader is entitled to relief.” Fed. R. Civ. P. 8(a). “Each allegation
must be simple, concise and direct.” Fed. R. Civ. P. 8(d).
117
 Case: 13-14590 Date Filed: 05/18/2017 Page: 118 of 284

opposed that motion, the plaintiff, having the burden of proof on the applicability

of preclusion, would introduce into evidence the part of the Engle proceedings that

established the foreclosure. The court would then examine such evidence and

rule.103

That things did not operate in this way suggests that Engle-progeny cases

have more to do with a change in substantive law than with an invocation of

traditional claim or issue preclusion. If, for example, Engle III established as a

substantive rule of tort law (1) that all cigarettes are defective, unreasonably

dangerous, and negligently made and (2) that one who smokes them can recover

damages for smoking-related injury—because it is conclusively presumed that the

defect or negligence caused the injury—then the way progeny cases have been

pleaded makes more sense. If that were the tort law,104 the class plaintiffs would

satisfy the Iqbal standard simply by alleging that they purchased a defendant’s

cigarettes, became addicted, and suffered injury as a result. That is precisely what

they have been allowed to do; they neither allege in their complaints nor proffer

evidence that the defendants wrongful conduct caused their injuries. Even under

this scenario, however, if a defendant moved to dismiss a claim under Rule

103
See supra Part II.B.
104
The Florida Supreme Court later confirmed in Douglas III that it had created such a
law in Engle III. See Douglas III, 110 So. 3d at 429 (When a plaintiff “prov[es] that addiction to
the Engle defendants’ cigarettes containing nicotine was a legal cause of the injuries alleged,”
“injury as a result of the Engle defendants’ conduct is assumed.”). As I explain later, this law is
unconstitutional. See infra note 141 and accompanying text.
118
 Case: 13-14590 Date Filed: 05/18/2017 Page: 119 of 284

12(b)(6), the court would have to determine whether Engle III actually

implemented such a change in substantive tort law—that it relieved the class

plaintiffs of the burden of proving injury causation. That determination, in turn,

would hinge on a review of the Engle proceedings.

In contrast to the plaintiffs’ deficient complaints, the Engle defendants’

answers have been pleaded in accordance with the Federal Rules of Civil

Procedure. Those answers admit or deny the plaintiffs allegations and assert

affirmative defenses. Below, I provide RJR’s answer from Brown I as a template

of a typical Engle-defendant answer. The answer begins with a Preliminary

Statement, which is followed by a response to each of the amended complaint’s

numbered paragraphs and thirty-four affirmative defenses.105

PRELIMINARY STATEMENT

...

Reynolds contends that the Florida Supreme Court’s decision
contains several errors of law and denies the Engle defendants their
due process rights.

...

First, the Florida Supreme Court invalidated certain Phase I
findings as being “nonspecific” and “inadequate to allow a subsequent
jury to consider individual questions of reliance and legal cause.” See
Engle [III], 945 So. 2d at 1246. The preserved Engle Phase I findings,

105
Those affirmative defenses included failure to state a claim for relief and federal
preemption. Answer, Defenses, and Jury Demand of Defendant R.J. Reynolds Tobacco Co. at
20–21, Brown I, 576 F. Supp. 2d 1328 (No. 3:07-cv-00761).
119
 Case: 13-14590 Date Filed: 05/18/2017 Page: 120 of 284

however, suffer from the same deficiencies. Those findings are also
so generalized and nonspecific that they are inadequate to support an
individualized determination of essential issues such as liability, legal
causation, and damages in this or any other subsequent individual
action. Nothing in the Phase I verdict identifies the misconduct
underlying the jury’s findings. Giving preclusive effect to these
findings in this or any other individual action would subject
defendants to liability for conduct that no one can determine the
Engle Phase I jury found to be tortious, thereby violating Florida law
and denying defendants due process and a fair trial. Moreover,
applying these generic findings in this or any other individual action
would mean no jury will make specific findings regarding these issues
as they relate to these Plaintiffs and/or Plaintiffs’ Decedents, thereby
depriving defendants of their Seventh Amendment right to a trial by
jury in this action. Specifically, the preserved Engle Phase I findings
are deficient for the following reasons:

 Engle Phase I findings numbers 3 (Strict Liability—that the
defendants placed cigarettes on the market that were defective
and unreasonably dangerous) and 6 (Breach of Implied
Warranty—that all of the defendants sold or supplied cigarettes
that were defective in that they were not reasonably fit for the
uses intended) are deficient because they do not identify the
product(s), defect(s), or manufacturing dates, brands, types, or
designs of cigarettes found to be defective. Accordingly, no
subsequent court or fact finder can determine whether any
product, brand, type, or design used by a particular plaintiff
was found defective (or not defective) by the Engle jury or
whether any such design characteristic found defective by the
Engle jury caused these Plaintiffs’ Decedents’ injuries or any
other plaintiff’s injury. . . .
 The Florida Supreme Court rejected Engle Phase I findings
numbers 4 (Fraud and Misrepresentation) and 5 (Civil
Conspiracy—Misrepresentation) because “fraud” was too
individualized a claim to allow the finding to be applied in
subsequent actions. Engle Phase I findings numbers 4(a)
(Fraud by Concealment—that the defendants concealed or
omitted material information not otherwise known or available
knowing that the material was false or misleading or failed to
disclose a material fact concerning the health effects or
120
 Case: 13-14590 Date Filed: 05/18/2017 Page: 121 of 284

addictive nature of smoking cigarettes) and 5(a) (Civil
Conspiracy—Concealment—that the defendants agreed to
conceal or omit information regarding the health effects of
cigarette smoking or the addictive nature of cigarette smoking
with the intention that smokers and the public would rely on
this information to their detriment) suffer from the same
deficiency. Findings 4(a) and 5(a) do not identify what
information was found to have been misrepresented or
concealed, or the date(s) that such information was
misrepresented or concealed. Therefore, no subsequent court or
fact finder can determine whether a particular plaintiff relied
upon a statement or omission found tortious by the Engle jury
or whether any statement or omission found to be tortious by
the Engle jury was a legal cause of injury to the plaintiff. . . .
 Engle Phase I finding number 7 (Breach of Express Warranty—
that all of the defendants sold or supplied cigarettes that, at the
time of sale or supply, did not conform to representations of
fact made by said defendants) is deficient because it does not
identify the specific representations of fact, what defendant
made the representations, when the representations were made,
the product(s), brands, or time of sale of the cigarettes that did
not allegedly conform to representations of fact, or how the
cigarettes did not conform to those representations as
determined by the Engle jury. Thus, no subsequent court or
fact finder can determine whether any particular plaintiff heard
any specific representations of fact or purchased cigarettes in
reliance upon those representations of fact, whether any
particular plaintiff’s cigarettes did not conform to the specific
representations of fact, or whether any breach of express
warranty as determined by the Engle jury was a legal cause of
injury to a particular plaintiff. . . .
 Engle Phase I finding number 8 (Negligence—that the
defendants failed to exercise the degree of care which a
reasonable cigarette manufacturer would exercise under like
circumstances) is deficient because it does not identify the
negligent conduct, or whether it was based on a failure to warn
or negligent design. No subsequent court or fact finder can
determine whether the acts or omissions alleged by these
Plaintiffs or any other whether any conduct found to be

121
 Case: 13-14590 Date Filed: 05/18/2017 Page: 122 of 284

negligent by the Engle jury was a legal cause of injury to a
particular plaintiff.

Second, the Engle Phase I findings cannot be given preclusive
effect in this or any other subsequent individual action because res
judicata requires a judgment on the merits that resolves claim or cause
of action. The Engle Phase I findings did not determine liability and
do not constitute a judgment that resolved any claim or cause of
action.

Third, application of the Engle Phase I findings in this or any
other subsequent individual action would violate the prohibitions set
forth in the Seventh Amendment to the United States Constitution and
Article I, § 22 of the Florida Constitution against re-examination by
one jury of issues decided by another jury because the generic and
nonspecific nature of those findings necessarily requires a subsequent
jury to reexamine the Phase I findings to determine what conduct the
Engle jury determined was tortious.

Fourth, the Engle Phase I findings cannot be used as a basis for
determining punitive damages because the Phase I findings do not
identify the conduct that the Engle jury found to be tortious or
unlawful, and due process requires that punitive damages be based
upon the wrongful conduct causing the injury to the plaintiff.

Fifth, the Florida Supreme Court retroactively changed the
basis for class certification from Rule 1.220(b)(3) to Rule
1.220(d)(4)(A), without allowing the defendants to argue the
impropriety of certifying the class under Rule 1.220(d)(4)(A), thereby
depriving defendants of their due process right to notice and an
opportunity to be heard concerning the proper procedure for having
the jury arrive at sufficiently specific Phase I findings.

Answer, Defenses, and Jury Demand of Defendant R.J. Reynolds Tobacco Co. at

2–7, Brown I, 576 F. Supp. 2d 1328 (No. 3:07-cv-00761) (emphasis added).

In deciding the preclusion issue, the District Court, sitting as a recognizing

court, and the parties drew on Florida’s res judicata doctrines, claim and issue
122
 Case: 13-14590 Date Filed: 05/18/2017 Page: 123 of 284

preclusion. In briefing the preclusion issue, plaintiffs’ counsel argued that the

Engle III Court issued four implied holdings. The first three holdings relate to the

tort claims pleaded in the class action complaint. The fourth relates to the duty of

recognizing trial courts in progeny cases.

Plaintiffs first argued that Engle III, by invoking “res judicata”—which the

plaintiffs interpreted as claim preclusion—implicitly held that the Phase I findings

established the elements of the plaintiffs’ tort claims.106 Plaintiffs Response to

Tobacco’s Rule 16(c) Motion at 4, Brown I, 576 F. Supp. 2d 1328 (No. 3:07-cv-

00761). Relatedly, the plaintiffs argued, the Engle III Court also implicitly held

that the Phase I findings foreclosed all of the tobacco companies’ defenses to the

plaintiffs’ tort claims.107 See id. (“The [Phase I] jury is . . . conclusively presumed

to have considered all issues related to the claims of defect, negligence, conspiracy

to defraud, and the other counts.”). Defendants could, of course, still contend that

a plaintiff was not a class member—because, for example, she was not addicted—

that her disease was not caused by smoking, that she was comparatively negligent,

106
Those claims included strict liability, breach of express warranty, breach of implied
warranty, conspiracy to fraudulently conceal, fraudulent concealment, and negligence.
107
It held this by relieving plaintiffs of their burden of (1) identifying the unreasonably
dangerous defect(s) and tortious conduct that caused their harm and (2) proving that such
defect(s) and tortious conduct caused their harm. With Engle III predetermining those issues,
class members had nothing left to litigate except the issues of addiction, damages, and
comparative fault, the resolution of which would turn in large part on the credibility of the
smoker’s testimony. As I detail below, the Florida Supreme Court later endorsed this
interpretation of Engle III. See Douglas III, 110 So. 3d at 429 (When a plaintiff “prov[es] that
addiction to the Engle defendants’ cigarettes containing nicotine was a legal cause of the injuries
alleged,” “injury as a result of the Engle defendants’ conduct is assumed.”).
123
 Case: 13-14590 Date Filed: 05/18/2017 Page: 124 of 284

and that her damages should be less than those demanded. Id. at 19. In sum, the

plaintiffs argued, Engle III held that progeny trials would be “‘damages’ trials”

rather than causation trials; plaintiffs had no need to prove that their injury was

caused by a defendant’s tortious conduct. Id. Instead, as plaintiffs argued Engle

III also held, the only issue of causation the plaintiffs needed to establish was “that

smoking cigarettes caused a plaintiff’s particular injury.”108 Id. (emphasis in

original).

Finally, the plaintiffs also argued that by commanding recognizing trial

courts to give the Phase I findings res judicata effect, the Engle III Court was

informing those courts that it had predetermined the preclusive effect of the Phase I

findings. See id. at 13 (“We know—from Engle itself—that Florida law permits a

verdict of this type to be given res judicata effect.” (emphasis in original)). Thus,

under Florida law, the plaintiffs contended, recognizing trial courts no longer had

any business evaluating whether the elements of preclusion—including the final-

judgment and actually decided requirements109—had been satisfied or whether due

108
Under this interpretation of Engle III, later endorsed by the Florida Supreme Court in
Douglas III, class members could have limited their allegations in their complaints to these: (1)
the plaintiff smoked the Engle defendant’s cigarettes, (2) the plaintiff became addicted, and (3)
the smoking caused a disease. The substantive law that Engle III / Douglas III created made it
unnecessary to allege one of the six Engle III-approved torts. Why? Because the Engle III /
Douglas III law empowered plaintiffs to hold defendants liable simply by proving class
membership—addiction and smoking-related injury. Proving addiction was easy—plaintiffs
merely had to present enough evidence to survive a motion for directed verdict. Thus, alleging
that the defendant had committed a specific tort law violation was mere window dressing.
109
These requirements are discussed in greater detail above. See supra Part II.A.
124
 Case: 13-14590 Date Filed: 05/18/2017 Page: 125 of 284

process had been afforded to Engle defendants in the rendering court. See id. at

12–13 (“[I]t is the law of Florida, deemed so by Florida’s highest court, that

whatever Phase I can be called or labeled, it is sufficient to be the basis of claims

preclusion.” (emphasis in original)). If due process had been denied, so be it.

Before it addressed the preclusive effect that should be afforded to the Phase

I findings, the District Court had to decide a preliminary question concerning its

jurisdiction. The plaintiffs argued that the Rooker-Feldman doctrine110 deprived

the Court of jurisdiction to “independently review[] the state court rulings.” Brown

I, 576 F. Supp. 2d at 1334. That is, the Court could not entertain the defendants’

argument that because it was “impossible to know what allegations formed the

basis of each [Phase I] finding, affording preclusive effect to the general Phase I

findings would be an arbitrary application of the common law rules of preclusion”

and a denial of due process of law. Id. at 1344. Due to the lack of jurisdiction, the

plaintiffs continued, the Court had to apply preclusion without evaluating the due

110
“The Rooker-Feldman doctrine prevents the lower federal courts from exercising
jurisdiction over cases brought by ‘state-court losers’ challenging ‘state-court judgments
rendered before the district court proceedings commenced.’” Lance v. Dennis, 546 U.S. 459,
460, 126 S. Ct. 1198, 1199, 163 L. Ed. 2d 1059 (2006) (per curiam) (quoting Exxon Mobil Corp.
v. Saudi Basic Indus. Corp., 544 U.S. 280, 284, 125 S. Ct. 1517, 1521–22, 161 L. Ed. 2d 454
(2005)).
125
 Case: 13-14590 Date Filed: 05/18/2017 Page: 126 of 284

process implications of doing so. Id. at 1334. The District Court rejected the

plaintiffs’ arguments and held Rooker-Feldman inapplicable.111

Next, the Court considered what preclusive effect it should give to the Phase

I findings. The defendants argued that it was “apparent that the Florida Supreme

Court intended that the findings function as issue preclusion (or collateral estoppel)

in subsequent proceedings,” id. at 1338, but that using the findings to establish

their liability in the instant case “would be an arbitrary application of the common

law rules of preclusion” and thus would deny them due process. Id. at 1344–45.

The plaintiffs’ response was that “the Engle findings should act as claim preclusion

(or res judicata) since the Supreme Court of Florida explicitly used the legal term

‘res judicata’ in its decision,” and the defendants had received all the process they

were due in the Engle proceedings. Id. at 1338.

The Court considered the plaintiffs’ argument “problematic.” Id. at 1339.

First, the Florida Supreme Court, as the

rendering court . . . may not decide the preclusive effect of its own
judgments. It is the duty of the second trial court—which knows both
what the earlier finding was and how it relates to a later case—to

111
The District Court recognized that the Rooker-Feldman doctrine “deprives a district
court of its subject matter jurisdiction to entertain claims that a final state court judgment violates
the federal rights of the state court loser,” but held the doctrine inapplicable. Id. at 1336. It
made that determination because Rooker-Feldman only applies where “a state court loser files a
mirror image case in federal court invoking the federal court’s federal question jurisdiction, 28
U.S.C. § 1331, asking the federal court to void or modify a state court judgment on grounds that
it is unconstitutional.” Id. at 1337.
126
 Case: 13-14590 Date Filed: 05/18/2017 Page: 127 of 284

independently determine what preclusive effect a prior judgment may
be given.112

Id. at 1339 (citations omitted). “Second, as acknowledged by the Florida Supreme

Court, the Phase I jury verdict did not establish liability as to any Defendant.” 113

112
The District Court was adamant about not allowing the Florida Supreme Court to
usurp its role as a recognizing court, devoting a long paragraph to the independence to which
recognizing courts are entitled:
Plaintiffs contend that this Court need not determine which preclusion doctrine
applies because the Florida Supreme Court's announcement that the Phase I
findings serve as “res judicata” forecloses the issue. This argument is problematic
in several respects. First, as a general proposition, the rendering court, or parallel
court system, may not decide the preclusive effect of its own judgments. It is the
duty of the second trial court—which knows both what the earlier finding was and
how it relates to a later case—to independently determine what preclusive effect a
prior judgment may be given. See Matsushita Elec. Indus. Co. v. Epstein, 516
U.S. 367, 396, 116 S. Ct. 873, 134 L. Ed. 2d 6 (1996) (Souter, J., concurring in
part, dissenting in part); Midway Motor Lodge v. Innkeepers' Telemanagement &
Equip. Corp., 54 F.3d 406, 409 (7th Cir.1995) (“In the law of preclusion . . . the
court rendering the first judgment does not get to determine that judgment's
effect; the second court is entitled to make its own decision.”); Teamsters Local
282 Pension Trust Fund v. Angelos, 762 F.2d 522, 525 (7th Cir.1985)
(Easterbrook, J.) (same); see 18 Charles Alan Wright et al., Federal Practice and
Procedure § 4413 (2d ed. 2002) (noting “general rule that a court cannot dictate
preclusion consequences at the time of deciding a first action,” except in limited
cases where it seeks to limit the decision's preclusive effect). Recognizing this
principle, Florida courts have required that parties seeking to assert either claim
preclusion or issue preclusion as a defense bear the burden of demonstrating that
the doctrine applies to the subsequent litigation. Campbell v. State, 906 So. 2d
293, 295 (Fla. 2d DCA 2004); State St. Bank & Trust Co. v. Badra, 765 So. 2d
251, 253 (Fla. 4th DCA 2000) (finding that party claiming benefit of res judicata
in second proceeding bears the burden of proving that the claim was previously
adjudicated); Meyers v. Shore Indus., Inc., 597 So.2d 345, 346 (Fla. 2d DCA
1992) (finding that party asserting collateral estoppel bears burden of
demonstrating its applicability).
Brown I, 576 F. Supp. 2d at 1339–40.
113
The Supreme Court did not explicitly acknowledge this as a holding. It merely quoted
the Third District in Engle II: “[T]he Phase I jury ‘did not determine whether the defendants
were liable to anyone.’” Engle III, 945 So. 2d at 1263 (emphasis in original) (quoting Engle II,
853 So. 2d at 450). The Third District made this comment in addressing the Phase II-B award of
punitive damages. The award was premature, the Third District said, because the tobacco
127
 Case: 13-14590 Date Filed: 05/18/2017 Page: 128 of 284

Id. at 1340. Thus, “the Phase I findings did not serve to merge the claims asserted

by Plaintiffs into an enforceable judgment” against the tobacco companies.114 Id.

Although claim preclusion was not viable due to the absence of a final

judgment, the District Court assessed whether plaintiffs could successfully assert

issue preclusion. Id. In conducting its assessment, the Court performed its

recognizing-court duties of evaluating the due process implications of preclusion

and determining whether the party asserting preclusion had carried the burden of

proving its elements.115 Id. at 1340–47. Issue preclusion, the Court determined,

could not be invoked because neither its Florida-law nor constitutional

requirements had been satisfied.

The Court found that the plaintiffs had not carried their burden of proving

issue preclusion’s actually decided element, a requirement that carries

companies had not been held liable to any of the class members except the three representative
plaintiffs. Engle II, 853 So. 2d at 452–53.
114
Claim preclusion carries a strict finality-of-judgment requirement. Restatement
(Second) of Judgments § 13 cmt. g (1982). Indeed, that “claim preclusion applies only where
there has been a prior final ‘judgment on the merits’” has long been a “cardinal rule” in Florida.
Douglas III, 110 So. 3d at 438 (Canady, J. dissenting) (quoting Juliano, 801 So. 2d at 105)
(citing Kimbrell v. Paige, 448 So. 2d 1009, 1012 (Fla. 1984)). In the context of claim preclusion,
the Florida Supreme Court has endorsed as “most comprehensive,” a definition of “[a] judgment
on the merits” as “one based on the legal rights and liabilities of the parties.” Allie v. Ionata, 503
So. 2d 1237, 1241 (Fla. 1987) (emphasis added) (citing 46 Am. Jur. 2d Judgments § 74 (1964)).
In addition, claim preclusion has traditionally required a final judgment that “leav[es] nothing
more to be done in the cause except execution.” Id. at 1240. This stringent final-judgment
requirement comports with common sense: claim preclusion “should be applied so as to give
rather than deny justice,” 18 Wright, supra, § 4415 n. 1 (citation omitted), and barring a cause of
action that was never fully litigated unjustly “blockades [an] unexplored path[] that may lead to
the truth.” Felsen, 442 U.S. at 132, 99 S. Ct. at 2210.
115
See supra Part II.B.
128
 Case: 13-14590 Date Filed: 05/18/2017 Page: 129 of 284

constitutional significance.116 According to the Court, the “jury form, and any

verdict delivered from the form” were “flaw[ed]” and “nonspecific[]” such that

“this Court ‘would have to embark on sheer speculation’ to determine what issues

were actually decided during the Phase I trial and how to apply them to the

individual claims before this Court.” Id. at 1342 (quoting Hoag v. New Jersey, 356

U.S. 464, 472, 78 S. Ct. 829, 829, 2 L. Ed. 2d 913 (1958)). The Court simply

could not determine “what acts or omission committed by what Defendant

breached what duty to which Plaintiff causing what injury.” Id. Accordingly, to

preclude defendants from litigating such issues would violate the Supreme Court’s

instruction “that courts not apply the doctrine of issue preclusion to prior

determinations unless the court ‘is certain that the precise fact was determined by

the former judgment.’” Id. at 1345 (quoting De Sollar v. Hanscome, 158 U.S. 216,

221, 15 S. Ct. 816, 39 L. Ed. 956 (1895)).

116
A party may only assert issue preclusion with respect to issues that were actually
litigated and determined in the first lawsuit. Florida Bar v. Clement, 662 So. 2d 690, 697–98
(Fla. 1995) (per curiam). Just as claim preclusion’s final-judgment requirement helps ensure that
parties have a fair opportunity to fully litigate a cause of action, issue preclusion’s actually
decided requirement ensures that parties have at least one opportunity to fully litigate each issue.
Given the actually decided requirement’s role in ensuring parties’ opportunity to litigate,
the Supreme Court has noted that requirement’s constitutional significance. As the Supreme
Court held in Fayerweather v. Ritch, 195 U.S. 276, 25 S. Ct. 58, 49 L. Ed. 193 (1904), a
recognizing court may not give preclusive effect to an issue determination unless the issue was
“distinctly put in issue . . . the parties presented their evidence, or at least had the opportunity to
present it, and . . . the question was decided” in the first suit. Id. at 299, 25 S. Ct. 58, 64
(emphasis added).
129
 Case: 13-14590 Date Filed: 05/18/2017 Page: 130 of 284

Moreover, “since it is impossible to determine the precise issues decided by

the Phase I jury . . . the traditional elements of issue preclusion—e.g., identicality,

criticality, and necessity to the prior determination—cannot be satisfied.” Id. at

1346. Accordingly, the Court concluded that it was “foreclosed from applying the

Phase I findings as establishing any part of Plaintiffs’ claims.” Id. (citations

omitted).

Because the plaintiffs were unsuccessful in invoking claim and issue

preclusion, the Engle defendants had the right to deny that their tortious conduct

caused the plaintiffs’ injuries. In its order rejecting the plaintiffs’ arguments that

Rooker-Feldman precluded it from deciding the due process issue the companies

had presented, the District Court certified that its rulings qualified for interlocutory

appeal under 28 U.S.C. § 1292(b). Id. at 1348. We agreed and granted the parties’

application to appeal. Notice of Interlocutory Appeal at 1, Brown I, 576 F. Supp.

2d 1328 (No. 3:07-cv-00761). Meanwhile, further proceedings in the Engle-

progeny cases in the Middle District were stayed pending our decision. E.g., Order

at 1–2, Waggoner v. R.J. Reynolds Tobacco Co., 835 F. Supp. 2d 1244 (2011) (No.

3:09-cv-10367).

* * *

Although Engle III sought to predetermine preclusion such that recognizing

courts would not consider whether the elements of preclusion had been satisfied or

130
 Case: 13-14590 Date Filed: 05/18/2017 Page: 131 of 284

whether applying preclusion would deny due process, the Brown I Court firmly

rejected this attempted usurpation of its recognizing-court responsibilities—“the

rendering court,” the Court affirmed, “may not decide the preclusive effect of its

own judgments.”117 Brown I, 576 F. Supp. 2d at 1339. In performing its

recognizing-court duties, the Court found that under both Florida law and the U.S.

Constitution, plaintiffs could invoke neither claim or issue preclusion.

B. In Brown II, We Upheld the District Court’s Decision as a Recognizing
Court to Apply Florida’s Traditional Issue-Preclusion Doctrine to the Phase
I Findings

On appeal, we affirmed the District Court’s rejection of the plaintiffs’

Rooker-Feldman argument for the reasons that Court gave and in light of the

Supreme Court’s recent decision in Exxon Mobil Corp. v. Saudi Basic Indus.

Corp., 544 U.S. 280, 125 S. Ct. 1517, 161 L. Ed. 2d 454 (2005).118 Brown II, 611

F.3d at 1330.

117
Nothing in the District Court’s dispositive order indicates that the Court understood
Engle III’s res judicata instruction as a revision of Florida preclusion law that would only apply
in progeny cases. Rather, the District Court applied Florida preclusion law as it stood before
Engle III. As I detail above, Florida law that delegates to recognizing courts the task of
determining the preclusive effect of prior adjudications is fixed by U.S. Supreme Court
precedent. See Hansberry v. Lee, 311 U.S. 32, 61 S. Ct. 115, 85 L. Ed. 22 (1940) (“[W]hen the
judgment of a state court, ascribing to the judgment of another court the binding force and effect
of res judicata, is challenged for want of due process it becomes the duty of [the recognizing
court] to examine the course of procedure in both litigations.”).
118
There, the Supreme Court
131
 Case: 13-14590 Date Filed: 05/18/2017 Page: 132 of 284

We also affirmed the District Court’s rejection of the plaintiffs’ argument

that claim preclusion, rather than issue preclusion, was what the Engle III Court

had in mind when it used the term “res judicata.” Although “the plaintiffs argued

before the district court and suggested in their brief to this Court that the Florida

Supreme Court was referring to claim preclusion in Engle III,” the plaintiffs, at

oral argument “clarified that their position [was] that the Phase I approved findings

are entitled to issue preclusive effect.” Id. at 1333 n.7. “[I]f the plaintiffs had

continued to argue for claim preclusion, we would have rejected that position”

because claim preclusion’s final-judgment requirement had not been satisfied. Id.

at 1332, 1333 n.7.

After the plaintiffs stipulated that claim preclusion was not viable, we

evaluated the viability of issue preclusion. “Issue preclusion,” we observed,

“operates more narrowly to prevent re-litigation of issues that have already been

decided between the parties in an earlier lawsuit.” Id. at 1332 (citations omitted).

Like the District Court before us, we recognized that Florida’s doctrine of issue

clarified the [Rooker-Feldman] doctrine and narrowed its application, noting that
“the doctrine has sometimes been construed [by lower federal courts] to extend
far beyond the contours of the Rooker and Feldman cases.” [Exxon, 544 U.S.] at
283, 125 S. Ct. at 1521. The Court held that it should be “confined to cases of the
kind from which the doctrine acquired its name: cases brought by state-court
losers complaining of injuries caused by state-court judgments rendered before
the district court proceedings commenced and inviting district court review and
rejection of those judgments.” Id. at 284, 125 S. Ct. at 1521–22.
132
 Case: 13-14590 Date Filed: 05/18/2017 Page: 133 of 284

preclusion carried an actually decided requirement.119 Id. (citing Rohan v. Trakker

Maps, Inc., 633 So. 2d 1176, 1177 (Fla. Dist. Ct. App. 1994). Unlike the District

Court, however, we did not take any position as to whether that requirement was

required under the U.S. Constitution, “because,” we assumed,120 “under Florida

law the findings could not be used” “to establish facts that were not decided by the

[Phase I] jury.” Id. at 1334.

The parties’ dispute, therefore, came down to what the Phase I jury decided.

The defendants argued that the jury decided only what it indicated on the Phase I

verdict form—“those [facts] framed by the specific factual issue set out in the

questions posed to them on the verdict form.” Id. The plaintiffs, on the other

hand, advocated a more expansive reading of the Phase I findings that relied on

“flesh[ing] out” “the jury’s answers” “using the record as a whole” and “going

outside the record.” Id. at 1335. This process of “fleshing out,” the plaintiffs

contended, would lead the Court to conclude that when the jury answered “yes” to,

for example, the verdict-form question about defendants “plac[ing] cigarettes on

the market that were defective and unreasonably dangerous,” it meant that “all

cigarettes the defendants sold were defective and unreasonably dangerous.” Id.

(emphasis added).

119
The District Court below had referred to this requirement as the “actually decided”
requirement. Brown I, 576 F. Supp. 2d at 1342.
120
We assumed that Engle III had not arbitrarily eliminated the actually decided
requirement from Florida’s issue-preclusion doctrine for purposes of Engle-progeny cases only.
133
 Case: 13-14590 Date Filed: 05/18/2017 Page: 134 of 284

Though we welcomed the plaintiffs to scour the trial record—without

looking beyond it—for proof of what the jury determined, we were skeptical that

such proof existed. Id. “[T]he plaintiffs have pointed to nothing in the record, and

there is certainly nothing in the jury findings themselves,” we observed, “to

support [the plaintiffs’] factual assertion” that the Phase I jury found that

defendants’ tortious conduct tainted all cigarettes.121 Id.

With the dispute over claim and issue preclusion resolved, and the due

process issue avoided, we remanded the case to the District Court to provide

plaintiffs an opportunity to prove,122 inter alia, that “the jury’s [unreasonably-

121
Against all odds, the Majority now claim to have discovered something we overlooked
in Brown II. They, like the Douglas III Court, invent a quote in the jury instructions, Ante at 22,
discuss the “common thrust” of the evidence, id. at 7, and conclude that the “Engle jury actually
decided common elements of the negligence and strict liability of R.J. Reynolds and Philip
Morris.” Id. at 20.
122
We made it clear that Florida law allocated the burden of proof to the plaintiffs:

Under Florida law the issue preclusion standard requires the asserting party to
show with a “reasonable degree of certainty” that the specific factual issue was
determined in its favor. The entire trial record may be considered for that
purpose, although the burden is on the asserting party to point to specific parts of
it to support its position.

Brown II, 611 F.3d at 1335 (emphasis added). The following cases will demonstrate that
the plaintiffs were never able to meet their burden of the proving that the jury actually
decided these issues. Nevertheless, the Majority relieve the plaintiffs of this burden and
purports to prove what the plaintiffs never could.
134
 Case: 13-14590 Date Filed: 05/18/2017 Page: 135 of 284

dangerous-defect finding] . . . establishes that all of the cigarettes that the

defendants sold” “were defective and unreasonably dangerous.”123 Id. at 1336.

* * *

Brown II, as a recognizing-court decision, became the Eleventh Circuit’s

controlling precedent regarding the preclusive effect of Engle III in the litigation of

progeny cases in the district courts. The Phase I findings resolved factual issues,

not causes of action. Id. at 1333 (“[F]actual issues and not causes of action were

decided in Phase I.”). Absent evidence that the Phase I jury decided more facts

than those it disclosed in its findings—which we were skeptical existed, but

welcomed plaintiffs to locate—plaintiffs could not rely upon the Phase I findings

to identify particular cigarette defect(s) and tortious conduct, let alone prove that

123
Although this question is not altogether irrelevant, it misses the most relevant point.
Even if plaintiffs could establish inter alia that all cigarettes were defective, unreasonably
dangerous, and negligently produced, they would still have to identify the unreasonably
dangerous defect(s) and negligent-and-otherwise-tortious conduct to prove that such defect(s)
and conduct caused their harm. Nevertheless, the Majority conclude that the Phase I jury
determined that the defendants acted “wrongfully toward all of the class members,” Ante at 21,
and that “all cigarettes the defendants placed on the market were defective and unreasonably
dangerous.” Id. (emphasis in original), without ever revealing what the defect or negligent
conduct is that the jury supposedly identified. Thus, with no defect or negligent conduct to point
to, it was impossible for Mr. Graham to prove that the defendant’s tortious conduct caused his
wife’s harm. Recognizing this fact, the District Court relieved Mr. Graham of this burden. The
Majority straightforwardly acknowledge that a class plaintiff need only prove that “smoking was
the proximate cause of her injury,” id. at 28, rather than proving that the defendant’s tortious
conduct was the proximate cause of her injury, as is required in all other tort cases in Florida and
the rest of the United States. Again, I have pointed out that the Supreme Court has explicitly
held that a state law such as this that establishes liability once a plaintiff merely proves that her
injury resulted from the defendant’s conduct, rather than proving that the injury results from the
defendant’s tortious conduct, is arbitrary and violates due process. See infra note 141 and
accompanying text. Because the Majority cannot explain their clear violation of precedent, they
respond with silence.
135
 Case: 13-14590 Date Filed: 05/18/2017 Page: 136 of 284

such defect(s) and conduct caused their harm. See id. at 1335 (“[T]here is certainly

nothing in the jury findings . . . to support [the plaintiffs’] factual assertion” “that

all cigarettes the defendants sold were defective and unreasonably dangerous.”).

By necessary implication, we held that a District Court judgment based solely on

the Phase I findings would deprive the defendant of its property without due

process of law.124

C. The Florida District Courts of Appeal Rejected Brown II on the Basis of
Engle III’s Instruction

Brown II was decided on July 22, 2010. After the mandate issued, the stays

were lifted in twelve “lead” Middle District of Florida cases, including Graham.125

The Court in Waggoner v. R.J. Reynolds Tobacco Co., 835 F. Supp. 2d 1244 (M.D.

Fla. 2011), was selected as the recognizing court for “determin[ing] which Engle

facts should be given preclusive effect under Florida law as the Eleventh Circuit

had outlined it” in Brown II.126 Id. at 1253. The Waggoner Court was unable to

undertake the assignment, however, until June 2011.

124
Litigants enjoy a “due process right to fully and fairly litigate each issue in their case.”
duPont, 771 F.2d at 880; see also Burson, 402 U.S. at 542, 91 S. Ct. at 1591 (“It is a proposition
which hardly seems to need explication that a hearing which excludes consideration of an
element essential to the decision . . . does not meet [the requirements of the Due Process
Clause].”).
125
The District Judges of the Middle District took this action jointly.
126
Plaintiffs bore the burden of proof on this question. Although, pursuant to Brown II,
611 F.3d at 1335, the plaintiffs could look at the entire record, they would presumably focus on
136
 Case: 13-14590 Date Filed: 05/18/2017 Page: 137 of 284

On June 2, 2011, the lawyers representing the parties in the twelve lead

cases appeared before the District Court for a Rule 16(c)127 management

conference. Id. at 1256. In the interim, between the issuance of the mandate in

Brown II and the commencement of the Rule 16(c) conference, the First District

Court of Appeal, in Martin II, 53 So. 3d 1060,128 had reached a preclusion holding

contrary to Brown II’s. Before the Rule 16(c) conference adjourned, the Fourth

District Court of Appeal, in Jimmie Lee Brown II agreed with Brown II’s analysis,

but reached the same result as the First District’s in Martin II, albeit with serious

reservations as to whether its decision would deny the tobacco companies of their

property without due process of law.

1. The Martin I Circuit Court Concluded That Engle III’s Instruction
Required It to Hold the Defendants Liable if the Plaintiff Simply
Proved Class Membership Irrespective of the Phase I Findings

The Martin case was brought on October 24, 2007, in the Circuit Court for

Escambia County, Florida. Beverly Martin sued RJR, Philip Morris USA, Inc.,

and Lorillard Tobacco to recover for the death of her husband, Benny Martin.

those aspects of the record that demonstrated what the Phase I jury decided: the Court’s jury
instructions, which informed the jury of the elements of the plaintiffs’ tort claims; the special
interrogatories, which asked the jury to answer “yes” or “no” to specific factual questions; and
the jury’s answers to the interrogatory questions.
127
See Fed. R. Civ. P. 16(c). The conference lasted until December 20, 2011.
128
The opinion in Martin issued on December 14, 2010. Rehearing was denied on
February 11, 2011.
137
 Case: 13-14590 Date Filed: 05/18/2017 Page: 138 of 284

Martin II, 53So. 3d at 1064 n.2. In her first amended complaint,129 she asserted

four of the Engle III-approved causes of action: strict liability, negligence, fraud

by concealment, and conspiracy to commit fraud by concealment. Id. at 1065. She

also sought punitive damages. Id. Her tort claims and the prayer for punitive

damages were based solely on the “Common Liability Findings” asserted in the

Brown I complaint. Amended Complaint at 4, Brown I, 576 F. Supp. 2d 1328 (No.

3:07-cv-00761). Those findings, the complaint alleged, were sufficient to

“conclusively establish” her tort claims. First Amended Complaint at ¶¶ 29–61,

Martin v. R.J. Reynolds Tobacco Co. (Martin I), No. 2007-CA-2520 (Fla. Cir. Ct.

2009), 2009 WL 6492304..

The defendants’ answers to the amended complaint raised the same due

process objection as their answers did in Brown I. Philip Morris USA Inc.’s

Answer, Martin I (No. 2007-CA-2520), 2008 WL 6722672 at *12. On August 25,

2008, their attorneys and those representing Engle defendants in the other progeny

cases pending in the Escambia County Circuit Court jointly moved the Circuit

Court, pursuant to Florida Rule of Civil Procedure 1.200,130 to determine, as a

129
The first amended complaint was filed on August 20, 2008. First Amended
Complaint, Martin v. R.J. Reynolds Tobacco Co. (Martin I), No. 2007-CA-2520 (Fla. Cir. Ct.
2009), 2009 WL 6492304.
130
Florida Rule of Civil Procedure 1.200 gives trial courts authority to schedule case
management conferences “to coordinate the process of the action if . . . complex litigation factors
. . . are present” and “determine other matters that may aid in the disposition of the action.” Fla.
R. Civ. P. 1.200 (a)(3), (a)(13).
138
 Case: 13-14590 Date Filed: 05/18/2017 Page: 139 of 284

recognizing court, what the Florida Supreme Court meant when it declared that the

Phase I findings would be given “res judicata effect” in progeny cases.

Defendants’ Rule 1.200 Motion at 1–3, In re: Engle Progeny Cases Tobacco

Litigation (Fla. Cir. Ct. Feb. 24, 2009) (No. 2008-CA-80000). The parties asserted

positions similar to those they asserted in Brown I. Id.; Order Denying

Defendants’ Rule 1.200 Motion at 1–3, In re: Engle Progeny Cases Tobacco

Litigation (No. 2008-CA-80000).

The Court ruled on the motion in an order entered on February 24, 2009.

Order Denying Defendants’ Rule 1.200 Motion, In re: Engle Progeny Cases

Tobacco Litigation (No. 2008-CA-80000). It could not say whether the Florida

Supreme Court had intended to invoke “res judicata, collateral estoppel, estoppel

by judgment, stare decisis, or some other mechanism.” Id. at 3. Whatever the

mechanism was, it was “unorthodox.” Id. at 2. The intended effect of the

mechanism, however, was clear: the Florida Supreme Court had intended to

facilitate rather than “void the class action litigation.” Id.at 3. The Phase I

findings “must [be] use[d]” even if they appeared useless. Id.

The Court’s Rule 1.200 ruling governed the trial in Martin I. All Ms. Martin

had to prove to hold RJR liable was Mr. Martin’s class membership—his addiction

139
 Case: 13-14590 Date Filed: 05/18/2017 Page: 140 of 284

to an RJR cigarette and a smoking-related injury.131 Martin II, 53 So. 3d at 1066.

Thus, the jury was not required to determine whether the cigarettes Mr. Martin

smoked were defective and unreasonably dangerous; whether RJR committed one

or more negligent acts that caused him to smoke; whether it concealed information

about the health effects or addictive nature of smoking that would have caused him

to stop smoking had he been aware of it; or whether any Engle defendant acted to

conceal such information pursuant to a conspiracy of which RJR was a member.132

131
The Majority argue, “Contrary to the dissent’s view, no tobacco company can be held
liable to any smoker without proof at trial that the smoker belongs to the Engle class, that she
smoked cigarettes manufactured by the company during the relevant class period, and that
smoking was the proximate cause of her injury.” Ante at 28 (emphasis in original) (citation
omitted). In arguing that my view differs from theirs on what must be proven, the Majority
neglect to consider the fact that the Engle class is defined as “[a]ll Florida citizens and
residents,” “and their survivors, who have suffered, presently suffer or have died from diseases
and medical conditions caused by their addiction to cigarettes that contain nicotine,” R.J.
Reynolds Tobacco Co. v. Engle, 672 So. 2d 39, 40–42 (Fla. 3d Dist. Ct. App. 1996) (emphasis
added), and thus requiring a progeny plaintiff to prove that she is a member of the class and
smoked the defendants’ cigarettes—as I explain a progeny plaintiff must prove—is precisely the
same as requiring the plaintiff to prove the three items the Majority list. The Florida Supreme
Court has recently confirmed this. See R.J. Reynolds Tobacco Co. v. Ciccone, 190 So. 3d 1028,
1031 (Fla. 2016) (“According to the framework for tobacco litigation established in Engle
[progeny litigation], [the plaintiff’s] case proceeded to a ‘Phase I’ trial, in which, if she
established Engle class membership, she would receive the benefit of res judicata effect of the
Engle jury’s ‘common core findings’ regarding the issue[] of liability.”). Additionally, the
Majority fail to recognize that their itemized list of issues a plaintiff must prove directly supports
my argument that a plaintiff need only prove that smoking was the proximate cause of her injury,
rather than proving that the defendant’s tortious conduct was the proximate cause her injury, as is
required in all other tort cases in Florida and every other state in the United States.
132
Under Martin II’s holding, the reasons that individual smokers chose to smoke are
superfluous to the determination of the tobacco companies’ liability. Many class members may
have smoked for a reason totally unrelated to the Engle defendant’s tortious conduct.
Nonetheless, Martin II’s conclusive presumption treats all class members as one, relieving all of
the burden of proving that the defendant’s tortious conduct caused their injury. It does not
matter what tort claim(s) the plaintiff chooses to assert, since each of the Engle III-approved
140
 Case: 13-14590 Date Filed: 05/18/2017 Page: 141 of 284

Id. at 1064–66; Jury Instructions, Martin I (No. 2007-CA-2520), 2009 WL

2599305. The jury found RJR liable on all four clams and assessed Ms. Martin’s

damages at $5 million. Martin II, 53 So. 3d at 1066. That amount was reduced to

$3.3 million based on the jury’s apportionment of fault. Id. Ms. Martin was also

awarded $25 million in punitive damages. Id. RJR appealed the judgment.

2. The First District Court of Appeal in Martin II Agreed That Engle
III’s Instruction Required It to Hold the Defendants Liable to all Class
Members Irrespective of the Phase I Findings

The appeal was “the first . . . ‘Engle progeny’ case to reach a district court of

appeal following the Florida Supreme Court’s decision” in Engle III. Id. at 1062.

The “crux” of RJR’s appeal, as the First District saw it, was “the extent to which

an Engle class member can rely upon the findings from the class action when she

individually pursues one or more Engle defendants for damages.” Id. In other

words, to what extent could Ms. Martin use the Engle findings to establish the

elements of her claims? Id. Reiterating its argument from previous cases, RJR

pointed out that the Phase I findings

facially prove only that RJR at some point manufactured and sold an
unspecified brand of cigarette containing an undefined defect; RJR
committed one or more unspecified negligent acts; RJR on some
occasion concealed unspecified information about the health effects of
smoking and the addictive nature of smoking; and RJR and several
other entities agreed to conceal said unspecified information.

claims is a key to the courthouse. Once there, all the plaintiff has to establish is that she is
addicted to the defendant’s cigarettes.
141
 Case: 13-14590 Date Filed: 05/18/2017 Page: 142 of 284

Id. Therefore, RJR contended, Ms. Martin should have been required to prove, and

RJR should have been allowed to contest, that the brand of cigarettes Mr. Martin

smoked was defective, unreasonably dangerous, and negligently produced.133 Id.

Further, Ms. Martin should have been required to identify the particular conduct

the jury deemed tortious and the particular product feature(s) the jury deemed

defective and prove that such conduct and feature(s) caused Mr. Martin’s injury.

Id. Because the Circuit Court had simply presumed that RJR’s tortious conduct

caused her husband’s injuries, it had violated RJR’s due process right to litigate

essential elements of its case.134 Id.

133
By the same token, Mrs. Martin should have been required to prove, and RJR should
have been allowed to contest, that Mr. Martin was injured by RJR’s concealment of information
and agreement to conceal information. I focus mainly on Mrs. Martin’s strict-liability and
negligence claims because they are the claims before us in this appeal.
134
Whether an Engle defendant’s tortious conduct caused any class member’s injury was
not an issue tried to the Phase I jury. None of the class plaintiffs, other than the class
representatives, testified at the Phase I trial, and the jury was not asked to specify unreasonably
dangerous defects or the way in which defendants failed to exercise due care. Under the original
trial plan, the issue of whether an Engle defendant’s product defect(s) and tortious conduct
caused a class member’s injury would not be decided until Phase III.
Because the Phase I findings did not specify unreasonably dangerous defects or tortious
conduct, and because causation was not litigated in Phase I, class members could not prove a
defendant’s liability under traditional tort law unless the parties were allowed to relitigate
conduct: Which brands were defective, unreasonably dangerous, and negligently produced? In
what ways were those brands defective and how had defendants breached their duty of care?
But Engle III made it clear that progeny courts were not supposed to entertain such
litigation. The only way, therefore, for plaintiffs to establish liability is if the traditional tort law
that had been in place at the beginning of the trial were replaced by law that presumed that (1)
every cigarette had an unreasonably dangerous defect and was negligently produced and (2) all
smoking-related injuries were caused by the manufacturer’s tortious conduct. In Douglas III, the
Florida Supreme Court confirmed that Engle III had indeed replaced traditional tort law,
implementing the conclusive presumptions plaintiffs would need to hold defendants liable. See
142
 Case: 13-14590 Date Filed: 05/18/2017 Page: 143 of 284

The First District rejected RJR’s characterization of the Phase I findings. In

doing so, it did not look to the jury instructions or the special interrogatories.

Instead, it looked to the Engle Omnibus Order for interpretational assistance. In

that order, Judge Kaye, who had tried Phases I and II, determined that “the

plaintiff[s] ha[d] presented evidence that could support [the Phase I findings].”

Friedrich v. Fetterman & Assocs., 137 So. 3d, 362, 365 (Fla. 2013) (emphasis

added). In other words, “There was more than sufficient evidence at trial to . . .

support the jury verdict.” But the First District cited the Omnibus Order, not for

what the Phase I jury could have determined, but for what it did determine. That a

properly instructed jury could have determined that the “findings encompassed all

brands” was, to the Martin II Court, proof that the jury made such a determination.

Id. at 1068. That a properly instructed jury could have “determined the defendants

. . . breached their duty [to all class members] by [negligently] selling [defective]

cigarettes” was proof that the jury determined that as well. Id. The First District

implemented this strange sufficiency-of-the-evidence standard throughout.135 See,

e.g., id. at 1069 (“[T]he record contains abundant evidence from which the jury

Douglas III, 110 So. 3d at 429 (When a plaintiff “prov[es] that addiction to the Engle
defendants’ cigarettes containing nicotine was a legal cause of the injuries alleged,” “injury as a
result of the Engle defendants’ conduct is assumed.”).
135
By asking what the jury could have determined, rather than what it actually
determined, the Court created a conclusive presumption based on evidence in the record and not
a jury finding. The Court thereby denied RJR’s constitutional right to have a jury decide
essential factual issues. See infra note 142 and accompanying text.
143
 Case: 13-14590 Date Filed: 05/18/2017 Page: 144 of 284

could infer Mr. Martin's reliance on pervasive misleading advertising campaigns

for the Lucky Strike brand in particular and for cigarettes in general.” (emphasis

added)).

Why did the First District, as a recognizing court, take upon itself the

plaintiff’s burden of proving what the Phase I jury decided? Why did it lighten

that burden from one of necessary inference136 to sufficiency of the evidence?

Why, in doing so, did it ignore the jury instructions and verdict form in favor of an

136
Determinations about what a jury actually decided must be made on the basis of
necessary inference. The Supreme Court made this clear in De Sollar v. Hanscome: “[I]f [the
instructions] left it open to the jury to find for the defendant upon either of the two propositions,
and the verdict does not specify upon which the jury acted, there can be no certainty that they
found upon one rather than the other” and preclusion is inappropriate. 158 U.S. 216, 222, 15 S.
Ct. 816, 818, 39 L. Ed. 956 (1895); see also Fayerweather, 195 U.S. at 302, 25 S. Ct. at 65
(When the basis upon which a judgment is rendered is unclear, it is “tantamount to a finding in
favor of the successful party of all facts necessary to sustain the judgment.” (emphasis added)).
In Brown II, we observed that Florida courts uphold this common-law protection:
[P]reclusive effect is not given to issues which could have, but may not have,
been decided in an earlier lawsuit between the parties. See, e.g., Acadia Partners,
L.P. v. Tompkins, 673 So. 2d 487, 488–89 (Fla. 5th DCA 1996) (holding that
jury’s verdict “for [the defendant]” in a breach of contract action did not establish
the absence of breach because the jury was instructed that it could find for the
defendant if it concluded that the defendant had not breached the contract or if the
defendant proved an affirmative defense); Allstate Ins. Co. v. A.D.H., Inc., 397
So. 2d 928, 929–30 (Fla. 3d DCA 1981) (concluding that subcontractor could not
show that general contractor was at fault and therefore not entitled to
indemnification based on jury’s “undifferentiated general verdict finding [the
general contractor] ‘negligent’” in an earlier lawsuit; the jury could have
determined that the general contractor was at fault or vicariously liable);
Seaboard, 260 So. 2d at 864–65 (finding that general verdict “in favor of the
defendant” could have been based on jury's conclusion that the defendant was not
negligent or that the plaintiff was contributorily negligent); see id. at 865 (“[I]t is
impossible to ascertain with any reasonable degree of certainty as to what issue
was adjudicated in the former suit except to say that the jury found in favor of [the
defendant]. Such uncertainty as to the effect of the prior adjudication renders the
doctrine of collateral estoppel inapplicable.”).
611 F.3d at 1334.
144
 Case: 13-14590 Date Filed: 05/18/2017 Page: 145 of 284

inapposite ruling on a motion for directed verdict? The First District’s

interpretation of Engle III provides insight: RJR’s characterization of the Phase I

findings had to be rejected because it would “nullify” “the supreme court’s [Engle

III] decision” and “district courts of appeal do not have the prerogative to overrule

Florida Supreme Court precedent.” Id. at 1066–67 (citation omitted). Engle III

had predetermined the res judicata question, and that was good enough for the First

District, “[n]o matter the wording of the findings on the Phase I verdict form.” Id.

at 1067.

Although the First District perceived that the Florida Supreme Court wanted

it to preclude defendants’ defenses, it, like the Circuit Court in its Rule 1.200

order, could not tell which preclusion doctrine the Supreme Court had intended to

invoke. The First District purported to “find it unnecessary to distinguish

between” “issue preclusion versus claim preclusion.” Id. Nevertheless, the Court

expressly disavowed Brown II’s assertion that plaintiffs, in accordance with issue

preclusion’s actually decided requirement, had to “trot out the class action trial

transcript to prove applicability of the Phase I findings.”137 Id.

Recall that the plaintiffs in Brown II had stipulated that the Florida Supreme

Court had invoked issue, rather than claim, preclusion in Engle III. Brown II, 611

137
By process of elimination, then, had the First District opted for claim preclusion, or
was it hinting that it interpreted Engle III as devising an entirely new preclusion doctrine? Did
that new preclusion doctrine operate only in Engle-progeny cases?
145
 Case: 13-14590 Date Filed: 05/18/2017 Page: 146 of 284

F.3d at 1333 n.7. Recall also that plaintiffs in that case had requested an

opportunity to “flesh out” the Phase I verdict form “using the record as a whole.”

Id. at 1335. Here, the First District rejected issue preclusion and its actually

decided requirement, because “[s]uch a requirement undercuts the supreme court’s

ruling in [Engle III].”138 Martin II, 53 So. 3d at 1067.

In sum, the driving force behind the First District’s unusual analysis was its

interpretation of Engle III. It upheld the Circuit Court because that Court

“correctly construed Engle [III] and instructed the jury accordingly on the

preclusive effect of the Phase I findings.” Id. at 1069.

RJR petitioned the Florida Supreme Court for review, but the Court declined

in an opinion stating,

This cause having heretofore been submitted to the Court on
jurisdictional briefs and portions of the record deemed necessary to
reflect jurisdiction under Article V, Section 3(b), Florida Constitution,

138
Now, seven years later, the Majority assert that the Martin II Court had it wrong.
They explain that if the Martin II Court had simply been willing to use the record as a whole, and
pay closer attention to the “common thrust” of the evidence, Ante at 7, and the “unmodified
noun[s]” on the verdict form, id. at 23, it would have recognized that the actually decided
requirement did not “undercut[] the supreme court’s ruling in [Engle III],” Martin II, 53 So. 3d at
1067, because the Phase I jury had in fact “actually decided common elements of the negligence
and strict liability of R.J. Reynolds and Philip Morris.” Ante at 20. According to the Majority,
not only did the Martin II Court err in failing to recognize on its own that the jury had actually
decided these issues, it erred in failing to recognize that the Florida Supreme Court had already
searched the record and determined that the jury actually decided these issues. See id. at 30
(“The Florida Supreme Court in Engle interpreted those findings to determine what the [Engle]
jury actually decided.”). Even though the Majority assume that the “‘actually decided’
requirement is a fundamental requirement of due process,” id. at 20, and that the Florida
Supreme Court had already discovered the treasure trove of useful jury findings, they conclude
that the Florida Supreme Court did not feel compelled to reveal its discovery and thus
accordingly declined RJR’s petition for review of Martin II’s constitutional error.
146
 Case: 13-14590 Date Filed: 05/18/2017 Page: 147 of 284

and the Court having determined that it should decline to accept
jurisdiction, it is ordered that the petition for review is denied.
No motion for rehearing will be entertained by the Court.

R.J. Reynolds Tobacco Co. v. Martin, 67 So. 3d 1050 (2011) (Table). The U.S.

Supreme Court denied RJR’s petition for a writ of certiorari.

* * *

In entertaining RJR’s appeal, the First District faced a compelling

constitutional argument. As RJR contended, the plaintiff had invoked “the

doctrine of res judicata . . . to prevent any jury determination of the critical facts on

which [the plaintiff’s] claims turn.” Reply Brief of Appellant at 1, Martin II, 53

So. 3d 1060 (No. 1D09-4934). RJR had been precluded from contesting, and Ms.

Martin had been spared the burden of proving, that RJR’s tortious conduct caused

her late husband’s injury. The Phase I findings on which such expansive

preclusion had been premised plainly “[did] not establish that there was a defect in

the Lucky Strike cigarettes smoked by Mr. Martin, let alone one that caused his

death.” Id. at 2. Nor did the findings “establish any negligent conduct, concealed

information, or conspiratorial conduct that caused Mr. Martin’s death.” Id. The

First District knew this. It knew it because the Phase I jury was not tasked with

determining whether an Engle defendant’s conduct caused a class member’s injury.

That determination, according to the original trial plan and Engle III, would be

made by the progeny juries. But the First District also knew that Engle III, by

147
 Case: 13-14590 Date Filed: 05/18/2017 Page: 148 of 284

declaring “res judicata,” had signaled or implicitly held that the Phase I findings

would assist class members in holding Engle defendants liable. See Martin II, 53

So. 3d at 1069 (“[W]e interpret the supreme court’s ruling in Engle to mean

individual class plaintiffs, when pursuing RJR and the other class defendants for

damages, can rely on the Phase I jury’s factual findings.”). The First District’s

dilemma, then, was to either acknowledge the worthlessness of the Phase I findings

and “essentially nullify” Engle III in the process, Id. at 1066, or ignore the

findings’ worthlessness and uphold Engle III. Concluding that “district courts of

appeal do not have the prerogative to overrule Florida Supreme Court precedent,”

the First District chose the latter option. Id. (citing Hoffman v. Jones, 280 So. 2d

431, 434 (Fla. 1973)).

In so choosing, the First District attempted to mitigate the worthless-findings

problem. The Phase I verdict form and jury instructions were transparently

worthless, so the First District looked to another source, the Engle Omnibus Order,

for interpretational assistance. In that order, Judge Kaye concluded that “the

plaintiff[s] ha[d] presented evidence that could support [the Phase I findings].”

Friedrich v. Fetterman & Assocs., 137 So. 3d, 362, 365 (Fla. 2013) (emphasis

added). In other words, “[t]here was more than sufficient evidence at trial to . . .

support the jury verdict.” But the First District cited the Omnibus Order, not for

what the Phase I jury could have determined, but for what it did determine. That a

148
 Case: 13-14590 Date Filed: 05/18/2017 Page: 149 of 284

properly instructed jury could have determined that the “findings encompassed all

brands” was, to the Martin II Court, proof that the jury did determine that. Id. at

1068. That a properly instructed jury could have “determined the defendants . . .

breached their duty [to all class members] by selling cigarettes” was proof that the

jury determined that as well. Id.

Ironically, in carrying out the Engle III Court’s implicit instruction to hold

defendants liable to all class members, the Martin II Court ignored the very explicit

instruction from which the implied instruction was derived. The Martin II Court

did not give “res judicata effect to certain Phase I findings” as Engle III directed.

Engle III, 945 So. 2d at 1254. Instead, in blatant disregard of the defendants’ jury-

trial rights,139 it gave res judicata effect to the evidence presented at the Phase I

trial.

139
Article I, section 22 of the Florida Constitution provides that “the right of trial by jury
shall be secure to all and remain inviolate.” Fla. Const. art. I, § 22. Parties have a jury-trial
right with respect to issues that are legal, as opposed to equitable, in nature. Yer Girl Tera Mia v.
Wimberly, 962 So. 2d 993, 996 (Fla. 5th Dist. Ct. App. 2007). This right “should not be
withdrawn from the jury’s consideration unless as a matter of law no proper view of the evidence
could possibly sustain” an alternative determination. Bourgeois v. Dade Cty., 99 So. 2d 575, 577
(Fla. Div. A 1956).
Here, the conduct elements of the class members’ causes of action presented factual
issues. The Phase I jury did not indicate the conduct it deemed tortious. To simply presume the
jury decided that all cigarettes were defective and unreasonably dangerous based on the fact that
a properly instructed jury could have decided that effectively disregards the jury’s role as
decision-maker and retroactively transforms a jury trial into a bench trial. Under the Florida
Constitution, a court cannot override a jury unless “no proper view of the evidence could
possibly sustain” an alternative determination. At Phase I, the Engle defendants presented a lot
of evidence disputing the idea that all cigarettes are defective and unreasonably dangerous.
Furthermore, the class representatives presented evidence upon which a jury could have found
149
 Case: 13-14590 Date Filed: 05/18/2017 Page: 150 of 284

This sleight of hand carried class plaintiffs only part of the way to

establishing RJR’s liability. Even if all cigarettes were defective and unreasonably

dangerous, and even if all defendants breached a duty to all class members,

plaintiffs still could not prove that an unreasonably dangerous defect or a tortious

act caused their harm unless they could identify the unreasonably dangerous defect

or tortious act.140

The First District bridged the remaining gap by changing the law rather than

the facts. Specifically, it implemented, without saying it was doing so, a

conclusive presumption under which class members were allowed to presume

rather than prove that defendants’ tortious conduct caused their injury. In doing so,

the Court disregarded Engle III’s holding that “individualized issues such as legal

that only some brands of cigarettes were defective and unreasonably dangerous. See supra note
61 and accompanying text.
140
To establish strict liability in the products-liability context, a plaintiff must prove that
a particular defect in a manufacturer’s product caused her injury. See, e.g., West v. Caterpillar
Tractor Co., Inc., 336 So. 2d 80, 87 (Fla. 1976) (“In order to hold a manufacturer liable on the
theory of strict liability in tort, the user must establish . . . the defect and unreasonably dangerous
condition of the product, and the existence of the proximate causal connection between such
condition and the user’s injuries or damages.” (citation omitted)); Restatement (Third) of Torts §
15 (1998) (noting that the defect itself must cause the plaintiff’s injury). Similarly, to establish
negligence, a plaintiff must prove that a specific negligent act or omission caused her injury. See
Williams v. Davis, 974 So. 2d 1052, 1056 (Fla. 2007) (stating that negligence liability requires “a
reasonably close causal connection between the nonconforming conduct and the resulting injury
to the claimant” (emphasis added) (quotation marks and citation omitted)); Sardell v. Malanio,
202 So. 2d 746, 747 (Fla. 1967) (A “direct” link must be established “between the negligent act
and the injury” so that “it can reasonably be said that but for the act the injury would not have
occurred.”); Restatement (Third) of Torts: Physical and Emotional Harm § 26 (2010) (“Tortious
conduct must be a factual cause of harm.” (emphasis added)); Restatement (Second) of Torts §
430 (1965) (same); Restatement (First) of Torts § 430 (1934) (same).
150
 Case: 13-14590 Date Filed: 05/18/2017 Page: 151 of 284

causation” would be litigated in progeny trials. Engle III, 945 So. 2d at 1268.

Perhaps the Court was interpreting a coded message embedded in Engle III that

departed from its explicit language. Or maybe the Court was breaking a lesser

commandment in order to keep a greater one. In any case, under Martin II’s

reasoning, legal causation would be presumed.

The Court explained its reasoning as follows: It acknowledged that “[t]he

Phase I jury . . . [did] not [determine] ‘whether any class members . . . were injured

by Tobacco’s conduct.’” Id. at 1067 (quoting Engle III, 945 So. 2d at 1256). It

held, however, that progeny plaintiffs need not establish that a defendant’s tortious

conduct caused their harm; rather, they only needed to show that addiction to

cigarettes caused their harm. Id. at 1069.

The conclusive presumption the Martin II Court implemented had serious

flaws. First, it violated RJR’s due process rights.141 The conclusive presumption

141
In Henderson, 279 U.S. at 643, 49 S. Ct. at 447, the Supreme Court held that a
defendant railroad company’s due process rights were violated where it was held liable even
though the plaintiff offered no evidence of a connection between tortious conduct and the injury
at issue. Id. at 640–44, 49 S. Ct. 445–48. Instead of presenting such evidence, the plaintiff
relied on a state-law presumption that “[t]he mere fact of collision between a railway train and a
vehicle . . . was caused by negligence of the railway company.” Id. at 642–43, 49 S. Ct. 445,
447. Because, as a factual matter, a collision could result from “negligence of the railway, or of
the traveler on the highway, or of both, or without fault of any one,” the Supreme Court struck
down the presumption as “unreasonable and arbitrary.” Id. at 644, 49 S. Ct. 445, 447.

Similarly here, Mrs. Martin never alleged or attempted to prove a specific connection
between RJR’s tortious conduct and Mr. Martin’s injury. All she proffered was a complaint that
cited Engle III and plead verbatim the Phase I findings. The Circuit Court, without articulating it
as such, applied a conclusive presumption under Florida common law, which provided that the
151
 Case: 13-14590 Date Filed: 05/18/2017 Page: 152 of 284

also violated RJR’s right to have a jury determine whether it engaged in the

wrongful conduct that caused Mr. Martin’s death.142

In its attempt to give effect to Engle III’s coded instructions, the First

District departed far from Engle III’s explicit language. Nothing in Engle III

foreshadowed Martin II’s reasoning. Nothing in Engle III suggests that all

cigarettes had been found defective and unreasonably dangerous or that all

defendants had been found to breach their duty to all class members. And nothing

in the opinion even hints at the conclusive presumption the Court created. In fact,

the conclusive presumption runs contrary to Engle III’s assertion that legal

causation would be litigated rather than presumed in progeny trials. In creating the

presumption, the Court substantially altered midstream the elements of the tort

claims asserted in the class action complaint and litigated in Phase I. Though the

Martin II Court may have been doing its best to give effect to an instruction it

mere fact of Mr. Martin’s addiction-related injury conclusively establishes that his injury was
caused by the defendants’ tortious conduct. Because his injury may very well have been caused
by cigarettes’ non-defective-and-unreasonably-dangerous features and by RJR’s nontortious
conduct, the presumption, which the First District upheld on appeal, is unreasonable and
arbitrary.
142
Article I, section 22 of the Florida Constitution provides that “the right of trial by jury
shall be secure to all and remain inviolate.” Fla. Const. art. I, § 22. Parties have a jury-trial
right with respect to issues that are legal, as opposed to equitable, in nature. Yer Girl Tera Mia v.
Wimberly, 962 So. 2d 993, 996 (Fla. 5th Dist. Ct. App. 2007). This right “should not be
withdrawn from the jury’s consideration unless as a matter of law no proper view of the evidence
could possibly sustain” an alternative determination. Bourgeois v. Dade Cty., 99 So. 2d 575, 577
(Fla. Div. A 1956).
Here, the causation elements of Mrs. Martin’s causes of action were legal issues. No jury
ever considered whether RJR’s tortious conduct caused Mr. Martin’s injuries.
152
 Case: 13-14590 Date Filed: 05/18/2017 Page: 153 of 284

interpreted as embedded in the Engle III opinion, the First District’s reasoning is so

removed from Engle III as to constitute a new substantive ruling. As will be

shown below, the Waggoner Court treated it as such.

3. The Fourth District Court of Appeal in Jimmie Lee Brown II Held
That Engle III’s Instruction Meant Issue Preclusion but That the
Plaintiff Did Not Need to Identify a Specific Defect or Negligent
Conduct

Jimmie Lee Brown v. R.J. Reynolds Tobacco Co. (Jimmie Lee Brown I), No.

4D09-2664 (Fla. Cir. Ct. 2010), 2009 WL 2493781 was tried to a jury on issues

framed by the parties’ pleadings. The plaintiff, Jimmie Lee Brown, filed suit on

behalf of Roger Brown, who was deceased, on March 1, 2007 in the Circuit Court

for Broward County, Florida. Complaint at 1, Jimmie Lee Brown I (No. 4D09-

2664). In it, he alleged, echoing the complaints in Brown I and Martin I, that the

Common Liability Findings conclusively established the elements of the plaintiff’s

tort claims. Id. at 5–9. The defendants’ answers, in turn, were similar to those

filed in Brown I and Martin I. Answer, Jimmie Lee Brown I (No. 4D09-2664).

The trial proceeded in two phases. Jimmie Lee Brown II, 70 So. 3d at 711.

In the first phase, the Court asked the jury to determine whether the decedent “was

a member of the Engle class, i.e. whether he was addicted to RJR cigarettes

containing nicotine; and, if so, was his addiction a legal cause of his death.” Id.

153
 Case: 13-14590 Date Filed: 05/18/2017 Page: 154 of 284

The jury found that Roger Brown was an Engle class member.143 Id. In the second

phase, a trial involving claims of strict liability and negligence,144 the Court gave

the jury binding instructions similar to those in Martin I. Jury Instructions, Jimmie

Lee Brown I (No. 4D09-1664), 2009 WL 2599305 at 1–12. The jury assessed Ms.

Brown’s damages at $1.2 million, which was reduced to $600,000 due to the

decedent’s fault. Jimmie Lee Brown II, 70 So. 3d at 714.

In appealing the judgment to the Fourth District, RJR repeated the

arguments it had made to the District Court in Brown I and to the First District in

Martin II: Engle III’s res judicata declaration did not relieve the plaintiff of “the

burden to prove that RJR committed particular negligent acts in a violation of a

duty of care owed to Mr. Brown.” Id. Nor did it relieve the burden “to prove that

the cigarettes Mr. Brown smoked contained a specific defect that injured Mr.

Brown.” Id. Because, RJR argued, “res judicata . . . necessarily mean[t] . . . issue

preclusion . . . post-Engle plaintiffs must demonstrate that the issues on which they

seek preclusion were ‘actually litigated’ in [Phase I].” Id.

143
The jury found that he was addicted to Camels, Pall Malls, Marlboros, and Winstons.
Id. at 712.
144
Like the other progeny cases, the complaint asserted claims of strict liability,
negligence, fraudulent concealment and conspiracy to fraudulently conceal. Id. at 711. At trial,
the Court directed a verdict for RJR on the two fraudulent-concealment claims because the
plaintiff failed to produce evidence of reliance. Id. at 711 n.6.
154
 Case: 13-14590 Date Filed: 05/18/2017 Page: 155 of 284

The Fourth District agreed. Citing our decision in Brown II, the

Court held that contrary to the First District’s interpretation,145 the Supreme

Court’s reference to the res judicata effect of the Phase I findings “necessarily

meant issue preclusion, not claim preclusion.” Id. at 715. Recall that the First

District in Martin II allowed Ms. Martin to hold RJR liable simply by proving class

membership—his addiction to an RJR cigarette and a smoking-related injury.

Martin II, 53 So. 3d at 1066. The Fourth District repudiated that approach, holding

instead that the “class membership” jury instruction could not be used “for the dual

purpose of satisfying the element of legal causation with respect to addiction and

legal causation on the underlying strict liability and negligence claims.” 146 Jimmie

Lee Brown II, 70 So. 3d at 714. “[P]laintiffs must prove more than mere class

membership and damages.” Id. at 715. They must, the Fourth District insisted,

prove “legal causation and damages.”147 Id. In doing so, they cannot use the Phase

I findings “to establish facts that were not actually decided by the jury.” Id.

(quoting Brown II, 611 F.3d at 1333).

145
“By equating the legal causation instruction used on the issue of addiction with a
finding of legal causation on the plaintiff's strict liability and negligence claims, the First District
[in Martin II] effectively interpreted the ‘res judicata’ language . . . in Engle III to mean claim
preclusion instead of issue preclusion.” Id. at 716.
146
In affirming the judgment against the defendants, the Majority disagree with the
Fourth District on this point.
147
The Fourth District recognized, and was disturbed, that the Martin II Court applied the
two conclusive presumptions I describe in notes 135, 136, and 142. The plaintiff had been
spared her burden of proving that the defendant’s tortious conduct caused harm, and that did not
sit well with the Fourth District.
155
 Case: 13-14590 Date Filed: 05/18/2017 Page: 156 of 284

For all its lip service to the defendants’ arguments and Brown II’s reasoning,

however, the Fourth District “[did] not go as far as Brown [II] to require trial

courts to evaluate whether . . . elements of . . . plaintiffs’ claims are established by

the Engle findings.” Id. Specifically, though the Court believed in the necessity of

issue preclusion’s actually decided requirement, it nevertheless held that plaintiffs

were not “required to point to a specific defect” or “specific tortious conduct.” Id.

at 717, 718. Although it believed in the necessity of proving legal causation, it

held that plaintiffs need not do so—they did not need to identify defect(s) or

tortious conduct, let alone prove that such defect(s) and conduct caused their harm.

Why the sharp disconnect between analysis and holding? Because the

Fourth District was “constrained by the Florida Supreme Court’s decision in Engle

III.” Id. at 715. That decision makes it clear that “conduct . . . was determined”

and was not to be litigated in progeny cases. Id. at 717. To require plaintiffs to

prove “that Tobacco committed particular negligent acts when asserting a

negligence claim . . . would render the Florida Supreme Court’s opinion in Engle

III meaningless.” Id. at 718 (emphasis added). The Florida Supreme Court had

issued a gag order, and the Fourth District had no choice but to obey.

Even the Fourth District’s constitutional concerns could not justify a

departure from Engle III’s mandate. It was “concerned the preclusive effect of the

Engle findings violates Tobacco’s due process rights.” Id. at 716. It was

156
 Case: 13-14590 Date Filed: 05/18/2017 Page: 157 of 284

concerned, specifically, that allowing plaintiffs to invoke issue preclusion without

its actually decided requirement constituted an “extreme application[] of the

doctrine . . . inconsistent with a federal right that is ‘fundamental in character.’”

Id. (quoting Postal Tel. Cable Co. v. City of Newport, 247 U.S. 464, 476, 38 S. Ct.

566, 62 L. Ed. 1215 (1918)). Nevertheless, the Court affirmed the Circuit Court’s

judgment in which plaintiffs were allowed to do just that and thereby hold

defendants liable without proving that defendants’ wrongful conduct caused harm.

Id. at 718. Thus, although the Fourth District disavowed Martin II, which it

interpreted as “effectively” implementing “claim preclusion instead of issue

preclusion,” its approach tracked closely to the First District’s—the Fourth District

similarly implemented an unconstitutional conclusive presumption that class

members’ harm was caused by tortious conduct.148 Id. at 716.

Although the Court followed the Engle III mandate, it emphatically voiced

its disapproval. In a special concurrence endorsed by the Court, id. at 716, Chief

Judge May noted the “confusion in the trial courts” stemming from a “struggle

with the extent to which [the Phase I] findings resolve ultimate issues in the trial of

individual claims,” Id. at 718 (May, J., concurring). He quoted from our Brown II

decision to highlight our concern that the “jury findings themselves” provide no

indication that “all cigarettes the defendants sold were defective and unreasonably

148
See supra note 141 and accompanying text.
157
 Case: 13-14590 Date Filed: 05/18/2017 Page: 158 of 284

dangerous.” Id. at 720 (quoting 611 F.3d at 1335). Likewise, he quoted Justice

Wells’s dissent from Engle III, lamenting the many questions Engle III left

unanswered, including, “How are the findings to be used in cases in which the

findings are used?” Id. at 719 (quoting Engle III, 945 So. 2d at 1284 (Wells, J.,

concurring in part and dissenting in part)). Such questions, Judge May, explained

left trial courts and litigants no choice but to play “a form of legal poker.” Id. at

720. One aspect of the game was clear: “the Engle factual findings are binding.”

Id. But “a lurking constitutional issue hovers over the poker game: To what extent

does the preclusive effect of the Engle findings violate the manufacturer’s due

process rights?” Id. With this constitutional question, along with many other

questions, lurking, “parties to the tobacco litigation [were left to] continue to play

legal poker, placing their bets on questions left unresolved by Engle.” Id.

In the wake of Chief Judge May’s special concurrence, the Florida Supreme

Court initially accepted RJR’s petition to review the Fourth District’s decision.

R.J. Reynolds Tobacco Co. v. Jimmie Lee Brown, 133 So. 3d 931, 931 (Fla. 2014)

(per curiam). But then it changed its mind: “Upon further consideration,” the

Court explained, “we have determined that we should exercise our discretion and

discharge jurisdiction. Accordingly, we hereby dismiss this review proceeding.

No motion for rehearing will be entertained by the Court.” Id. The Supreme Court

denied review because while RJR’s petition for review was pending, it agreed to

158
 Case: 13-14590 Date Filed: 05/18/2017 Page: 159 of 284

answer a certified question from the Second District Court of Appeal in Philip

Morris v. Douglas (Douglas II), 83 So. 3d 1002, 1011 (Fla. 2d Dist. Ct. App.

2012), which asked the Florida Supreme Court to determine whether its method of

affording res judicata to the Phase I findings denied the Engle defendants’ due

process rights.

* * *

Jimmie Lee Brown II provides the sharpest illustration of the dilemma facing

Florida District Courts of Appeal. Engle III issued a mandate: use the Phase I

findings—use them. The final-judgment, actually decided, and due process

inquiries that recognizing courts ask before affording preclusive effect to prior

adjudications149 were not necessary in progeny cases because the Engle III Court

had already predetermined the res judicata effect of the Phase I findings. The

District Courts of Appeal understood this much. What they did not understand,

however, was how to execute. What preclusion doctrine were they supposed to use

when plaintiffs could not satisfy all the elements of either claim or issue

preclusion? What were they supposed to say in response to the defendants’

legitimate due process concerns?

The First and Fourth Districts took different approaches. The First District

dutifully accepted the Engle III mandate and jumped through hoops—distorting

149
See supra Part II.
159
 Case: 13-14590 Date Filed: 05/18/2017 Page: 160 of 284

facts, disregarding its recognizing-court duties, and remaking Florida tort law—in

an attempt to make the mandate work. The Fourth District was less

accommodating. It rejected Engle III’s invitation to disregard its recognizing-court

duties, identifying preclusion-law elements the plaintiffs had not satisfied and the

due process deprivations foisted upon defendants. But the Court could do no more

than make note of such concerns. It was “constrained” by the state’s highest court

to implement the Engle III mandate. Jimmie Lee Brown II, 70 3d. at 715. Thus,

despite its resistance, the Fourth District begrudgingly upheld two unconstitutional

conclusive presumptions and violated Engle defendants’ jury-trial rights just like

its more cooperative sister court did in Martin II.150

D. In Light of Martin II and Jimmie Lee Brown II, the Middle District of
Florida in Waggoner Ruled That the Preclusive Application of the Phase I
Findings to Hold the Defendants Liable Would Not Violate Due Process

The Waggoner Court faced the same “legal-poker-game” questions that

previous courts had faced. How should a recognizing court respond to an

invitation to abandon its recognizing-court duties? To what extent should it

preclude defendants from litigating their case on the basis of Engle III and the

Phase I findings? In addressing these questions, the Court was presented with a

150
See supra notes 141–47 and accompanying text. The two presumptions were (1) the
cigarettes the plaintiffs smoked were defective, unreasonably dangerous, and negligently
produced and (2) the defendants’ product defect(s) and negligent conduct caused the decedent’s
injury.
160
 Case: 13-14590 Date Filed: 05/18/2017 Page: 161 of 284

complaint that the plaintiffs had amended to take advantage of Martin II’s

favorable holding.151 In their complaint, the plaintiffs asserted the six tort claims

Engle III had “approved,” basing the claims solely on the Phase I findings.152

As they had from the beginning of the Rule 16(c) management conference,

the Engle defendants continued to contend that “the Due Process Clause of the

U.S. Constitution bars Plaintiff from using the Engle findings to establish the

wrongful-conduct elements of her claims, because she cannot show that any issue

as to which she seeks preclusion was actually decided by the Engle jury.”

Defendants’ Rule 16(c) Motion at 1, Waggoner, 835 F. Supp. 2d 1244 (M.D. Fla.

2011) (No. 3:09-CV-10367). They cited our Brown II opinion to support their

argument that Engle-progeny cases “are clearly governed by issue preclusion and

not claim preclusion principles.” Id. at 1 n.2.

Recognizing that the District Court, bound by Brown II, would require them

to “flesh[] out” the Phase I findings,153 Brown II, 734 F.3d at 1335, the plaintiffs

151
The Waggoners amended their complaint with a first amended complaint on February
18, 2011, seven days after the Martin II Court denied rehearing. First Amended Complaint at 1,
Waggoner v. R. J. Reynolds Tobacco Co., 835 F. Supp. 2d 1244 (M.D. Fla. 2011) (No. 3:09-CV-
10367). They amended again with a second amended complaint on March 3, 2011. Second
Amended Complaint at 1, Waggoner, 835 F. Supp. 2d 1244 (No. 3:09-CV-10367).
152
Those claims included the following: Count I, Strict Liability; Count II, Breach of
Express Warranty; Count III, Breach of Implied Warranty; Count IV, Civil Conspiracy to
Fraudulently Conceal; Count V, Fraudulent Concealment; Count VI, Negligence/Gross
Negligence. Id. at 6–8.
153
Recall that the Martin II and Jimmie Lee Brown II Courts had not required plaintiffs to
flesh out the Phase I findings. Those Courts deemed as sufficient the plaintiffs’ proffer of the
Engle III opinion along with the conclusory allegations of their complaints, which consisted of
161
 Case: 13-14590 Date Filed: 05/18/2017 Page: 162 of 284

filed a “Brown [II] Proffer purporting to provide Engle record evidence” that

elucidated what the Phase I jury decided. Waggoner, 835 F. Supp. 2d at 1256.

Like the Martin II Court, they relied heavily on the Omnibus Order—which

evaluated what a properly instructed jury could have determined—as evidence of

what the Phase I jury actually determined.154 Id.; Plaintiffs’ Response to

Defendants’ Rule 16(c) Motion at 18, Waggoner, 835 F. Supp. 2d 1244 (M.D. Fla.

2011) (No. 3:09-CV-10367). Adopting Martin II’s reasoning, they argued that not

only did their class representatives present enough evidence at the Phase I trial that

a properly instructed jury could have found that all cigarettes were defective,

unreasonably dangerous, and negligently produced, they presented enough

evidence that “the jury necessarily determined that all the tobacco companies’

cigarettes were defective.” Waggoner, 835 F. Supp. 2d at 1265 (emphasis added)

(citation and quotation marks omitted). The defendants resisted this argument,

reminding the Court “that the [Phase I] jury was presented with . . . many differing

and contradictory theories,” id. at 1266, and that class representatives had

quotes from the Phase I findings. In Martin II, the First District took the plaintiffs’ burden upon
itself and looked to the Omnibus Order in a strange sufficiency-of-the-evidence inquiry. 50 So.
3d at 1068. In Jimmie Lee Brown II, the Fourth District also relieved the plaintiff of her burden
to prove what the Phase I jury decided. 70 So. 3d at 715. The Majority now claim that the
Florida Supreme Court, behind closed doors, identified the implied jury findings that these
progeny courts believed would be too onerous for the plaintiffs to discover. See Ante at 30 (“The
Florida Supreme Court in Engle interpreted those findings to determine what the jury actually
decided.”).
154
See supra note 135 and accompanying text.
162
 Case: 13-14590 Date Filed: 05/18/2017 Page: 163 of 284

specifically rejected as “[a] major fiction” the theory that all cigarettes were

defective.155 Defendants’ Rule 16(c) Motion at 18, Waggoner, 835 F. Supp. 2d

1244 (M.D. Fla. 2011) (No. 3:09-cv-10367) (citation omitted).

The plaintiffs adopted Martin II’s sufficiency-of-the evidence reasoning to

identify what the Phase I jury decided, but they achieved an even greater level of

precision than the Martin II Court. Martin II states that “the Phase I jury findings

encompassed all the brands,” 53 So. 3d at 1068, but that opinion never actually

identifies the defendants’ unreasonably dangerous defect(s) or tortious act(s).156

The Waggonner plaintiffs figured out what the Martin II Court could not: it all

came down to nicotine. Phase I “was, at bottom, a case about addiction to

cigarettes [containing] nicotine.” Waggoner, 835 F. Supp. 2d at 1265 (citation and

quotation marks omitted). “[I]t was the presence of [nicotine] in every cigarette . .

. that made all of Defendants’ products defective,” unreasonably dangerous, and

negligently produced. Id. In sum, the plaintiffs interpreted a “year-long [Phase I]

trial in which myriad defect, negligence and fraud theories were vigorously

litigated,” id. at 1276 (emphasis in original), as producing a record so one-sided

and straightforward that a jury necessarily must have adopted a single theory of

155
“The plaintiffs had good reason for not proceeding on that theory. Florida courts have
rejected such a broad imposition of liability (Liggett Group, Inc. v. Davis, 973 So. 2d 467, 472
(Fla. 4th [Dist. Ct. App.] 2007)), and . . . federal law would preempt such a result.” Defendants’
Rule 16(c) Motion at 18, Waggoner, 835 F. Supp. 2d 1244 (M.D. Fla. 2011) (No. 3:09-CV-
10367).
156
See supra note 140 and accompanying text.
163
 Case: 13-14590 Date Filed: 05/18/2017 Page: 164 of 284

liability, one that the class representatives disavowed at trial and one that could

have been litigated in days or weeks.

To the defendants, the Phase I record revealed a very different story, one of

hopeless complexity. “[D]ue to the generality of the Phase I findings and the

multiple theories of liability advanced in the trial record,” the defendants argued,

“plaintiffs simply can’t get there (due process) from here (a Brown [II] Proffer).”157

Id. at 1266.

In light of the defendants’ argument that even “the most thorough Brown [II]

Proffer imaginable . . . would not satisfy . . . federal due process,” the District

Court declined to review the plaintiffs’ Proffer in detail. Id. at 1266–67.

Evaluating the Proffer would not “decide the issue before the Court.” Id. at 1267.

Instead, the Court examined the threshold question of whether an actually decided

inquiry was even required under the U.S. Constitution. The District Court in

Brown I held that it was. Brown I, 576 F. Supp. 2d at 1345 (citing De Sollar v.

Hanscome, 158 U.S. 216, 221, 15 S. Ct. 816, 39 L. Ed. 956 (1895)). We avoided

the question in Brown II because we assumed that Engle III had not sub silentio

amended Florida’s issue-preclusion doctrine to eliminate its actually decided

157
The defendants echoed the Brown I Court in this regard: “[I]t is impossible to
determine the precise issues decided by the Phase I jury.” Brown I, 576 F. Supp. 2d at 1346.
164
 Case: 13-14590 Date Filed: 05/18/2017 Page: 165 of 284

requirement.158 Brown II, 611 F.3d at 1334. After Martin II and Jimmie Lee

Brown II, our assumption appeared faulty,159 and the District Court thus felt the

need to address the constitutional question.

The District Court observed that “[s]tate courts are generally free to develop

their own rules for protecting against the relitigation of common issues or the

piecemeal resolution of disputes.” Waggoner, 835 F. Supp. 2d at 1267 (quoting

Richards v. Jefferson County, 517 U.S. 793, 797, 116 S. Ct. 1761, 1765, 135 L. Ed.

2d 76 (1996)). Federal courts, pursuant to the Full Faith and Credit Act, 28 U.S.C.

§ 1738, must apply such law so long as it conforms to due process. Id. at 1260,

1267 (citing Richards, 517 U.S. at 797, 116 S. Ct. at 1765). Due process, the

Court observed, “protects those rights ‘so rooted in the traditions and conscience of

our people as to be ranked as fundamental.’” Id. at 1267 (quoting Snyder v.

Massachusetts, 291 U.S. 97, 105, 54 S. Ct. 330, 332, 78 L. Ed. 674 (1934)). The

question the District Court examined, then, was whether “a state’s strict adherence

to the boundaries of traditional preclusion law . . . is a ‘fundamental federal right.’”

Id. The Court thus focused its attention on a substantive due process question. In

158
Our assumption was reasonable: “[T]his Court does not intentionally overrule itself
sub silentio.” F.B. v. State, 852 So. 2d 226, 228 (Fla. 2003) (quoting Puryear v. State, 810 So. 2d
901, 905 (Fla. 2002).
159
“The Florida First District Court of Appeal . . . [held] that the [Phase I] findings may
be used to establish elements of a progeny plaintiff’s claims even in the absence of a record-
based showing that the [Phase I] jury actually decided those issues.” Defendants’ Rule 16(c)
Motion at 2, Waggoner, 835 F. Supp. 2d 1244 (M.D. Fla. 2011) (No. 3:09-CV-10367).
165
 Case: 13-14590 Date Filed: 05/18/2017 Page: 166 of 284

doing so, it neglected the procedural due process question with which defendants

were primarily concerned: Are defendants denied their property without due

process of law when a plaintiff is permitted “to use the [Phase I] findings to

establish [for example] that the cigarettes she smoked were defective, in the face of

a possibility that no jury ever found that fact”? Defendants’ Rule 16(c) Motion at

4, Waggoner, 835 F. Supp. 2d 1244 (M.D. Fla. 2011) (No. 3:09-CV-10367).

In tackling its substantive due process question, the Court found the

defendants unable to identify any fundamental-right deprivations. The Court

implicitly acknowledged that Florida courts had changed the state’s preclusion law

in a significant way for Engle-progeny cases.160 Specifically, it assumed, as we did

in Brown II, that the Florida Supreme Court meant issue preclusion when it said

“res judicata,”161 and it interpreted the Florida courts as eliminating issue

preclusion’s actually decided requirement. Eliminating the actually decided

requirement did not constitute a due process deprivation because defendants did

not have a fundamental right “to a strict application of traditional preclusion law.”

Waggoner, 835 F. Supp. 2d at 1268–69.

160
The Court did not consider the equal protection and due process implications of
creating—after material liability aspects of the case had been tried and the case was on appeal—
a special preclusion doctrine that applied in favor of the plaintiffs and against a few unpopular
defendants and, in its application, materially changed—in the plaintiffs’ favor—the substantive
tort law that had been in place when the Phase I trial was held.
161
The Florida Supreme Court later insisted in Douglas III that it actually meant claim
preclusion. 110 So. 3d at 432.
166
 Case: 13-14590 Date Filed: 05/18/2017 Page: 167 of 284

Although the Court conceded that the defendants had fundamental rights to

an opportunity to be heard and against arbitrary deprivations of property, it held

that such rights had not been violated in Engle-progeny cases like Waggoner. Id.

at 1272–77. In reaching this conclusion, it first observed that defendants had been

afforded an opportunity to be heard at Phase I: “Defendants had every reason to

litigate each potential theory of liability to the fullest extent possible.” Id. at 1276.

It acknowledged that Phase I did not afford defendants “their day in court on . . .

legal causation, comparative fault, and damages,” id., but the Court assumed that

such a day would come later because “the Phase I jury ‘did not determine whether

the defendants were liable to anyone.’”162 Id. at 1272 (quoting Engle III, 945 So.

2d at 1263).

After Phase I, “defendants continue to vigorously litigate each and every

remaining issue in each and every progeny suit”; thus, “the preclusive application

of the Phase I approved findings in no way [arbitrarily] deprives them of property.”

Id. What were the “remaining issues” to which the Court referred? First,

plaintiff’s “addict[ion] to one of Defendants’ cigarettes containing nicotine”;

second, “that such addiction was the legal cause of [plaintiff’s harm]”; third, “that

[plaintiff’s harm] manifested before the class membership cut-off date”; and

162
The Florida Supreme Court later disavowed this statement in Douglas III and cut off
defendants’ day in court on legal causation. See Douglas III, 110 So. 3d at 433–34
(“Respectfully, the Engle judgment was a final judgment on the merits.”).
167
 Case: 13-14590 Date Filed: 05/18/2017 Page: 168 of 284

fourth, “that no other procedural bar prevents any aspect of her claim.” Id. at 1273

(emphasis added). Whether defendants’ tortious conduct caused plaintiffs’ harm

did not make the District Court’s list of remaining issues on which defendants

deserved an opportunity to be heard.163

Returning to its core substantive due process reasoning, the Waggoner Court

again emphasized that defendants did not have a fundamental right to traditional

preclusion law. Florida courts needed “flexibility to accommodate the due process

interests of both the Defendants and the thousands of Engle progeny plaintiffs.”

Id. at 1277. The Court did not expound on the limits of such flexibility and did not

mention any concerns regarding Florida courts’ changing the state’s preclusion

doctrines with respect to only one group of unpopular defendants after the parties

had already litigated part of their lawsuit.

* * *

The Waggoner Court failed to directly engage with the defendants’ due

process arguments. It reframed as substantive their procedural due process

concerns, and refused to endorse the proposition that due process prevents state

courts from changing their preclusion doctrines. Naturally, courts sometimes

163
The Court hinted that, under Florida law, plaintiffs had to prove legal causation
separately from class membership. Waggoner, 835 F. Supp. 2d at 1274. But Martin II conflates
the two. See Martin II, 53 So. 3d at 1066 (affirming the district court in allowing Mrs. Martin to
hold RJR liable simply by proving Mr. Martin’s class membership—his addiction to an RJR
cigarette and a smoking-related injury). The Waggoner Court’s confusion in this regard may
help explain its due process ruling.
168
 Case: 13-14590 Date Filed: 05/18/2017 Page: 169 of 284

deviate from or change established procedures and “not all [such] deviations . . .

result in constitutional infirmity.” Honda Motor Co. v. Oberg 512 U.S. 415, 430,

114 S. Ct. 2331, 2339, 129 L. Ed. 2d 336 (1994). But query whether courts deny

due process when they abrogate common-law practice so as to descend

significantly below the level of protection afforded at common law. Id. at 430–33,

114 S. Ct. 2331, 2339–42. Query also whether state courts can change preclusion

law with respect to a few unpopular defendants in a lawsuit that has already been

partially litigated. Waggoner does not say.

The Waggoner Court’s silence on these matters may, in part, be explained

by its thorough misunderstanding of Florida preclusion law. The Court assumed it

was dealing with issue preclusion. It was not.164 It interpreted Martin II as

requiring Engle-progeny plaintiffs to prove legal causation separately from class

membership. That interpretation was incorrect.165 Finally, the Court assumed the

Florida Supreme Court meant what it said when it noted in Engle III that “the

Phase I jury ‘did not determine whether the defendants were liable to anyone.’” Id.

at 1272 (quoting Engle III, 945 So. 2d at 1263). The Florida Supreme Court later

disavowed that statement in Douglas III. See Douglas III, 110 So. 3d at 433–34

(“Respectfully, the Engle judgment was a final judgment on the merits.”).

164
See supra note 161 and accompanying text.
165
See supra note 131 and accompanying text.
169
 Case: 13-14590 Date Filed: 05/18/2017 Page: 170 of 284

The upshot of all this confusion is that the Waggoner Court never evaluated

whether defendants were deprived of property without due process of law when

they were held liable despite plaintiffs never proving, and defendants never having

an opportunity to contest, that defendants’ tortious conduct caused plaintiffs’

harm.166 The answer to that question, in my estimation, is simple.

E. The Second District Court of Appeal in Douglas II Accepted Martin II’s
Reasoning, But Certified the Due Process Question to the Florida Supreme
Court

As noted earlier, Waggoner was the lead case among the cases awaiting trial

in the Middle District of Florida. With the Waggoner preclusion decision in hand,

the judges presiding over the remaining cases were free to proceed. Moving

forward, progeny courts would accord the Phase I findings preclusive effect in

accordance with the Martin II formulation of res judicata. The threshold question

for progeny juries would be whether the plaintiff proved addiction to a defendant’s

cigarettes; if so, the trial would focus, not on the causal connection between the

defendant’s tortious conduct and the plaintiff’s injury, but on damages arising from

the plaintiff’s smoking-related disease and the comparative negligence of the

plaintiff and any other Engle defendant involved.

166
Recall that the question of whether defendants’ tortious conduct caused the class
members any harm was not tried in Phase I. Under the original Engle trial plan, class members
would litigate that issue in the Phase III litigation of their individual tort claims. Walker II, 734
F.3d at 1281.
170
 Case: 13-14590 Date Filed: 05/18/2017 Page: 171 of 284

Earl Graham’s case against RJR and Philip Morris was one of the lead

progeny cases handled under the Waggoner umbrella. Whether it would be tried in

accordance with preclusion law as set forth in Martin II, though, would depend on

the Florida Supreme Court’s ruling on a question the Second District Court of

Appeal had certified in Douglas II, 83 So. 3d at 1011, regarding the due process

implications of the way in which progeny courts had been applying the Phase I

findings.

The case had been brought in the Circuit Court for Hillsborough County,

Florida, by James Douglas as the representative of the estate of his late wife,

Charlotte Douglas. Its complaint presented the six Engle III-approved claims and,

as in Brown I, Martin I, Jimmie Lee Brown I, and Waggoner, cited the Engle III

opinion and the Common Liability Findings as conclusive proof of the elements of

the claims.167 Third Amended Complaint at 7–8, Douglas v. Philip Morris USA

Inc. (Douglas I), No. 08-CA-008108 (Fla. Cir. Ct. Mar. 12, 2010). The defendants

answered the complaint, denying liability and raising the same due process issues

they had raised in those cases. Answer to Amended Complaint, Douglas I, No. 08-

CA-008108.

167
The claims were “strict liability, negligence, breach of express and implied warranty,
fraudulent concealment, and conspiracy to fraudulently conceal.” Douglas II, 83 So. 3d at 1003.
171
 Case: 13-14590 Date Filed: 05/18/2017 Page: 172 of 284

Douglas I was tried on the plaintiffs’ claims of strict liability and fraudulent

concealment.168 The plaintiff produced evidence of his wife’s addiction to the

defendants’ cigarettes169 and that smoking caused her death. After the defendants

rested, the Court instructed the jury, first, that the plaintiff had to prove

membership in the Engle class; that is, that the decedent was addicted to and

injured by cigarettes containing nicotine. Douglas II, 83 So. 3d at 1004. Next, as

in Martin II and Jimmie Lee Brown II, the Court informed the jury that it had to

“accept the eight Phase I Engle findings as proven fact.” Id. at 1005. Accordingly,

if it found that the decedent’s death was caused by cigarettes (rather than the

defendants’ tortious conduct), the jury had only to determine the percentage of the

cigarettes she smoked that were of the respective defendants’ brands. Id. at 1003.

The jury [found] each of the named defendants strictly liable for Mrs.
Douglas' death, apportioning fault as follows: 50% to Mrs. Douglas,
18% to Philip Morris, 5% to R.J. Reynolds, and 27% to Liggett.
Additionally, the jury found against Mr. Douglas on the issue of Mrs.
Douglas' detrimental reliance on concealment or omissions by the
Tobacco Companies.

168
The remaining claims were withdrawn before the case was submitted to the jury. The
Supreme Court, in answering the certified question in Douglas III, 110 So. 3d at 419, indicated
that the negligence claim was not withdrawn. Rather, the Second District had rejected the claim
because it thought “causation instructions and findings beyond those required by Engle” were
necessary. Id. at 422.
169
The decedent smoked the following brands: Lark, Benson & Hedges, Virginia Slims,
Winston, and Salem. Reply Brief for Petitioners, Douglas III, No. SC12-617, 2012 WL 3078034
at *8.
172
 Case: 13-14590 Date Filed: 05/18/2017 Page: 173 of 284

Id. The Court entered judgment in accordance with the jury’s verdict, and the

tobacco companies appealed, claiming once again that allowing a class plaintiff to

recover against an Engle defendant solely on proof of addiction to the defendant’s

cigarettes and resulting injury constituted a denial of due process. Id. at 1010.

In addressing the due process issue, the Second District reviewed Brown II,

Martin II, and Jimmie Lee Brown II. The Court noted that we had concluded in

Brown II that Engle III’s “res judicata” statement meant issue preclusion. Id. at

1006 (citing Brown II, 611 F.3d at 1333). However, the Court also noted that we

had “pointed out that the parties disagree as to what issue preclusion meant”170 and

quoted the following passage from Brown II to substantiate:

Question 3 on the verdict form asked the jury: “Did one or more of
the Defendant Tobacco Companies place cigarettes on the market that
were defective and unreasonably dangerous?” The jury answered
“yes[ ]” for every time period for every defendant except Brooke
Group, Ltd., Inc. Under the defendants' view, the only fact that the
jury found was that they sold some cigarette that was defective and
unreasonably dangerous during the time periods listed on the verdict
form. That would mean that the finding may not establish anything
more specific; it may not establish, for instance, that any particular
type or brand of cigarette sold by a defendant during the relevant time
period was defective and unreasonably dangerous. Under the
plaintiffs' broader view[,] the jury's finding must mean that all
cigarettes the defendants sold were defective and unreasonably

170
The plaintiffs agreed with the defendants that the words “res judicata” in Engle III
meant issue preclusion, not claim preclusion, conceding the point during oral argument before
this Court. See Brown II, 734 F.3d at 1333 n.7 (“[A]t oral argument, the plaintiffs clarified that
their position is that the Phase I approved findings are entitled to issue preclusive effect. . . . [I]f
the plaintiffs had continued to argue for claim preclusion, we would have rejected that
position.”).
173
 Case: 13-14590 Date Filed: 05/18/2017 Page: 174 of 284

dangerous because there is nothing to suggest that any type or brand
of cigarette is any safer or less dangerous than any other type or
brand.

Id. at 1006–07 (quoting Brown II, 734 F.3d at 1333). The Court then said that, we

went on to observe that the plaintiffs had not pointed to anything in
the transcript of the Engle trial that showed that the jury made such
specific findings[,] . . . advised the trial court that the findings were
entitled to res judicata effect as to the factual issues that were litigated
specifically resolved in the record [and] instructed the trial court on
remand to determine what particular issues were litigated and resolved
in Phase I and then to preclude the defendant tobacco companies from
relitigating those issues.

Id. at 1007.

The Second District then compared the First District’s decision in Martin II,

which interpreted res judicata as claim preclusion, with the Fourth District’s

decision in Jimmie Lee Brown II, which interpreted res judicata as issue preclusion.

Id. at 1007–11. The Court found a substantial difference in the two Courts’

analyses.

The First District, citing the statements in the Engle trial judge’s Omnibus

Order “as conclusive on each of the elements of the [plaintiff’s] causes of action,”

id. at 1008, concluded that the Circuit Court “properly relied on the Phase I

findings and that there was no need for the plaintiff class members to

‘independently prove up those elements or demonstrate the relevance of the

findings to their lawsuits.’” Id. at 1008 (quoting Martin II, 53 So. 3d at 1069).

174
 Case: 13-14590 Date Filed: 05/18/2017 Page: 175 of 284

The Fourth District, on the other hand, believed that the Phase I findings

“preclusively establish the conduct elements of the strict liability and negligence

claims,” but not the causation element of those claims. Id. at 1009 (quoting Jimmie

Lee Brown II, 70 So. 3d at 715) (emphasis added). “Legal causation,” the Fourth

District concluded, “and damages, must be proven in the second phase of trial.

Additionally, the Fourth District gave lip service to the idea that legal causation for

negligence and strict liability should be distinguishable from the causation that

proves addiction resulting from class membership.” Id. (citing Jimmie Lee Brown

II, 70 So. 3d at 715). Specifically, the Jimmie Lee Brown II Court reasoned, rather

than merely proving a causal connection between cigarettes and injury, progeny

plaintiffs should be required to show a causal connection between a defendant’s

tortious conduct and the plaintiff’s injury: “post-Engle III plaintiffs must show ‘(i)

[that] the defendant’s failure to exercise reasonable care was a legal cause of

decedent’s death[] and (ii) [that] the defective and unreasonably dangerous

cigarettes were a legal cause of decedent’s death.’” Id. (quoting Jimmie Lee Brown

II, 70 So. 3d at 715).

After concluding its review of Martin II and Jimmie Lee Brown II, the

Douglas II Court opted for Martin II’s reasoning, finding no violation of due

175
 Case: 13-14590 Date Filed: 05/18/2017 Page: 176 of 284

process.171 The Court concluded, however, that the due process issues the Engle

defendants had been raising were of such significance that the Florida Supreme

Court should address them. The Court therefore certified the following question to

the Florida Supreme Court:

DOES ACCEPTING AS RES JUDICATA THE EIGHT PHASE I
FINDINGS APPROVED IN ENGLE V. LIGGETT GROUP, INC.,
945 So. 2d 1246 (Fla. 2006), VIOLATE THE TOBACCO
COMPANIES' DUE PROCESS RIGHTS GUARANTEED BY THE
FOURTEENTH AMENDMENT OF THE UNITED STATES
CONSTITUTION?

Id. at 1011.

IV.
THE FLORIDA SUPREME COURT IN DOUGLAS III HELD THAT THE
ENGLE III COURT HAD (1) IMPLICITLY DETERMINED THAT THE PHASE
I FINDINGS WERE FULL-BLOWN LIABILITY DETERMINATIONS AND (2)
IMPLICITLY ENTERED JUDGMENT AGAINST ALL DEFENDANTS ON
BEHALF OF ALL CLASS PLAINTIFFS

The Florida Supreme Court accepted the certification and entertained the

certified question wearing two hats. First, it wore the hat of the rendering Engle III

Court, attempting to recall what it had in mind when it decided Engle III seven

years before. By invoking “res judicata,” did it intend to invoke claim preclusion?

In other words, did it implicitly enter judgment pursuant to the Phase I findings,

171
In so finding, the Second District did not explicitly address the two conclusive
presumptions on which progeny plaintiffs had been relying to establish their claims. See supra
notes 134, 141, and accompanying text.
176
 Case: 13-14590 Date Filed: 05/18/2017 Page: 177 of 284

which it interpreted as establishing the Engle defendants’ liability to all class

members? Or did it intend to invoke issue preclusion, interpreting the Phase I

findings as factual findings upon which future courts would enter judgment?

Second, it wore the hat of a recognizing court, applying state preclusion law

and evaluating whether doing so denied the defendants of their property without

due process of law. I consider each of these two hats in turn.

* * *

The Second District, in its opinion certifying the constitutional question,

explained the preclusive effect the Phase I findings would have if they represented

factual determinations of whether the defendants engaged in tortious conduct, on

the one hand, or full-blown liability determinations, on the other. It pointed to our

opinion in Brown II and the Fourth District’s in Jimmie Lee Brown II as examples

of courts interpreting the Phase I findings as factual findings and the First District’s

opinion in Martin II as an example of a court portraying the Phase I findings as

liability determinations as to each of the Engle III-approved tort claims. Douglas

II, 83 So. 3d at 1006–11.

Under Brown II’s reasoning, the Phase I findings were factual

determinations that foreclosed litigation in progeny lawsuits over whether each of

177
 Case: 13-14590 Date Filed: 05/18/2017 Page: 178 of 284

the defendants engaged in six kinds of tortious conduct.172 The findings did not

represent a jury determination that the defendants were liable to all class plaintiffs

for all six torts. Thus, litigation of other essential facts—causation (which requires

identifying tortious conduct)173 and damages—would not be foreclosed or limited.

Indeed, consistent with issue preclusion’s actually decided requirement,174 the

findings would be “given effect to the full extent of, but no farther than, what the

172
The “conduct elements” for the six torts are as follows: (1) selling a defective and
unreasonably dangerous product for strict liability, (2) breaching a duty of care for negligence,
(3) breaching an express warranty for breach of express warranty, (4) breaching an implied
warranty for breach of implied warranty, (5) agreeing to conceal or omit material information for
civil conspiracy to fraudulently conceal, and (6) concealing or omitting material information for
fraudulent concealment.
173
The appropriate causation inquiry is whether the particular tortious conduct identified
in Phase I caused the plaintiff’s injury. See, e.g., West v. Caterpillar Tractor Co., Inc., 336 So.
2d 80, 87 (Fla. 1976) (“In order to hold a manufacturer liable on the theory of strict liability in
tort, the user must establish . . . the defect and unreasonably dangerous condition of the product,
and the existence of the proximate causal connection between such condition and the user’s
injuries or damages.” (citation omitted)); Restatement (Third) of Torts § 15 (1998) (noting that
the defect itself must cause the plaintiff’s injury); Williams v. Davis, 974 So. 2d 1052, 1056 (Fla.
2007) (stating that negligence liability requires “a reasonably close causal connection between
the nonconforming conduct and the resulting injury to the claimant” (emphasis added) (quotation
marks and citation omitted)); Sardell v. Malanio, 202 So. 2d 746, 747 (Fla. 1967) (a “direct” link
must be established “between the negligent act and the injury” so that “it can reasonably be said
that but for the act the injury would not have occurred”); Restatement (Third) of Torts: Physical
and Emotional Harm § 26 (2010) (“Tortious conduct must be a factual cause of harm.”
(emphasis added)); Restatement (Second) of Torts § 430 (1965) (same); Restatement (First) of
Torts § 430 (1934) (same).
174
A party may only assert issue preclusion with respect to issues that were actually
litigated and determined in the first lawsuit. Florida Bar v. Clement, 662 So. 2d 690, 697–98
(Fla. 1995) (per curiam). Just as claim preclusion’s final-judgment requirement helps ensure that
parties have a fair opportunity to fully litigate a cause of action, issue preclusion’s “actually
decided” requirement ensures that parties have at least one opportunity to fully litigate each
issue. The actually decided requirement has two logical derivatives: the asserting party in the
second lawsuit must both identify the issue that the rendering court allegedly decided and prove
that it was actually decided. 18 Wright, supra, § 4420 (noting recognizing courts’ “need to
discover what it was that has been actually decided.”). See supra Part II.A for a more thorough
explanation of issue preclusion’s actually decided requirement.
178
 Case: 13-14590 Date Filed: 05/18/2017 Page: 179 of 284

jury found.” Brown II, 611 F.3d at 1333. This approach was consistent with what

the Phase I findings said on their face and Engle III’s holding that “individualized

issues such as legal causation [and] comparative fault” would be litigated in

progeny trials. Engle III, 945 So. 2d at 1268. The Fourth District, in Jimmie Lee

Brown II, agreed with this limitation (even if it was “constrained” not to implement

it): the findings clearly were not, and should not be treated as, liability

determinations; rather, they should operate to foreclose litigation only as to facts

the Phase I jury actually found. 70 So. 3d at 714–15. Thus, every progeny

plaintiff should be required to prove, and defendants should be able to contest, that

the conduct deemed tortious in Phase I caused the plaintiff’s harm.175 Id. at 715.

In Martin II, the First District purported to “generally agree[]” with the

Brown II approach, but concluded that determining what the Phase I jury actually

decided, as issue preclusion requires, was entirely unworkable and would

“undercut” the Engle III plan.176 The First District interpreted the Engle III plan as

one designed to enhance class members’ chances of prevailing against Engle

defendants in progeny cases; issue preclusion was inconsistent with that plan

175
This is precisely what the original trial plan dictated would take place in Phase III.
Walker II, 734 F.3d at 1281.
176
Requiring class plaintiffs to “trot out the class action transcript to prove [the]
applicability of the Phase I findings,” in order to identify the defendants’ tortious conduct and
prove that such conduct caused their harm would “undercut the supreme court’s” plan. Martin,
53 So. 3d at 1067.
179
 Case: 13-14590 Date Filed: 05/18/2017 Page: 180 of 284

because progeny courts would be unable to determine exactly what conduct the

Phase I jury identified as tortious.

What would be consistent, though, is if the Phase I findings could somehow

be portrayed as liability determinations rather than mere conduct findings. The

Martin II Court accomplished such a portrayal simply by proclaiming it, insisting

that the Phase I findings established the elements of the progeny plaintiffs’ tort

claims such that they “need not independently prove up those elements or

demonstrate the relevance of the findings to their lawsuits.” Martin II, 53 So. 3d at

1069. “No matter the wording of the findings on the Phase I verdict form.” Id. at

1067. And never mind that the Phase I jury was instructed not to consider

liability,177 and that Engle III makes clear that “Phase I yielded no determination as

177
The relevant jury instructions read as follows:
In a typical lawsuit, the jury hears testimony from the litigants, including the
plaintiffs. . . . Ordinarily, the plaintiffs would testify about their claims and
describe their medical conditions and their damages. However, this case is not
typical, because it is not brought on behalf of one or two individuals, but rather on
behalf of a group of . . . individuals. . . .
This trial did not address issues as to the conduct or damages of individual
members of the Florida class. Those issues are not relevant during this trial.
The Court has determined that the conduct of class members is not
relevant to the issues presented in this common liability trial, and therefore , you
did not hear testimony from any members of the Florida class who are plaintiffs
bringing the action . . . .
You will not determine any issues regarding the conduct of individual
class members of the Florida class, including any issues as to compensatory
damages for individual class members. . . .
It is your duty as jurors to decide the issues, and only the issues that I
submit for determination by your verdicts.

180
 Case: 13-14590 Date Filed: 05/18/2017 Page: 181 of 284

to the defendants’ liability to any individual class member. Id. at1064 (citing

Engle III, 945 So. 2d at 1263). Never mind, also, that the Phase I Circuit Court

rejected the plaintiffs’ proposed verdict-form questions about whether the

defendants’ wrongful conduct was a “legal cause of damage, injury or death” so as

to preserve litigation of legal/proximate cause for later phases of the Engle

litigation.178 On the basis of its counterfactual portrayal of the Phase I findings, the

Martin II Court concluded that progeny plaintiffs needed only to prove class

membership—harm caused by smoking a defendant’s cigarettes; no need to

identify particular tortious conduct or prove that such conduct caused harm.

The Florida Supreme Court agreed. It endorsed Martin II’s reasoning that

interpreting “res judicata” as issue preclusion—and thus interpreting the Phase I

findings merely as establishing facts that might assist the class members in proving

178
The defendants objected to such instructions as follows:
[L]egal cause, of course, is not being determined here. Proximate cause is not
being determined here. So that portion of the instruction should not be given to
the jury. This jury cannot determine whether any particular claim is the legal
cause of injury because neither proximate cause connected to an individual class
member nor injury from an individual class member has been put into play or
evidence put on about it at this time.
The defendants also noted that
the absence of the individual in Phase I is why you can’t have this type of
instruction, because, in the normal case, you have a witness, you have a plaintiff
who’s in the box, and somebody comes in and looks not just at a bunch of
statistics, but they look at the specific medical information relating to that
individual.
181
 Case: 13-14590 Date Filed: 05/18/2017 Page: 182 of 284

their tort claims —was unacceptable because it would render those findings

“useless”:

[T]o decide here that we really meant issue preclusion even though we
said res judicata in Engle would effectively make the Phase I findings
regarding the Engle defendants’ conduct useless in individual actions.
See Martin [II], 53 So. 3d at 1067 (concluding that individual
plaintiffs are not required to “trot out the class action trial transcript to
prove applicability of the Phase I findings” because “[s]uch a
requirement undercuts the supreme court's ruling” in Engle [III]).

Douglas III, 110 So. 3d at 433 (emphasis added). Portraying the findings as

liability determinations was essential because proving a causal connection between

tortious conduct and injury in progeny cases would necessarily require identifying

the defendants’ tortious conduct; simply proving that a defendant’s conduct, which

may or may not have been tortious, caused harm would be insufficient.179 Despite

a yearlong trial, the Phase I findings provided no information about the particular

conduct the jury had deemed tortious. Thus, merely allowing those findings to

stand would leave plaintiffs in about the same position in which they would have

been had Phase I never taken place—“defendants [would be] permitted to relitigate

matters pertaining to their conduct.” Id. at 429.

179
Under traditional Florida tort law, a plaintiff alleging strict liability in the products-
liability context must prove inter alia (a) that the product in question was defective179 and (b)
that the “defect caused the injury or harm alleged.” Aubin v. Union Carbide Corp., 177 So. 3d
489, 513 (Fla. 2015). Similarly, under traditional Florida tort law, a plaintiff alleging negligence
must prove inter alia (a) that the defendant breached a duty of care owed to her and (b) that the
defendant’s breach caused her harm. Williams v. Davis, 974 So. 2d 1052, 1056 (Fla. 2007).
182
 Case: 13-14590 Date Filed: 05/18/2017 Page: 183 of 284

Whereas issue preclusion left class members empty handed, claim

preclusion assisted them because it necessarily reinterpreted the Phase I findings as

“conclusive not only as to every matter which was offered and received to sustain

or defeat the claim, but as to every other matter which might with propriety have

been litigated and determined in that action.”180 Id. at 432 (quoting Kimbrell v.

Paige, 448 So. 2d 1009, 1012 (Fla. 1984)). Assisting the plaintiffs, the Court

concluded, is what it had in mind when it wrote its opinion in Engle III.

Justice Canady, in dissent, rejected the Douglas III Court’s reinterpretation

of the Phase I findings as establishing the defendants’ liability to hundreds of

thousands of class plaintiffs who were unknown and not present at Phase I. He

pointed to Question 3181 from the Phase I verdict form as an example of a finding

that was “a much too slender reed to support the imposition of liability on the

defendants.” Id. at 436 (Canady, J. dissenting)

The finding is sufficient to establish that the defendants sold some
cigarettes that were defective and unreasonably dangerous. But it is
not sufficient to establish that all of the cigarettes sold by the

180
The question of whether the defendants’ tortious conduct caused the individual class
members’ injury could not have been litigated and determined in Phase I. Pursuant to the trial
plan, the purpose of Phase I was to litigate “common issues relating exclusively to the
defendants’ conduct and the general health effects of smoking.” Engle III, 945 So. 2d at 1256.
“[I]ndividual causation,” whether the Engle defendants’ allegedly tortious conduct caused an
individual class member’s injury, was to be determined by new juries in Phase III because it was
“highly individualized.” Id. at 1254 (“[I]ndividual causation and apportionment of fault among
the defendants are highly individualized and do not lend themselves to class action treatment.”).
181
Question 3 asked, “Did one or more of the Defendant Tobacco Companies place
cigarettes on the market that were defective and unreasonably dangerous?” The Phase I jury
answered “yes” as to all defendants.
183
 Case: 13-14590 Date Filed: 05/18/2017 Page: 184 of 284

defendants were defective and unreasonably dangerous. Nor is it
sufficient to establish that the particular brands of cigarettes
consumed by Mrs. Douglas were defective and unreasonably
dangerous. The plaintiffs pursued their claims in Phase I based on
several alternative theories of defect, some of which applied only to
certain brands and designs. Given this context, it is unreasonable to
read the jury's finding that the defendants “placed cigarettes on the
market that were defective and unreasonably dangerous” as a finding
that all of the cigarettes placed on the market by the defendants were
defective and unreasonably dangerous.

Id. at 436–37. Justice Canady felt that all the findings were too general to identify

the defendants’ tortious conduct, let alone establish that such conduct caused every

class members’ harm.

His counterparts in the majority disagreed: “[T]he Phase I jury already

determined that the defendants’ conduct subjects them to liability to Engle class

members under [strict liability and negligence] theor[ies].” Id. at 430. To bolster

its portrayal of the Phase I findings, the Douglas III majority, like the First District

in Martin II, focused not on the Phase I findings themselves, but on the evidence

that was before the Phase I jury. Its inquiry centered on whether that evidence was

sufficient to withstand a motion for directed verdict.182 Because, Douglas III

182
That the matter boiled down to a sufficiency-of-the-evidence inquiry explains why the
Court looked to the evidence the Phase I trial judge relied on in his Omnibus Order denying the
defendants’ motion for judgment in accordance with their motion for directed verdict made prior
to the conclusion of the Phase I trial. Douglas III quotes from the Omnibus Order as follows:

There was more than sufficient evidence at trial to satisfy the legal requirements
of this [c]ount and to support the jury verdict that cigarettes manufactured and
placed on the market by the [Engle] defendants were defective in many ways
184
 Case: 13-14590 Date Filed: 05/18/2017 Page: 185 of 284

reasons, the Phase I record contained “more than sufficient evidence” upon which

a properly instructed jury could have found the defendants liable to all plaintiffs,

“[t]hat the . . . jury did not make detailed findings . . . [was] immaterial.” Id. at

423, 433 (emphasis added).

By adopting such a portrayal, the Court disavowed its earlier statement in

Engle III in which it made clear that “the Phase I jury ‘did not determine whether

the defendants were liable to anyone.’” Engle III, 945 So. 2d at 1263. It also

disregarded its Engle III holding that “individualized issues such as legal

causation” would be litigated in progeny trials. Id. at 1268. Such issues were now

“immaterial.” Douglas III, 110 So. 3d at 433. Although the majority agreed with

Justice Canady that the findings are “useless” for the purposes of identifying the

conduct the Phase I jury deemed tortious and proving that such conduct caused

including the fact that the cigarettes contained many carcinogens, nitrosamines,
and other deleterious compounds such as carbon monoxide. That levels of
nicotine were manipulated, sometime[s] by utilization of ammonia to achieve a
desired “free basing effect” of pure nicotine to the brain, and sometime[s] by
using a higher nicotine content tobacco called Y–1, and by other means such as
manipulation of the levels of tar and nicotine. The evidence more than
sufficiently proved that nicotine is an addictive substance which when combined
with other deleterious properties, made the cigarette unreasonably dangerous. The
evidence also showed some cigarettes were manufactured with the breathing air
holes in the filter being too close to the lips so that they were covered by the
smoker thereby increasing the amount of the deleterious effect of smoking the
cigarette. There was also evidence at trial that some filters being test marketed
utilize glass fibers that could produce disease and deleterious effects if inhaled by
a smoker.

Id. at 423–24.

185
 Case: 13-14590 Date Filed: 05/18/2017 Page: 186 of 284

harm183—the majority conceded that Phase I jury “did not make detailed findings

about what evidence it relied upon to make the Phase I common liability

findings”—such uselessness did not matter because the jury had determined

liability instead. Id.

With defendants’ liability to all class members established, all that remained

for progeny plaintiffs to prove was (1) their class membership (by proving

addiction to a defendant’s cigarettes and a smoking-related injury184) and (2) their

damages. Engle defendants, therefore, could contend that a plaintiff was not a

class member because her injury was not caused by smoking,185 but they were

183
That a defendant sold some negligently produced, defective, and unreasonably
dangerous cigarettes of an unspecified brand at an unspecified point in time was not probative as
to whether Ms. Douglas’s injuries were caused by the defendant’s negligent conduct or
unreasonably dangerous product defect(s). Id. at 433.
184
Douglas III treats this single inquiry as though it were two separate inquiries:
“plaintiffs must establish (i) membership in the Engle class; (ii) individual causation, i.e., that
addiction to smoking the Engle defendants’ cigarettes containing nicotine was a legal cause of
the injuries alleged.” 110 So. 3d at 430. To be clear, when a progeny plaintiff proves
“individual causation” as Douglas III defines that term, she is really just proving her membership
in the Engle class. To prove membership in the Engle class, a plaintiff must show addiction to a
defendant’s cigarettes and an injury caused by smoking. Thus, to contest class membership, a
defendant would deny that a plaintiff’s injury was caused by smoking, “for example, by proving
that the disease at issue was the result of a genetic predisposition, exposure to an occupational
hazard, or something unrelated to the plaintiff's addiction to smoking the Engle defendants’
cigarettes.” Id. Similarly, to contest “individual causation,” a defendant would deny that a
plaintiff’s injury was caused by smoking.
185
Defendants “may defend against the establishment of individual causation, for
example, by proving that the disease at issue was the result of a genetic predisposition, exposure
to an occupational hazard, or something unrelated to the plaintiff's addiction to smoking the
Engle defendants’ cigarettes.” 110 So. 3d at 428. Engle defendants can also deny that the
plaintiff smoked its cigarettes and that the plaintiff was addicted to nicotine. They could also
contend that their conduct was not the sole cause of a plaintiff’s injury; for example, that the
plaintiff also smoked another manufacturer’s cigarettes.
186
 Case: 13-14590 Date Filed: 05/18/2017 Page: 187 of 284

precluded by “res judicata” from contending that a plaintiff’s smoking-related

injury was not caused by tortious conduct—when a plaintiff “prov[es] that

addiction to the Engle defendants’ cigarettes containing nicotine was a legal cause

of the injuries alleged,” “injury as a result of the Engle defendants’ conduct is

assumed.” Id. at 429 (emphasis added).

In sum, the Florida Supreme Court, wearing its rendering-court hat in

Douglas III, insisted that it knew when it wrote Engle III that the Phase I findings

This, however, is the extent of evidence that the parties could introduce for comparative
fault because the record would be devoid of any specific evidence on the defendants’ conduct or
the defect in their cigarettes. After all, under the conclusive presumptions established in Martin
II, see supra note 141 and accompanying text, and adopted in Douglas III, see infra note 195 and
accompanying text, the plaintiff would only have to plead the Phase I jury findings that the
defendants had at some point and in some way been negligent and at some point sold cigarettes
that were defective and unreasonably dangerous in some way. The jury would lack any evidence
against which to weigh the plaintiff’s negligence other than this vague assurance that the
defendants had done something tortious. This is, needless to say, not how comparative fault
inquiries are supposed to work. See Rosenfeld v. Seltzer, 993 So. 2d 557, 560 (Fla. 4th Dist. Ct.
App. 2008) (“Under comparative negligence . . . the jury is not instructed to absolve the
defendant of negligence based upon the plaintiff's negligence but to weigh the evidence of both
and provide for a proportionate recovery based upon the defendant’s percentage of negligence.”).
Putting aside the fact the assessment of comparative negligence in progeny cases is a
farce, the Majority imply that the defendant’s opportunity to present evidence on comparative
fault in the District Court forecloses a determination that due process has been violated in this
case. See Ante at 28 (“Every tobacco company must also be afforded the opportunity to contest
the smokers’ pleadings and evidence and to plead and prove the smokers’ comparative fault.
Indeed, in this appeal, after the district court instructed it, the jury reduced Graham’s damages
award for his deceased spouse’s comparative fault.”). This is an outlandish suggestion.
Comparative fault is assessed after a defendant has already been held to be at fault. The
defendants’ opportunity to reduce the amount they owe to a plaintiff does nothing to correct the
lack of due process in finding them at fault in the first place, just as the opportunity to present
evidence in sentencing does nothing to correct an unconstitutional conviction. Due process may
be flexible, but precedent is not. Under the Majority’s new standard, any arbitrary redistribution
of property announced by state tort law complies with due process as long as the defendants have
an opportunity to present evidence of comparative fault.
187
 Case: 13-14590 Date Filed: 05/18/2017 Page: 188 of 284

were “useless” under an issue-preclusion regime, which is why it invoked claim

preclusion instead.186 See 110 So. 3d at 433 (“[W]e used the correct term when we

gave the Phase I findings ‘res judicata effect,’ signifying that relitigation of the

class’s cause of action established by the Phase I findings would be barred.”). The

Florida Supreme Court thus knew all along that progeny plaintiffs would be unable

to decipher what the Phase I jury had decided—“the Engle jury did not make

detailed findings for which evidence it relied upon to make the Phase I . . .

findings.”187 Id. Though the Court always knew that the Phase I findings were

useless for issue-preclusion purposes, it nevertheless intended those findings to be

binding for claim-preclusion purposes. The same findings that admittedly could

not be relied upon to identify something the Phase I jury had been asked to

186
The findings would be relevant for impeachment purposes if, for example, a tobacco-
company executive testified that the company never sold a defective cigarette or never acted
negligently.
187
We also highlighted, in the vacated panel opinion in this case, Graham v. R.J.
Reynolds Tobacco Co., 782 F3d. 1261, 1269–70 (11th Cir. 2015), that it would be impossible for
a progeny jury to determine the facts on which the Phase I jury relied in answering “yes” to the
Phase I verdict-form questions:
[W]hen the jury said that all defendants placed cigarettes on the market that were
defective and unreasonably dangerous, was that because the defendants sold
cigarettes containing ammoniated tobacco? Or was it because the defendants sold
cigarettes containing glass filter fibers? The jury could have answered “yes” to
the first question for some defendants and “yes” to the second question for the
others; “yes” to the first question and “no” to second; or “no” to the first question
and “yes” to the second—the answer to the special interrogatory would have been
the same. Under all three scenarios, the jury would have concluded that all
defendants sold defective and unreasonably dangerous cigarettes. But no one
could ever know which defendants produced which brand or brands of cigarettes
with what defect or defects. And that result, the tobacco companies contended,
stretched any application of res judicata past its constitutional breaking point.
188
 Case: 13-14590 Date Filed: 05/18/2017 Page: 189 of 284

determine—tortious conduct—could be relied upon to establish something that

jury was asked not to determine—causation.188

In addition to taking issue with the Court’s preposterous portrayal of the

Phase I findings, Justice Canady also argued that the Court could not apply claim

preclusion “[b]ecause the judgment that emerged from Engle was not a final

judgment on the merits.” Id. at 436 (Canady, J. dissenting).

The majority recites the requirement of claim preclusion for a final
judgment on the merits but then fails to apply that requirement to the
circumstances presented by this case. Here, of course, the Engle
litigation did not result in a final judgment on the merits with respect
to the members of the class. In Engle [III]—stating the obvious—we
specifically acknowledged that “the Phase I jury ‘did not determine
whether the defendants were liable to anyone.’” Engle [III], 945 So.2d
at 1263. The Phase I findings of the jury were determinations of fact
on particular issues; the jury's verdict did not fully adjudicate any
claim and did not result in a final judgment on the merits. The
application of claim preclusion in such circumstances is a radical
departure from the well established Florida law concerning claim
preclusion. And the majority has cited no authority—either within or
outside the class action context—holding that a judgment that
adjudicates only a portion of a claim is entitled to claim-preclusive
effect.

Id. at 438–39 (emphasis added except for “not”).

Again, the Court disagreed: “[T]he Engle judgment was a final judgment on

the merits.” Id. at 433. By invoking claim preclusion, the Douglas III Court

reasoned, the Engle III Court implicitly entered judgment pursuant to the Phase I

188
The Phase I jury was instructed that it was not determining causation because, under
the original Engle trial plan, individual class members would allege and prove in Phase III that
their injuries were caused by defendants’ product defects and/or tortious conduct.
189
 Case: 13-14590 Date Filed: 05/18/2017 Page: 190 of 284

“common liability findings” and “necessarily decided that the approved Phase I

findings are specific enough.” Id. at 429 (citing Engle III, 945 So. 2d at 1255).

Because the defendants’ liability had been properly adjudicated, the Court

reasoned, it was not unusual for “the jury’s findings in the first trial [to be] binding

in the second even if the first trial does not result in a money judgment.” Id. at 434

(citing 3 A. Conte & H. Newberg, Newberg on Class Actions § 9:53 (4th ed. 2012),

which points out that “[n]ot infrequently, actions filed as class actions present

predominating common issues of liability, while proof of damages may remain as

individual issues for the several class members”).

* * *.

After interpreting its opinion in Engle III, the Florida Supreme Court moved

on to its recognizing-court tasks of applying state preclusion law and evaluating

whether doing so would deprive the defendants of their property without due

process of law. Progeny courts had read Engle III as predetermining the res

judicata effect of the Phase I findings, thereby foreclosing them from carrying out

their constitutional duty to examine the Engle procedures to determine whether

those procedures had denied the tobacco companies due process and whether the

party asserting preclusion had established its elements. See, e.g., Martin II, 53 So.

3d at 1067 (refusing to perform recognizing-court tasks because Engle III had

predetermined the res judicata question, and that was good enough for the First

190
 Case: 13-14590 Date Filed: 05/18/2017 Page: 191 of 284

District, “[n]o matter the wording of the findings on the Phase I verdict form.”).

Prodded by the defendants’ briefs, e.g., Reply Brief for Petitioners at 14 –15,

Douglas III, 110 So. 3d 419 (No. SC12-617), the Douglas III Court reintroduced

the recognizing court’s duties under Florida preclusion law and the U.S.

Constitution in progeny cases, observing that

when the judgment of a state court, ascribing to the judgment of
another court the binding force and effect of res judicata is challenged
for want of due process, it becomes the duty of this Court to examine
the course of procedure in both litigations to ascertain whether the
litigant whose rights have thus been adjudicated has been afforded
such notice and opportunity to be heard as are requisite to the due
process which the Constitution prescribes.

Douglas III, 110 So. 3d at 431 (quoting Hansberry, 311 U.S. at 40, 61 S. Ct. at

117). In conducting such an examination, Douglas III explains, a recognizing

court must ascertain whether the litigant was denied “the basic common law

protection against an arbitrary deprivation of property . . . due process [requires].”

Id. at 431 (citing Honda Motor Co., Ltd. v. Oberg, 512 U.S. 415, 432, 114 S. Ct.

2331, 2340–41, 129 L. Ed. 2d 336 (1994)).

The Florida Supreme Court’s first step in its examination was to evaluate

whether defendants were afforded adequate notice. It found that the original trial

plan provided some notice: “The class action trial plan put the Engle defendants

on notice that if the Phase I jury found against them, the conduct elements of the

class's claims would be established, leaving only plaintiff-specific issues for

191
 Case: 13-14590 Date Filed: 05/18/2017 Page: 192 of 284

individual trials.” Id. at 429 (emphasis added). The Court was right, of course,

that the defendants had notice that conduct elements of the plaintiffs’ claims would

be decided in Phase I.189 The Court did not, however, identify any point at which

the defendants were afforded notice that causation and liability would be decided

in Phase I or, as it were, in the Engle III review of the issues decided in Engle II.

The Court also did not comment on whether the defendants were afforded notice

that it was considering overruling its holding in Engle III that “individualized

issues such as legal causation” would be litigated, rather than presumed, in

progeny trials. Engle III, 945 So. 2d at 1268.

Next, the Florida Supreme Court examined whether the defendants were

given an opportunity to be heard. The Court answered that question with a

resounding “yes”—with respect to conduct:

As illustrated by hundreds of witnesses, thousands of documents and
exhibits, and tens of thousands of pages of testimony, the Engle
defendants had notice and the opportunity to defend against all
theories of liability for each of the class's claims in the year-long
Phase I trial. And, as we held in Engle, the Phase I jury's verdict fully
settled all arguments regarding the Engle defendants' conduct. See
Waggoner, 835 F. Supp. 2d at 1273–74 (recognizing the “Phase I trial
was conducted for the explicit purpose of determining issues related to
the [Engle d]efendants’ conduct which were common to the entire

189
That the Court decided to devote attention here is strange because the defendants have
never contended that they were denied notice and opportunity to be heard during the Phase I
trial. In that trial, they enjoyed the same rights the plaintiff class had to present evidence, cross-
examine witnesses, raise objections, and address the jury in opening statement and closing
argument.
192
 Case: 13-14590 Date Filed: 05/18/2017 Page: 193 of 284

class, meaning [they] had every reason to litigate each potential
theory of liability to the fullest extent possible”).

Douglas III, 110 So. 3d at 431.

The Court also found that the defendants had been afforded an opportunity

to be heard, in a generic sense, in the Douglas I trial and on appeal below:

As illustrated by the Douglas trial record, which is tens of thousands
of pages long, individual plaintiffs do not simply walk into court, state
that they are entitled to the benefit of the Phase I findings, prove their
damages, and walk away with a judgment against the Engle
defendants. Instead, to gain the benefit of the Phase I findings in the
first instance, individual plaintiffs must prove membership in the
Engle class. As in this case, proving class membership often hinges
on the contested issue of whether the plaintiff smoked cigarettes
because of addiction or for some other reason (like the reasons of
stress relief, enjoyment of cigarettes, and weight control argued
below). Once class membership is established, individual plaintiffs
use the Phase I findings to prove the conduct elements of the six
causes of action this Court upheld in Engle; however, for the strict
liability and negligence claims at issue here, they must then prove
individual causation and damages.[190] If an individual plaintiff
receives a favorable verdict, it is then subject to appellate review.
Therefore, the Engle defendants receive the same process as any civil
defendant. See Waggoner, 835 F. Supp. 2d at 1273–74 (recognizing
that giving the Phase I findings res judicata effect does not arbitrarily
deprive the Engle defendants of their property because, to gain the
benefit of these findings, individual plaintiffs must first prove class
membership and then, after clearing that hurdle, must prove the
remaining elements of a prima facie case, all of which is subject to
judicial review).

190
Note that here, again, the Court continues to claim that “individual causation” is a
distinct inquiry from proving class membership. It is not. See Martin II, 53 So. 3d at 1066
(allowing the plaintiff to hold the defendant liable simply by proving class membership—his
addiction to an RJR cigarette and a smoking-related injury).
193
 Case: 13-14590 Date Filed: 05/18/2017 Page: 194 of 284

Id. at 432 (emphasis added). Though the Court was satisfied that the defendants

had an opportunity to be heard on whether they had committed tortious acts, it was

silent as to whether the defendants were ever, at any stage, afforded an opportunity

to be heard on the causal connection between their tortious acts and class

members’ injuries. The Court was also silent as to whether the defendants were

afforded an opportunity to be heard on the matters it said were decided in Engle

III, or on whether it should overrule its holding from Engle III that “individualized

issues such as legal causation” would be litigated in progeny trials. Engle III, 945

So. 2d at 1268.

The defendants felt that the Florida Supreme Court, in considering whether

they had been given notice and opportunity to be heard with respect to conduct in

Phase I, had entirely missed the point of their concerns. They were concerned

about notice and opportunity to be heard with respect to causation, not conduct.

They contended that progeny plaintiffs were being “improperly excuse[d] . . . from

having to prove that the Engle defendants’ conduct was a legal cause of their

injuries.” Douglas III, 110 So. 3d at 430. They were also concerned about the

lack of notice and opportunity to be heard afforded to them in Engle III. In that

case, they argued, the Florida Supreme Court acted on its own initiative, without

194
 Case: 13-14590 Date Filed: 05/18/2017 Page: 195 of 284

notifying the parties or allowing them an opportunity to be heard.191 For that

reason, after the Engle III opinion issued, the defendants moved the Court for

rehearing. The Court denied that motion in a one-line order, which it chose not to

publish.

To its credit, the Court did engage, at least in part, with one of the

defendants’ due process concerns. It examined the procedures afforded to the

defendants to determine whether they had been denied any “basic procedural

protections of the common law.” Oberg, 512 U.S. at 430, 114 S. Ct at 2340.

Specifically, it examined the defendants’ claim that the Constitution requires that

issues be actually decided before they may be given preclusive effect. As

mentioned above, the Court responded to that argument by insisting that

recognizing progeny courts apply claim preclusion to preclude causes of action not

issues. “[C]laim preclusion, unlike issue preclusion,” the Court reasoned, “has no

‘actually decided’ requirement.’” Douglas III, 110 So. 3d at 435. Although the

Court addressed the absence of the actually decided protection, it failed to address

191
Consider, for example, the Engle III Court’s decision to certify an issues class
pursuant to Fla. R. Civ. P. 1.200(d)(4)(A), and then instruct the trial courts to give the Phase I
findings res judicata effect in the progeny cases. Those issues were not before the Court in Engle
III. Certifying a class under that rule is a job belonging to the trial courts, not the appellate
courts. Aside from that, the plaintiff class had not raised the certification issue in its petition for
review or, after the Court granted review, in its brief on the merits. Nor had the defendants
raised it in their answer brief. Nor did the Court broach the subject at oral argument. Moreover,
certifying the class was not needed to enable the Court to dispose of the petition for review. The
issue did not surface until the Court took the case under submission following oral argument and
then issued the Engle III opinion.
195
 Case: 13-14590 Date Filed: 05/18/2017 Page: 196 of 284

whether the defendants had been denied basic common-law protections when the

rendering Engle III Court dictated the preclusive effect of the Phase I findings to

recognizing progeny courts; when class plaintiffs were allowed to assert claim

preclusion—an affirmative defense under Florida law192—offensively; or when

class plaintiffs had been spared their common-law burden of proof on the elements

of both preclusion and causation.

* * *

In wearing both rendering- and recognizing-court hats in Douglas III, the

Florida Supreme Court found itself in a conflict-of-interest position, which called

into question its ability to be impartial. By pronouncing the “res judicata effect” of

the Phase I findings in Engle III, the Court had signaled that those findings would

be useful to class plaintiffs such that they would not have to relitigate the

defendants’ conduct. Under the reasoning of Brown II and Jimmie Lee Brown II, it

had not turned out that way—in order to prove that the defendants’ tortious

conduct caused a plaintiff’s injury, the plaintiff would have to identify the tortious

conduct. The useless Phase I findings provided no way to do that, so litigation

over conduct would begin anew. The Florida Supreme Court was thus faced with a

choice: stick with what Engle III said, engage head on with the defendants’ due

process concerns, and admit mistake; or insist that Engle III meant something

192
Fla. R. Civ. P. 1.110.
196
 Case: 13-14590 Date Filed: 05/18/2017 Page: 197 of 284

different from what it said, dodge the defendants’ due process concerns, and

pretend the mistake never happened. Unfortunately, the Court chose the latter

option.

This conflict, in turn, gave rise to another. In defending its portrayal of the

Phase I findings, the Douglas III Court proffered evidence, on behalf of the

plaintiff, from the Phase I trial record. It then evaluated that evidence using a

lower standard of review than the law requires193 to support the conclusion it had

presupposed. But juries make findings by answering questions, not by looking at

evidence, so the very premise of the Douglas III Court’s sufficiency-of-the-

evidence evaluation was flawed.

The Douglas III Court’s conflicts caused it to disregard Engle III while

purporting to interpret it. Engle III says that the Phase I findings are entitled to res

judicata effect. Those findings, that Court made clear, “did not determine whether

the defendants were liable to anyone.” 945 So. 2d at 1262–63 (quoting Engle II,

853 So. 2d at 450). That liability, according to the Engle trial plan, would be

established in the Phase III trials, where “the remaining issues, including individual

causation and apportionment of fault among the defendants,” would be litigated.

Id. at 1254. That the Phase I Circuit Court rejected proposed verdict-form

193
Determinations about what a jury actually decided must be made on the basis of
necessary inference, not, as the Douglas III Court seems to believe, on the basis of sufficiency of
the evidence. See supra note 136 and accompanying text.
197
 Case: 13-14590 Date Filed: 05/18/2017 Page: 198 of 284

questions about proximate causation corroborates the obvious proposition that it

was both inappropriate and impossible for the Phase I jury to determine whether

the defendants were liable to hundreds of thousands of absent class members.194

Consistent with the reality of the Phase I trial and the necessary implications

of Engle III’s no-liability statement, Brown II and Jimmie Lee Brown II interpreted

Engle III’s res judicata statement as an instruction to recognizing courts to give

issue preclusive effect to the Phase I factual findings. As Brown II had discovered,

and as Douglas III acknowledged, however, those factual findings were “useless.”

Douglas III, 110 So. 3d at 433. Had the Florida Supreme Court in Engle III made

an embarrassing mistake by signaling otherwise?

Martin II’s reasoning seemed to provide an escape. The Engle III Court had

clearly intended to throw the plaintiffs a bone, so a bone they would get. If

progeny courts could preclude the defendants from defending their entire cause of

action, Martin II reasoned, the uselessness of the findings would be rendered

irrelevant. Claim preclusion and reinterpreting the Phase I findings were the key.

So Douglas III—faced with the prospect of an embarrassing mea culpa—

adopted Martin II’s reasoning along with its conclusive presumptions.195 But, as

Justice Canady observed, that reasoning is only half baked—claim preclusion

194
See supra note 178 and accompanying text.
195
See supra note 141 and accompanying text.
198
 Case: 13-14590 Date Filed: 05/18/2017 Page: 199 of 284

requires a final judgment, and Engle III makes it clear that plaintiffs do not have

one. In fact, Engle III holds that “individualized issues such as legal causation”

had yet to be litigated. 945 So. 2d at 1268. Douglas III simply overrules that

holding through the sheer force of its own authority as the state’s highest court:

“[t]he Engle judgment” is “a final judgment on the merits” because “the class jury

resolved . . . the Engle defendants’ common liability to the class under several

legal theories.”196 Id. at 433–34.

But even this stunning reversal did not get the Florida Supreme Court out of

Dodge. The defendants had some compelling due process concerns that were

made even more compelling by the Court’s reversal and reliance on claim

preclusion. Why, for example, were they not entitled to an opportunity to be heard

as to whether their tortious conduct caused the plaintiffs’ injuries? Why were they

never given notice before the Phase I trial or before the Engle III decision was

handed down that the Phase I findings would hold them liable to all class

plaintiffs? Why was it okay to relieve plaintiffs of their common-law burden of

proof and allow them to use an affirmative defense offensively?

Luckily, for the Florida Supreme Court, it was in a position to avoid such

uncomfortable questions. When the defendants said “no notice and opportunity to

196
The contradiction between Engle III on its face (and as interpreted by Brown II) and
Engle III as interpreted by Douglas III illustrates the analytical and substantive differences
between issue preclusion, which Engle III appears to invoke, and claim preclusion, which the
Douglas III court claims Engle III invoked.
199
 Case: 13-14590 Date Filed: 05/18/2017 Page: 200 of 284

be heard on causation,” the Court simply redirected: “notice and opportunity to be

heard on conduct.” When the defendants said “issues must be actually decided,”

the Court said, “What issues? All we can see are causes of action litigated to

completion.”

Thus, to the Douglas III Court, Engle III was nothing but a code. Where

Engle III says “no final judgment,” it means “final judgment.” Where Engle III

says “res judicata to factual findings,” it means “res judicata to causes of action

litigated to completion.” Where Engle III holds that “legal causation would be

litigated in progeny trials,” it means “legal causation would be presumed in

progeny trials.” The Martin II Court understood the code and ran with it. The

Jimmie Lee Brown II Court begrudgingly accepted it. We, in Brown II, naively

believed that Engle III meant what it said.

Why did the Florida Supreme Court resort to a code in Engle III as opposed

to simply saying what it meant? Was it attempting to punish unpopular defendants

and benefit sympathetic plaintiffs while concealing the constitutional shortcuts it

took to do so? Was it attempting to legislate a ban on cigarettes while cloaking the

resulting preemption problems?

Consider, for a moment, what may have been running through the minds of

the justices that comprised the Engle III majority as they contemplated how to draft

200
 Case: 13-14590 Date Filed: 05/18/2017 Page: 201 of 284

their majority opinion. Douglas III tells us that the Engle III majority197 knew that

the Phase I findings were “useless.” See id. at 433 (insisting that the Engle III

majority deliberately opted for claim preclusion over issue preclusion because it

knew the Phase I findings would be “useless in individual actions”). Despite the

reams of evidence that plaintiffs had presented against the defendants, the jury had

not been properly instructed, the Phase I verdict form had not been properly

formulated, and the yearlong Phase I trial had accordingly been a waste. But why,

the majority may have thought, should plaintiffs suffer for the trial judge’s

incompetence?198 Defendants should be held accountable, and we should do what

we can to avoid such an embarrassing failure in a high-profile Florida state-court

case.

But animating litigation that had so badly faltered was not an easy task. The

majority did not even try, as some progeny courts would, to breathe life into the

197
Three of the justices in the Douglas III majority, Justices Lewis, Pariente, and Quince,
were members of the four-justice Engle III majority.
198
I do not wish to be overly hard on the trial judge. Any incompetence resulted from
following plaintiffs’ counsel’s lead at every turn, especially in framing the generic and
nonspecific special interrogatories to the jury. See supra Part I.B.1. It was the judge’s
responsibility to think structurally about this litigation. While plaintiffs may have wanted
broader special interrogatories to increase their chances of success, the trial judge knew that the
answers to these interrogatories would have to be used in subsequent proceedings by different
juries and should have acted accordingly.
201
 Case: 13-14590 Date Filed: 05/18/2017 Page: 202 of 284

useless Phase I findings by creatively interpreting them in light of the Omnibus

Order199 or the Phase I trial record.200

Further, the Phase I jury, in accordance with the trial plan, was instructed not

to evaluate the defendants’ liability to all class members.201 To nevertheless hold

that the jury’s findings established the defendants’ liability to all class members

would too obviously run roughshod over the defendants’ jury-trial rights.202 It

would also too obviously implicate due process concerns: The parties had filed

their briefs and orally argued their positions on the issues the Supreme Court

accepted for review in Engle III. Those issues did not include adjudicating liability

at that stage of the case, and parties did not argue, nor did they have an opportunity

to argue, whether adjudicating liability to all class members was appropriate.

199
The Martin II Court took this approach. See supra Part III.C.2.
200
The Majority in this case take this approach. See infra Part VI.
201
See supra note 177 and accompanying text.
202
Article I, section 22 of the Florida Constitution provides that “the right of trial by jury
shall be secure to all and remain inviolate.” Fla. Const. art. I, § 22. Parties have a jury-trial
right with respect to issues that are legal, as opposed to equitable, in nature. Yer Girl Tera Mia v.
Wimberly, 962 So. 2d 993, 996 (Fla. 5th Dist. Ct. App. 2007). This right “should not be
withdrawn from the jury’s consideration unless as a matter of law no proper view of the evidence
could possibly sustain” an alternative determination.” Bourgeois v. Dade Cty., 99 So. 2d 575,
577 (Fla. Div. A 1956).
Here, the causation elements of the class members’ causes of action presented factual
issues. No jury ever considered whether the Engle defendants’ tortious conduct caused all class
members’ injuries. The trial plan called for the Phase III juries to decide the issue, in what
ultimately became the progeny cases. This included the present case, Graham. Whether the
defendants’ tortious conduct caused the plaintiff’s injuries would have been hotly contested in
every case. The issue could not be withdrawn from the jury’s consideration without denying the
defendants their jury-trial right.
202
 Case: 13-14590 Date Filed: 05/18/2017 Page: 203 of 284

Requesting briefs on the question would only shed light on obvious constitutional

barriers to the majority’s desired outcome.

So whether they did so intentionally or not, the Engle III majority achieved

surreptitiously what they could not have achieved openly. Their res judicata

instruction, which they issued without input from the parties,203 expertly toed the

line between subtlety and clarity. To start, “res judicata” is an ambiguous term that

embraces both issue and claim preclusion.204 Using that term thus obscured that

the Court had invoked claim preclusion and had thereby implicitly adjudicated

defendants’ liability to all class members.205 The Court further obfuscated this idea

by observing that “the Phase I jury ‘did not determine whether defendants were

liable to anyone.’”206 Engle III, 945 So. 2d at 1263 (citing Engle II, 853 So. 2d at

450). Moreover, its instruction was dicta,207 and this allowed the class

representatives to successfully argue in their brief in opposition to the Engle

defendants’ petition for a writ of certiorari that the defendants’ “due process

203
Requesting briefs on this issue would have shed unwanted light on common-law and
constitutional impediments to the majority’s desired outcome.
204
A court’s “limit[ing] the res judicata phrase so as to exclude the doctrines of issue
preclusion or collateral estoppel. . . . is potentially confusing, and it is better to use res judicata in
its broader sense to encompass both sets of doctrine.” 18 Charles Alan Wright & Arthur R.
Miller, Federal Practice and Procedure § 4402 (3d ed. 2016).
205
So effective was this obfuscation that the Majority apparently still believe that the
Engle III Court invoked issue preclusion rather than claim preclusion, despite Douglas III’s
contrary clarification. See infra Part VI.
206
The Engle III Court did not say whether it, as opposed to the Phase I jury, had
determined that the defendants were liable to class members.
207
See supra notes 77, 89, and accompanying text.
203
 Case: 13-14590 Date Filed: 05/18/2017 Page: 204 of 284

concerns are premature and not ripe for review.” Brief in Opposition, R.J.

Reynolds Tobacco Co. v. Engle, 552 U.S. 941, 128 S. Ct. 96, 169 L. Ed. 2d 244

(2007) (No. 06-1545), 2007 WL 2363238, at *1. As the class representatives

argued, “[N]o court has yet [applied] the findings” and recognizing courts

ordinarily decide for themselves whether to give res judicata effect to a rendering

court’s findings.208 Id.

Amidst all of its ambiguity and doublespeak, Engle III was still clear enough

to carry a binding message to progeny courts: adjudicate class members’ claims

without requiring them to relitigate the issues litigated in Phase I. In other words,

hold defendants liable to all class members.209 The Fourth District felt

“constrained” by Engle III’s message, Jimmie Lee Brown II, 70 So. 3d at 715, and

the First District embraced it, Martin II, 53 So. 3d at 1066–67. Even our court

deferred to it. Walker II, 734 F.3d at 1289.

Did the Engle III majority really intentionally mastermind such a deviously

clever opinion? In doing so, did they deliberately forego asking the parties for

briefs because they did not want to shed light on the common-law and

constitutional obstacles that stood in the way of their objective? If they did—and

Douglas III invites that conclusion—the Engle III Court, in a troublingly

208
See supra note 77 and accompanying text.
209
Because causation cannot be properly litigated without identifying tortious conduct,
see supra note 140 and accompanying text, progeny courts, tasked with carrying out Engle III’s
mandate, would have to treat both conduct and causation as already adjudicated.
204
 Case: 13-14590 Date Filed: 05/18/2017 Page: 205 of 284

calculated way, deprived the defendants of their property without due process of

law and entered a final judgment in a controversy that did not exist, one the Court

itself contrived. Had the Court acknowledged this in Engle III, is there any doubt

the U.S. Supreme Court would have granted certiorari review? If the Engle III

majority had the Douglas III result in mind, they could not let it see the light of

day. As Justice Brandeis remarked, “Sunlight is said to be the best of

disinfectants; electric light the most efficient policeman.”210

Whatever result Engle III had in mind, Douglas III ushered in a new Florida

law211 under which tobacco manufacturer’s liability would now be established

differently from that of all other tort defendants.212 The “highly individualized”

210
Louis D. Brandeis, Other People’s Money and How the Bankers Use It 92 (1914).
211
It is obvious that Douglas III is not a mere interpretation of Engle III. Douglas III
stripped the Engle III opinion of its dicta and in its place rendered a judgment banning all
cigarettes under Florida tort law. Engle III and Douglas III shared the same record—the Phase I
trial transcript. In Engle III, the Supreme Court was sitting as a rendering court. It had
jurisdiction to review issues the Third District had decided in Engle II. Whether and to what
extent the class members should have the benefit of the facts depicted in the Phase I findings in
prosecuting their individual tort actions was not one of the issues litigated in Engle II, and thus
was not before the Court in Engle III. That issue would be decided in the first instance by a
progeny court carrying out its recognizing-court duties under Hansberry and Florida preclusion
law. In consequence, the Engle III statement that the Phase I findings “will have res judicata
effect” was, under Florida preclusion law, absolutely meaningless.
212
In asserting that the “Due Process Clause does not require a state to follow the federal
common law of res judicata and collateral estoppel,” Ante at 26, the Majority suggest that my
position presumes otherwise. Assuming this suggestion is an oversight rather an intentional red
herring, the Majority fail to recognize that my extensive overview of the law of res judicata and
collateral estoppel in Parts II.A and II.B details the common aspects of these doctrines in every
jurisdiction including Florida, rather than the federal common law. The Majority also reason that
the array of uniformly followed principles that Florida’s new preclusion law eradicates were not
sufficiently engrained in the common law such that their removal constitutes a violation of due
process under Oberg 512 U.S. at 430, 114 S. Ct. at 2339. I do not contend that every uniformly
205
 Case: 13-14590 Date Filed: 05/18/2017 Page: 206 of 284

differences between smokers, which were not amenable to “class action

treatment,” Engle III, 945 So. 2d at 1263, and were to be dealt with in the class

members’ individual Phase III actions, would no longer matter. Nor would it

matter what brands the smoker used. It would not matter which of the defendants’

advertisements she had seen or not seen. In fact, it would not matter if she had

never seen a manufacturer’s advertisement. It would not matter if she grew up in a

house of smokers or smoked because her friends always smoked. Most simply, it

would not matter why the smoker chose to smoke. Likewise, it would not matter if

the defendants breached a duty of care to the smoker. And it would not matter

whether the defendants’ (unidentified) breach of duty caused the smoker any harm.

Now, under Florida tort law as set out in Douglas III, a plaintiff can recover

damages from an Engle defendant by merely by proving her status as a class

member by establishing that she contracted a smoking-related illness.213 She does

not have to prove that the cigarettes she smoked were defective or negligently sold,

nor that the defect or the act of negligence caused her harm. That cause is

assumed conclusively. Douglas III, 110 So. 3d at 429 (When a plaintiff “prov[es]

followed feature of preclusion law is necessitated by due process. However, I do contend that
the principles elucidated in Oberg are yet another reason why the Majority should give more
thought to the due process question at hand than it does with its simple reassurance that the
principle of due process negates “any concept of inflexible procedures.” Ante at 27 (citation
omitted).
213
Addiction is a requirement of class membership; it is not a requirement of the tort law
Douglas III created, under which class members establish a defendant’s liability merely by
proving that she smoked the defendant’s cigarettes and suffered injury as a result.
206
 Case: 13-14590 Date Filed: 05/18/2017 Page: 207 of 284

that addiction to the Engle defendants’ cigarettes containing nicotine was a legal

cause of the injuries alleged,” any “injury as a result of the Engle defendants’

conduct is assumed.”) (emphasis added)). Thus, Douglas III’s conclusive

presumption treats all class members as one, relieving all of the burden of proving

that the defendant’s tortious conduct caused their injury.

In sum, Douglas III did three things. First, it established an

unconstitutional214 conclusive presumption215 that all of the cigarettes smoked by

214
See supra note 6 for a description of Henderson, the applicable Supreme Court case.
Here, state common law set forth by the Florida Supreme Court allows progeny plaintiffs
to hold defendants liable despite never presenting evidence that the cigarettes they smoked were
inter alia defective, unreasonably dangerous, and negligently produced. Because plaintiffs’
injuries may have been caused by cigarettes that were never deemed defective, unreasonably
dangerous, or negligently produced, the state-law presumption that allows plaintiffs to presume
otherwise is unreasonable and arbitrary.
215
The Majority argue that the District Court’s application of an irrebuttable presumption
of liability to hold the defendants liable to Mr. Graham is appropriate since “‘state proceedings
need do no more than satisfy the minimum procedural requirements’ of due process to receive
full faith and credit.” Ante at 29 (citing Kremer, 456 U.S. at 481).
The Majority apply the wrong legal standard. It was not until Douglas III that the Florida
Supreme Court implemented and ordered recognizing courts to apply a baseless presumption of
liability and causation in Engle-progeny cases as a matter of Florida law. See supra notes 214–
25 and accompanying text. Florida preclusion law—both claim and issue preclusion—requires
mutuality of parties. Florida Dep’t of Transp. v. Juliano, 801 So. 2d 101, 105 (Fla. 2001).
Because Ms. Graham was not a party to Douglas III, the Majority cannot give full faith and
credit to that decision. Instead, when the Majority rely on Douglas III, they are applying it as “a
matter of state law,” Ante at 30, under Erie. However, “Erie [only] mandates that a federal court
sitting in diversity apply the substantive law of the forum State, absent a federal statutory or
constitutional directive to the contrary.” Salve Regina College v. Russell, 499 U.S. 225, 226, 111
S.Ct. 1217, 113 L.Ed.2d 190 (1991) (emphasis added). The substance of the law that the
Majority apply is an irrebutable presumption that the defendants’ breached their duty of care to
Ms. Graham, and an irrebutable presumption that such unidentified breach of duty caused Ms.
Graham’s harm. Such baseless presumptions did not comply with due process in 1929, see
supra note 141, and they do not comply with due process now. Therefore, we may not apply the
unconstitutional state law “articulated in Douglas [III],” Ante at 17, under Erie. See also
207
 Case: 13-14590 Date Filed: 05/18/2017 Page: 208 of 284

the class members were defective and unreasonably dangerous and negligently

sold.216 The Court did this not on the basis of the useless Phase I findings, but on

the basis of the evidence adduced during the Phase I trial (as depicted in the Engle

trial judge’s Omnibus Order).217 Indulging this presumption had the same effect as

deciding that all cigarettes were defective and unreasonably dangerous and

negligently sold as a matter of law. However, since the evidence on those issues

was at variance, the Court deprived the defendants of their constitutional right to a

jury trial. See supra note 142 and accompanying text.

Marbury v. Madison, 5 U.S. 137, 177–80 (1803) (“If two laws conflict with each other, the
courts must decide on the operation of each. . . . [A] law repugnant to the constitution is void.”).
216
The Court established a conclusive presumption regarding the elements of all of the
Engle III-approved tort claims.
217
It is clear that the Phase I jury did not find that all of the cigarettes the Engle
defendants sold were defective and unreasonably dangerous and were sold due to the defendants’
negligence. The jury did not find those facts because, as Justice Canady pointed out, the
plaintiffs’ counsel prevented the jury from finding them. Addressing the plaintiffs’ strict liability
claim, Justice Canady said this:
The plaintiffs pursued their claims in Phase I based on several alternative theories of
defect, some of which applied only to certain brands and designs. Given this context, it is
unreasonable to read the jury's finding that the defendants “placed cigarettes on the
market that were defective and unreasonably dangerous” as a finding that all of the
cigarettes placed on the market by the defendants were defective and unreasonably
dangerous.
The plaintiffs very easily could have sought such a broad, all-encompassing
finding by proposing a slightly altered jury verdict form which referred to all of the
cigarettes placed on the market by the defendants. The plaintiffs failed, however, to do
so. Whether that failure was inadvertent or calculated, it was the plaintiffs' responsibility
and cannot be laid at the door of the defendants. The attempt to lay it at the defendants'
door by way of the doctrine of claim preclusion is ill-conceived.

Douglas III, 110 So. 3d at 437 (Canady, J., dissenting).

208
 Case: 13-14590 Date Filed: 05/18/2017 Page: 209 of 284

Second, Douglas III utilized this conclusive presumption as the foundation

for a second unconstitutional218 conclusive presumption—that the Engle

defendants’ tortious conduct (presumed present in all cases) caused the class

members’ injuries—in order to enable the class members to establish the causation

element of their tort claims.219 Without the presumption, only the Phase I

findings—useless in resolving the causation issue—supported the claims. Why?

Because the Phase I jury was instructed that its role was not to decide whether the

defendant’s tortious conduct, if any, caused a class member’s injury. Rather, its

role was limited to answering the special interrogatories, all of which dealt with the

218
As detailed above, in Henderson, 279 U.S. at 643, 49 S. Ct. at 447, the Supreme Court
held that a defendant railroad company’s due process rights were violated where it was held
liable even though the plaintiff offered no evidence of a connection between tortious conduct and
the injury at issue. Id. at 640–44, 49 S. Ct. 445–48. Instead of presenting such evidence, the
plaintiff relied on a state-law presumption that “[t]he mere fact of collision between a railway
train and a vehicle . . . was caused by negligence of the railway company.” Id. at 642–43, 49 S.
Ct. 445, 447. Because, as a factual matter, a collision could result from “negligence of the
railway, or of the traveler on the highway, or of both, or without fault of any one,” the Supreme
Court struck down the presumption as “unreasonable and arbitrary.” Id. at 644, 49 S. Ct. 445,
447.

Similarly here, class plaintiffs have not been required to allege and have not attempted to
prove a specific connection between the Engle defendants’ tortious conduct and their injury. All
they have proffered is a complaint that cites Engle III and pleads verbatim the Phase I findings.
Progeny trial courts have applied a conclusive presumption under Florida common law, which
provides that the mere fact of a plaintiff’s smoking-related injury conclusively establishes that
his injury was caused by the defendants’ tortious conduct. Because plaintiffs’ injury may very
well have been caused by cigarettes’ non-defective-and-unreasonably-dangerous features and by
the defendants’ nontortious conduct, the presumption is unreasonable and arbitrary.
219
The Florida Supreme Court in Engle III had described causation as being one of those
“individualized issues” that made “continued class action treatment for Phase III of the trial plan
. . . not feasible.” 945 So. 2d at 1268. Yet, Douglas III declares that causation is to be presumed
in all progeny cases. See supra Part IV. Are we still to believe that Douglas III is an
interpretation of Engle III?
209
 Case: 13-14590 Date Filed: 05/18/2017 Page: 210 of 284

conduct alleged in the class action complaint. A progeny jury would determine

whether the defendant’s tortious conduct caused a class member’s injury. And as

the defendants noted over a decade ago, it is impossible to properly determine

causation because the findings do not identify the defect or tortious conduct that

the jury found. In approving Douglas II’s affirmance of the Circuit Court’s use of

a conclusive presumption to establish the cause of Ms. Douglas’ death, the Florida

Supreme Court denied the defendant’s right to submit the causation issue to a jury.

Third, in creating these conclusive presumptions, Douglas III materially

altered the laws of products liability and negligence as they relate to the

manufacture and sale of cigarettes, ultimately making the sale of cigarettes

unlawful. Douglas III also overruled Engle III’s holding that “individualized

issues such as legal causation [and] comparative fault” would be litigated in

progeny trials. Engle III, 945 So. 2d at 1268.

The effect of Douglas III’s holdings is that tobacco companies now have a

tort-law duty not to sell cigarettes in the state of Florida. As sanctions for

breaching this duty, an Engle defendant must pay damages—and possibly punitive

damages—to any class member who can satisfy a jury that she is addicted to its

cigarettes220 and suffered a smoking-related disease.221

220
Addiction to cigarettes is merely the jurisdictional hook—the substance of the
operative tort law makes the Engle defendants liable for every smoking-related injury. I see no
210
 Case: 13-14590 Date Filed: 05/18/2017 Page: 211 of 284

V.
THE WALKER PANEL EFFECTIVELY REWROTE AND THEN GAVE FULL
FAITH AND CREDIT TO DOUGLAS III BEFORE ISSUING A NEW OPINION
THAT GAVE FULL FAITH AND CREDIT TO ENGLE III, YET LEFT THE
ORIGINAL OPINION’S INAPPOSITE REASONING INTACT

Like the progeny plaintiffs in Waggoner, the plaintiffs in Walker I222—suing

on behalf of deceased relatives—amended their complaints in the immediate

aftermath of Martin II to capitalize on its claim-preclusion holding. Mr. Walker’s

second amended complaint characterized Engle III’s res judicata dicta223 as a

“mandate” that “provided Plaintiff the opportunity to complete unresolved

damages claims.”224 Second Amended Complaint at 2, 3, Walker v. R.J. Reynolds

Tobacco Co. (Walker I), No. 3:09-cv-10598-RBD-JBT (M.D. Fla. Mar. 3, 2011)

(emphasis added). The plaintiffs assumed, in accordance with Martin II, that all

that remained to be resolved in his case was damages; the defendants were

reason why litigants outside of the Engle class should be unable, under Florida law, to sue any
tobacco manufacturer for any smoking-related injury.
221
It could be argued that Douglas III’s determination that all cigarettes are
presumptively defective and negligently sold would render an Engle defendant liable to anyone
who contracts disease caused by smoking. In the Court’s mind, cigarettes are unreasonably
dangerous due to the serious health problem they create. Smoking causes disease to addicted and
nonaddicted smokers alike. Addiction is relevant here because the class that was certified was
limited to smokers who could not quit.
222
In Walker II, “R.J. Reynolds challenged . . . the jury verdicts in favor of two plaintiffs,
Alvin Walker and George Duke III. 734 F.3d. at 1286.
223
See infra notes 77, 89, and accompanying text.
224
Mr. Duke’s complaint was basically identical. Second Amended Complaint, Duke v.
R.J. Reynolds Tobacco Co., No 3:09-cv-10104-TJC-JBT (M.D. Fla. Mar. 3, 2011).
211
 Case: 13-14590 Date Filed: 05/18/2017 Page: 212 of 284

precluded by the “mandate” in Engle III from litigating both (1) that the cigarettes

smoked by the decedents were defective, unreasonably dangerous, and negligently

produced and (2) that the defendants’ tortious conduct caused the decedents’ death.

Id. Accordingly, they pleaded only class membership—addiction and a smoking-

related injury. Id. at 4. To support their assertion of claim preclusion, they

proffered only the Engle III opinion and the “Common Liability Findings”

approved by the Engle III Court. Id. at 5–9.

In their answers, the defendants took issue, as they had in previous progeny

cases, with the plaintiffs’ meager pleadings. RJR insisted that the Phase I findings

were not liability findings at all; rather, they “are so generalized and nonspecific

that they are inadequate to support an individualized determination of essential

issues such as liability, legal causation, and damages in this or any other

subsequent individual action.” Answer, Defenses and Jury Demand of Defendant

R.J. Reynolds Tobacco Co at 2, Walker I, No. 3:09-cv-10598-RBD-JBT. In

particular, RJR argued that the Phase I finding related to product defect(s) was

deficient because it provided no basis on which a

subsequent court or fact finder [could] determine whether any
product, brand, type, or design used by a particular plaintiff was found
defective (or not defective) by the [Phase I] jury or whether any such
design characteristic found defective by the [Phase I] jury caused this
Plaintiff’s and/or Plaintiff’s Decedent’s injuries.

212
 Case: 13-14590 Date Filed: 05/18/2017 Page: 213 of 284

Id. at 3. Similarly, with respect to the Phase I negligent-conduct finding, RJR

contended that

[n]o subsequent court or fact finder can determine whether the acts or
omissions alleged by this Plaintiff . . . were found negligent . . . by the
[Phase I] jury or whether any conduct found to be negligent by the
[Phase I] jury was a legal cause of injury to a particular plaintiff.

Id. at 5–6. Philip Morris and Lorillard echoed such sentiments, maintaining that

proving a connection between tortious conduct and injury on the basis of the Phase

I findings was “impossible.” Philip Morris USA Inc.’s Answer to Plaintiff’s

Second Amended Complaint and Demand for Trial by Jury at 16, 18, Walker I, No.

3:09-cv-10598-RBD-JBT; Lorillard Tobacco Co.’s Answer to Plaintiff’s Second

Amended Complaint and Demand for Jury Trial at 12, 14, Walker I, No. 3:09-cv-

10598-RBD-JBT.

As in previous cases, these objections triggered the District Courts’

Hansberry obligation under Florida law and the U.S. Constitution to examine the

proceedings to ensure that applying Florida claim preclusion would not violate the

tobacco companies’ due process rights.225 See supra Part II.B. Relying on

225
Recall that the District Courts in this posture were required to examine both whether
the defendants were provided with sufficient notice and opportunity to be heard on the elements
of their claims that were purportedly decided by the Engle III Court, and whether applying state
preclusion law to foreclose the defendants from contesting these elements would deny them due
process of law in the recognizing court. As explained in Part II.B, Florida law imposes the same
duties on recognizing courts as the U.S. Constitution. See Douglas III, 110 So. 3d at 430–31
(explaining that when a due process objection is raised it becomes the “duty of [the recognizing
court] to examine the course of procedures in both litigations to ascertain whether the litigant
whose rights have thus been adjudicated has been afforded . . . due process” (quoting Hansberry,
213
 Case: 13-14590 Date Filed: 05/18/2017 Page: 214 of 284

Waggoner’s pretrial ruling, the District Courts concluded that an additional

examination was unnecessary.226 They gave preclusive effect to Engle III and

accepted the plaintiffs’ claim-preclusion proffer of Engle III and its approved

findings as adequate to establish that the Engle defendants were liable to all class

members, including Walker and Duke. Philip Morris and Lorillard later settled,

but the plaintiffs’ damages cases against RJR went to a jury. In accordance with

Martin II and the plaintiffs’ pleadings, juries in both cases were instructed to hold

RJR liable if the plaintiffs proved their membership in the Engle class—“addiction

to smoking the defendant’s cigarettes and resulting injury.” Martin II, 53 So. 3d at

1069. The jury in Walker I “found in favor of Walker on the claims of strict

liability and negligence” and the jury in Duke “found in favor of Duke only on the

claim of strict liability.” Walker II, 734 F.3d at 1286. RJR appealed the District

Court judgments to our Court.

While these appeals were pending, the Florida Supreme Court decided

Douglas III, which, as explained above, portrayed the Phase I findings as liability

determinations and endorsed Martin II’s retroactive substitution of claim

311 U.S. at 40, 61 S. Ct. at 117)); Id. at 431 (explaining that a recognizing court must ascertain
whether the litigant was denied “the basic common law protection against an arbitrary
deprivation of property . . . due process [requires]” (citing Oberg, 512 U.S. at 432, 114 S. Ct. at
2340–41)).
226
The panel noted that the Martin II Court, applying claim preclusion, “disagreed with
our decision in Brown [II],” and that that the Court’s new characterization “supplanted our
interpretation of Florida law” in Waggoner under Erie. Walker II, 734 F.3d at 1284.
214
 Case: 13-14590 Date Filed: 05/18/2017 Page: 215 of 284

preclusion for issue preclusion as the Engle litigation model.227 Before it addressed

RJR’s arguments, the Walker panel thus had to decide between two basic

portrayals of the Phase I findings—Brown II’s depiction of the findings as factual

determinations regarding the defendants’ conduct and Douglas III’s depiction of

the findings as full-blown liability determinations.228 Acknowledging clear

language on the Phase I verdict form and in Engle III that contradicted Douglas

III’s portrayal, the panel adopted Brown II’s basic depiction.229 See Walker v. R.J.

Reynolds Tobacco Co. (Walker 0) at 18, Nos. 12-13500, 12-14731 (11th Cir. Sept.

6, 2013) vacated and superseded by Walker II, 734 F.3d 1278 (“During [Phase I],

the jury considered only ‘common issues relating exclusively to the defendants’

conduct . . . but did not decide whether the tobacco companies were liable to any of

the class representatives or members of the class.” (citing Engle III, 945 So. 2d at

1256, 1263)).230

227
In doing so, the Court starkly rejected the application of issue preclusion in progeny
cases, as the “useless” Phase I findings made it impossible to deduce what the Engle jury
“actually decided.” See supra Part IV.
228
Under Florida tort law, a defendant is not liable “unless the act complained of is the
proximate cause of the injury.” 55 Fla. Jur 2d Torts §2 (2017). Thus, for the Phase I findings to
be portrayed as liability determinations, the Phase I jury must have determined duty, breach, and
causation.
229
In doing so, the panel avoided addressing the constitutional problems associated with
the Douglas III Court’s significant changes in Florida law. See supra Part IV.
230
As I will explain in detail below, the panel initially issued its Walker 0 opinion in
September 2013. Following RJR’s petition for rehearing, the panel sua sponte vacated Walker 0
and issued Walker II in October 2013. A copy of Walker 0 can be found as an attachment to
RJR’s Petition for Panel Rehearing Walker 0 (Nos. 12-13500, 12-14731).
215
 Case: 13-14590 Date Filed: 05/18/2017 Page: 216 of 284

Upon rejecting Douglas III’s portrayal of the Phase I findings as liability

determinations, the Walker panel could no longer apply claim preclusion in

accordance with Florida law. It could, however, reach Douglas III’s outcome

under the issue-preclusion framework set forth in Brown II, if it could portray the

Phase I conduct findings as “specific enough to apply in favor of every class

plaintiff.” Id. at 21. In other words, if “the tobacco companies acted wrongfully

toward all plaintiffs” because all cigarettes are defective, unreasonably dangerous,

and negligently produced because they “contain nicotine are addictive and produce

dependence,”231 then any error that might have occurred in the Walker I and Duke

trials could arguably be characterized as harmless. Id. That the District Courts

precluded RJR from contesting whether the cigarettes Walker and Duke smoked

were defective, unreasonably dangerous, and negligently produced did not seem

problematic because those issues would have been actually decided by a previous

fact finder. Likewise, that the District Courts precluded RJR from contesting

whether its tortious conduct caused Walker’s and Duke’s injuries seemed

231
The panel hinted at, but did not commit to, this portrayal of the Phase I findings.
Instead, its opinions generally refer to the Phase I findings by using unintelligible generalities.
For example, the panel said, “[T]he Supreme Court of Florida later ruled that the findings of the
jury in the class action have res judicata for common issues decided against the tobacco
companies” without ever defining the term “common issues.” Walker 0, 1; Walker II, 734 F.3d
at 1279.
Without identifying the defect that taints all cigarettes, the panel necessarily denied the
defendants their Seventh Amendment right to a jury determination on a contested element of
their claim, and they adopt, in violation of Henderson, an unconstitutional presumption that
smoking-related injuries are caused by tortious conduct. See supra note 6 and accompanying
text.
216
 Case: 13-14590 Date Filed: 05/18/2017 Page: 217 of 284

acceptable because Walker and Duke had to prove their status as class members by

proving that (supposedly defective and negligently produced) cigarettes caused

their injuries.

The Walker panel did not interpret the “ambiguous [Phase I] jury verdict[s]”

as “specific enough” factual findings. Id. at 23. Rather than simply look at the

questions the Phase I jury was instructed to answer, it relied on the Douglas III

Court to interpret the findings. The Douglas III Court, the panel insisted, “looked

past the ambiguous jury verdict[s] to decide [a] question of fact”; namely, the

Douglas III Court decided that “the approved Phase I findings are specific

enough.” Id. at 23–24. Granted, the panel “disagree[d] with [Douglas III’s]

holding about what the jury in Phase I decided.” Id. at 18. Nevertheless, the panel

concluded that it “could not refuse to give full faith and credit” to the Douglas III

Court’s supposed “merely erroneous” factual determination. Id. (citation omitted).

In its petition for rehearing, RJR rejected the panel’s portrayal of Douglas

III.232 It maintained that the Court in Douglas III never searched the record to

determine what the Phase I jury actually decided. Id. In fact, RJR argued, the

Court conceded that performing such a search was futile: “[A]pplying issue

preclusion—and its ‘actually decided’ requirement—‘would effectively make the

Phase I findings . . . useless in individual actions.’” Id. at 12 (quoting Douglas III,

232
RJR’s primary objection, as explained below, was that the panel’s review of Douglas
III was irrelevant altogether.
217
 Case: 13-14590 Date Filed: 05/18/2017 Page: 218 of 284

110 So. 3d at 423 (Fla. 2013)). RJR elaborated, “Precisely because the court could

not determine exactly what the Engle jury had decided, the court went on to

develop—and defend at length—its novel version of claim preclusion.” Id. Not

only did Douglas III make clear that the “useless” Phase I findings lack evidentiary

value, it distinguished RJR’s “lead due-process precedent as a case about issue

preclusion, and it emphasized the ‘specific importance to this case’ of the fact ‘that

claim preclusion, unlike issue preclusion, has no “actually decided” requirement.’”

Id. at 13 (quoting Douglas III, 110 So. 3d at 435). The Walker panel disregarded

these concerns.

By portraying Douglas III as determining that the Phase I findings were

“specific enough” factual findings, the panel also disregarded Douglas III’s

reliance on claim preclusion and the plaintiffs’ corresponding claim-preclusion

proffer, opting for issue preclusion instead. In light of RJR’s due process

objections, the Walker panel, as a recognizing court, had a duty under both Florida

law and the U.S. Constitution to “examine the course of procedures in both” (1) the

Walker I and Duke trials and (2) the case to which it gave full faith and credit,

Douglas III. Douglas III, 110 So. 3d at 430–31 (quoting Hansberry, 311 U.S. at

40, 61 S. Ct. at 117). The panel found no due process deprivation in Walker I and

Duke because, as explained above, it believed any errors that occurred in those

trials—including the misinterpretation of the Phase I findings and the application

218
 Case: 13-14590 Date Filed: 05/18/2017 Page: 219 of 284

of claim preclusion instead of issue preclusion—were harmless. The panel also

found that the Douglas III decision “did not arbitrarily deprive R.J. Reynolds of

property without due process of law” because the Douglas III Court “looked

through the jury verdict entered in Phase I to determine what issues the jury

decided.”233 Walker 0, at 19. According to the panel, the Douglas III Court

“concluded that the jury was asked only to ‘determine all common liability issues

for the class,’ not brand specific defects,”234 and the panel held that the Douglas III

233
The panel created an elaborate narrative to support its claim that Douglas III
performed fact finding. The first step in Douglas III’s fact finding, according to the panel,
involved a “review of . . . the [Phase I] jury instructions,” Walker 0, at 18; Walker II, 734 F.3d at
1287, and a determination as to what those instructions said. Although the instructions did not
direct the jury to “determine all common liability issues,” the Douglas III Court, the panel
explained, concluded that the instructions did direct the jury to “determine all common liability
issues.” Walker 0, at 19; Walker II, 734 F.3d at 1287. Although the panel did not explain why
such a facially erroneous portrayal of the jury instructions was not arbitrary, it did note again that
the Douglas III Court “was entitled to look beyond the jury verdict to determine what issues the
jury decided.” Walker 0, at 19; Walker II, 734 F.3d at 1287. According to the Walker panel,
after the Douglas III Court revised the jury instructions, it applied the presumption that a “jury
followed its instructions” to make its factual determination that “the jury found only issues of
common liability.” Walker 0, at 20; Walker II, 734 F.3d at 1288.
234
The Walker panel did not accurately portray the Phase I Circuit Court’s jury
instructions, stating that the “trial court instructed the jury that ‘all common liability issues
would be tried before [the jury].’” Walker 0, at 5–6; Walker II, 734 F.3d at 1287 (alteration in
original). But the Majority’s portrayal is even more off base. They assert (like the Court in
Douglas III) that the “trial court instructed the jury to determine all common liability issues for
the class concerning the conduct of the tobacco industry.” Ante at 22 (internal quotation marks
omitted). The trial court provided no such instruction. What the Majority appear to be alluding
to is a reference in the Circuit Court’s explanation to the jurors that the class action lawsuit was
unlike traditional lawsuits in which a “jury hears testimony from the . . . parties bringing the
lawsuit.” In this explanation, the Circuit Court explained aspects of the three-phase trial plan,
and stated, “Because the size of the Florida class and the complexity of the issues, the Court has
determined that all common liability issues would be tried before a jury in a single trial.”
The Circuit Court never defined the term “common liability.” Further, this sentence, of
course, was not an instruction that the jury should determine only “common liability” issues. To
the contrary, shortly after this general description of the trial, the Circuit Court stated, “Members
219
 Case: 13-14590 Date Filed: 05/18/2017 Page: 220 of 284

Court was “entitled to look beyond the jury verdict[s] to determine what issues the

jury decided.” Id. (citations omitted). The panel did not explain whether the

Douglas III Court afforded the parties before it with notice that it was going to

“look beyond the jury verdict[s]” to interpret them. Id. Nor did the panel

comment on whether the Douglas III Court gave the parties an opportunity to be

heard on the appropriate interpretation of Phase I findings. Rather, in a cursory

response to RJR’s due process argument, the panel noted that “if due process

requires a finding that an issue was actually decided, then the [Douglas III Court]

made the necessary finding.” Walker 0, at 24.

After its portrayal and inspection of Douglas III, the Walker panel noted its

clever reconciliation of the disparate Brown II and Douglas III approaches. It

observed that in Brown II, “we stated that, although the jury verdict in Phase I was

ambiguous on its face, members of the Engle class should be allowed an

opportunity to establish that the jury in Phase I actually decided particular issues in

their favor.” Id. at 20. According to the panel, the Douglas III Court took the

plaintiffs’ burden upon itself and decided erroneously, but not arbitrarily, when it

concluded the jury findings were specific enough. Id.

of the jury, I shall now instruct on the law you must follow in reaching your verdict.” (emphasis
added). It continued, “It is your duty as jurors to decide the issues, and only the issues that I
submit for determination by your verdict.” The issues that the Court submitted to the jury—none
of which involved the term “common liability”—are those that are detailed extensively in Part
I.B.1; that is, the Phase I findings.
220
 Case: 13-14590 Date Filed: 05/18/2017 Page: 221 of 284

For all its cleverness, though, the panel lost track of fundamentals. It

reached its creative solution sua sponte,235 never affording the parties an

opportunity to brief it. As RJR argued in its petition for rehearing, neither party

had “argued that the panel was bound by Douglas [III] on any matter: not on its

legal reasoning . . . and certainly not on its supposed factfinding.” RJR Petition for

Panel Rehearing at 29, Walker 0 (Nos. 12-13500, 12-14731).

The panel’s failure to solicit input from the parties had consequences. As

RJR put it, the panel’s full-faith-and-credit analysis was “demonstrably erroneous.”

Id. at 11. RJR explained,

By its terms, the Full Faith and Credit Act accords state judicial
decisions only “the same full faith and credit” in federal court as they
would have “in the courts of [the rendering] State.” 28 U.S.C. § 1738
(emphasis added). . . . [I]t is hornbook Florida law that preclusion
requires complete mutuality of parties. . . . Because Duke and Walker
were not parties to Douglas [III], Florida law would afford Douglas
[III] no preclusive effect in these cases, and the Act requires the
federal courts to follow suit.

235
The panel faulted RJR for failing to identify a case that contradicted the panel’s
erroneous holding on an issue it sua sponte injected into the case without notice to the parties,
stating the following:
[RJR does not] identify any other court that has declined to give full faith and
credit to a judgment of a state court about what issues were actually decided in a
prior litigation on the ground that the state court decision was so wrong that it
amounted to a violation of due process.
Walker 0, at 22–23. Aside from the fact that a case of the nature the panel apparently
thought RJR should produce would be irrelevant to the actual inquiry at hand, pause to consider
how profoundly unusual the procedural posture of a case would need to be to even allow a court
to pass on the question the panel felt it was confronted with addressing.
221
 Case: 13-14590 Date Filed: 05/18/2017 Page: 222 of 284

Id. Because Alvin Walker and James Duke were not in privity with James L.

Douglas, they could not enforce the judgment entered in Douglas III. E.C. v. Katz,

731 So. 2d 1268, 1269 (Fla. 1999) (“[U]nless both parties are bound by the prior

judgment, neither may use it in a subsequent action.”). Nor could they rely on the

opinion’s factual findings. Forman v. Florida Land Holding Corp., 102 So. 2d

596, 598 (Fla. 1958) (“Stare decisis relates only to the determination of questions

of law, [and] . . . has no relation whatever to the binding effect of determinations of

fact.”). In sum, Douglas III was not a rehearing of Engle III. It was merely

another recognizing court interpreting the preclusive effect of the Phase I findings

to one specific plaintiff. Thus, not only did the panel err in giving full faith and

credit to Douglas III, its evaluation of the process afforded to the parties in that

case was totally irrelevant.

“After [RJR’s] first petition for rehearing [in response to Walker 0]

explained that Florida law requires mutuality of parties, which Douglas [III] and

[Walker] lack,” the panel had no choice but to correct its error. Petition for Writ of

Certiorari at 18 n.2, R.J. Reynolds Tobacco Co. v. Walker, 134 S. Ct. 2727 (No. 13-

1193). But rather than abandon its analysis, which had been premised upon a

fundamental misunderstanding of Florida law, the panel stuck to its inapposite

guns. Instead of “giv[ing] full faith and credit to the decision in Douglas [III],”

Walker 0, at 17, the panel “[gave] full faith and credit to the decision in Engle [III],

222
 Case: 13-14590 Date Filed: 05/18/2017 Page: 223 of 284

as interpreted in Douglas [III],” Walker II, 734 F.3d at 1287. Never mind that the

panel had determined that the Douglas III Court, not the Engle III Court, had been

the one to determine the Phase I findings were specific enough “based on [the

Douglas III Court’s] review of the class action trial plan and the jury

instructions.”236 Walker 0, at 19 (citing Douglas III, 110 So. 3d at 423). And

never mind that the panel had conducted its Hansberry inquiry on Douglas III, not

Engle III.237 The panel simply changed a few words in its opinion, and then

reissued its opinion with all of its original analysis intact. In fact, the panel made

substantive changes to only four sentences in its twenty-six page opinion.238 For

ease of reference, the changes are identified in Table 1.

236
In other words, the dispositive questions for the Walker panel were not whether the
Phase I jury “actually decided” that “R.J. Reynolds acted wrongfully in connection with . . . all
of its brands of cigarettes,” and whether RJR had adequate notice that these issues were to be
decided by the rendering court in Engle III. Instead, the dispositive question was whether the
Douglas III Court’s supposed factual determination that the Phase I jury “actually decided” this
issue was so manifestly incorrect and so lacking in notice that its judgment amounted to an
arbitrary deprivation of property in violation of due process.
237
The panel was explicit in its decision not to review the propriety of the proceedings in
Engle III, noting “R.J. Reynolds argues that we should conduct a searching review of the Engle
class action . . . but we lack the power to do so.” Walker 0, at 18. This conclusion was
nonsense. The panel not only had the power, it had a constitutional obligation to review the
procedures (1) that the Engle trial court used to produce the Phase I findings and (2) that the
Florida Supreme Court employed in instructing progeny courts to give them preclusive effect.
238
The panel also added the following two sentences at the start of the opinion: “We sua
sponte vacate and reconsider our original opinion in this matter. We substitute the following
opinion for our original opinion.” Walker II, 734 F.3d at 1280.
223
 Case: 13-14590 Date Filed: 05/18/2017 Page: 224 of 284

Table 1

Alterations From Walker 0 to Walker II

Walker 0 Walker II
“These principles require that we give “These principles require that we give
full faith and credit to the decision in full faith and credit to the decision in
Douglas [III] so long as it ‘satisf[ies] the Engle, as interpreted in Douglas [III], so
minimum procedural requirements’ of long as it ‘satisf[ies] the minimum
due process.” Walker 0, at 17 (quoting procedural requirements’ of due
Kremer, 456 U.S. at 481, 102 S. Ct. at process.” Walker II, 734 F.3d at 1286
1897). (quoting Kremer, 456 U.S. at 481, 102
S. Ct. at 1897).
“Our inquiry is a narrow one: whether “Our inquiry is a narrow one: whether
giving full faith and credit to the giving full faith and credit to the
decision in Douglas [III] would decision in Engle, as interpreted in
arbitrarily deprive R.J. Reynolds of its Douglas [III], would arbitrarily
property without due process of law.” deprive R.J. Reynolds of its property
Walker 0, at 18. without due process of law.” Walker
II, 734 F.3d at 1287.
“And we cannot refuse to give full “And we cannot refuse to give full
faith and credit to the decision in faith and credit to the decision in
Douglas [III] because we disagree Engle because we disagree with the
with its holding about what the jury in decision in Douglas [III] about what
Phase I decided.” Walker 0, at 18. the jury in Phase I decided.” Walker
II, 734 F.3d at 1287.
“Nor does R.J. Reynolds identify any “Nor does R.J. Reynolds identify any
other court that has declined to give other court that has declined to give
full faith and credit to a judgment of a full faith and credit to a judgment of a
state court about what issues were state court as later interpreted by the
actually decided in a prior litigation same state court on the ground that
on the ground that the state court the later state court decision was so
decision was so wrong that it wrong that it amounted to a violation
amounted to a violation of due of due process.” Walker II, 734 F.3d
process.” Walker 0, at 22–23. at 1289.

224
 Case: 13-14590 Date Filed: 05/18/2017 Page: 225 of 284

Aside from these changes, every aspect of the panel’s opinion remained

precisely the same. For example, it was still, in the panel’s estimation, the Court in

Douglas III—in its role as a recognizing court— that “was entitled to look beyond

the jury verdict to determine what issues the jury decided.” Walker 0, at 19;

Walker II, 734 F.3d at 1287. And it was still, according to the Walker panel, the

Court in Douglas III that fulfilled its role as a recognizing court, and “looked past

the ambiguous jury verdict to decide this question of fact.” Walker 0, at 23;

Walker II, 734 F.3d at 1289. Further, it was still the Court’s supposed factual

decision in Douglas III that the Walker panel concluded was not arbitrary. Walker

0, at 20; Walker II, 734 F.3d at 1288. And, “if due process requires a finding that

an issue was actually decided,” then it was still, in the panel’s opinion, the Court in

Douglas III that “made the necessary finding.”239 Walker 0, at 20; Walker II, 734

F.3d at 1289.

RJR took issue with the panel’s “incomplete” and “semantic” revisions to its

initial opinion. RJR Petition for Panel Rehearing or Rehearing En Banc at 4, 15,

Walker II (Nos. 12-13500, 12-14731). In its petition for rehearing Walker II, RJR

239
Walker II’s reliance on Douglas III is so clear that even the Majority acknowledge that
the Walker panel concluded that the “actually decided” requirement was fulfilled in Walker I and
Duke based on a purported nonbinding factual determination made in the Douglas III, an opinion
issued after the judgments in Walker I and Duke were entered. See Ante at 16 (“[In Walker II],
[w]e concluded that, even if due process requires that an issue be actually decided, the Florida
Supreme Court ruled in Douglas that the approved findings from Phase I concerned conduct that
is common to all class members and established negligence and defect elements of the class
members’ claims.” (citing Walker II, 734 F.3d at 1289)).
225
 Case: 13-14590 Date Filed: 05/18/2017 Page: 226 of 284

explained, “The panel’s cosmetic revisions to the original opinion” “do not even

fix the original decision’s sua sponte—and obviously erroneous—application of

the Full Faith and Credit Act to Douglas [III].” Id. at 14. RJR continued, “By

revising its decision at all, the panel presumably recognized that Douglas [III]

itself is entitled to no preclusive effect in these cases.” Id. That is so because

(1) the plaintiffs here were not parties in Douglas; (2) Florida law
requires the “same parties” in both cases for preclusion to apply, see,
e.g., Brown, 611 F.3d at 1332–33 (citing many cases); and (3) the Full
Faith and Credit Act accords state judicial decisions only “the same
full faith and credit” in federal court as they would have “in the courts
of [the rendering] State,” 28 U.S.C. § 1738.

Id. Yet, despite the panel’s apparent understanding of this fact, RJR noted, it still

“defer[ed] to the decision in Douglas [III]” and still failed to conduct even a

cursory review of the decision made by the rendering Engle III Court.240 Id. at 15

(quoting Walker II, 734 F.3d at 1289). Moreover, in purporting to defer to a

factual finding Douglas III never made, it also rejected Douglas III’s portrayal of

the Phase I findings as nonspecific liability determinations that are “useless” under

an issue-preclusion framework. Douglas III, 110 So. 3d at 433. Thus, according

240
Though Walker II failed to evaluate the process afforded to litigants in Engle III, it
followed Douglas III’s lead in evaluating the process afforded to litigants during the Phase I trial.
RJR protested that such an inquiry was a “red herring” and that “due-process rights must be
evaluated not against the process in [Phase I] itself, but against the process in [progeny] cases.”
Consolidated Reply Brief of Appellant R.J. Reynolds at 33, Walker II, 734 F.3d 1278 (Nos. 12-
13500, 12-141731). The panel ignored this argument, and, just as Douglas III had, 110 So. 3d at
431, determined that Engle defendants “had a full and fair opportunity to litigate . . . in Phase I,”
Walker 0, at 21; Walker II, 734 F.3d at 1288.

226
 Case: 13-14590 Date Filed: 05/18/2017 Page: 227 of 284

to RJR, the panel “concot[ed] and then ‘deferr[ed]’ to an indefensible reading” of

Douglas III. Petition for Writ of Certiorari at 29–30, R.J. Reynolds Tobacco Co. v.

Walker, 134 S. Ct. 2727 (No. 13-1193).

* * *

The Walker panel erred in at least eight ways.241 First, it disregarded the

plaintiffs’ claim-preclusion proffers, replacing them with its own issue-preclusion

proffer. In doing so, the panel not only shifted the plaintiffs’ evidentiary burden,242

it improperly carried that burden for the plaintiffs.243 The panel neglected to

241
These errors are relevant because the Majority “reaffirm” Walker II. Ante at 3.
242
Parties asserting claim preclusion under Florida law must establish the following four
elements: (1) “a final judgment on the merits”; (2) a “decision . . . rendered by a court of
competent jurisdiction”; (3) “the same cause of action . . . involved in both cases”; and (4) “the
parties, or those in privity with them, are identical in both suits.” Baloco v. Drummond Co., 767
F.3d 1229, 1246 (11th Cir. 2014). In contrast, parties asserting issue preclusion under Florida
law must establish an entirely different set of elements: (1) identical parties, (2) identical
issue(s), (3) full litigation of the particular matter, (4) determination of the particular matter, and
(5) a “final decision” in the prior proceeding by a court of competent jurisdiction. Dadeland
Depot, Inc. v. St. Paul Fire and Marine Ins. Co., 945 So. 2d 1216, 1235 (Fla. 2006) (quoting
Dep’t of Health & Rehab. Servs. v. B.J.M., 656 So. 2d 906, 910 (Fla. 1995)).
243
The party claiming preclusion bears the burden of proving its elements. 18 Wright,
supra, § 4405 (“[T]he burden of establishing preclusion is placed on the party claiming it.”).
The Florida Supreme Court made this clear more than a hundred years ago in Prall: “If there is
any uncertainty as to the matter formerly adjudicated, the burden of showing it with sufficient
certainty . . . is upon the party who claims the benefit of the former judgment.” 50 So. at 870;
see also Bagwell, 14 So. 2d at 843 (“The burden of proof to establish a former adjudication, by
law, was cast on the defendant below.”). At least until Douglas III, modern Florida courts were
still consistently hearkening to this common-sense principle. See, e.g., Campbell v. State, 906
So. 2d 293, 295 (Fla. 2d Dist. Ct. App. 2004) (“The party claiming the benefit of res judicata has
the burden of establishing with sufficient certainty, by the record or by extrinsic evidence, that
the matter was formerly adjudicated”); State St. Bank & Trust Co. v. Badra, 765 So. 2d 251, 253
(Fla. 4th Dist. Ct. App. 2000) (“[T]o establish res judicata . . . the party claiming the benefit of
the former adjudication has the burden of establishing, with sufficient certainty by the record or
by extrinsic evidence, that the matter was formerly adjudicated.”); Meyers v. Shore Inds. Inc.,
597 So. 2d 345, 346 (Fla. 2d Dist. Ct. App. 1992) (“The party asserting the defense of estoppel
227
 Case: 13-14590 Date Filed: 05/18/2017 Page: 228 of 284

consider whether its abrogation of two common-law protections rooted in

fundamental fairness—the party asserting preclusion’s burden of proof and the

presumption against preclusion—violated RJR’s rights to due process.

Second, after injecting its new theory of preclusion into the case on behalf of

the plaintiffs, the panel failed to provide RJR with an opportunity to be heard on

the theory’s applicability to its case before entering judgment. The panel failed to

consider the due process implications of denying RJR its right to be heard on a

dispositive issue in the case against it.

Third, as the panel only hinted at, but did not commit to, what it believed is

the defect that taints all cigarettes,244 it sanctioned an unconstitutional conclusive

presumption that all smoking-related injuries are caused by the manufacturer’s

tortious conduct.245

Fourth, in sanctioning this conclusive presumption, the panel denied RJR’s

Seventh Amendment right to a jury trial on a contested and material element of the

claims against them.

by judgment has the burden of demonstrating with sufficient certainty through the record or
extrinsic evidence that the issue was adjudicated fully.”).
244
The Florida Supreme Court has recently rejected the defect at which the panel
hinted—that all cigarettes are defective because they cause disease. See infra note 257 and
accompanying text.
245
See supra note 6 and accompanying text.
228
 Case: 13-14590 Date Filed: 05/18/2017 Page: 229 of 284

Fifth, the panel appeared to act as both advocate and arbiter. In doing so, it

failed to consider whether this denied RJR of its due process right to an impartial

decision maker

Sixth, in advocating for the plaintiffs, the panel effectively rewrote Douglas

III in a strained attempt to reconcile it with our Brown II precedent.246 To its

credit, the panel rejected Douglas III’s preposterous portrayal of the Phase I

findings as liability determinations. Yet, the panel then adopted an equally

preposterous portrayal of the Phase I findings based, in turn, on a preposterous

portrayal of Douglas III.247 As interpreted by the panel, when the Douglas III

Court said that the Phase I findings would be “useless” under issue preclusion, 110

So. 3d at 433, it really meant that the Phase I findings were dispositive under issue

preclusion, like special verdicts that establish all the elements of liability under

various tort theories.248 When the Douglas III Court said that “the Engle jury did

not make detailed findings for which evidence it relied upon to make the Phase I

246
Ironically, in the panel’s attempt to reconcile Douglas III and Brown II, it attributed to
Douglas III a portrayal of the Phase I findings that both Douglas III and Brown II reject.
247
That the panel relied on the Douglas III Court to interpret the Phase I findings rather
than simply looking at the questions the Phase I jury was instructed to answer is, in itself, strong
evidence that the Court’s supposed interpretation is inconsistent with what the jury actually
decided.
248
If the Douglas III Court had found that the findings were dispositive under issue
preclusion, it would not have adopted Martin II’s rationale. Instead, it would have criticized the
Martin II Court for failing to require the plaintiffs in that case to “trot out” the portions of the
Phase I trial record that established that the Phase I jury determined that all cigarettes are
defective, unreasonably dangerous, and negligently made because they contain nicotine and
cause disease.
229
 Case: 13-14590 Date Filed: 05/18/2017 Page: 230 of 284

common liability findings,” id., it really meant that it had “looked past the

ambiguous jury verdict[s]” to identify (but not disclose) precisely the evidence

upon which the Phase I Engle jury relied.249 Walker 0, at 21; Walker II, 734 F.3d

at 1289. After its absurd portrayal of Douglas III was in place, the panel then

“disagree[d]” with Douglas III’s supposed factual determinations, because such

determinations were “erroneous.”250 Walker 0, at 18; Walker II, 734 F.3d at 1287.

The panel nevertheless gave full faith and credit to them. Walker 0, at 18; Walker

II, 734 F.3d at 1287. The panel failed to evaluate whether its absurd depiction of

Douglas III amounted to an arbitrary denial of RJR’s right to due process.251

249
The Walker panel appears to have been confused by the Douglas III Court’s assertion
that the Phase I findings were “specific enough.” Douglas III, 110 So. 3d at 428. The Douglas
III Court made clear it believed that the findings are “specific enough” because they establish the
defendants’ liability to all class members, not because they reveal (1) the particular conduct the
Phase I jury deemed tortious and (2) that such conduct necessarily caused all class members’
injuries. See id. at 430, 433 (“[T]he Phase I jury already determined that the defendants’ conduct
subjects them to liability to Engle class members under [strict liability and negligence] theor[ies].
. . . [T]o decide here that we really meant issue preclusion . . . would effectively make the Phase I
findings . . . useless in individual actions.”).
250
Recall that the Brown II panel observed that “the plaintiffs have pointed to nothing in
the record, and there is certainly nothing in the jury findings themselves, to support [the
plaintiffs’] factual assertion” that the Phase I jury found that the defendants’ tortious conduct
tainted all cigarettes. 611 F.3d at 1335.
251
Did the panel resort to its fantastic portrayal of Douglas III because it recognized that
the framework for claim preclusion simply was not present? Perhaps the panel could overlook
that Douglas III disregarded Engle III’s observation that in “Phase I, the jury decided issues
related to Tobacco’s conduct but did not consider whether any class members . . . were injured
by Tobacco’s conduct.” Engle III, 945 So. 2d at 1263. And maybe the panel did not realize that
the Phase I jury was explicitly instructed not to consider whether the Engle defendants’ conduct
caused all class members’ injuries. See supra note 177 and accompanying text. The panel
apparently could not, however, ignore the simple facts that not a single class plaintiff was present
in Phase I and that the defendants “did not have their day in court on the broader questions
230
 Case: 13-14590 Date Filed: 05/18/2017 Page: 231 of 284

This, in turn, gave rise to the panel’s seventh error—giving full faith and

credit to an opinion the plaintiffs did not proffer and upon which they could not

rely under Florida law. This error thoroughly tainted the panel’s Walker 0 opinion.

In that opinion, the panel not only erroneously gave full faith and credit to Douglas

III, it conducted a totally irrelevant Hansberry inquiry on Douglas III, concluding

that the Douglas III Court appropriately “look[ed] beyond the [Phase I] jury

verdict[s]” to interpret them. Walker 0, at 19.

When the panel attempted to hastily correct its seventh error by vacating

Walker 0 and issuing Walker II, it committed its most indefensible error of all. The

panel left intact all of its tainted and inapposite reasoning from Walker 0. Walker

II, therefore, gives full faith and credit to Engle III, yet it inexplicably reviews the

process afforded to litigants in Douglas III, Walker II, 734 F.3d at 1287, while

conducting no inquiry whatsoever on what was decided in Engle III or the process

involving the causes of action the class asserted,” which were left to be resolved in later phases.
Brown II, 611 F.3d at 1333.
Does the panel’s recasting of Douglas III suggest that it knew Douglas III was rendered
in violation of the defendants’ due process rights? The panel did not evaluate, at least openly,
whether another host of issues in Douglas III also amounted to violations of due process,
including: (1) whether the Court’s usurpation of the plaintiffs’ burden of proving preclusion
amounted to an abrogation of the common-law protections against the arbitrary deprivation of
due process; (2) whether the Court’s reversal of Florida’s presumption against preclusion
amounted to an abrogation of the common-law protection against the arbitrary deprivation of due
process; (3) whether the Court’s adjudication of a position that it advanced on behalf of the
plaintiffs amounted to a violation of the defendants’ due process right to an impartial decision
maker; (4) whether the Court’s endorsement of a conclusive presumption—that smoking-caused
injuries are presumptively caused by the defendants’ tortious conduct—violated the defendants’
rights to due process.
231
 Case: 13-14590 Date Filed: 05/18/2017 Page: 232 of 284

afforded to litigants in that proceeding.252 For example, the panel does not say

whether the issues it precluded the defendants from litigating were decided in

Engle III. Nor does it indicate whether any such issues were before the Court and

decided with adequate notice and opportunity to be heard.253

Moreover, Walker II still claims that it was the Douglas III Court that

“look[ed] beyond the [Phase I] jury verdict[s]” to interpret them. Id. Thus, the

Walker II panel gave full faith and credit to Engle III, yet the relevant factual

determination was made by the Douglas III Court. Thus, Walker II, issued by a

federal appellate court tasked with interpreting the U.S. Constitution, holds that “if

due process requires a finding that an issue was actually decided, then the

[Douglas III] Court made the necessary finding.” Walker II, 734 F.3d at 1289.

Federal courts cannot discharge their constitutional duties by deferring to

inapplicable state-court opinions.

Why did the panel choose to publish such a transparently nonsensical

opinion? Was it because it had already backed itself into a corner with Walker 0?

Walker 0 makes clear that the panel saw only “ambigu[ity]” when it looked at the

252
For example, Walker II makes no reference to what issues were before the Engle III
Court, what the parties briefed, and what Engle III decided.
253
As the panel did not even evaluate whether the issues it attributed to Engle III as
deciding were rendered with adequate notice and opportunity to be heard, it also did not evaluate
any of the other due process problems that may have occurred in that proceeding. For example,
it did not assess whether the Engle III Court (1) disguised a judgment as dicta so as to minimize
the defendants’ chance at certiorari review by the U.S. Supreme Court or (2) improperly usurped
the role of recognizing progeny courts.
232
 Case: 13-14590 Date Filed: 05/18/2017 Page: 233 of 284

Phase I findings. Walker 0, at 23. It “disagree[d] with [Douglas III’s] [supposed]

holding about what the jury in Phase I decided” and believed that holding was

“erroneous.” Id. at 18. The panel thus could not, in the immediate aftermath of

Walker 0, reasonably contend in Walker II that it “looked beyond the jury verdict”

for itself “to determine what issues the jury decided.”254 Walker 0, at 19. Nor

could the panel claim, with a straight face, that the rendering Engle III Court had

been the one to “look beyond the jury verdict” after it had just finished explaining

that the recognizing Douglas III Court had been the one to do so.255 Did the panel

therefore resign to change a few words, muddy the waters a bit by referring to

“Engle [III], as interpreted in Douglas [III],” and hope that this Court would not

notice, or at least not fully grasp the implications of, its inapposite reasoning when

faced with the defendants’ petition for rehearing en banc? Walker II, 734 F.3d at

1286.

254
Nor could the panel say that the District Courts in Duke and Walker I secretly
conducted such a search and failed to reveal it to the parties. The panel had just admitted in
Walker 0 that the District Courts’ each relied entirely on the “decision in Waggoner” to fashion
their instructions to the jury that “Phase I conclusively established the tortious-conduct elements
of the plaintiffs’ claims.” Walker 0, at 16
255
Recall that recognizing courts, not rendering courts, perform actually decided
inquiries. See supra Part II.
233
 Case: 13-14590 Date Filed: 05/18/2017 Page: 234 of 284

VI.
THE MAJORITY REPEAT AND ADD TO THE WALKER PANEL’S ERRORS

Mr. Graham’s complaint was nearly identical to the plaintiffs’ complaints in

Walker I and Duke. Just like those complaints,256 Mr. Graham’s alleged that Engle

III was a “mandate” that “provided Plaintiff the opportunity to complete

unresolved damages claims.” Second Amended Complaint at 2, 3, Graham v. R.J.

Reynolds Tobacco Co., No. 3:09-cv-13602-MMH-HTS (M.D. Fla. Dec. 5, 2011)

(emphasis added). Just like Walker and Duke, Mr. Graham assumed, in

accordance with Martin II, that all that remained to be resolved in his case was

damages; the defendants were precluded by the “mandate” in Engle III from

litigating both (1) that the cigarettes smoked by Ms. Graham were defective,

unreasonably dangerous, and negligently produced and (2) that the defendants’

tortious conduct caused Ms. Graham’s death. Id. Accordingly, Mr. Graham

pleaded only class membership—Ms. Graham’s addiction and a smoking-related

injury. Id. at 4. To support his assertion of claim preclusion, he proffered only the

Engle III opinion and the “common liability findings” approved by the Engle III

court. Id. at 5–9.

In response, the defendants contended that Mr. Graham failed to sufficiently

plead claim preclusion. Philip Morris USA Inc.’s Answer to Plaintiff’s Second

Amended Complaint at 9, Graham I, No. 3:09-cv-13602-MMH-HTS; Answer,

256
See supra note 224 and accompanying text.
234
 Case: 13-14590 Date Filed: 05/18/2017 Page: 235 of 284

Defenses and Jury Demand of Defendant R.J. Reynolds Tobacco Co at 23,

Graham I, No. 3:09-cv-13602-MMH-HTS. Moreover, they argued that Mr.

Graham could not rely on either claim or issue preclusion because the Phase I

findings are “inadequate to support an individualized determination of liability,

legal causation, and damages in this or any other subsequent individual action.”

Answer, Defenses and Jury Demand of Defendant R.J. Reynolds Tobacco Co at 2,

Graham I, No. 3:09-cv-13602-MMH-HTS; see also Philip Morris USA Inc.’s

Answer to Plaintiff’s Second Amended Complaint at 10, Graham I, No. 3:09-cv-

13602-MMH-HTS (making the same argument).

The District Court, relying on Waggoner’s pretrial ruling, accepted Mr.

Graham’s claim-preclusion proffer of Engle III and its approved findings as

adequate to establish that RJR and Philip Morris were liable to all class members,

including Mr. Graham. Accordingly, the District Court precluded RJR and Philip

Morris from contesting, and did not instruct the jury to determine, (1) that the

cigarettes Ms. Graham smoked were defective, unreasonably dangerous, and

negligently produced and (2) that the defendants’ tortious conduct caused Ms.

Graham’s death. Ante at 17. The jury thus ruled in Mr. Graham’s favor because

he proved, to their satisfaction, his status as a class member. In sum, the Graham

trial played out in much the same way as the Walker I and Duke trials.

235
 Case: 13-14590 Date Filed: 05/18/2017 Page: 236 of 284

Thus, we are confronted with the same question with which the Walker

panel was confronted: Did the District Court deprive RJR and Philip Morris of

property without due process of law by applying claim preclusion so as to preclude

the defendants from litigating (1) that the cigarettes Ms. Graham smoked were

defective, unreasonably dangerous, and negligently produced and (2) that the

defendants’ tortious conduct caused Ms. Graham’s death?

In answering this question, the Majority implicitly acknowledge that the

District Court erred by applying claim preclusion. They “reaffirm” Walker II, Ante

at 3, in which the panel rejected Douglas III’s portrayal of the Phase I findings as

liability determinations. See Walker II, 734 F.3d at 1280 (“[T]he [Phase I] jury did

not decide whether the tobacco companies were liable for damages to individual

members of the class.”).

Nevertheless, following the Walker panel’s basic analytical framework, the

Majority conclude that the District Court’s error was harmless. Like the Walker

panel, the Majority reach this conclusion by portraying the Phase I findings as

“specific enough” factual findings—a portrayal Mr. Graham neither advanced nor

proffered evidence to support. Ante at 21. That is, they view the Phase I findings

as establishing that all cigarettes are defective, unreasonably dangerous, and

negligently sold. See id. (“The Florida Supreme Court rejected [the] argument”

that “the jury did not necessarily find that all cigarettes the defendants placed on

236
 Case: 13-14590 Date Filed: 05/18/2017 Page: 237 of 284

the market were defective and unreasonably dangerous.”). In adopting such a

portrayal, they attempt to correct for the District Court’s barring the defendants

from litigating whether the cigarettes Ms. Graham smoked were defective,

unreasonably dangerous, and negligently sold. The Majority also suggest that the

unreasonably dangerous defect that taints all cigarettes is that they “cause disease

and are addictive.”257 Id. at 23. In hinting at this portrayal,258 the Majority attempt

to correct for the District Court’s barring the defendants from litigating whether

their tortious conduct caused Ms. Graham’s death.259

Whereas the Walker panel imputed its portrayal of the Phase I findings as

“specific enough” factual findings to Douglas III, the Majority attribute their

257
The Florida Supreme Court has repeatedly rejected this suggestion. See, e.g., R.J.
Reynolds Tobacco Co v. Marotta, No. SC16-218, 2017 WL 1282111, at *9 (Fla. Apr. 6, 2017)
(“[T]he inherent characteristics of all cigarettes did not form the sole basis for liability. Rather,
the case was premised on the allegation that the Engle defendants intentionally increased the
amount of nicotine in their products.”).
258
The Majority hint at, but do not commit to, this portrayal. Instead, their opinion
generally refers to the Phase I findings by using unintelligible generalities. For example, the
Majority say, “[T]he Engle jury actually decided common elements of the negligence and strict
liability of R.J. Reynolds and Philip Morris” without ever defining the term “common elements.”
Ante at 20. They also insist that “Phase I established . . . [that] the companies acted wrongfully
toward all of the class members.” Id. at 21.
Without identifying the defect that taints all cigarettes, the Majority necessarily deny the
defendants their Seventh Amendment right to a jury determination on a contested element of
their claim, and they adopt, in violation of Henderson, an unconstitutional presumption that
smoking-related injuries are caused by tortious conduct. See supra note 6 and accompanying
text.
259
The Majority’s logic seems to be that when class members prove their class
membership by proving a smoking-related disease, they also necessarily prove that a defendant’s
tortious conduct caused their disease because the tendency of cigarettes to cause disease is the
defect that taints all cigarettes and makes the sale of cigarettes negligent.
237
 Case: 13-14590 Date Filed: 05/18/2017 Page: 238 of 284

identical portrayal to Engle III.260 They assert, without offering evidence to

support their assertion, that the Engle III Court was the one to “interpret[] [the

Phase I] findings to determine what the jury actually decided.”261 Id. at 30. In

doing so, they sidestep the mutuality obstacle that doomed the Walker 0 opinion.

See supra Part V.

In support of their portrayal, which they attribute to Engle III, the Majority

mine the Phase I trial record and proffer excerpts for the plaintiff to establish a

theory of preclusion the plaintiff did not advance. Id. at 5–9, 23. On the basis of

their own evidentiary proffer, they appear to conclude that Engle III’s supposed

260
The Majority also do not explicitly repeat some of the Walker panel’s mistakes. For
example, the Majority do not explicitly claim that the Douglas III Court “looked past the
ambiguous jury verdict[s].” Walker II, 734 F.3d at 1289. Nor do they contradict Florida law by
giving full faith and credit to Douglas III’s supposed factual determination, as the Walker panel
did in its vacated Walker 0 opinion.
261
The Majority do not explain why the plaintiffs did not allege such an interpretation in
their briefs to this Court. Nor do they explain why Brown I, Brown II, Martin II, and Jimmie Lee
Brown II all similarly failed to recognize interpretation, and why, rather than correct those
Courts’ misunderstanding, the Florida Supreme Court refused to accept jurisdiction. Nor do they
explain why they did not mention such an interpretation in Walker II, which they affirm. Other
than assuring that their “terminology . . . was unorthodox,” Ante at 25, they also do not explain
why the Douglas III Court, which included three of the four justices in the Engle III majority,
explicitly rejected making such an interpretation, and called the Phase I findings “useless.” Nor
do they explain why the Florida Supreme Court has repeatedly rejected making such an
interpretation in Engle III, and why as recently as last month, the Florida Supreme Court made
clear that the Phase I jury did not premise its finding on its determination that the defendants’
cigarettes are defective because of their potential to cause disease. See Marotta, at *9 (“[T]he
inherent characteristics of all cigarettes did not form the sole basis for liability. Rather, the case
was premised on the allegation that the Engle defendants intentionally increased the amount of
nicotine in their products.”).
238
 Case: 13-14590 Date Filed: 05/18/2017 Page: 239 of 284

portrayal of the Phase I findings is not arbitrary262 because a properly instructed

jury could have made such findings.263 See, e.g., id. at 23 (“[T]he jury’s answers

on the verdict form, when read together with the entire record, were consistent with

the general theories that the tobacco companies’ cigarettes are defective and the

sale of their cigarettes is negligent because all of those cigarettes cause disease and

are addictive.” (emphasis added)).

Though the Majority give full faith and credit to Engle III, id. at 3, 16, they

do not perform their recognizing-court inquiries as Florida law and the U.S.

262
The Majority also seem to back away from the Walker panel’s assertion that their
portrayal of the Phase I findings—which the Walker panel attributed to the Douglas III Court,
and the Majority attribute to the Engle III Court—is erroneous.
263
The Majority opinion reads, in places, as though the Majority intend to give full faith
and credit directly to the Phase I jury findings rather than the Engle III judgment. See Ante at 30
(“[W]e . . . give full faith and credit to the jury findings in Engle.).” That, of course, cannot
happen. Jury findings are not a judgment, and findings without a judgment are inadequate as a
basis for either issue or claim preclusion. Armellini Express Lines, Inc. v. Sexton, 384 So. 2d
310, 310 (Fla 5th Dist. Ct. App. 1980); see also Fla. Dep’t of Transp. v. Juliano, 801 So. 2d 101,
105 (Fla. 2001) (listing “a final judgment on the merits” as an essential element of claim
preclusion). “Courts reduce their opinions and verdicts to judgments precisely to define the
rights and liabilities of the parties. . . . A prevailing party seeks to enforce . . . the court’s
judgment.” Jennings v. Stephens, 135 S. Ct. 793, 799, 190 L. Ed. 2d 662 (2015) (emphasis in
original) (citations omitted).
If the Majority’s approach really is to give full faith and credit directly to the findings,
irrespective of what the Engle III Court said about those findings, not only is their very premise
flawed, but their execution is flawed as well. As detailed extensively in Part II of this dissent,
under the law followed in every jurisdiction, including Florida, courts determine what a jury
“actually decided” based on necessary inference. Juries make decisions by answering questions,
not by looking at evidence. Thus, the “common thrust” of the Phase I evidence, Ante at 7, and
the “consisten[cy]” of such evidence with a particular theory of negligence or strict liability, id.
at 22, are irrelevant.
239
 Case: 13-14590 Date Filed: 05/18/2017 Page: 240 of 284

Constitution require.264 Accordingly, the Majority do not evaluate whether the

Engle III Court violated the defendants’ due process rights by making a dispositive

determination about the Phase I findings secretly—that is, without revealing that it

had done so—and without affording the parties notice or opportunity to be

heard.265

264
See Douglas III, 110 So. 3d at 430–31 (explaining that when a due process objection
is raised it becomes the “duty of [the recognizing court] to examine the course of procedures in
both [rendering- and recognizing-court] litigations to ascertain whether the litigant whose rights
have thus been adjudicated has been afforded . . . due process” (quoting Hansberry, 311 U.S. at
40, 61 S. Ct. at 117)).
265
One excerpt of the Majority opinion reads as follows: “Douglas [III] decided a matter
of state law when it explained the preclusive effect of the Engle jury’s Phase I findings. We are
bound by the decisions of state supreme courts on matters of state law when we exercise
diversity jurisdiction, subject to the constraints of due process.” Ante at 30. In isolation, this
statement suggests that the Majority affirm the District Court’s judgment not because they give
preclusive effect to findings the Engle III Court supposedly made, but because they are applying,
under Erie, a substantive law set forth in Douglas III, which dictates how cases against the Engle
defendants should proceed.
If the Majority mean to suggest that Douglas III instituted an “unorthodox” and “novel
notion of res judicata,” Ante at 25, 27, our duty as a recognizing court, under Florida law and the
U.S. Constitution, would shift to evaluating whether the newly created law “eliminate[s] the
basic common law protection against an arbitrary deprivation of property.” Douglas III, 110 So.
3d at 431 (citing Oberg, 512 U.S. at 432, 114 S. Ct. at 2339). Does it, for example, allow
plaintiffs to preclude the Engle defendants from litigating matters—such as whether their tortious
conduct caused an individual plaintiff’s injuries—on which they have never had an opportunity
to be heard? Douglas III leaves no doubt. See 110 So. 3d at 429 (holding that when a plaintiff
“prov[es] that addiction to the Engle defendants’ cigarettes containing nicotine was a legal cause
of the injuries alleged,” “injury as a result of the Engle defendants’ conduct is assumed”
(emphasis added)).
Moreover, the law to which the Majority hint would be a state law enacted by the Florida
Supreme Court that applies to the unique detriment of a single group of unpopular defendants.
The law would also be one that directs recognizing progeny courts to hold the unpopular
defendants liable to all individuals harmed by their products even if they do not receive a jury
trial on contested elements of their claim and even if it is not proven that they committed a
tortious act or that such tortious act caused the individual’s harm. This irrebuttable presumption
of liability, as explained above, is unconstitutional and cannot be applied under Erie. See supra
note 6 and accompanying text.
240
 Case: 13-14590 Date Filed: 05/18/2017 Page: 241 of 284

* * *

Sitting in the same posture as the Walker panel, the Majority replicate and

add to the constitutional violations the panel committed in its blind affirmance of

judgments entered in Duke and Walker I. First, like the Walker panel, the Majority

disregard the plaintiff’s claim-preclusion proffer, replacing it with their own issue-

preclusion proffer.266 In doing so, the Majority not only shift the plaintiffs’

evidentiary burden,267 they improperly carry that burden for the plaintiffs.268 Just

Absent a clearer statement by the Majority saying that their affirmance is entirely based
on the application of this state law, I assume they also affirm on the basis of their
unconstitutional misapplication of the full faith and credit act that I have detailed above.
266
In support of his request for claim preclusion, Mr. Graham proffered little more than
the Engle III decision, contending that it barred RJR and Philip Morris from contesting (1) that
the cigarettes Ms. Graham smoked were defective, unreasonably dangerous, and negligently
produced and (2) that their tortious conduct caused Ms. Graham’s death. Thus, our first question
as a recognizing court should have been whether the plaintiff’s proffer satisfies the elements of
Florida claim preclusion.
The traditional elements of Florida claim preclusion include (1) “a final judgment on the
merits”; (2) a “decision . . . rendered by a court of competent jurisdiction”; (3) “the same cause
of action . . . involved in both cases”; and (4) “the parties, or those in privity with them, are
identical in both suits.” Baloco v. Drummond Co., 767 F.3d 1229, 1246 (11th Cir. 2014). Engle
III says nothing about reaching a final judgment as to the defendants’ liability to all class
members. In fact, Engle III held that “[i]t was . . . error for the Phase I jury to consider whether
[the Engle defendants were] liable for punitive damages” because it had not yet been
“determine[d] whether the defendants were liable to anyone.” Engle III, 945 So. 2d at 1263
(emphasis added) (citation omitted). Mr. Graham’s proffer therefore fails to establish the
elements of traditional Florida claim preclusion. Because the plaintiffs do not argue that
Douglas III altered those traditional elements of claim preclusion for this one case (an argument
that would be wrought with due process problems), the District Court’s judgement should have
been reversed on that basis.
267
Parties asserting claim preclusion under Florida law must establish the following four
elements: (1) “a final judgment on the merits”; (2) a “decision . . . rendered by a court of
competent jurisdiction”; (3) “the same cause of action . . . involved in both cases”; and (4) “the
parties, or those in privity with them, are identical in both suits.” Baloco v. Drummond Co., 767
241
 Case: 13-14590 Date Filed: 05/18/2017 Page: 242 of 284

as the Walker panel did, the Majority leave unanswered whether their abrogation of

two common-law protections rooted in fundamental fairness—the party asserting

preclusion’s burden of proof and the presumption against preclusion—deny the

defendants’ rights to due process.

Second, following the Walker panel’s lead, after injecting their own theory

of preclusion into the case in place of the plaintiff’s, the Majority fail to provide

the defendants with an opportunity to be heard on the theory’s applicability to their

case before entering judgment. Despite the Majority’s demonstrated willingness to

F.3d 1229, 1246 (11th Cir. 2014). In contrast, parties asserting issue preclusion under Florida
law must establish an entirely different set of elements: (1) identical parties,267 (2) identical
issue(s), (3) full litigation of the particular matter, (4) determination of the particular matter, and
(5) a “final decision” in the prior proceeding by a court of competent jurisdiction. Dadeland
Depot, Inc. v. St. Paul Fire and Marine Ins. Co., 945 So. 2d 1216, 1235 (Fla. 2006) (quoting
Dep’t of Health & Rehab. Servs. v. B.J.M., 656 So. 2d 906, 910 (Fla. 1995)).
268
The party claiming preclusion bears the burden of proving its elements. 18 Wright,
supra, § 4405 (“[T]he burden of establishing preclusion is placed on the party claiming it.”).
The Florida Supreme Court made this clear more than a hundred years ago in Prall: “If there is
any uncertainty as to the matter formerly adjudicated, the burden of showing it with sufficient
certainty . . . is upon the party who claims the benefit of the former judgment.” 50 So. at 870;
see also Bagwell, 14 So. 2d at 843 (“The burden of proof to establish a former adjudication, by
law, was cast on the defendant below.”). At least until Douglas III, modern Florida courts were
still consistently hearkening to this common-sense principle. See, e.g., Campbell v. State, 906
So. 2d 293, 295 (Fla. 2d Dist. Ct. App. 2004) (“The party claiming the benefit of res judicata has
the burden of establishing with sufficient certainty, by the record or by extrinsic evidence, that
the matter was formerly adjudicated”); State St. Bank & Trust Co. v. Badra, 765 So. 2d 251, 253
(Fla. 4th Dist. Ct. App. 2000) (“[T]o establish res judicata . . . the party claiming the benefit of
the former adjudication has the burden of establishing, with sufficient certainty by the record or
by extrinsic evidence, that the matter was formerly adjudicated.”); Meyers v. Shore Inds. Inc.,
597 So. 2d 345, 346 (Fla. 2d Dist. Ct. App. 1992) (“The party asserting the defense of estoppel
by judgment has the burden of demonstrating with sufficient certainty through the record or
extrinsic evidence that the issue was adjudicated fully.”).
242
 Case: 13-14590 Date Filed: 05/18/2017 Page: 243 of 284

allow the parties to file supplemental briefs,269 they refuse to provide the

defendants their constitutional right to be heard on the accuracy of the Majority’s

novel factual determination that the Engle III Court determined that the Phase I

findings were “specific enough” factual findings.270

Third, in a notice-and-opportunity-to-be-heard double whammy, the

Majority conduct no Hansberry examination into Engle III to evaluate whether the

“interpret[tation]” of the Phase I findings they attribute to Engle III, and to which

they give full faith and credit, was rendered with adequate notice and opportunity

to be heard.271 They do not, for example, mention the proceedings in Engle II or

describe what issues were before the Court in Engle III.272 Had the Majority

269
We allowed the parties to file supplemental briefs regarding the impact on this case of
the Florida Supreme Court’s recent Marotta opinion.
270
As I discuss above, if the Majority do not believe that the defect in the cigarettes is
their potential to cause disease, then the Majority need to define what they believe the defect is
that Engle III identified, or explain why applying an irrebuttable presumption that the
unidentified defect caused Ms. Graham’s harm does not violate the defendants’ Seventh
Amendment and due process rights. Or they should provide the defendant with an opportunity to
be heard on the factual assertion that Engle III determined that the Phase I jury determined (a)
that the cigarettes Ms. Graham smoked were defective and negligently produced, and (b) that the
unidentified defect and negligent conduct caused Ms. Graham’s harm.
271
The Walker panel’s failure to conduct a Hansberry inquiry into Engle III can be
explained by its mistaken belief that Douglas III was the operative judgment. The Majority,
which appear to understand that Engle III is the operative judgment, cannot rely on such an
excuse. Did they fail to conduct the inquiry because they do not like what a proper inquiry
would show?
272
As the Majority did not even evaluate whether the issues they attributed to Engle III as
deciding were rendered with adequate notice and opportunity to be heard, they also did not
evaluate any of the other due process problems that may have occurred in that proceeding. For
example, it did not assess whether the Engle III Court (1) disguised a judgment as dicta so as to
minimize the defendants’ chance at certiorari review by the U.S. Supreme Court or (2)
243
 Case: 13-14590 Date Filed: 05/18/2017 Page: 244 of 284

conducted such an evaluation, they would have recognized that the parties in that

case were not on notice that the Engle III Court was going to make a determination

about how progeny courts should interpret the Phase I findings.273 They would

have recognized that the parties had no opportunity to be heard on such a

determination, and it is therefore not entitled to preclusive effect.274

improperly usurped the role of recognizing progeny courts by dictating the res judicata effect of
its own findings.
273
The one belief the Majority, the Florida Supreme Court, and I appear to share about
the Florida Supreme Court is that it has repeatedly acted deceptively. The Majority portray the
Engle III Court as (1) making a sua sponte and secret determination that the Phase I findings
were “specific enough” factual findings and then (2) refusing to review District Court of Appeal
decisions that failed to recognize that. The Florida Supreme Court in Douglas III portrayed the
Engle III Court as (1) making a sua sponte and secret determination that the Phase I findings
were nonspecific liability determinations and then (2) refusing to review District Court of Appeal
decisions that failed to recognize that. I believe the Florida Supreme Court acted deceptively in
Engle III by sending a message to progeny courts, disguised as dicta, instructing them to prop up
the Phase I findings by whatever rationale they could devise. I also believe the Florida Supreme
Court acted deceptively in Douglas III, when it adopted the Martin II rationale—which had been
carefully crafted to circumvent both our Brown II opinion and various defendant objections—and
pretended like it had that rationale in mind when it issued its Engle III opinion seven years
earlier.
274
The Engle III judgment depicted by the Majority involved the Court searching the trial
record and dictating the meaning of the Phase I findings without affording the parties notice or
opportunity to be heard. Recall that the issues before the Florida Supreme Court in Engle III
were those briefed by the parties following the Court’s acceptance of conflict jurisdiction after
Engle II. These issues were whether it was error for the Third District Court of Appeal to (1)
reverse the Circuit Court’s final judgment based on plaintiff’s counsel’s unprofessional conduct
and (2) vacate the punitive-damages award and decertify the class.
“[A] State may not grant preclusive effect in its own courts to a constitutionally infirm
judgment,” Kremer, 456 U.S. at 482–83, 102 S. Ct. at 1898, and “Section 1738 requires federal
courts to give the same preclusive effect to state court judgments that those judgments would be
given in the courts of the State from which the judgments emerged.” Id. at 466, 1889. Because
the Majority suggest that Engle III disposed of matters that were not pleaded, briefed, or raised in
any way by either party, such determinations—had they been made by the Engle III Court—
would not be judgments entitled to full faith and credit.
244
 Case: 13-14590 Date Filed: 05/18/2017 Page: 245 of 284

Fourth, in acting both as advocate and arbiter and attributing to the Engle III

Court a portrayal of the Phase I findings that it expressly declined to make, the

Majority violate the defendants’ due process right to an impartial decision maker.

Fifth, as the Majority only hint at, but do not commit to, what they believe is

the defect that taints all cigarettes,275 they sanction an unconstitutional conclusive

presumption that all smoking-related injuries are caused by the manufacturer’s

tortious conduct.276

Sixth, in sanctioning this conclusive presumption, the Majority deny the

defendants’ Seventh Amendment right to a jury trial on a contested and material

element of the claims against them.

In engaging in their interpretational endeavors and causing or sanctioning

the constitutional violations described above, was the Majority attempting to reach

a particular outcome? Was that outcome to ensure that Engle-progeny plaintiffs

Even more simply, we cannot give full faith and credit or preclusive effect to a judgment
when doing so would deprive a party of her property without due process of law. As detailed
above, litigants enjoy a “due process right to fully and fairly litigate each issue in their case.”
DuPont 771 F.2d at 880; see also Burson, 402 U.S. at 542, 91 S. Ct. at 1591 (“It is a proposition
which hardly seems to need explication that a hearing which excludes consideration of an
element essential to the decision . . . does not meet [the requirements of the Due Process
Clause].”). RJR and Philip Morris were never afforded an opportunity to litigate (1) whether the
cigarettes Ms. Graham smoked were defective, unreasonably dangerous, and negligently
produced and (2) whether RJR’s and Philip Morris’s tortious conduct caused Ms. Graham’s
death. We therefore cannot sanction the District Court’s deprivation of their property.
275
The Florida Supreme Court has recently rejected the defect at which they hint—that
all cigarettes are defective because they cause disease. See supra note 257 and accompanying
text.
276
See supra note 6 and accompanying text.
245
 Case: 13-14590 Date Filed: 05/18/2017 Page: 246 of 284

secure the same result in federal courts as they would achieve in Florida courts so

as to avoid unequal treatment resulting from the accident of diversity jurisdiction?

If so, the Majority fail to recognize the many constitutional impediments to their

desired outcome. They also fail to consider—as the Walker panel did when it

issued its Walker 0 opinion—the impact of Florida’s complete-mutuality

requirement. The Majority’s conclusion that the District Court did not deny RJR’s

and Philip Morris’s due process rights by precluding them from contesting (1) that

the cigarettes Ms. Graham smoked were defective, unreasonably dangerous, and

negligently produced and (2) that their tortious conduct caused her harm hinges on

the Majority’s (demonstrably false) factual determination that the Engle III Court

determined that the Phase I findings of fact were “specific enough” to establish that

Ms. Graham’s death was caused by unreasonably dangerous defects in RJR’s and

Philip Morris’s cigarettes and by RJR’s and Philip Morris’s negligent conduct.

Class plaintiffs who are not present in this case will be unable to rely on the

Majority’s fact findings.277 Future district courts attempting to implement the

Majority’s reasoning will thus need to make their own determination that the

plaintiff before them made a sufficient issue-preclusion proffer, including proof to

277
As explained above in Part II.A, Florida preclusion law—both issue and claim
preclusion—does not allow parties to successfully assert preclusion unless the recognizing-court
litigants are identical to the rendering-court litigants. In accordance with Engle III, due to the
“highly individualized” issues being litigated, the plaintiffs differ in each progeny case. 945 So.
2d at 1263; see also Forman v. Florida Land Holding Corp., 102 So. 2d 596, 598 (Fla. 1958)
(“Stare decisis relates only to the determination of questions of law, [and] . . . has no relation
whatever to the binding effect of determinations of fact.”)
246
 Case: 13-14590 Date Filed: 05/18/2017 Page: 247 of 284

“a reasonable degree of certainty” that the Engle III Court determined that the

Phase I findings are “specific enough” to preclude the defendants from contesting

essential elements of the claims against them. As recognizing courts with

Hansberry duties of their own, they will also need to make their own determination

about whether the Engle III Court made such a factual determination with adequate

notice and opportunity to be heard. In short, the only proposition for which the

Majority’s opinion stands is that the judgments rendered against RJR and Philip

Morris and in favor of Mr. Graham are affirmed.

VII.
THE FUNCTIONAL BAN ON CIGARETTES IS PREEMPTED BY FEDERAL
LAW

In addition to holding that the Engle defendants’ due process rights were not

violated, the Majority also hold that federal law does not preempt Florida’s

functional ban on cigarettes. Consistent with its due process analysis, the Majority

“constru[e] the [Phase I] findings as embracing a theory that all cigarettes

manufactured by the tobacco companies are defective and the sale of all cigarettes

is negligent.” Ante at 32. Nevertheless, the Majority hold that the “six tobacco-

specific [federal] laws that are relevant to this appeal” do not “reflect[] a federal

objective to permit the sale or manufacture of cigarettes.” Id. at 32, 36. Thus,

because banning cigarettes would not be preempted, Florida law “regulat[ing]

247
 Case: 13-14590 Date Filed: 05/18/2017 Page: 248 of 284

cigarette sales” by “impos[ing] tort liability on cigarette manufacturers” is not

preempted. Id. at 42.

Not only do the Majority’s and the Douglas III’s Court’s preclusion regimes

both engender severe due process violations, both are preempted by federal law.

Under the Majority’s regime, every cigarette is defective and unreasonably

dangerous and the very act of selling cigarettes is a breach of a duty of care. Under

the Douglas III Court’s regime, tobacco manufacturers are presumed liable for any

smoking-related injury. Either way, under Douglas III’s claim-preclusion

framework or the Majority’s issue-preclusion framework, cigarettes have

effectively been banned. Though the Majority and I agree on this point, we

disagree about whether such a ban is preempted by federal law. I believe that it is.

Our constitutional system contemplates “that both the National and State

governments have elements of sovereignty the other is bound to respect.” Arizona

v. United States, 567 U.S. __, __, 132 S. Ct. 2492, 2500, 183 L. Ed. 2d 351 (2012).

When state and federal law “conflict or [otherwise work] at cross-purposes,” id.,

the Supremacy Clause commands that federal law “shall be the supreme Law of

the Land . . . any Thing in the Constitution or Laws of any State to the Contrary

notwithstanding.” U.S. Const. art. VI, cl. 2. That is, state laws that “interfere with,

or are contrary to,” federal law cannot hold sway—they “must yield.” Gibbons v.

Ogden, 22 U.S. (9 Wheat.) 1, 211, 6 L. Ed. 23 (1824).

248
 Case: 13-14590 Date Filed: 05/18/2017 Page: 249 of 284

Federal law may preempt state law in three ways. First, Congress has the

authority to expressly preempt state law by statute. Crosby v. Nat’l Foreign Trade

Council, 530 U.S. 363, 372, 120 S. Ct. 2288, 2293, 147 L. Ed. 2d 352 (2000).

Second, even in the absence of an express preemption provision, “[t]he scheme of

federal regulation may be so pervasive as to make reasonable the inference that

Congress left no room for the States to supplement it.” Rice v. Santa Fe Elevator

Corp., 331 U.S. 218, 230, 67 S. Ct. 1146, 1152, 91 L. Ed. 1447 (1947). Third,

federal and state law may impermissibly conflict, for example, “where it is

impossible for a private party to comply with both state and federal law,” Crosby,

530 U.S. at 372, 120 S. Ct. at 2294; or when the state law at issue “stands as an

obstacle to the accomplishment and execution of the full purposes and objectives

of Congress,” Hines v. Davidowitz, 312 U.S. 52, 67, 61 S. Ct. 399, 404, 85 L. Ed.

581 (1941).278 It is this last subcategory of conflict preemption—obstacle

preemption—the Court faces here.279

278
In surveying this taxonomy, however, we must keep in mind that “[c]ategories and
labels are helpful, but only to a point, and they too often tend to obfuscate instead of illuminate.”
Fla. State Conference of the NAACP v. Browning, 522 F.3d 1153, 1167 (11th Cir. 2008); see
also English v. Gen. Elec. Co., 496 U.S. 72, 79 n.5, 110 S. Ct. 2270, 2275 n.5, 110 L. Ed. 2d 65
(1990); cf. Caleb Nelson, Preemption, 86 Va. L. Rev. 225, 264 (2000).
279
It is well-established that a lack of express preemption “does not bar the ordinary
working of conflict pre-emption principles.” Geier v. Am. Honda Motor Co., 529 U.S. 861, 869,
120 S. Ct. 1913, 1919, 146 L. Ed. 2d 914 (2000); see also This That & The Other Gift &
Tobacco, Inc. v. Cobb Cnty., 285 F.3d 1319, 1323 n.1 (11th Cir. 2002) (“The existence of an
express preemption clause, however, neither bars the ordinary working of conflict preemption
principles nor by itself precludes a finding of implied preemption.”). With this in mind, I will
address only obstacle preemption. Cf. Hillman v. Maretta, 569 U.S. __, __, 133 S. Ct. 1943,
249
 Case: 13-14590 Date Filed: 05/18/2017 Page: 250 of 284

A. Obstacle Preemption

Obstacle preemption leaves the tobacco companies with a tough row to hoe.

Supreme Court precedent teaches that “a high threshold must be met if a state law

is to be preempted for conflicting with the purposes of a federal Act.” Chamber of

Commerce v. Whiting, 563 U.S. 582, 607, 131 S. Ct. 1968, 1985, 179 L. Ed. 2d

1031 (2011) (quotation marks omitted). Indeed, “[i]mplied preemption analysis

does not justify a freewheeling judicial inquiry into whether a state statute is in

tension with federal objectives.” Id. (quotation marks omitted). That is because

“such an endeavor would undercut the principle that it is Congress rather than the

courts that preempts state law.” Id. (quotation marks omitted).

In addition to overcoming this “high threshold,” the tobacco companies must

also confront the presumption against preemption—namely, that “we start with the

assumption that the historic police powers of the States were not to be superseded

by [federal law] unless that was the clear and manifest purpose of Congress.”

Rice, 331 U.S. at 230, 67 S. Ct. at 1152.280 The presumption is a “cornerstone[] of

1949, 186 L. Ed. 2d 43 (2013) (holding a state law invalid under obstacle preemption without
discussing the scope of the federal statute’s express-preemption clause).
280
It is unclear whether Whiting applies the presumption against preemption, albeit sub
silentio, or whether it imposes an additional hurdle, above and beyond the presumption, to
making a successful obstacle-preemption argument.
250
 Case: 13-14590 Date Filed: 05/18/2017 Page: 251 of 284

our pre-emption jurisprudence.”281 Wyeth v. Levine, 555 U.S. 555, 565, 129 S. Ct.

1187, 1194, 173 L. Ed. 2d 51 (2009). And its logic carries particular force when,

as here, “federal law is said to bar state action in fields of traditional state

regulation.” N.Y. State Conference of Blue Cross & Blue Shield Plans v. Travelers

Ins. Co., 514 U.S. 645, 655, 115 S. Ct. 1671, 1676, 131 L. Ed. 2d 695 (1995). We

must recognize, therefore, “the historic primacy of state regulation of matters of

health and safety,” which can be enforced through state statutes and state tort law

alike.282 Medtronic, Inc. v. Lohr, 518 U.S. 470, 485, 116 S. Ct. 2240, 2250, 135 L.

Ed. 2d 700 (1996). Given the “great latitude” that states possess “under their

281
The presumption against preemption has been hotly debated, particularly when
applied to issues of statutory interpretation in cases involving express preemption. Compare,
e.g., PLIVA, Inc. v. Mensing, 564 U.S. 604, 622, 131 S. Ct. 2567, 2580, 180 L. Ed. 2d 580 (2011)
(Thomas, J.) (plurality opinion) (“[C]ourts should not strain to find ways to reconcile federal law
with seemingly conflicting state law.”), with id. at 641, 131 S. Ct. at 2591 (Sotomayor, J.,
dissenting) (“In the context of express pre-emption, we read federal statutes whenever possible
not to pre-empt state law.”). In the absence of an express preemption provision, however, the
presumption appears to rest on less contested ground, at least for the time being. Wyeth v.
Levine, 555 U.S. 555, 589 n.2, 129 S. Ct. 1187, 1208 n.2, 173 L. Ed. 2d 51 (2009) (Thomas, J.,
concurring) (“Because it is evident from the text of the relevant federal statutes and regulations
themselves that the state-law judgment below is not pre-empted, it is not necessary to decide
whether, or to what extent, the presumption should apply in a case such as this one, where
Congress has not enacted an express-pre-emption clause.”). That said, the presumption has a
tendency to make sporadic appearances in the Supreme Court’s preemption jurisprudence;
among the five preemption cases decided during the 2011 Term, for example, not one discussed
the presumption. Ernest A. Young, “The Ordinary Diet of the Law”: The Presumption Against
Preemption in the Roberts Court, 2011 Sup. Ct. Rev. 253, 331.
282
“[C]ommon-law damages actions . . . are premised on the existence of a legal duty . . .
. [I]t is the essence of the common law to enforce duties that are either affirmative requirements
or negative prohibitions. . . . At least since Erie R. Co. v. Tompkins, 304 U.S. 64, 58 S. Ct. 817,
82 L. Ed. 1188 (1938), we have recognized the phrase ‘state law’ to include common law as well
as statutes and regulations.” Cipollone v. Liggett Grp., Inc., 505 U.S. 504, 522, 112 S. Ct. 2608,
2620, 120 L. Ed. 2d 407 (1992) (plurality opinion) (interpreting an express preemption provision
contained in the Public Health Cigarette Smoking Act of 1969, Pub. L. No. 91-222, § 5(b), 84
Stat. 87 (codified at 15 U.S.C. § 1334(b)).
251
 Case: 13-14590 Date Filed: 05/18/2017 Page: 252 of 284

police powers to legislate as to the protection of the lives, limbs, health, comfort,

and quiet of all persons,” id. at 475, 116 S. Ct. at 2245 (quotation marks omitted),

we will not ascribe to Congress the intent “cavalierly [to] pre-empt state-law

causes of action,” id. at 485, 116 S. Ct. at 2250. To do otherwise would ignore

altogether that “[t]he allocation of powers in our federal system preserves the

integrity, dignity, and residual sovereignty of the States.” Bond v. United States,

564 U.S. 211, 221, 131 S. Ct. 2355, 2364, 180 L. Ed. 2d 269 (2011).

Finally, the lodestar of any preemption inquiry is congressional intent.

Retail Clerks Int’l Ass’n v. Schermerhorn, 375 U.S. 96, 103, 84 S. Ct. 219, 223, 11

L. Ed. 2d 179 (1963). In assessing the extent to which state law “stands as an

obstacle to the accomplishment and execution of the full purposes and objectives

of Congress,” Hines, 312 U.S. at 67, 61 S. Ct. at 404, “[w]hat [constitutes] a

sufficient obstacle is a matter of judgment, to be informed by examining the

federal statute as a whole and identifying its purpose and intended effects,”

Crosby, 530 U.S. at 373, 120 S. Ct. at 2294. To begin, then, “we must first

ascertain the nature of the federal interest.” Hillman, 569 U.S. at __, 133 S. Ct. at

1950.

252
 Case: 13-14590 Date Filed: 05/18/2017 Page: 253 of 284

B. Federal Regulation of Tobacco is Premised on Consumers’ Ability to
Choose

By my count, Congress has enacted at least seven statutes regulating tobacco

products in the past fifty years.283 I examine their text and structure, which provide

the most reliable indicia of what Congress has resolved itself to achieve. CTS

Corp. v. Waldburger, 575 U.S. __, __ 134 S. Ct. 2175, 2185, 189 L. Ed. 2d 62

(2014). This amounts to the “classic judicial task of reconciling many laws

enacted over time, and getting them to ‘make sense’ in combination.” United

States v. Fausto, 484 U.S. 439, 453, 108 S. Ct. 668, 676–77, 98 L. Ed. 2d 830

(1988).

I start with first principles. Congress possesses the constitutional authority

to ban cigarettes. See U.S. Const., art. I, § 8, cl. 3. It has never done so. This,

despite an ever-growing body of research documenting the health risks associated

with smoking. In 1964, for example, the Surgeon General issued a report

concluding that “[c]igarette smoking is a health hazard of sufficient importance in

the United States to warrant appropriate remedial action.” Advisory Comm. to the

Surgeon Gen. of the Public Health Serv., U.S. Dep’t of Health, Educ., & Welfare,

Smoking and Health 33 (1964), available at

283
The Majority assert that “Congress has enacted six tobacco-specific laws that are
relevant to this appeal,” excluding from its count the Family Smoking Prevention and Tobacco
Control Act of 2009 (“the TCA”). Ante at 32–34. Though I agree that the TCA does not directly
control this case, I nonetheless find it to be at least relevant in understanding the state of federal
law.
253
 Case: 13-14590 Date Filed: 05/18/2017 Page: 254 of 284

http://profiles.nlm.nih.gov/ps/access/NNBBMQ.pdf. The report warned “that

cigarette smoking contributes substantially to mortality from certain specific

diseases and to the overall death rate.” Id. at 31.

These findings spurred legislative action. Congress’s first attempt to address

cigarette smoking and its consequences came in the Federal Cigarette Labeling and

Advertising Act (the “Labeling Act”), Pub. L. No. 89-92, 79 Stat. 282 (1965)

(codified as amended at 15 U.S.C. §§ 1331–1341). The Labeling Act aimed to

“establish a comprehensive Federal program to deal with cigarette labeling and

advertising.” Id. § 2. Central to this comprehensive program was a requirement

that all cigarette packages display the warning statement, “Caution: Cigarette

Smoking May Be Hazardous to Your Health.” Id. § 4.

For our purposes, the Labeling Act is instructive because it encapsulates the

competing interests Congress has sought to reconcile when regulating cigarettes.

On one hand, Congress has recognized that smoking can cause serious physical

harm, even death. On the other hand, Congress has also acknowledged the

important role tobacco production and manufacturing plays in the national

economy. Congress has carefully calibrated these policy considerations by

promoting full disclosure to consumers about the attendant risks tobacco products

carry, thereby permitting consumers to make to a free but an informed choice. The

plain language of the Labeling Act summarizes well this approach:

254
 Case: 13-14590 Date Filed: 05/18/2017 Page: 255 of 284

It is the policy of the Congress . . . [that]

(1) the public may be adequately informed that cigarette
smoking may be hazardous to health by inclusion of a warning
to that effect on each package of cigarettes; and
(2) commerce and the national economy may be (A) protected
to the maximum extent consistent with this declared policy and
(B) not impeded by diverse, nonuniform, and confusing
cigarette labeling and advertising regulations . . . .

Id. § 2284

Since the Labeling Act’s passage, Congress’s basic goals have remained

largely unchanged. For example, Congress has tinkered with the text of the

warning labels affixed to cigarette packages in an effort to arm consumers with

more complete and accurate information. Public Health Cigarette Smoking Act of

1969, Pub. L. No. 91-222, § 4, 84 Stat. 87 (codified as amended at 15 U.S.C.

§ 1333); Comprehensive Smoking Education Act, Pub. L. No. 98-474, § 4, 98 Stat.

2200 (1984) (codified at 15 U.S.C. § 1333). To promote transparency, Congress

284
Senator Neuberger (D-OR), who introduced a version of the Labeling Act in the
Senate, put it this way:
I do not carry around with me a pair of scissors to cut off burning
cigarettes in the mouths of those I meet. I have never attacked a cigarette stand
with a hatchet. I have never equated smoking with sin. Abstention from tobacco
is not a condition of employment with my staff. I have never introduced
legislation nor have I ever delivered a speech calling for the abolition of
cigarettes. . . .
What have I advocated, then? Briefly, I believe there are four general
sectors of Government activity in which remedial action is justified: first,
education of both the presmoking adolescent and the adult smoker; second,
expanded research into the technology of safer smoking; third, reform of cigarette
advertising and promotion; and fourth, cautionary and informative labeling of
cigarette packages.
111 Cong. Rec. S13899 (daily ed. June 16, 1965) (statement of S. Neuberger).
255
 Case: 13-14590 Date Filed: 05/18/2017 Page: 256 of 284

has required the Secretary of Health and Human Services to issue a report to

Congress every three years regarding the “addictive property of tobacco.” Alcohol

and Drug Abuse Amendments of 1983, Pub. L. No. 98-24, 97 Stat. 175. Congress

has stepped in also to regulate smokeless tobacco products. Comprehensive

Smokeless Tobacco Health Education Act of 1986, Pub. L. No. 99-252, 100 Stat.

30. And Congress has even incentivized states to prohibit the sale of tobacco

products to minors by conditioning block grants on the creation of programs “to

discourage the use of . . . tobacco products by individuals to whom it is unlawful to

sell or distribute such . . . products.” Alcohol, Drug Abuse, and Mental Health

Administration Reorganization Act, Pub. L. No. 102-321, § 202, 106 Stat. 323

(1992) (codified at 42 U.S.C. § 300x-22).

All this, but Congress has enacted no ban on the sale of cigarettes to adult

consumers. No ban even though over the last fifty years a scientific consensus has

emerged that smoking can kill. The Surgeon General has reaffirmed this, at least

twice. Office of the Surgeon Gen., U.S. Dep’t of Health & Human Servs., The

Health Consequences of Smoking: Nicotine Addiction (1988), available at

http://profiles.nlm.nih.gov/ps/access/NNBBZD.pdf; Office of the Surgeon Gen.,

U.S. Dep’t of Health & Human Servs., The Health Consequences of Smoking—50

Years of Progress (2014), available at

http://www.surgeongeneral.gov/library/reports/50-years-of-progress/full-

256
 Case: 13-14590 Date Filed: 05/18/2017 Page: 257 of 284

report.pdf. The Environmental Protection Agency has classified secondhand

smoke as a known human carcinogen. Office of Health & Envtl. Assessment,

Respiratory Health Effects of Passive Smoking: Lung Cancer and Other Disorders

4 (1992), available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=2835.

The Food and Drug Administration (the “FDA”) has published research indicating

that “[t]he pharmacological processes that cause [nicotine addiction] are similar to

those that cause addiction to heroin and cocaine.” FDA, Jurisdictional

Determination, 61 Fed. Reg. 44619, 44631 (Aug. 28, 1996). These are, of course,

but a few examples.

In short, Congress has known about the dangers of cigarettes for many years.

Congress has regulated cigarettes for many years. But it has never banned them.

Indeed, regulation of cigarettes rests on the assumption that they will still be sold

and that consumers will maintain a “right to choose to smoke or not to smoke.”

H.R. Rep. No. 89-449 (1965), reprinted in 1965 U.S.C.C.A.N. 2350, 2352.

The Supreme Court has so concluded, holding that the FDA lacked

jurisdiction to regulate cigarettes because it would have otherwise been required by

statute to prohibit their sale. FDA v. Brown & Williamson Tobacco Co., 529 U.S.

120, 161, 120 S. Ct. 1291, 1315–16, 146 L. Ed. 2d 121 (2000). This result, the

Court determined, would have contravened the intent of Congress, given that “the

257
 Case: 13-14590 Date Filed: 05/18/2017 Page: 258 of 284

collective premise of these statutes is that cigarettes and smokeless tobacco will

continue to be sold in the United States.” Id. at 139, 120 S. Ct. at 1304.

And although Congress has since overruled this decision, granting the FDA

regulatory authority over cigarettes in 2009, Congress nonetheless stated that the

FDA “is prohibited from” “banning all cigarettes” or “requiring the reduction of

nicotine yields of a tobacco product to zero.” Family Smoking Prevention and

Tobacco Control Act (“the TCA”), Pub. L. No. 111-31, § 907(d)(3)(A)–(B), 123

Stat. 1776 (2009) (codified at 21 U.S.C. § 387g). To be sure, the TCA does not

“affect any action pending in Federal . . . court” prior to its enactment—including

this one. Id. § 4(a)(2); see Engle III, 945 So. 2d at 1277 (noting that Engle-

progeny cases must be filed within one year of the issuance of the case’s mandate).

It merely makes textually explicit what was already evident by negative

implication: Congress never has intended to prohibit consumers from purchasing

cigarettes. To the contrary, it has designed “a distinct regulatory scheme” to

govern the product’s advertising, labelling, and—most importantly—sale. Brown

& Williamson, 529 U.S. at 155, 120 S. Ct. at 1312.

C. Florida Has Imposed a Duty Not to Sell Cigarettes Contrary to Federal Law

I now turn to how these federal objectives interact with state law. Federal

law can expressly or impliedly preempt a state tort suit. E.g., Geier v. Am. Honda

258
 Case: 13-14590 Date Filed: 05/18/2017 Page: 259 of 284

Motor Co., 529 U.S. 861, 120 S. Ct. 1913, 146 L. Ed. 2d 914 (2000) (finding

implied preemption of state tort suit); Cipollone v. Liggett Grp., Inc., 505 U.S. 504,

112 S. Ct. 2608, 120 L. Ed. 2d 407 (1992) (plurality opinion) (finding express

preemption of certain state tort suits); see generally Williamson v. Mazda Motor of

Am., Inc., 562 U.S. 323, 330, 131 S. Ct. 1131, 1136, 179 L. Ed. 2d 75 (2011)

(collecting cases). A tort is “a breach of a duty that the law imposes on persons

who stand in a particular relation to one another.” Tort, Black’s Law Dictionary

1626 (10th ed. 2014). As such, successful tort actions “are premised on the

existence of a legal duty.” Cipollone, 505 U.S. at 522, 112 S. Ct. at 2620 (plurality

opinion); see also Geier, 529 U.S. at 881, 120 S. Ct. at 1925 (characterizing a

successful tort action as “a state law—i.e., a rule of state tort law imposing . . . a

duty”). Strict-liability and negligence claims like those at issue here are no

exception. Mut. Pharm. Co. v. Bartlett, 570 U.S. __, __, 133 S. Ct. 2466, 2474 n.1,

186 L. Ed. 2d 607 (2013) (“[M]ost common-law causes of action for negligence

and strict liability . . . exist . . . to . . . impose affirmative duties.”); Samuel

Friedland Family Enters. v. Amoroso, 630 So. 2d 1067, 1068 n.3 (Fla. 1994)

(recognizing, in the strict-liability context, that “[o]ne who sells any product in a

defective condition unreasonably dangerous to the user or consumer . . . is subject

to liability for physical harm thereby caused” even though “the seller has exercised

all possible care in the preparation and sale of his product” (quoting Restatement

259
 Case: 13-14590 Date Filed: 05/18/2017 Page: 260 of 284

(Second) of Torts § 402A)); Curd v. Mosaic Fertilizer LLC, 39 So. 3d 1216, 1227

(Fla. 2010) (noting that a negligence claim requires identification of “[a] duty, or

obligation, recognized by the law, requiring the [defendant] to conform to a certain

standard of conduct, for the protection of others against unreasonable risks”

(citation omitted) (second alteration in original)).

These duties, moreover, can stand as just as much of an obstacle to the

purposes and objectives of Congress as a state statute or administrative regulation.

E.g., Williamson, 562 U.S. at 329, 131 S. Ct. at 1136; Geier, 529 U.S. at 886, 120

S. Ct. at 1928. That is because, like any statute, common-law duties amount to

“either affirmative requirements or negative prohibitions.” Cipollone, 505 U.S. at

522, 112 S. Ct. at 2620 (plurality opinion). This Court’s job, then, is to determine

whether the legal duties underpinning Graham’s strict-liability and negligence

claims stand impermissibly as an obstacle to the achievement of federal

objectives—here, regulating, but not banning, the sale of cigarettes. To

accomplish this task, we must once again return to Engle and its progeny.

State laws, like that created by Douglas III, are broadly applicable, not

restrained by mutuality rules or class membership. Cf. Palsgraf v. Long Island

R.R. Co., 162 N.E. 99 (1928) (expounding as a matter of state law the rule that in

tort a duty is owed only to foreseeable plaintiffs). In other words, under Douglas

III, anyone who is addicted to cigarettes can hold any tobacco company liable for

260
 Case: 13-14590 Date Filed: 05/18/2017 Page: 261 of 284

damages simply by proving addiction and injury—“injury as a result of . . .

conduct is assumed.”285 Douglas III, 110 So. 3d at 429.

That private litigants, rather than executive agencies, are enlisted to enforce

the ban does not diminish its potency. Although no executive agency intervenes to

prevent tobacco companies from continuing to sell cigarettes while paying the

resulting damages, “pre-emption cases do not ordinarily turn on such compliance-

related considerations as whether a private party in practice would ignore state

legal obligations—paying, say, a fine instead—or how likely it is that state law

actually would be enforced.” Geier, 529 U.S. at 882, 120 S. Ct. at 1926; cf.

Cipollone, 505 U.S. at 521, 112 S. Ct. at 2620 (plurality opinion) (noting that state

regulation “can be as effectively exerted through an award of damages as through

some form of preventive relief. The obligation to pay compensation can be, indeed

is designed to be, a potent method of governing conduct and controlling policy.”

(quotation marks omitted)).

Admittedly, how compliance-related considerations should factor into

preemption analysis—if at all—remains something of an open question. “The

Court has on occasion suggested that tort law may be somewhat different, and that

related considerations—for example, the ability to pay damages instead of

285
If Douglas III holds as a matter of law that addictiveness is an unreasonably
dangerous defect, query whether such a law creates a cause of action against other companies
whose products are addictive and can cause negative health consequences: alcohol distillers,
prescription-drug distributers, coffee roasters, and chocolatiers.
261
 Case: 13-14590 Date Filed: 05/18/2017 Page: 262 of 284

modifying one’s behavior—may be relevant for pre-emption purposes.” Geier,

529 U.S. at 882, 120 S. Ct. at 1926.286 We do not write on a blank slate, however.

Justice Blackmun’s opinion for himself and two other Justices in Cipollone

forcefully contends that tort law should be treated differently from positive

enactments for preemption purposes. 505 U.S. at 536–37, 112 S. Ct. at 2627–28

(Blackmun, J., concurring in part and dissenting in part) (“The effect of tort law on

a manufacturer’s behavior is necessarily indirect. . . . The level of choice that a

defendant retains in shaping its own behavior distinguishes the indirect regulatory

effect of the common law from positive enactments such as statutes and

administrative regulations.”). But Justice Blackmun lost that argument: his

opinion did not command a majority. And critically, his logic was called into

question by a majority of the Court in Geier. 529 U.S. at 882, 120 S. Ct. at 1926

(“[T]his Court’s pre-emption cases ordinarily assume compliance with the state-

law duty in question.”). Absent more specific guidance from the Supreme Court,

286
For this proposition, Geier relies on a trio of cases relating to field preemption and the
Atomic Energy Act, which are far removed, both factually and legally, from this appeal. English
v. Gen. Elec. Co., 496 U.S. 72, 110 S. Ct. 2270, 110 L. Ed. 2d 65 (1990); Goodyear Atomic
Corp. v. Miller, 486 U.S. 174, 108 S. Ct. 1704, 100 L. Ed. 2d 158 (1988); Silkwood v. Kerr-
McGee Corp., 464 U.S. 238, 104 S. Ct. 615, 78 L. Ed. 2d 443 (1984). As such, these three cases
are far too thin a reed to rely upon. And in any event, Geier itself clearly places a thumb on the
scale in favor of assuming compliance with the duties imposed through a successful state tort
suit. 529 U.S. at 882, 120 S. Ct. at 1926.

262
 Case: 13-14590 Date Filed: 05/18/2017 Page: 263 of 284

this Court must follow Geier’s lead in assuming that the tobacco companies will

comply with whatever state-law duties Florida may impose.

Nor is it convincing to argue that Congress, well aware of state tort litigation

against the tobacco companies, would not have intended to preempt state-law

claims similar to the two at issue here. See Wyeth v. Levine, 555 U.S. 555, 574–75,

129 S. Ct. 1187, 1200, 173 L. Ed. 2d 51 (2009) (“If Congress thought state-law

suits posed an obstacle to its objectives, it surely would have enacted an express

pre-emption provision . . . . Its silence on the issue, coupled with its certain

awareness of the prevalence of state tort litigation, is powerful evidence that

Congress did not intend [to preempt state tort suits.]”); cf. Bonito Boats, Inc. v.

Thunder Craft Boats, Inc., 489 U.S. 141, 166–67, 109 S. Ct. 971, 986, 103 L. Ed.

2d 118 (1989). That proposition may be true at a high level of generality. But as I

have explained in great detail, Graham’s is not a run-of-the-mill tort suit. If it

were, our analysis would be radically different. Make no mistake: this opinion

should not be taken to mean that I believe Congress intended to insulate tobacco

companies from all state tort liability. To the contrary, there is nothing in the text,

structure, or legislative history of the federal statutes examined above to support

such a far-reaching proposition.

I merely conclude that, having surveyed both federal and state law, it is clear

that Congress would have intended to preempt Graham’s strict-liability and

263
 Case: 13-14590 Date Filed: 05/18/2017 Page: 264 of 284

negligence claims, rooted as they are in a broadly applicable state law set forth by

the Florida Supreme Court that deems all cigarettes defective, unreasonably

dangerous, and negligently produced. I therefore express no opinion as to other

state-law suits that may rest on significantly narrower theories of liability.

D. The Majority Misinterpret the Statutory Framework of Tobacco Regulation

At bottom, the Majority and I disagree over how to understand the federal

statutory framework regulating tobacco in place at the time of Engle III and

Douglas III and how to understand the Supreme Court’s decision in Brown &

Williamson interpreting that framework. Though the importance of our

disagreement should not be minimized—and given the uncertainty surrounding this

particular issue and preemption generally, I would once again urge the Supreme

Court to clarify the hazy state of preemption law—I end by highlighting the narrow

scope of our disagreement. Like the Majority, I agree that the Tenth Amendment

guarantees that “State governments retain their historic police powers to protect

public health.” See ante at 41. Also like the Majority, I fully embrace the “happy

incident” of Our Federalism that States are generally free to serve as the proving

grounds for “novel social and economic experiments” of how best to further public

health. See id. (quoting New State Ice Co. v. Liebmann, 285 U.S. 262, 311, 52 S.

Ct. 371, 386–87, 76 L. Ed. 747 (1932) (Brandeis, J., dissenting)). And like the

264
 Case: 13-14590 Date Filed: 05/18/2017 Page: 265 of 284

Majority, I see no reason to question the constitutionality of Tennessee’s ability to

ban the sale of cigarettes at the end of the 19th century, long before Congress

began regulating tobacco nationally in any significant fashion. See id.; Austin v.

Tennessee, 179 U.S. 343, 21 S. Ct. 132, 45 L. Ed. 224 (1900).

Our disagreement is simply this. I understand the federal statutory

framework regulating tobacco in place at the time of Engle III and Douglas III, as

the Supreme Court’s decision in Brown & Williamson confirms, to allow states

wide leeway to concurrently regulate tobacco while prohibiting states from

banning the sale of cigarettes outright. As a result, we cannot give effect to the

Florida Supreme Court’s decisions in a manner that operates as a ban on the sale of

cigarettes without elevating state law over federal law, which the Supremacy

Clause forbids. The Majority come to the opposite conclusion. Because I believe

the Majority err by doing so, I must dissent from the Majority’s preemption

holding as well.

CONCLUSION

In 2003, the Third District Court of Appeal effectively ended the Engle

litigation. It decertified the Engle class because the Phase I proceeding had failed

to achieve its purpose—the Phase I jury was not instructed to make “specific

findings as to any act by any defendant at any period of time,” much less

“determine whether defendants were liable to anyone.” Engle II, 853 So. 2d at

265
 Case: 13-14590 Date Filed: 05/18/2017 Page: 266 of 284

450, 467 n.8. With no useful findings on which to rely, plaintiffs could sue

tobacco manufacturers in individual lawsuits, but they would have to start from

scratch.

Three years later, the Florida Supreme Court, lamenting the demise of the

Engle litigation, sua sponte crafted a “pragmatic solution” designed to rejuvenate

it. Engle III, 945 So. 2d at 1269. Its solution entailed “retaining [most of] the

jury’s Phase I findings” —the only things that were left standing after the Court’s

decertification of the class —and puzzlingly declaring, in dicta, that those findings

“will have res judicata effect in [progeny] trials.” Id. In retaining these useless

findings and directing the hundreds of thousands of class members to file claims

within the year, the Court sent a signal to progeny courts that they should attempt

to develop some rationale for propping up the plaintiffs’ cases and allowing them

to recover.

The result, to put it mildly, “was some confusion among the courts.” R.J.

Reynolds Tobacco Co. v. Marotta, __ So. 3d __, No. SC16-218, 2017 WL

1282111, at *9 (Fla. Apr. 6, 2017). One Judge compared litigating and

adjudicating a progeny case to “play[ing] legal poker, placing . . . bets on questions

left unresolved by Engle [III].” Jimmie Lee Brown II, 70 So. 3d at 720 (May, J.

concurring). When this Court reviewed an Engle-progeny case in Brown II, we

placed our bet on a belief that the Florida Supreme Court had not attempted to

266
 Case: 13-14590 Date Filed: 05/18/2017 Page: 267 of 284

secretly transform the useless Phase I findings into a hefty jackpot for Engle class

members. See supra Part III.B.

Recognizing that our bet effectively ended the game for progeny plaintiffs,

the First District Court of Appeal, in Martin II, made a different bet. It sua sponte

interpreted Engle III as implicitly holding sua sponte that the Phase I findings were

far from useless—“[n]o matter the wording of the findings on the Phase I verdict

form.” Martin II, 53 So. 3d at 1067. Rather than being useless, the findings

established the defendants’ liability to all class members. Engle III’s holding to

this effect (to say nothing of its secret and sua sponte nature) was not a problem,

the First District reasoned, because sufficient evidence was presented at trial upon

which a properly instructed jury could have found that the defendants were liable

to all class members.

Hoping to avoid stepping into the fray and upsetting a rationale advanced by

the First District, the Florida Supreme Court denied certiorari review of Martin II.

But it soon became clear that the Florida Supreme Court would need to reenter the

scene to sort out the reckless betting it set in motion. Shortly after Martin II was

issued, the Second and Fourth Districts placed slightly different bets. They

understood, as Martin II did, that they were supposed to hold the defendants liable

by “some mechanism.” They acquiesced in that respect, as they felt “constrained,”

Jimmie Lee Brown II, 70 So. 3d at 715, to achieve the result the Engle III Court

267
 Case: 13-14590 Date Filed: 05/18/2017 Page: 268 of 284

seemed to desire. However, they had serious reservations about doing so, and

could not coherently explain how “the smokers would prove causation in

individual cases,” Ante at 13, when the Phase I findings do not even reveal the

basis for the supposed defect in the defendants’ cigarettes. In the midst of such

bewilderment, the Second District asked the Florida Supreme Court for help with a

certified question.

To settle “the confusion among the courts,” Marotta, at *9, the Douglas III

Court reluctantly accepted the certified question, and chose the rationale it liked

best. Preferring the Martin II outcome, it declared the First District’s bet a winner

and ours a loser. Endorsing Martin II’s analysis, the Florida Supreme Court

concluded that it had sua sponte secretly ruled in 2006 that the Phase I findings

were nonspecific liability determinations—“useless” under an issue-preclusion

framework, but dispositive under a claim-preclusion framework. Douglas III, 110

So. 3d at 433. Progeny plaintiffs thus had nothing to do but plead claim

preclusion, proffer the Engle III opinion and the Phase I findings, prove their class

membership and damages, and defend against claims of comparative fault.

Granted, it was outlandish for the Douglas III Court to suggest that Engle III

adjudicated the claims of all the absent and unidentified class members, but in the

Engle-progeny poker parlor, the house always wins. Under Martin II’s claim-

preclusion rationale, progeny courts could be saved from the embarrassing and

268
 Case: 13-14590 Date Filed: 05/18/2017 Page: 269 of 284

impossible task of explaining how the unidentified defect in the Phase I findings

caused each class plaintiff’s harm, because this aspect of the plaintiff’s claim

would be treated as having been established in Engle III. Eager to stop the

progeny courts from asking questions, Douglas III endorsed the rationale.

When the Walker panel entered the parlor, it was faced with the question of

whether such a preclusion regime violated the Engle defendants’ due process

rights. In tackling this question, the panel rejected the Douglas III Court’s

portrayal of the Phase I findings as nonspecific liability determinations, noting

language in Engle III that contradicted that portrayal. Instead, it adopted a

portrayal of the Phase I findings that the plaintiffs before it had not advanced and

that no court had adopted previously: the Phase I findings were factual

determinations that were “specific enough” to identify the conduct the Phase I jury

deemed tortious. Walker 0¸at 20. The panel then imputed to the Douglas III

Court that portrayal and sua sponte gave full faith and credit to it, not realizing that

Florida law forbade giving full faith and credit to Douglas III. Upon realizing its

mistake, the panel was forced to vacate its opinion and issue another one. But

instead of rethinking its counterfactual rationale, the panel simply gave full faith

and credit to Engle III instead of Douglas III and left all of its inapposite reasoning

intact. The panel, therefore, never evaluated what Engle III decided or whether the

Engle III Court violated the defendants’ due process rights by making a dispositive

269
 Case: 13-14590 Date Filed: 05/18/2017 Page: 270 of 284

determination about the Phase I findings secretly and without affording the parties

notice or opportunity to be heard.

The Majority now double down on the Walker panel’s misplaced bet. Like

the Walker panel, the Majority reject the Douglas III Court’s portrayal of the Phase

I findings as liability determinations. Instead, echoing the Walker panel, they

adopt a portrayal of the Phase I findings that Mr. Graham neither advanced nor

proffered evidence to support. This time, instead of imputing their “specific

enough” portrayal to Douglas III, they impute it to Engle III. In support of their

portrayal, the Majority mine the trial record and proffer excerpts for the plaintiff,

concluding that Engle III’s supposed portrayal of the Phase I findings is not

arbitrary because a properly instructed jury could have made such findings.

Although the Majority did not make the mistake of giving full faith and credit to

Douglas III, they still fail to evaluate, as Florida law and the U.S. Constitution

require, whether the Engle III Court violated the defendants’ due process rights by

making a dispositive determination about the Phase I findings secretly and without

affording the parties notice or opportunity to be heard.

In short, Engle III sent a signal to progeny courts to develop a rationale for

holding the defendants liable to class plaintiffs. The Florida courts, most explicitly

in Douglas III, then developed a rationale that the Walker panel and the Majority,

correctly, albeit implicitly, recognize is unconstitutional. Yet, instead of simply

270
 Case: 13-14590 Date Filed: 05/18/2017 Page: 271 of 284

refusing to apply the Florida courts’ unconstitutional rationale, the Walker panel

and now the Majority, develop their own rationale that is similarly sullied with

constitutional errors.

If one lesson can be learned from this chaotic poker game it is that we

should stick to our day jobs. Rather than act as advocates for the plaintiff, we

should saddle him with the burden the law tasks him with carrying, and assess,

impartially, whether the plaintiffs have established the elements of proving

preclusion in the manner the law demands. On the record before us now, the

plaintiff clearly has not, and the District Court’s judgment should be reversed.

271
Case: 13-14590 Date Filed: 05/18/2017 Page: 272 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 273 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 274 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 275 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 276 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 277 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 278 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 279 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 280 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 281 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 282 of 284
Case: 13-14590 Date Filed: 05/18/2017 Page: 283 of 284
 Case: 13-14590 Date Filed: 05/18/2017 Page: 284 of 284

WILSON, Circuit Judge, dissenting:

At its most fundamental level, the Due Process Clause guarantees an

aggrieved party notice and “the opportunity to present his case and have its merits

fairly judged.” See Logan v. Zimmerman Brush Co., 455 U.S. 422, 428, 433, 102

S. Ct. 1148, 1153, 1156 (1982). The defendants have no doubt been provided

notice and some degree of opportunity to be heard in court, but like Judge Tjoflat, I

am not content that the use of the Engle jury’s highly generalized findings in other

forums meets “the minimum procedural requirements of the . . . Due Process

Clause in order to qualify for . . . full faith and credit.” See Kremer v. Chem.

Constr. Corp., 456 U.S. 461, 481, 102 S. Ct. 1883, 1897–98 (1982). Because such

a violation requires remand, I see no need to determine whether the Engle findings

are preempted by federal law.

272