Cato Institute
1000 Massachusetts Ave, NW
Washington DC 20001-5403
 Phone (202) 842-0200
Fax (202) 842-3490
Contact Us
Cato Policy Analysis No. 275 June 20, 1997

Policy Analysis

Tobacco Medicaid Litigation:
Snuffing Out the Rule of Law

by Robert A. Levy

Robert A. Levy is senior fellow in constitutional studies at the Cato Institute.

Executive Summary

Tobacco is under siege. No fewer than 39 states are suing the industry to recover Medicaid outlays for smoking-related illnesses. Florida has led that effort with a new statute, allegedly resting on principles of equity, that strips tobacco companies of their traditional rights and puts in their place a shockingly simple rule of law: the state needs money; the industry has money; so the industry shall give and the state shall take.

During the past 40 years, not a single smoker received a single dollar of damages from tobacco companies, as juries repeatedly concluded that smokers are responsible for their own behavior and their own losses. Yet under the new regimen, if a smoker happens to be a Medicaid recipient, individual responsibility is out the window. The same tobacco company selling the same person the same product that results in the same injury is, magically, liable, not to the smoker, but to the state. By legislative fiat, liability hinges on a smoker's Medicaid status, a fortuity totally unrelated to any misdeeds of the industry.

And it gets worse. The state is not even required to show that a particular party was harmed by his use of tobacco. Instead, causation may be proven by statistics alone. One would think that the industry could at least investigate whether patients suffering from "smoking-related illnesses" ever smoked. Wrong. Incredibly, the industry will be allowed to depose only 25 of 400,000 Medicaid claimants.

These lawsuits retroactively eradicate settled doctrine and deny due process to an industry singled out for its deep pockets and public image, not its legal culpability. The mark of a free society is how it treats not its most but its least popular members. The rule of law must be steadfastly upheld today, or none of us will be secure tomorrow.

Introduction

Hardly a day goes by that the newspapers, magazines, radio stations, and television networks don't bring us the latest battlefield update in the tobacco wars. If it's not the estate of a three-pack-a-day smoker contending that cigarettes are a defective product, it's a class action suit by airline flight attendants claiming injury from secondhand smoke, a nationwide settlement exchanging legal immunity for big bucks and more regulation, a proposal by an administrative agency to restrict cigarette advertising, or a court ruling on the question of whether tobacco is a drug. With tobacco so much in the news, it is easy to conclude that the underlying issues are industry specific. That would be a big mistake. We are dealing here with moral, political, and legal questions that transcend any single industry. What's at stake is nothing less than our principle of individual responsibility, our choice between legislation and adjudication, and our constitutional right to due process--in short, the rule of law. To illustrate, consider this short hypothetical, which today is all too close to reality, as we shall see.

Imagine that Joe, an avid and experienced skier on vacation in the Colorado Rockies, puts on his skis, heads down the mountain, loses control, and suffers a badly broken leg that requires extensive medical care. Back home in Florida, Joe incurs $3,500 in doctor and hospital bills. Five years later he sues the manufacturer of the skis, alleging that they were defectively designed. More precisely, because he no longer owns the skis and cannot identify the particular manufacturer, Joe sues all of the manufacturers that sell skis in Florida. Never mind that he doesn't recall whether the skis were purchased in Florida; never mind that the accident occurred in Colorado. Because his medical bills were paid in Florida, Joe insists that all ski manufacturers must be held liable in proportion to their revenues in that state. And when asked how he knows that the skis were defective, Joe points to statistics showing that there is a significantly higher probability of breaking a leg if you ski than if you don't. Indeed, observes Joe, using skis exactly as they are designed to be used often leads to injury. Joe is not persuaded by data indicating that a large majority of skiers never break their legs.

Predictably, the judge dismisses Joe's claim, then sanctions his attorney for bringing a frivolous lawsuit. Not only was Joe's litigation time barred--five years after the event--but Joe also failed to produce any evidence that would link his injury to the skis he used. Moreover, even if Joe had filed on time, proven causation, and identified the ski manufacturer, he would undoubtedly have lost the case because he knowingly and voluntarily assumed the risk of injury. He was aware that skiing was risky, yet he elected of his own free will to assume that risk.

Suppose instead that Joe has an insurance policy that covers his $3,500 medical expense. Like virtually all such policies, Joe's contains a subrogation clause, which allows the insurance company (the subrogee) to substitute itself for Joe and sue any party that Joe could have sued. That way, if Joe has a lawful claim against someone who injures him, the wrongdoer ultimately pays and the insurance company is reimbursed. In return for agreeing to the subrogation clause, Joe gets the benefit of a much lower premium. So, in this instance, it would be not Joe but his insurer that could sue the ski manufacturers. The rules of subrogation are crystal clear, however. The insurer bears the same burden of proof that Joe would have borne, and the ski manufacturers can assert the same defenses against the insurer that they could have against Joe himself. However, given the frivolous nature of Joe's claim, no insurance company would file a lawsuit.

Before switching our focus to tobacco, let's change the facts of Joe's accident ever so slightly. Assume if you will that Joe qualifies for Medicaid reimbursement, that the state of Florida pays the $3,500 bill, and that the state--not Joe and not his insurance company--then attempts to recover its loss by suing the ski manufacturers. What result? "Same result," you say. After all, Medicaid is just an insurer by another name. Part of Joe's Medicaid "contract" requires that he assign any lawful claims to the state--exactly like subrogation. The rules of subrogation allow Florida, as Joe's insurer, to step into his shoes and litigate on his behalf if it thinks that someone else is liable for his injury. Joe would have lost; his private insurer would have been too embarrassed even to file suit. Thus, Florida loses. Right?

Wrong. Hard as it may be to believe, the Florida Medicaid Third-Party Liability Act, a 1978 statute amended in 1990 and 1994, [1] establishes a simple new rule: the state needs money; the manufacturers have money; so the manufacturers shall give and the state shall take. That overstates the matter only slightly. Joe would get nowhere if he sued the ski manufacturers directly; yet if he happens to be a Medicaid recipient, then the same defendants selling the same person the same product that results in the same injury are, magically, liable to the state. By legislative fiat, liability hinges on the injured party's Medicaid eligibility--a quirk that bears absolutely no rational relationship to the defendants' alleged wrongdoing.

Under the act, Florida is not limited to stepping into the shoes of the injured party. The statute gives the state a totally new, direct cause of action on the theory that the injury to Joe has independently harmed the state. Florida need not be concerned about waiting too long to sue; the statute abolishes all time limits. [2] Moreover, Florida can collect from the manufacturers even when losses are the injured party's own fault; the statute abrogates all of the manufacturers' affirmative defenses, including assumption of risk. The state does not even have to show a link between the harm to any particular party and his use of the manufacturers' product; the statute says that generalized causation may be proven by statistical analysis and Florida can seek reimbursement from all manufacturers that sell the product within the state--even if a person who retired to Florida purchased and used the product elsewhere.

And it gets worse. One would think that in preparing their defense the manufacturers would be allowed to discover--by deposition, interrogatories, physical examination, and other methods of gathering evidence--whether the claimed injuries were real or fraudulent, serious or trivial, and whether the patient ever used the manufacturers' product. Wrong again. Incredibly, the act originally provided that Florida was not required to identify the individual recipients of Medicaid payments; instead, the state could seek recovery for all recipients, anonymously, as a group. That section of the statute was too much for the Florida Supreme Court. In striking it, the court remarked, "It is illogical and unreasonable to call this a fair process. . . . [I]t is violative of the due process provisions of our constitution." [3]

The application of the Florida act to Joe's injury may seem bizarre, but astonishingly, it is the law. At present, ski manufacturers selling in Florida can rest easy, not because the act doesn't apply to them--it does--but because Gov. Lawton Chiles has shown little interest in enforcing it against them. Instead, he has focused on two industries he holds in special contempt. In March 1995 he issued an executive order directing state officials to pursue only those industries "responsible for disease and death caused by tobacco products and those responsible for disease and death caused by the sale and consumption of illegal drugs." [4] Unless we can envision a court compelling the Medellin cartel to reimburse Florida for crack-related Medicaid costs, the tobacco companies are the real targets. Yet, in principle and in fact, Florida's other industries could be on the hook in time: the governor could add to the list; a future governor might have bigger ideas; and a court has already slapped down the executive order by holding, not that the act is overbroad, but that the governor had no authority to limit its application to tobacco companies. [5] Thus, as a matter of prosecutorial discretion, only tobacco is presently under siege in Florida.

Across the nation, the theory behind the Florida approach is very much alive. In fact, as of mid-June 1997, 39 state attorneys general had filed Medicaid recovery suits against the industry. [6] Similar litigation has been filed by Puerto Rico, New York City, San Francisco, and Los Angeles and by private citizens in Alabama and Ohio. Only Florida has codified its legal theory by enacting legislation that strips tobacco companies of their traditional rights; the other states, for the moment, are proceeding under alleged common law and equitable principles that in most respects replicate the Florida statute.

Five states--Alabama, Ohio, Georgia, Nevada, and Colorado--have expressed varying degrees of skepticism about the current brand of Medicaid suits. [7] Presumably, the major tobacco-growing states will also decline to adopt the Florida theory. Yet, even if no more states do so, the potential liability could threaten the survival of some or all of the nation's tobacco companies--a remedy arguably more painful and protracted, and surely less certain and effective, than simply declaring the product illegal.

But it is not the durability of an industry that matters most. Far more weighty is the sustenance of a legal system that maximizes freedom while demanding accountability--no small assignment when ambitious politicians stalk an unpopular industry and when anti-smoking zealots join forces with elected state judges and plaintiffs' lawyers driven by huge contingency fees.

The analysis that follows covers several of the more important aspects of the Medicaid reimbursement suits, using the Florida Medicaid Third-Party Liability Act as the primary, though not exclusive, point of reference. After a brief historical review, I examine the principal legal mutations arising under the Florida statute: abrogation of affirmative defenses, statistical proof of causation, and market share liability. I next consider recoverable damages and contrast the health care costs for which the states seek indemnity with the various tax benefits and cost savings that accrue to the states from the sale of tobacco products. Finally, the analysis focuses on contingency fees, then compares public and private law approaches to liability.

It is important to establish at the outset that this study is not a defense of the tobacco industry. It is a defense of the rule of law. Litigation for Medicaid reimbursement hinges on a freshly minted cause of action, created out of whole cloth by states filling the dual and conflicting roles of lawmaker and plaintiff. The result is the retroactive abrogation of settled principles of law and denial of due process to a single industry selected more for its wealth and public image than for its legal culpability. It should be noted, however, that the industry is also being sued under conventional doctrines such as common law fraud and misrepresentation, antitrust, unjust enrichment, and public nuisance. Such suits are inventive and probably meritless, but at least they do not rely on gutting the industry's ability to defend itself. If the plaintiffs can prove any of their claims--the March 1997 Liggett settlement suggests that fraud and misrepresentation are the most likely bets [8]--then the industry will rightly be held accountable for its transgressions, but only after a trial in which commonly accepted rules are evenhandedly applied--the same rules that would apply to any other defendant and any other plaintiff.

Historical Background

In over 40 years of litigation by smokers and their families, there has not been one final adjudication of damages against a tobacco company. [9] Only after disclosure of embarrassing documents suggesting that the industry knew long ago that its products were addictive did a Florida jury in August 1996 award $750,000 to a victim of lung cancer; [10] but even that single victory may not be upheld on appeal. In May 1997, the same attorney, fresh from his triumph nine months earlier, represented another lung cancer victim before a different Florida jury. Despite the highly publicized admissions in Liggett, the jury refused to award either compensatory or punitive damages. [11]

Industry defendants invariably prevail, first, by disputing that tobacco was shown to be the cause of a particular person's ailment and, second, by demonstrating that the plaintiffs knowingly and voluntarily assumed the risk of smoking, in which case they are responsible for the consequences of their own behavior. Further, when Congress enacted the Federal Cigarette Labeling and Advertising Act of 1965, as amended in 1969, [12] the tobacco companies were able to argue that the act preempted some common law claims because of the obligatory warning labels on cigarette packs--a position that the Supreme Court finally affirmed in the Cipollone case. [13] After Cipollone, plaintiffs could no longer contend in state court that cigarette advertising or promotion failed to warn that smoking carries serious health risks. [14] Allegations of fraud, breach of express warranty, and design defect were still viable, but they had never convinced a jury.

That set the stage for some legal legerdemain--a fresh wave of litigation by state attorneys general, beginning in 1994, grounded in a new cause of action independent of the rights of any aggrieved cigarette smoker. By amending its Medicaid Third-Party Liability Act in 1990 and 1994, Florida not only codified the new claim but also corrupted the rules of engagement in a manner that would guarantee victory if the state had to litigate. A year later, when the legislature voted to repeal the new provisions, Governor Chiles vetoed the repeal. [15]

While the Florida statute still requires that the state prove either negligence or a defective product, causation, and damages, in six essential respects the amendments tilt the playing field hopelessly against the industry. Indeed, the express language of the statute leaves little room for doubt.

According to Florida, those provisions merely effectuate the intent of the liability scheme under the federal Medicaid statute, [16] which directs Medicaid-eligible individuals to assign to the state all rights to payment for medical care from third parties. The state can then sue the third party, and any amount collected reimburses the state to the extent of its Medicaid expenditures on behalf of the covered individual. A portion, which depends on the federal percentage participation in the state plan, of the reimbursed expenditures goes to the federal government, and any remainder goes to the patient.

Until the new dispensation, however, the state could assert only those theories of recovery that the patient could assert. The same rules of evidence, the same standards of responsibility, and the same burden of proof applied to the state, suing under subrogation, as to the patient suing on his own behalf. If a tobacco company would not have been liable in a direct suit by a cancer victim, then it would not have been liable in a similar suit by the state of Florida standing in the victim's shoes. Those fundamental principles of law have now been cavalierly dismissed by states intent on picking the deep pockets of the tobacco industry. The potential for abuse is manifest as self-interested legislators determine the legal regime under which a few friendless, nonvoting corporations will be fleeced.

At a minimum, one would have expected that the courts would strictly scrutinize that type of opportunistic legislation. [17] That was not the case. In a June 1996 suit challenging the constitutionality of the Medicaid Third-Party Liability Act, the Florida Supreme Court upheld, by a vote of four to three, the bulk of the act, calling it "a rational response to a public need." [18] That reasoning is so broad that it can justify almost anything. The majority noted that legislators frequently alter common law rights, and judicial review is typically deferential. Only the dissent seemed aware of the quantum difference between altering the common law and eviscerating a company's ability to dispute the charges brought against it. And only the dissent comprehended the affront to due process that occurs when a state fattens its own coffers by applying special rules--rules that operate solely when the state is the plaintiff and, in the naked words of the statute itself, "to the extent necessary to ensure full recovery."

To its modest credit, the Florida Supreme Court refused to foreclose the possibility of later challenges to specific applications of the act. The court held that the statute was not invalid on its face, but it reserved judgment on the question whether the statute could be enforced against an actual defendant in an actual case without violating the Constitution. Furthermore, the court did strike or modify three provisions of the act "to avoid offending the due process guarantees of the Florida Constitution." First, the court modified the state's abolition of the statute of limitations so as to reduce its retroactive impact; Florida will not be permitted to resurrect a claim that was already time barred when the Medicaid Third-Party Liability Act was enacted. Second, the court rejected Florida's bid to use both market share liability (whereby damages are apportioned by tobacco sales in the state) and joint and several liability (whereby the state can pursue any defendant for the entire damage award), finding that the two theories of recovery are "fundamentally incompatible." Still, in any given case, Florida will be able to use either theory, just not the two in combination. Third, and most important, the court threw out the utterly fantastic section of the act that authorized the state to sue without identifying a single Medicaid patient. That such a shameless provision was ever drafted, much less passed, speaks volumes about the state's aims and its legislators' regard for the rule of law.

On March 17, 1997, the U.S. Supreme Court declined to review the Florida Supreme Court's decision. [19] To many observers, the high court's reluctance to take the case was not surprising. Florida is the only one among the 39 states seeking recovery of Medicaid outlays that is suing under a statute that has undergone judicial scrutiny. Until there is a split among the various jurisdictions, or an especially egregious application of the statute in a particular case, it is unlikely that the U.S. Supreme Court will intercede. Because the tobacco companies were unsuccessful in their attempt to have the act invalidated, Florida's suit against the industry for Medicaid reimbursement can now move forward. Trial is scheduled to begin this August; lawyers expect the battle to last several months. [20]

The dubious honor of being the first state to have a reimbursement action go to trial will probably belong to Mississippi. Its Medicaid recovery case is due for trial in July, having withstood a state supreme court challenge by Gov. Kirk Fordice, who argued unsuccessfully that the attorney general was not authorized to file suit without the governor's approval. [21] Over the industry's objection, the Mississippi trial will not be a jury trial--yet another advantage procured by the state, one that raises serious constitutional questions. And in Texas, where $16 billion in damages is at stake, lawyers are sifting through 600 boxes of documents in preparation for a trial expected to begin in late September, [22] despite polls showing that Texans disapprove of the suit by more than a two-to-one ratio. [23]

The irony--or outrage, depending on your perspective--is that those trials were instigated by states to recover costs that they themselves incurred by voluntarily participating in the federal Medicaid program--costs allegedly attributable to a product that the states have refused to declare illegal. As Professor Richard Epstein put it, "The government can decide that it's not going to provide . . . this kind of insurance. . . . It might even . . . decide that it's going to ban tobacco. . . . But the one alternative that is absolutely unacceptable is . . . [to] retroactively impose a system of liability which nobody dreamed of." [24] Yet that is precisely what Florida, Mississippi, Texas--39 states in all--and assorted other governmental units have done; they have retroactively imposed a system of liability that nobody dreamed of. Here's how they did it.

Abrogation of Affirmative Defenses

In tort law, an affirmative defense is one that answers the plaintiff's allegations by stating, in effect, that even if the allegations are true, the defendant should prevail because the plaintiff has no legal claim. [25] To take an obvious example from criminal law, if a prosecutor contends that A murdered B, A might answer by asserting self-defense. Affirmative defenses include such claims as duress; statute of limitations; contributory negligence; and, most important to tobacco litigation, assumption of risk. In a nutshell, the assumption-of-risk doctrine, which until now has immunized the industry from liability, says, "If the user or consumer . . . is aware of the danger, and nevertheless proceeds unreasonably to make use of the product and is injured by it, he is barred from recovery." [26] Thus, under a traditional tort regime, if a consumer knows about the risks of smoking yet still smokes and thereby contracts a tobacco-related illness, he has no more claim against the tobacco manufacturer than would a nonsmoker who suffers the same illness caused by another agent. Florida may thereafter provide no assistance, some assistance, or complete assistance to the victim, but the amount of assistance has no effect on the responsibility of the tobacco manufacturer to the individual or to the state. [27]

That centuries-old rule of morality and law was revoked by the Florida Medicaid Third-Party Liability Act. Procedurally, the act lets the state circumvent the rules of subrogation and sue the industry directly. Substantively, because the act gives a direct claim to the state, which therefore no longer stands in the shoes of a Medicaid patient, the assumption-of-risk defense ordinarily available against the patient no longer makes sense. In justifying those changes, Florida officials attempted to square them with the traditional law by claiming that tobacco companies fraudulently misrepresented the risks of smoking, targeted underage customers with their ads, and concealed the addictive nature of cigarettes. Consequently, the story goes, smokers could not have made an informed judgment about the dangers inherent in smoking, so the assumption-of-risk defense is not applicable. In short, smokers cannot assume a risk they know nothing about. Let's see whether Florida's claims support that logic.

Fraudulent Misrepresentation

If fraudulent misrepresentation deprived consumers of the opportunity to make an informed judgment, one might expect that benefits under the act would extend to the parties supposedly deceived. Instead, the act applies not to consumers but only to the state in its role as plaintiff. Moreover, fraudulent misrepresentation, assuming it were proven, trumps any defense based on assumption of risk. A plaintiff who could demonstrate fraud would not have assumed the risk of smoking. Accordingly, abrogation of the defense would be entirely unnecessary.

Of equal importance, by filing its Medicaid recovery suit as a direct action under the new act, instead of standing in the shoes of each patient, Florida relinquished its right as a subrogee to the claims of Medicaid recipients. Florida's lawsuit is not for damages suffered by each injured smoker but for Medicaid costs paid out of the public purse. Consequently, a declaration by the state that tobacco companies defrauded smokers is irrelevant. Instead, because Florida is a direct claimant rather than a subrogee, the state now has to demonstrate that its officials, or its taxpayers, or perhaps the public at large--not just smokers--relied on the industry's misrepresentations as a material inducement, not to smoke, but to participate in the federal Medicaid program.

Not only has there been no such demonstration, none is likely to materialize. In the first instance, the state has portrayed itself as an inanimate entity, incapable of such sentient behavior as assumption of risk [28] or reliance. Second, any suggestion that Florida would have excluded coverage for smoking-related illnesses under Medicaid were it not for the industry's deception is wholly incompatible with public pronouncements to the contrary. Consider this statement by Richard Scruggs, a private attorney hired to represent the state: "[S]tates and governments must treat their indigent sick. They have no choice but to do that . . . regardless of whether they smoke, drink, or lead other risky lifestyles." [29]

Indeed, Florida distinguishes itself from a private insurance company by characterizing its Medicaid obligation as "imposed by law or statute" and funded by payroll taxes--unlike private insurers who receive premiums designed to produce a profit in return for which they voluntarily subject themselves to anticipated risks. For that reason, the state insists that it is entitled to the equitable remedy provided under the act--a direct cause of action sheltered from the assumption-of-risk defense and its corollary requirement that the state demonstrate fraud. [30] Yes, Florida agrees that subrogation may be appropriate for a private insurer, because the insurer has already received consideration for bearing the risk; but the state itself, unimpeded by legal niceties, should be allowed to pursue indemnification without subrogation in light of its obligation to pay damages caused by the acts of another. [31]

That is nothing but sophistry, of course. First, the state has no legal obligation to participate in the federal Medicaid program; the Florida legislature elects to do so, voluntarily. Second, the distinction between profit and nonprofit, between premium funded and tax funded, is completely extraneous to the assumption-of-risk issue. Medicaid is an insurer, whether or not it makes money. Both taxpayers and policyholders provide funding so that their "chosen" insurer can cover anticipated health costs. Both state and private insurers sometimes pay for treatment that arguably would not have been necessary but for the misconduct of a third party. Yet in determining whether such misconduct has occurred, why should the state alone be allowed to bypass the usual subrogation process and hold an outsider liable even when a victim assumes the risk of his own injury? Why should Florida's Medicaid system, but not a private insurance company, be able to impose its costs on a tobacco company that would have been exculpated but for the happenstance that a particular smoker qualified for public assistance?

Advertising

From those unanswered questions, we turn next to advertising and inquire whether smokers--especially underage smokers--were so misled that they could not have assumed the risk of smoking. Beginning as early as 1920 numerous epidemiological and experimental studies on the health hazards of smoking were reported by the media, and by 1962 more than 7,000 publications were examining the connection between smoking and health. [32] A 1954 Gallup poll indicated that 90 percent of Americans had heard or read that cigarettes can cause cancer. [33] For the past 30 years a conspicuous health warning has appeared on every package of cigarettes lawfully sold in the United States. Even children--allegedly brainwashed by crafty ads--were aware of the risks. As early as 1960 a nationwide poll by Senior Scholastic magazine found that only 2.6 percent of 10,000 high school students thought smoking had no connection with cancer. [34]

Given the vast amount of health-related information available from independent sources, advertising by the tobacco companies could not have undermined the assumption-of-risk defense. Studies suggest that such advertising influences the choice of brand more than the initial decision to smoke. [35] For example, focus groups conducted by the National Cancer Institute with African-American teenagers in five U.S. cities indicated that black youngsters are cynical, even resentful, about cigarette advertising. That is particularly striking because activists have long complained that the industry targets inner-city kids, despite evidence indicating that the percentage of such kids who smoke regularly has plummeted over the past decade. [36] Moreover, while marijuana ads are indisputably illegal, teenage use of that drug is accelerating rapidly. [37]

Any link between advertising and aggregate consumption is further eroded by a recent analysis of data for 1964 through 1990 on 22 of the 24 member countries of the Organization for Economic Cooperation and Development. By 1990, 6 of those 22 countries had banned all forms of tobacco advertising. Examining all of the countries and controlling for variables such as price, age profile, and long-term trends, the study concludes that the average effect on per capita tobacco sales of the ban in those six countries was a small increase--not quite statistically significant, but clearly contrary to the hypothesis that ad bans will reduce overall consumption. [38] The author suggests that the higher tobacco sales may have been due to the disappearance of health warnings that would otherwise have appeared in the ads. Because those warnings probably had a deterrent effect, their disappearance could have promoted aggregate consumption. [39]

A domestic version of that theory, holding that restrictions on U.S. tobacco advertising by the Federal Trade Commission perversely raised the level of cigarette smoking, goes like this: If advertising were unregulated, newer and smaller tobacco companies would vigorously seek to carve out a bigger market share by emphasizing health claims that might bolster brand preference. [40] But in 1950 the FTC foreclosed health claims--such as "less smoker's cough"--as well as tar and nicotine comparisons for existing brands. To get around that prohibition, aggressive companies then created new brands, which they promoted with an avalanche of health claims. Filter cigarettes grew from roughly 1 percent to 10 percent of domestic sales within four years. Over the same period, industry per capita sales declined for the first time since 1931, by 9 percent--perhaps dampened by the focus in cigarette ads on tobacco-related illnesses. The only company to post higher revenues was Brown & Williamson Tobacco Corporation, a small firm that concentrated on filter brands.

In 1954 the FTC tightened its restrictions by requiring scientific proof of health claims, even for new brands. The industry returned to advertising taste and pleasure, and aggregate sales rebounded almost immediately. By 1957 scientists had confirmed that low-tar cigarettes pose less danger. A new campaign of "Tar Derby" ads quickly emerged, and tar and nicotine levels collapsed 40 percent in two years. Determined to shut down the flow of health claims, the FTC next demanded that they be accompanied by epidemiological evidence, of which none existed. The commission then negotiated a "voluntary" ban on tar and nicotine comparisons.

Not surprisingly, the steep decline in tar and nicotine ended in 1959. Seven years later, apparently alerted to the bad news, the FTC reauthorized tar and nicotine data but continued to proscribe associated health claims. Then in 1970 Congress banned all radio and television ads. Overall consumption has declined slowly since that time. In today's climate the potential gains from health-related ads are undoubtedly greater than ever--for both aggressive companies and health-conscious consumers. Thanks in good part to ill-advised government regulation, however, those gains will not be realized. Instead of "healthy" competition for market share, we can probably look forward to more imagery and personal endorsements--the very format that anti-tobacco partisans decry.

To be sure, if it can be demonstrated that tobacco companies--however they are permitted to advertise--are aiding and abetting children who attempt to obtain cigarettes illegally, those companies should be prosecuted under current law. But an appearance by Tiger Woods on television wearing apparel emblazoned with Joe Camel--the predictable outcome of current FTC and congressional restrictions on health claims--is hardly proof of targeted advertising, much less aiding and abetting. In fact, to whatever extent it has targeted nonsmoking children, the tobacco industry seems to have been remarkably ineffective in persuading them to smoke, particularly if we examine the years surrounding the debut of Joe Camel in 1987. According to the U.S. Department of Health and Human Services, the proportion of youngsters between the ages of 12 and 17 that smoked declined from 29.4 percent in 1985, to 22.4 percent in 1990, to 20.2 percent in 1995. [41] The average age at which smokers first regularly use cigarettes trended slightly upward from 1962 through 1994. [42]

Addictiveness

If advertising has not misinformed smokers, thereby negating the assumption-of-risk defense, neither has alleged addiction rendered the defense inapplicable. Notwithstanding the assertion that many smokers are gripped by a dependence similar to drug addiction, millions of people quit smoking cold turkey; they experience only mild discomfort and, unlike drug addicts, do not require hospitalization for the effects of withdrawal. [43] Nor do smokers commit violent crimes to sustain their habit. "I'm not a doctor [but] forty million people have quit," concedes Don Beyer, the Democratic candidate for governor of Virginia, who nonetheless embraces the Clinton administration's plan to regulate tobacco as a drug. [44] Florida itself, in court papers in an unrelated 1995 case, acknowledges that "whether nicotine is addictive or not is a gray area. You have as many in the medical field that say it is as that say it isn't." [45] According to the state, breaking the habit may produce side effects, but they "last a week at most [and] can be effectively alleviated by over-the-counter remedies, such as cough syrup and analgesics, as well as by deep breathing and drinking more fluids." [46]

Author Richard Kluger, no fan of the tobacco industry, speaks bluntly about addiction:

While new evidence had emerged [since the Cipollone case] showing that Philip Morris and B&W, among others, had done research on the addictive nature of nicotine and had neither disclosed it to the public nor warned against the addicting potency, many similar findings by investigators outside the industry had long since been made and published. Public-health advocates, moreover, had for years advised that nicotine was as addicting as heroin and cocaine. . . . [W]hether one categorized smoking as a practice, a habit, an indulgence, a vice, a dependency, or an addiction, it was commonly known--and had been for decades--to be hard to stop once begun. Nor could anyone say for certain how much of a daily dose served to induce addiction; tolerances differed from person to person, and the industry had in fact made available brands with extremely low dosages. How, then, to justify a claim that the cigarette makers had massively imposed an intentionally addicting product on an innocent public that had little knowledge or choice in the matter? [47]

Kluger's point is that, given the mass of other data, concealment by or misinformation from the tobacco companies may not be sufficient either to prove fraud or to defeat an assumption-of-risk defense. In particular, to override the assumption-of-risk defense, the "addicted" smoker must show that he relied on the misinformation. Thus, even though the industry withheld or fabricated data, if the smoker had access to correct information from other sources, he may not have altered his behavior notwithstanding the misconduct of the tobacco companies.

Of course the central purpose of Florida's Medicaid Third-Party Liability Act is to guarantee that the state is not susceptible to an assumption-of-risk defense. That goal is accomplished by substituting a direct, so-called equitable cause of action for the traditional subrogation process. Before amending its act, Florida had relied on subrogation as its means of recouping Medicaid expenditures. By law, the availability and adequacy of a common law remedy like subrogation obviates the need for equitable relief. [48] But Florida has redefined "adequate." Simply put, an adequate remedy is now one that is guaranteed to win--even if that means inventing new law, applying it retroactively, and making a mockery of due process. The upshot is a bizarre action in equity--supposedly based on what is fair and just in a given situation, not bound by the traditional rules of common law. Thus liberated from the normal restraints, the state has replaced well-settled legal principles with a discriminatory and despotic statute tailored specifically to engorge the state's treasure chest.

Unclean Hands

Even without recourse to their usual assumption-of-risk defense, the tobacco companies thought they could rely on an analogous defense known as "unclean hands." That doctrine, which is applicable in equitable suits, simply says that the plaintiff's own fault is relevant to the remedy, if any, to which he is entitled. [49] In other words, if Florida engaged in activities that had the effect of exacerbating any harms attributable to tobacco, the state might not be able to recover for those harms. Like assumption of risk, unclean hands takes into account the extent to which alleged damages may have arisen or increased as a result of the plaintiff's behavior. Also like assumption of risk, the doctrine of unclean hands is rooted in the principle that holds parties accountable for the consequences of their own conduct.

Let's assess Florida's conduct to see if its hands are unclean. For starters, if the correctness of Florida's position was so apparent, why did it take 30 years after the surgeon general's initial warnings for the state to press its claims? Why didn't Florida opt out of the Medicaid program, or ask the federal government to exclude smoking-related illnesses from its list of covered treatments? Why didn't the state lower the nicotine content of cigarettes, or raise the tax rate, or ban their sale outright? On the other hand, if Florida wanted to preserve the pleasures of smoking for those of its citizens who paid their own medical bills, why didn't the state simply make it illegal for Medicaid beneficiaries to purchase cigarettes? Courts generally have let states decide when to treat groups differently on the basis of wealth, [50] and legislatures are reasonably free to require that recipients of government largesse act or refrain from acting in certain ways. [51] That solution would have curtailed Medicaid outlays, preserved the bulk of the cigarette tax base, protected the less affluent from their own misguided health habits, retained free choice for the large majority of smokers, and maintained some semblance of due process for the industry.

Despite their protestations of innocence, state officials could not have been oblivious to the barrage of criticism heaped upon the tobacco industry, dating from nearly 400 years ago. In his authoritative history of the industry, Richard Kluger reports that King James I, in 1604, castigated smoking as "a custom loathsome to the eye, hateful to the nose, harmful to the brain, dangerous to the lung, and the black stinking fume thereof, nearest resembling the horribly Stygian smoke of the pit that is bottomless." [52] By the early 1900s, goaded by the National Anti-Cigarette League, 14 states had prohibited the sale of cigarettes. [53] When Tennessee's ban was challenged on constitutional grounds, the state supreme court minced no words in upholding the statute, railing against cigarettes as

wholly noxious and deleterious to health. Their use is always harmful; never beneficial. They possess no virtue, but are inherently bad, and bad only. They find no true commendation for merit or usefulness in any sphere. On the contrary, they are widely condemned as pernicious altogether. Beyond question, their every tendency is toward the impairment of physical health and mental vigor. There is no proof in the record as to the character of cigarettes, yet their character is so well and so generally known to be that stated above, that the Courts are authorized to take judicial cognizance of the fact. No particular proof is required in regard to those facts which, by human observation and experience, have become well and generally known to be true. [54]

In the 1950s epidemiologists Richard Doll and A. Bradford Hill published a series of articles documenting the increased risk of lung cancer among smokers. They found that death from lung cancer was 10 times more likely for persons who smoked 10 to 20 cigarettes a day than for those who did not smoke at all. [55]

That Florida officials were mindful of the dangers of smoking is incontrovertible. Their awareness is no less damaging to the state's "equitable" case than a smoker's assumption of risk would have been under the common law. When the plaintiff is at fault, unclean hands can frustrate relief. As the state's hired gun, attorney Richard Scruggs recounted the law:

[A]ll of the states have, as the basis of their cases against the tobacco industry, these equitable theories. It doesn't mean that the tobacco industry is defenseless. They can show that the state has unclean hands, that the state has participated in the activity somehow, licensed it, or received tax revenues from it. [56]

Predictably, however, Scruggs's recitation of the law is one thing, but Florida's adherence to the rules is quite another. When push came to shove, here's what occurred.

Starting in 1972 the Florida prison system manufactured unfiltered cigarettes and distributed them at no cost to the inmates. To generate income, the state also sold some of its cigarettes to local governments. Those activities continued for roughly a decade despite widespread discussion in the legislature about the health hazards of smoking and the addictive power of nicotine. [57] Unclean hands? You bet. But when Philip Morris filed a pretrial motion to ensure that the evidence would be admissible, Florida successfully opposed the motion and persuaded Judge Harold J. Cohen that evidence of the state's shared responsibility for tobacco-related illness was just another "affirmative defense," forbidden by the Medicaid Third-Party Liability Act. [58] So much for Scruggs's contention that the doctrine of unclean hands was still alive and well.

Florida has pulled out all the stops to deny the tobacco companies any vestige of a defense. Governor Chiles, his attorney general, and the legislature, without much resistance from the judiciary, have exploited the laws they like, repealed the ones they don't, and concocted novel theories of liability whenever necessary to stack the deck. And no wonder. The state's hands are worse than unclean; they are downright filthy. Although Florida purportedly seeks recovery for all smoking-related disease, the state didn't sue companies that use tobacco leaf to fashion high-grade cigars; perhaps that's because Florida is home to the nation's leading producer of premium cigars. [59] More damaging still to the state's moral crusade, a recent report reveals that roughly $825 million of Florida's pension assets was invested in tobacco stocks. [60] Again, while the state contends that in 1992 alone "28,350 Floridians died from tobacco-related causes, representing 77 preventable deaths every day or 20 percent of all deaths in the State," [61] 10 of Florida's 23 congressional representatives guaranteed that federal subsidies for the industry would survive. Their 10 votes were more than enough to defeat the Durbin amendment, which would have prohibited the use of federal funds for certain tobacco programs. On June 12, 1996, the House of Representatives rejected the amendment by a two-vote margin. [62]

Perhaps most appalling of all, in an act of unparalleled hypocrisy, Florida actually embraces assumption of risk and personal responsibility in defending against a lawsuit brought by a prisoner seeking access to nicotine patches and other treatment for an alleged addiction to smoking. [63] In its October 1995 motion to dismiss the case, the state unabashedly asserted that "any future harm which Plaintiff may potentially suffer . . . is a direct result of Plaintiff himself choosing to buy and smoke cigarettes. Defendants are not responsible for Plaintiff's decision to purchase cigarettes any more than they would be responsible for Plaintiff buying a candy bar at the canteen." Continuing in the same vein: "Plaintiff is in no way entitled to medical intervention to 'cure' a habit which Plaintiff himself continues to indulge, and over which Plaintiff has ultimate control." Florida, it seems, has absolutely no compunction about summoning the principle of personal responsibility whenever its own money is at stake.

That two-faced application of the law should convince anyone that the state has no moral or equitable claim against the tobacco industry. Not only did Florida reject obvious remedies that would have reduced tobacco consumption and abated the associated Medicaid outlays, but it also nurtured cigarette manufacturers with its investment capital and its approval of federal giveaway programs. Indeed, the state was itself a cigarette manufacturer for almost a decade. For their part, few if any Florida Medicaid recipients could have eluded the endless warnings and incessant sermonizing about the health risks of smoking. So whether the label is "assumption of risk" or "unclean hands," smokers who receive and states that dispense Medicaid assistance are the parties principally accountable for the public health costs associated with the use of tobacco.

Even the federal government, which under President Clinton has pursued an aggressive anti-smoking campaign, acknowledges the role of personal responsibility when the question is whether the government itself should be held liable for its role in encouraging smoking. Thus, the Clinton administration now seeks congressional support for legislation that would overturn a 1993 opinion by senior lawyers in the Department of Veterans Affairs. That opinion suggests that the federal government may be liable for tobacco-linked illnesses contracted by millions of military personnel who smoked while on active duty. Sounding much like a tobacco executive, VA secretary Jesse Brown declared that it would be a "borderline absurdity" for the government to pay "for death or disability resulting from veterans' personal choice to engage in conduct damaging to their health." [64]

The military plainly encouraged its employees to smoke. Not only were cigarettes sold on bases at a huge discount, but soldiers in combat zones were given free cigarettes as part of their rations. That is surely more of an inducement than tobacco companies ever offered their customers. And no one has ever accused the tobacco industry of preying on youngsters under enormous stress by plying them with free cigarettes. "We know there's a dichotomy," rationalized a Pentagon spokeswoman, "but you have to balance the need for a healthy fighting force and an individual's right to use tobacco." [65] Rather than blame the government, Secretary Brown, who stopped smoking 35 years ago, volunteered this controlling principle: "If you choose to smoke, you are responsible for the consequences of your act." [66] Clearly, if that principle is sufficient to render the government immune from liability, it is sufficient to render private companies immune as well.

Statistical Proof of Causation

In mass tort cases, statistical evidence combined with other corroborating evidence can "supply a useful link in the process of proof." [67] But statistics alone are not enough. "An essential element of the plaintiff's cause of action for negligence, or for that matter for any other tort, is that there be some reasonable connection between the act or omission of the defendant and the damage which the plaintiff has suffered." [68] Even in a class action, assuming that the class is certified and the named plaintiffs prevail on the issue of liability, the remaining members of the class generally cannot recover in the damages phase of the trial unless they demonstrate individualized harm caused by the defendant. [69]

Those classical precepts of tort law are steadfastly applied by courts across the nation; but Florida, in order to minimize any uncertainty surrounding its prospect for success, substituted its own rendition of the law of causation.

In any action brought under this subsection, the evidence code shall be liberally construed regarding the issues of causation and of aggregate damages. The issue of causation and damages in any such action may be proven by use of statistical analysis. . . . [If] the number of recipients for which medical assistance has been provided by Medicaid is so large as to cause it to be impractical to join or identify each claim, the agency shall not be required to so identify the individual recipients for which payment has been made, but rather can proceed to seek recovery based upon payments made on behalf of an entire class of recipients. [70]

In other words, under the Medicaid Third-Party Liability Act, the state could sue the tobacco industry for health expenditures attributable to smoking by Medicaid recipients without disclosing (1) the names of recipients, (2) such recipients who actually smoked, or (3) such smokers who had illnesses allegedly caused by tobacco. All that had to be disclosed were aggregate statistics for Florida Medicaid recipients as a group.

In response to a due process challenge, the Florida Supreme Court upheld the part of the act that authorized statistical proof but excised the provision permitting nondisclosure of Medicaid recipients. The court could "find no way in which this subsection would allow a defendant to challenge improper payments made to individual recipients [or prove] that its product was never used by the recipient. . . . It is illogical and unreasonable to call this a fair process." [71] Trial court judge Cohen in West Palm Beach then directed the state to disclose the names of all Medicaid recipients who were treated for tobacco-related diseases. [72]

Speculation was that the state's litigation might die under the ponderous weight of 40,000 to 60,000 individual investigations. But Florida general counsel Dexter Douglas thought otherwise; he predicted that the court would limit the tobacco companies to an investigation of just enough randomly selected cases to authenticate the statistical evidence. [73] That prediction proved to be close to the mark. In February of this year, Judge Cohen gave the industry access to the names of 400,000 Medicaid patients--including burn victims (suggesting the industry may be liable if a smoker falls asleep with a lit cigarette) and patients with other injuries rarely associated with tobacco use, such as mental illness. From those 400,000 names, lawyers for the tobacco companies may select those of 25 persons. The state will then make available the medical records of those persons and the industry can investigate those cases, and no others, to confirm that the aggregate data proffered by the state are reliable. [74]

Reflecting his cynicism, or perhaps his bias, Judge Cohen cautioned that his order allows the industry "to pick 25 names. It's not to go out and hound 400,000 recipients." [75] Naturally, the court is concerned about privacy, but the 400,000 patients relinquished their claim to complete privacy when they requested Medicaid assistance. Certainly, if an insurance company were filing subrogation actions, each claimant would be required to disclose his medical record. Still unclear is how the tobacco companies, without prior access to the medical records, are to choose the 25 cases to be investigated. Moreover, since medical records do not necessarily contain smoking histories, even if the tobacco industry had them in advance, without further discovery, they would be of little or no use. Aggregate statistics on smoking-related illnesses cannot possibly be confirmed if the 25-person sample consists mostly of nonsmokers.

By wresting such absurd restrictions from the trial judge, the state has again flouted the rule of law despite unequivocal instruction from its highest court: tobacco companies must be able to investigate the legitimacy of Medicaid claims [76] and explore such questions as whether and how much each claimant smoked. Ironically, the Florida Supreme Court apparently had more influence with Federal District Judge D. Lowell Jensen in San Francisco than with its own lower court in West Palm Beach. In his 32-page opinion this past February dismissing most counts of San Francisco's Medicaid recovery complaint, Judge Jensen explicitly repudiated the Florida lower court opinion.

[I]n a direct suit by a smoker to recover his or her smoking-related medical expenses, the Court could inquire into any other health problems which may have exacerbated the costs of health care for that smoker. Likewise, the Court could ascertain from an individual smoker the amount of information he had regarding the risks associated with smoking. In the present suit, on the other hand, . . . it will be difficult, if not impossible, to explore these and other relevant issues. [77]

Judge Jensen concluded, "[I]n order to recover monies spent on health care for individual smokers, plaintiffs will be required to prove that each of those smokers' injuries were actually caused by smoking." [78] Less than two weeks earlier, a West Virginia trial judge had dismissed most of that state's case against the tobacco industry on similar grounds. [79] And in May of this year, Maryland Judge Roger W. Brown ruled that the state had no right "to assert claims in its own name . . . for the harms allegedly caused to third party smokers, unless such claims are made in the name of each of the individually injured third party Medicaid program recipients under the equitable doctrine of subrogation." [80] A trial judge in Washington State issued a similar decision this June. [81] So in at least four jurisdictions only those specific illnesses shown to have been caused by tobacco will be compensable.

That approach seems eminently reasonable. Indeed, the Florida Medicaid Third-Party Liability Act artfully mandates that the state prove causation. But the mandate is empty of meaning. Unlike any other tort action by any other plaintiff, Florida's showing of causation need not be injury specific. Instead, the state may show generalized causation based on no more than aggregate epidemiological data--notwithstanding authoritative and nearly universal acknowledgment that "[e]pidemiology cannot prove causation; causation is a judgment issue." [82] More precisely, causation can only be legally "established" by a coherent underlying theory--supported, but not proven, by empirical data. Because this point is pivotal, let's look at the drawbacks when mass tort plaintiffs rely exclusively on statistical evidence.

Epidemiology is concerned with the incidence, distribution, and etiology of diseases. In a legal context, the objective is to determine whether exposure to an agent caused a harmful effect. The focus is on general causation (whether the agent is capable of causing the disease) rather than specific causation (whether the agent caused the disease in an individual). Epidemiological studies may be able to pinpoint the agents that are associated with a disease, quantify the increased risk, and profile the persons prone to contract the disease. However, like those of all statistical analyses, the results are sensitive to the accuracy of the data and the validity of the research methods.

Most important, epidemiology cannot demonstrate that a statistical association between agent A and disease B signifies that A caused B. It is conceivable, perhaps likely, that the apparent association between A and B arises from the fact that both are highly correlated with one or more other factors, which statisticians call confounding variables. For example, there seems to be a close association between math scores and shoe size. Yet nobody would suggest that big feet enhance mathematical ability, or that math skills cause one's feet to grow. The obvious confounding variable is age. As people grow older, they learn more about math and they wear larger shoes. Similarly, in assessing the correlation between smoking and cancer, it is essential to control for a long list of factors (e.g., weight, age, diet, other lifestyle choices, family history, intensity and duration of smoking, and exposure to other causal agents).

Ideally, there should be experimental controls for confounding variables. That is, persons to be tested should be identical in all relevant respects except that one randomly selected group (cohort) is instructed to smoke a specified amount and a second such cohort is instructed not to smoke. Then the medical histories of the individuals in the two groups are compiled over time and compared. That process may well be impermissible on moral grounds, and it's impracticable in any event. Latency periods are lengthy and varied, matched individuals are difficult (or impossible) to identify, and smokers are self-selected not randomly selected. So statisticians must rely on observational data, as contrasted with the experimental data that are generated when investigators are able to choose the individuals who will be exposed to an agent and those who will not. As a result, the only effective controls for confounding variables are those that can be implemented statistically, and therein lie many problems.

Not least of the problems is that multiple explanatory variables, if they are correlated with one another, will mask the separate effect of the variable at issue in the litigation. Suppose we were to hypothesize that a particular disease might be due to smoking, consumption of fatty foods, race, and income. If the four explanatory variables were significantly intercorrelated and each had an effect on the incidence of the disease, then we could not reliably determine the impact of smoking alone. Even more basic, in all of the Medicaid cases, every smoker for whom recovery is now sought was a Medicaid recipient. The observational data were not limited to that subset of the population, however. Accordingly, if Medicaid recipients have different characteristics than nonrecipients--due to their financial circumstances, for example--and if the differences are associated with tobacco-related diseases, the data could be fundamentally flawed. Moreover, observational studies often suffer from "recall bias"--the proven tendency of individuals who have contracted a disease to recall more readily their past exposure to an allegedly causal agent.

Here is another, maybe simpler, way of looking at the considerable dilemma of confounding variables. Causation is most direct and straightforward when the agent is both a necessary and a sufficient antecedent of the disease (nonsmokers do not get cancer, all smokers get cancer). The situation is somewhat more complicated when the agent is necessary but not sufficient (nonsmokers do not get cancer, not all smokers get cancer) or when the agent is sufficient but not necessary (all smokers get cancer, nonsmokers can also get cancer). The most complex relationship, and the one most difficult to sustain with statistics, occurs when the agent is neither necessary nor sufficient (nonsmokers can get cancer, not all smokers get cancer). It is that fourth category that we must confront in the Medicaid recovery suits.

Furthermore, statistical tests are designed to address a question that is quite different from the question the court faces. The court needs to determine, in light of the observed disparity in disease rates between a sample of smokers and a sample of nonsmokers, what the probability is that smoking has no real effect. But a test of statistical significance reasons in the opposite direction: assuming that smoking has no real effect, what is the probability that a disparity as large as the one observed between smokers and nonsmokers could have arisen just because we studied a sample and not the entire population? While those two questions are interrelated, they are not equivalent. By analogy, assuming a fair coin, statistics tells us that the chance of two flips in a row coming up heads is one in four. That is not the same as saying that, given two heads in a row, the odds are only one in four that the coin is fair.

Put somewhat differently: When the difference in disease rates between smokers and nonsmokers is deemed to be significant, a statistician will reject the hypothesis that the difference was due to sampling error. He has not proven, however, that the real underlying disease rates are higher for smokers. Instead, he has simply ruled out one possible explanation (sampling error); alternative explanations for the observed difference--such as confounding variables, nonrandom selection, recall bias, and inaccurate data--are still possible. Those complexities make it essential that individualized confirmation of causation supplement any broad-based empirical data. In short, the issues are too complicated and the statistical techniques too untrustworthy to exclude case-by-case corroborative or exculpatory evidence.

That general-purpose warning takes on still greater importance in the context of the Medicaid reimbursement suits. For those suits, it is not sufficient for the state to demonstrate that tobacco is more probably than not a cause of various diseases. Indeed, the "more probable than not" legal standard ineluctably leads to the wrong damages. To illustrate: Suppose statistics "prove" that the disease rate among nonsmokers is 10 cases per 10,000, the disease rate among smokers is 16 per 10,000, and the entire increase of 6 cases is attributable to smoking. Thus, 6 of every 16 cases among smokers, or 37.5 percent, are "caused" by tobacco. Epidemiologists call that measure the "attributable portion of risk" or APR. No single smoker could show statistically that his disease was more probably than not due to tobacco. Consequently, the industry that caused 37.5 percent of the cases would escape without liability if the "more probable than not" standard were rigorously applied. On the other hand, if the APR were, say, 60 percent (i.e., if tobacco caused an increase from 10 cases among nonsmokers to 25 among smokers) then every smoker could show statistically that his disease was more probably than not due to tobacco. The industry would be liable for 100 percent of the damages even though it caused only 60 percent of the cases.

Because neither outcome makes much sense, the damage awards to the states, if any, would likely be based on the APR percentage itself. If the APR is 37.5 percent and the Medicaid expenditures for all tobacco-related diseases, including those not attributable to smoking, are $1 billion, then the industry will be liable for $375 million. While that procedure does seem more equitable, it raises an unsettling problem. When the APR is used directly to determine the assignable damages, it is imperative that the percentage be correct. That is, the actual magnitude of the APR replaces a dichotomous "more probable than not" standard, under which a defendant is either fully liable (if the APR is any number above 50 percent) or not liable at all (if the APR is any number below 50 percent). With that substitution, if the value of the APR is imprecise--even if the error is only a few percentage points--the aggregate impact on damages could be in the billions of dollars in 39 states.

There is one characteristic of epidemiological evidence that we know with absolute certainty--it is inescapably imprecise. It measures association rather than causation; it must deal with latency periods, intercorrelated confounding variables, recall bias, selection bias, sampling error, and bad data. Yet when the state is the plaintiff and loathsome tobacco companies are the defendants, Florida courts will be relying exclusively on statistics, without a shred of corroborating evidence that cigarettes caused a particular disease in a particular case. Apparently, the rule of law in Florida is sufficiently malleable to serve whatever objective the state can portray as beneficent.

Somehow the soundness of that argument eluded at least one Florida columnist.

[N]ot one of these states knows how many of its Medicaid patients even smoke or ever have smoked. Not one is prepared to produce a sick Medicaid patient and prove, with scientific evidence, that the patient's illness is a direct result of cigarette smoking. . . . The states have no such evidence. They are like a state suing General Motors on the grounds that automobiles are involved in accidents and that victims in accidents have run up the cost of Medicaid. [83]

To be blunt, it is impossible to demonstrate a causal link between an individual smoker and a particular tobacco company using epidemiological data alone. To circumvent that problem, Florida elected to base damages on general evidence of causation rather than specific injury. Because the state rather than an injured party is the sole plaintiff, the court would not have to determine the amount of damages to be distributed to each Medicaid recipient. But it still faces the redoubtable task of allocating the damages among the various tobacco companies. To simplify matters, the Medicaid Third-Party Liability Act provides that the state can "proceed under a market share theory, provided that the products involved are substantially interchangeable among brands, and that substantially similar factual or legal issues would be involved in seeking recovery against each liable third party individually." [84]

Under the market share doctrine, each tobacco company defendant would be liable for a portion of the aggregate damages--determined by dividing the company's Florida tobacco revenues by the corresponding revenues for all defendants combined. The market share theory originated with the diethylstilbestrol (DES) cases, in which a drug taken by pregnant women to prevent miscarriages allegedly caused medical problems in their children decades later. Because DES was manufactured by a number of companies but sold generically, neither the mother nor the child was able to name the manufacturer at the time the problems finally arose. Some courts allowed the children to recover merely by proving that DES caused their injury, without identifying the specific manufacturer.

Only six states have accepted the market share doctrine; many others have concluded that they could not justify holding companies liable for damages they may not have caused. [85] Even the few states that have applied the doctrine to DES litigation have not been willing to expand its coverage to nongeneric products, like cigarettes, or to products consumed closer to the time of injury. [86] By common law, Florida limited market share liability to those cases in which the plaintiff was "inherently unable" to determine which company manufactured the injurious product. [87] Moreover, the plaintiff had to meet a "due diligence" requirement; that is, he had to demonstrate that he had "made a genuine attempt to locate" the culpable party. [88]

The obligation to locate the culpable party no longer exists under the Florida Medicaid recovery statute, nor do the other requirements for using market share liability that had applied under Florida's now-defunct common law. The plaintiff state need not show that the Medicaid patient smoked, that smoking caused his illness, or even that a particular tobacco company produced the type of cigarettes that he consumed. Nor can the company defend itself by proving that it did not sell a specific type of cigarette in the relevant market during the relevant time frame. In fact, no one seems to know just what the relevant time frame is. Unlike DES, tobacco has been on the market continuously since well before the birth of Medicaid; some patients stopped smoking years before they contracted cancer; others persist in smoking despite their illness. Yet those obstacles to meeting the plaintiff's burden of proof have been expunged--to be replaced by a statistical tabulation showing only the total tobacco-related sales by each company in the state of Florida over who knows what period.

Forget about demonstrating that the supposed causal agent is a generic product; the state must only represent that cigarette brands are substantially interchangeable and that each patient's injury involves similar factual and legal issues. Are there no interbrand distinctions by price, size of cigarette, filtering mechanism, nicotine and tar content? Are there no individualized factual questions to resolve such as the number of cigarettes smoked per day, the overall time span, the state in which the smoking occurred, whether the patient was ill before coming to Florida, or the size and type of cigarette smoked? Isn't it plausible that cigarette consumption by Medicaid recipients differs in many respects (e.g., choice of brand) from consumption by the aggregate population on which the market share statistics are based?

Notwithstanding those obvious and formidable pitfalls, Florida (the only state to have done so) authorized by statute the imposition of market share liability. [89] In effect, the legislature--apparently seeking a painless substitute for fiscal discipline--simply rigged the outcome of litigation and thereby effectuated a quasi-tax on out-of-state companies. Whatever the motivation, by permitting state courts to allocate damages on the basis of no more than market share statistics, the Florida legislature blithely assigns liability without causation and blame without fault. Then, flagrantly misusing those statistics, as if cigarette brands were indistinguishable and injuries were factually equivalent, the state drives a final wedge between its Medicaid recovery suits and any rational view of corrective justice.

Damages

For those politicians eager to forge new law in the service of ends that they find congenial, legislation to recover public expenditures on behalf of poor people suffering from smoking-related illnesses may at first blush seem irresistible. But before too hastily enacting a punitive statute like the Florida Medicaid Third-Party Liability Act, legislators should have confirmed that there is indeed a pot of gold at the end of the rainbow.

Two questions arise: First, if a state's coffers have already been replenished by levying excise taxes on tobacco products in order to reclaim smoking-related public health expenditures, are the courts likely to permit double recovery? Second, even if there are recoverable damages not offset by excise tax receipts, will those damages be sufficient to compensate private contingency lawyers, pay litigation expenses, and reimburse the federal government for its share of Medicaid outlays?

Taking the second, less complicated question first--assuming no offset for excise tax receipts--those states that prevail in a Medicaid recovery suit will likely have few dollars to show for their efforts after funneling most of the damages back to Washington, then paying attorneys' fees and other legal costs. The federal Medicaid statute requires that any expenditures recovered by the state be remitted to the federal government in proportion to the federal participation in the state's Medicaid program. In Florida, the feds pick up about 56 percent of Medicaid costs; so they must receive 56 percent of any damage awards. Federal participation rates range from 50 percent to 79 percent; the lower the state's median income, the higher the federal outlay. [90] Mississippi receives 79 percent of its Medicaid funding from Washington.

To make matters worse, the federal government pays only half of any litigation costs, regardless of the federal participation rate. The state must pay the other half. Accordingly, in a state like Florida, where private lawyers are to receive a 25 percent contingency fee plus reimbursement of their expenses (say, another 5 percent), the state will retain only 29 cents out of each dollar of damages. The 71-cent balance goes to the federal government (56 cents) and to cover half of the legal fees and expenses (15 cents). In Mississippi, the state will retain only 6 cents of each dollar. Moreover, the state must still pay its own out-of-pocket expenses associated with discovery and trial.

Justifiably concerned that the federal rake-off would diminish states' incentive to litigate, Sen. Frank Lautenberg (D-N.J.) introduced a bill in June 1996 that would have allowed states to keep more of their winnings. The Lautenberg proposal would have reduced the federal remittance by 33 percent if Medicaid damages resulted from a lawsuit against tobacco companies. [91] More important than the bill itself, which did not pass, is what it says about the lengths to which some politicians were willing to go in their anti-tobacco obsession. With state legislators redefining the law and federal legislators trying to sweeten the pot, the signs of a moral crusade are pervasive.

Ironically, however, unless they are able to extort a financial settlement from the tobacco companies, the states so fervent in their rush to punish the industry may find that the huge payoff they're awaiting is illusory. For even if tobacco companies are held liable for all smoking-related public health costs--including publicly funded medical care, group life insurance, sick leave, nursing home care, pensions, and lost tax receipts that pay for retirement programs--courts are not likely to condone damage awards if cigarette excise taxes already generate net revenues to the states in excess of Medicaid costs. We come then to the first question posed above. As we shall see, recent authoritative studies confirm that the relevant excise taxes equal or exceed the relevant health expenditures.

In its 1985 report, the Office of Technology Assessment of the U.S. Congress estimated that the personal and social cost of smoking was $2.17 per pack of cigarettes sold. [92] However, OTA made no attempt to distinguish between "internal" costs borne directly by a smoker (e.g., the wages lost due to his tobacco-related illness or premature death) and "external" costs imposed on other persons by a smoker (e.g., public health expenditures for tobacco-related illnesses). Only the latter are relevant for purposes of the Medicaid recovery suits. When appropriate adjustments are made to isolate the external component of the cost, the revised estimates range from 15 cents per pack using 1986 data to 33 cents per pack using 1993 data. [93] Both estimates are far below the excise taxes collected on a pack of cigarettes.

Smokers are able to impose costs on nonsmokers because the government has decided, first, to insure the health costs of low-income and elderly persons and, second, to fund the insurance in a manner that does not distinguish between presumably high-risk smokers and lower-risk nonsmokers. If each smoker paid for his own illness, all costs would be internalized. If smokers insured themselves against tobacco-related illness and premiums reflected the health risk implicit in their smoking habits, smokers as a group would not impose external costs on nonsmokers as a group, although those individuals who contracted diseases would still impose external costs on those who did not. Apart from the quite separate dispute about injury due to secondhand smoke, externalities between smokers and nonsmokers arise only when the government both compels health insurance and prevents the tailoring of premiums so that they match the risk profile of the insured. Setting aside the possibility of fraud, the conduct of the tobacco companies is simply irrelevant to the problem of externalities.

States are suing the tobacco industry to recover external costs that (a) are imposed by smokers on nonsmokers, (b) could not have been imposed without the complicity of the state itself, (c) are unaffected by the conduct of the industry, and (d) are more than offset by the excise taxes flowing into state treasuries. In that context, we can review the three key studies on the social costs of smoking.

The first comprehensive analysis of the external cost of smoking was published in the Journal of the American Medical Association in 1989 by a team of researchers from the RAND Corporation. [94] The RAND study established the framework for subsequent research and, for our purposes, set forth three essential principles: First, if a smoker does not die from a smoking-related illness, he will die from something else. Accordingly, the relevant social cost is, not the entire amount spent on his illness, but the difference between the amount spent and the amount that would have been spent if he had not smoked. Second, premature death from smoking can produce long-term external benefits in the form of lower retirement costs [95] and reduced nursing home care. Those benefits are an offset to the near-term outlays for medical care, sick leave, and group life insurance. Third, because the near-term expenses of a smoking-related illness will necessarily be incurred before any longer term savings in retirement and nursing home costs are realized, it is important to compare apples with apples using present values adjusted for the timing of any receipts and disbursements.

Perhaps an oversimplified illustration might be helpful. Consider the social costs attributable to one smoker, an unmarried Medicaid recipient, who contracts lung cancer and dies today at age 65. His publicly funded medical costs are $20,000, paid at the date of death. If our hypothetical victim had not smoked, he would have retired at age 65 and lived to age 70; his publicly funded medical costs would have been $8,000, paid at the later date of death; and his Social Security benefits would have been $2,500 annually beginning on his 66th birthday and terminating at death. By smoking, the deceased Medicaid recipient imposed certain costs on the public, but he also enabled the public to avoid other outlays. The year-by-year comparison of those hypothetical costs and benefits, from the taxpayers' perspective, is given in Table 1.

Table 1
Costs and Benefits to Taxpayers

Year Item Costs &
(Benefits)		 Present Value
at 3%		 Present Value
at 5%
0 Medical costs $20,000 $20,000 $20,000
1 Social Security (2,500) (2,427) (2,381)
2 Social Security (2,500) (2,356) (2,268)
3 Social Security (2,500) (2,288) (2,160)
4 Social Security (2,500) (2,221) (2,057)
5 Social Security (2,500) (2,157) (1,959)
5 Medical costs (8,000) (6,901) (6,268)
Total   ($500) $1,650 $2,907

 

Although our smoker died prematurely and the Medicaid fund expended $20,000 as a result of his lung cancer, the government avoided $20,500 in other expenses--$8,000 in medical costs that would have been incurred if he had survived another five years and $12,500 in Social Security benefits that would have been paid over that time. On balance, purely in dollars and cents--that's what these lawsuits are all about--the public is ahead by $500 because the victim died at age 65 rather than age 70. However, the net savings disappear if we adjust for the timing of the inflows and outflows. The public's costs ($20,000) are paid when the smoker dies. The benefits are not fully realized for five years thereafter. Because a dollar today is worth more than a dollar in the future, all of the costs and benefits should be expressed in present-value terms. Using a 3 percent discount rate--roughly equal to the long-term real return (net of inflation) on risk-free government securities--the $500 savings is converted into a $1,650 cost. And if we raise the discount rate to 5 percent, the public cost escalates to $2,907. The higher the discount rate, the less weight is given to the benefits that arise in later years.

To be conservative, the researchers at the RAND Corporation discounted for present value at a 5 percent real rate. They concluded that the external cost of a pack of cigarettes in 1986 dollars was 15 cents. In a later study for the Congressional Research Service, that estimate was updated to 33 cents, expressed in 1993 dollars. [96] Yet smokers were then paying an average of 53 cents per pack in taxes--60 percent more than the costs they were imposing. In a separate study, also using 1993 dollars, Duke University economist W. Kip Viscusi reworked the RAND data and found that the medical care, sick leave, and group life components totaled approximately 51 cents per pack [97]--still lower than the excise tax rate, even without offset for retirement and nursing home savings. With all expenditures and savings factored in, the total external cost per pack, according to Viscusi, was 25.3 cents--less than half of the prevailing 53-cent tax rate. [98]

Those studies conclude, then, that total taxes from tobacco far exceed public health costs of tobacco-related illnesses. Thus, the public sector--state and federal together--has no claim for damages. In order for any given state to have a legitimate claim, it would have to receive a much smaller portion of federal-state aggregate excise taxes than it pays as a percentage of health costs. Yet the typical state collects 28.6 cents per pack whereas the federal government collects 24 cents. [99] So the state's share of revenues, on average, is over 54 percent. By comparison, states' contributions to the costs of the Medicaid program range from 50 percent in the wealthier states to only 21 percent in the poorer states. [100] On balance, states clearly are receiving more than they are paying.

Another argument floated by the states is that excise tax revenues cannot be counted as an offset to Medicaid outlays because the revenues may not be earmarked for the Medicaid program; that is, states can and do use the revenues for other purposes. The argument is wholly fatuous. First, we are concerned here, not with sales taxes that are assessed against virtually all products, but with added excise taxes that are directly traceable to, and would not exist but for, the purchase of cigarettes. Accordingly, those incremental revenues must necessarily enter into any cost-benefit calculus involving tobacco. Second, dollars are fungible; any appropriation of cigarette excise taxes to finance general expenditures merely releases those funds that would otherwise have been used for that purpose. The released funds are available to the Medicaid program. Third, to the extent that tobacco taxes are unrestricted as to use, they are even more valuable to the state than restricted funds. Fourth, it is the state itself that determines whether tax receipts are earmarked and which dollars are ultimately spent for what purpose. It would be utterly absurd for a state to be able to avoid an offset for cigarette tax receipts by the transparent expedient of redirecting those revenues to pay for, say, highway construction.

Naturally, the tobacco companies will want to know all about federal and state excise taxes collected from smokers, as well as any pension, retirement, and nursing home savings due to premature deaths from tobacco-related illnesses. Equally certain, the states will ask the courts not to allow either the discovery of such information or its proffer at trial. In making that argument, the states will probably invoke some version of the "collateral source rule," a common law tort doctrine by which a plaintiff's receipt of benefits from a source wholly independent of the defendant cannot be introduced as evidence in support of reduced damages. For example, if the victim of a car accident incurs medical bills that are covered by his medical insurance policy, the negligent driver who hit him cannot ask the jury to exclude those costs from the damage award on the ground that they have already been paid by a third party. The insurance policy is considered a separate contractual arrangement between the policyholder, who pays premiums, and the insurer, who bears the financial risk of injury.

Undoubtedly, the states will insist that excise taxes are paid not by tobacco companies but by consumers--a wholly independent source. But a tax on sales of a product that is the subject of litigation and sold by companies that are the defendants, collected by states that are the plaintiffs, paid by consumers who were subjected to harm, and imposed specifically because of that harm could hardly be regarded as independent. Moreover, economists would certainly reject the notion that consumers will be stuck with the entire tax bill. Whether a firm can pass along higher taxes to its customers depends on the price elasticity of demand for the firm's product--a measure of the change in sales volume when price is raised or lowered. Ordinarily, consumers will pay the full amount of a tax increase only if demand is infinitely inelastic, but economists know that tobacco sales do decline when the price is hiked. [101]

There is another objection, of course, to the invocation by the states of the collateral source rule. The rule is, after all, one of many common law tort principles that the states have rejected in litigating their Medicaid recovery suits. Now the states would like to have it both ways. They abandoned their traditional claims under tort law in favor of a novel direct cause of action. They repudiated assumption of risk and specific causation, each of which is a pillar of the conventional law. Yet, whenever it seems advantageous to rely selectively on time-tested doctrines like the collateral source rule, the states are not at all bashful. That manipulative, selective approach to the rules cannot be tolerated. Even if the excise taxes received by the states were an independent and unrelated benefit--which they are not--the collateral source rule is an integrated component of tort law; its application in these lawsuits is unprincipled and offensive.

Consequently, when it comes to reimbursing the public treasury for costs associated with tobacco-related diseases, the essential premise of these lawsuits is wrongheaded. To be sure, smokers impose burdens on publicly funded health care resources, but claims for reimbursement cannot focus exclusively on the expenditure side of the ledger. The states must also take into account, first, the excise tax receipts that already compensate for smoking-related public health outlays and, second, the costs that the public would have incurred if the smoker had not smoked. Mindful of those offsetting savings, the RAND Corporation concluded, "Although nonsmokers subsidize smokers' medical care and group life insurance, smokers subsidize nonsmokers' pensions and nursing home payments. On balance, smokers probably pay their way at the current [1986] level of excise taxes on cigarettes." [102] The RAND findings, since reinforced by both the Congressional Research Service and Professor Viscusi, are even more compelling in light of today's higher excise tax rates and mushrooming costs for geriatric care.

In short, even if the attorneys general somewhere, someday, somehow persuade a jury that the tobacco industry bears the blame for smoking-related illnesses, they will have won the battle and lost the war. Quite simply, the states have suffered no monetary damage.

The Choice between Public and Private Law

The state of Florida and 38 other states, faced with ballooning Medicaid costs, have decided that litigation against the tobacco industry is a comparatively painless path to fiscal health. Couching their claims in the lofty language of corrective justice, they have sought to recover Medicaid expenditures on behalf of taxpayers who bore the financial burden. The states could have raised taxes on cigarettes, of course. That option would have been simpler, less expensive, and quicker; and the parties who would have paid the tax (smokers and tobacco companies) are the same parties that ultimately would pick up the bill for court-awarded damages. Why then are lawsuits being substituted for taxation?

In an ideal world, this is how politicians would respond to that question: Efficiency in ensuring that a state is fiscally sound cannot be the paramount objective. Even if it is more efficient to raise revenue by legislation than by litigation, when particular parties are subject to sanctions for wrongdoing, the courtroom and not the legislature is where they have an opportunity to defend themselves. The court is the proper forum for evidence, for fact-finding, and for a verdict following a trial before a jury of peers. As a just society, we cannot suffer our legislature to embark on a punitive crusade against a disfavored industry without guaranteeing due process of law.

Lamentably, in the real world, the reason that legislatures have opted for Medicaid recovery suits instead of tax hikes has everything to do with politics and nothing to do with due process. Both taxes and tobacco companies are immensely unpopular; so lawmakers have taken the politically safe course of action by avoiding the one and attacking the other. To achieve that two-part objective, the states came up with an indirect tax, camouflaged as court-ordered damages. It was far easier to indict a "greedy" and baneful industry than to tack still another user's fee onto a product already bloated with government exactions. But there was one large fly in the ointment. For more than four decades juries had consistently endorsed the twin ideals of individual liberty and personal responsibility: smokers are free to purchase and use a product they know to be harmful, but, if they do so, they and not the tobacco companies are accountable for the consequences.

Those twin ideals had to be set aside. So, to lubricate the wheels of justice, the attorneys general concocted a new cause of action; abandoned the rules of subrogation; abolished the assumption-of-risk defense, which had blocked damages for 40 years; and eliminated any requirement that plaintiffs prove they were harmed by cigarettes.

Actually, it was a handful of sharp private attorneys--later to be hired at contingency fees ranging from 10 to 30 percent of the recovered damages--who were responsible for the novel legal theorizing that became the Florida Medicaid Third-Party Liability Act. [103] Those members of the plaintiffs' bar are now hopelessly conflicted, serving as government contractors with financial incentives proportionate to their hoped-for conquest. The sword of the state is brandished by private counsel with a direct pecuniary interest in the litigation. On the one hand, they are driven by the contemplation of a huge payoff; on the other hand, they fill a quasi-prosecutorial role in which their overriding objective is supposedly to seek justice. How could such lawyers possibly evaluate with impartiality the prospect of a settlement, say, or the tradeoff between injunctive and monetary relief?

States are not poor, unable to afford salaried attorneys. Nonetheless, state prosecutors are doling out multi-billion-dollar contingency fee contracts to private trial lawyers. What is worse, those contracts are awarded without competitive bidding to attorneys who are often bankrolling state political campaigns. [104] Indeed, in states like Texas and Louisiana, where judges are elected, trial lawyers give campaign contributions to the very judges who preside over their cases. [105] In Mississippi, Attorney General Mike Moore selected his number-one campaign contributor, Richard Scruggs, to lead the Medicaid recovery suit that goes to trial this July. Scruggs also received a $2.4 million contingency fee for a state asbestos lawsuit in 1992, after contributing over $20,000 to Moore's reelection campaign the year before. [106] In Texas, Attorney General Dan Morales chose five firms for the state's multi-billion-dollar tobacco litigation scheduled for trial in September; four of the five firms contributed a total of nearly $150,000 to Morales from 1990 to 1995. [107]

In West Virginia, tobacco defendants successfully challenged the state's contingency fee contract. [108] Attorney General Darrell McGraw had handpicked six lawyers, without competitive bidding, and declined to specify his selection criteria. [109] He did say, however, that "the State and her citizens stand only to benefit. The State has no exposure. There are no lawyer hourly fees. There are no costs. The taxpayers are thus fully protected." [110] He could have propounded a similar argument if the state had hired private lawyers to prosecute criminal cases, and only paid for convictions. Fortunately, we understand--even if McGraw does not--that defendants as well as taxpayers must be protected. The Supreme Court reminds us that an attorney for the state "is the representative not of an ordinary party to a controversy, but of a sovereignty whose obligation to govern impartially is as compelling as its obligation to govern at all." [111]

Our government is the single entity that is authorized under narrowly defined circumstances to wield coercive power against private citizens. When that government functions as prosecutor or plaintiff in a legal proceeding in which it also dispenses punishment, adequate safeguards against state misbehavior are essential. That is why we need the protections of the Fourth, Fifth, Sixth, and Eighth Amendments; that is why we demand proof beyond reasonable doubt in criminal proceedings; and that is why in civil litigation we rely primarily on private remedies with redress sought by, and for the benefit of, the injured party and not the state. Only when private remedies are deemed to be impracticable or ineffective--for example, when deterrence or punishment rather than corrective justice is the predominant goal (e.g., violent crimes), when culpable defendants are typically judgment proof (e.g., many other crimes), or when the ex ante risk of certain conduct is so high as to require restrictions before actual injury (e.g., nuclear plant safety)--do we condone public prosecution, often driven by statute rather than common law.

None of those conditions is present in the Medicaid recovery suits that the state attorneys general are now litigating. The primary goal purports to be corrective justice; the defendants are not judgment proof; and the ex ante risks are minimal (some would say zero), because no consumer is compelled to use tobacco products. To the extent that the taxpayer is alleged to have been harm