Profiting from smokers.*.

In the late 1990s more than 40 state attorneys general sued the major tobacco companies for the excessive medical costs imposed on the states by smoking. In November 1998 this litigation was concluded with the companies agreeing to pay approximately $9 billion a year, to be adjusted for inflation, in damages and lawyers' fees. The deal was incredibly corrupt; had it been made in any other industry it would surely have been declared illegal. And the tobacco companies were not the only bad guys in this story. The trial lawyers, the politicians, and even the public health officials and antismoking advocates who believed that any means were appropriate to achieve their desired ends of massive fees, political victories, and higher cigarette prices were the ones who most abused the system.

W. Kip Viscusi, an economist and professor at Harvard Law School, focuses on the merits of the state tobacco litigation, principally whether the companies were guilty of inadequately communicating the risk of their products to smokers, which links to youth smoking, and the economic damages the states suffered because of smoking. Viscusi, who has worked on these issues for many years, was hired by Philip Morris to testify on these two topics in the litigation. Nevertheless, he makes a credible case on both issues. Viscusi is substantially less interested in the actual structure and implications of the negotiated deal and spends only a limited amount of time exploring the policy implications of today's tobacco politics, though he does spend a concluding chapter advocating improved information about tobacco for smokers.

1. Risk Perception and Youth Smoking

Risk perception is inextricably linked to youth smoking, the focus of most recent tobacco politics. The companies are accused of promoting underage smoking and thereby addicting kids who are both poorly informed and incapable of making rational decisions about cigarettes. When these kids become adults they are already addicted and many find it difficult to quit. A further argument is that the warnings on packages were inadequate, particularly because they did not warn smokers about addiction.

But Viscusi claims that for many years, at least, kids have shown a remarkable degree of public awareness of the relationship between smoking and lung cancer--much greater in Gallup polls than the awareness that the earth revolves around the sun. As with the general public, kids overestimate the risk of a smoker dying from lung cancer by a factor of four. Indeed, 97% of all California 10th graders believe that breathing secondhand smoke is unhealthy. That said, even if kids have been well informed, those who chose to smoke may have made poor decisions that were influenced by tobacco company marketing programs, which were surely aimed at young adults if not at illegal youth smokers.

I believe a more interesting test is to compare youth smoking with other risk behaviors for 9th-12th graders. In 1995 the Centers for Disease Control estimated that the probability that a kid has smoked at least one cigarette in 20 of the past 30 days is roughly equal to the probability that he or she has driven drunk in the past month, and it is about twice the probability of having carried a gun to school in the past month or attempted suicide in the past year. Smoking kills, but perhaps more attention should be given to reducing other risk behaviors.

A second issue that receives scant attention is racial differences in youth smoking. African American kids are half as likely to smoke as Caucasian kids. But African American adults are far less successful at quitting. Perhaps the link between youth and adult smoking is not terribly simple. We do not know whether a reduction in Caucasian youth smoking would lead to a proportional reduction in adult smoking or if those who would least likely be long-term smokers would be most likely not to start, thereby making the Caucasian statistics look more like the African American statistics. Also, perhaps the public health emphasis, as opposed to the political emphasis, should be on adult smokers.

The political reason for the focus on youth smoking is that it is much harder to make the argument that we should be protecting adults against themselves, and if we recognize that most smokers are low income and addicted it becomes harder to justify regressive cigarette taxes, except with the youth smoking rationale. But as Viscusi points out, people aged 12-18 account for about 1% of cigarette purchases and perhaps 4% of all consumption; therefore, if the real focus is on youth smoking, then a tax increase is a remarkably poor instrument. (Some researchers claim that youth demand is as much as twice as price elastic as adult demand, but given kids' small share of consumption the point remains.)

Furthermore, there is the issue of who should have the incentive to reduce youth smoking. The federal government writes the warnings on cigarette packages. Antismoking ads are produced by antismoking groups. Enforcement of the youth smoking laws is the responsibility of state and local governments rather than the companies, who sell to 770 wholesalers who, in turn, sell to 283,000 retailers. Parental behavior has been convincingly shown to be a key determinant of youth smoking. And then there is the question of the law itself--even as recently as 1990 one third of the states permitted smoking by kids under 18. Yet the Tobacco Deal's division of the spoils gives states no financial incentive to reduce youth smoking.

Finally, studies indicate that lung cancer risk is highly nonlinear; it is much better (in terms of expected deaths) for five persons to smoke for 15 years than for one person to smoke for 30 years. (1) As the eminent epidemiologist Sir Richard Peto argues, tobacco deaths over the next 50 years will be affected much more by the number of adults who manage to stop smoking than by the number of adolescents who start smoking. Therefore, the emphasis on youth smoking may make more political sense than health sense.

2. State Medical Costs

Viscusi makes some good points in his analysis of the cost of smoking to the states. First, he reminds us that the lawsuits were on behalf of the states, not the smokers; in fact, the smokers ended up footing the bill. Therefore, the appropriate measure of damages would be any excessive costs borne by the states. The implication is that only the states' share of excessive Medicare and Medicaid costs should count, not the federal share, and adjustments should be made for the savings regarding early deaths. Lawsuits on behalf of smokers, not the states, are the appropriate venue for settling claims of wrongful death.

Viscusi argues that nursing home cost savings from smokers' early mortality typically exceed increases in a state's medical costs. Similarly, social security cost savings exceed medical costs borne by the federal government. Viscusi argues that the evidence on environmental tobacco smoke (ETS), or "secondhand smoke," is extremely weak; well over 95% of estimated ETS deaths are caused by smokers within the family, perhaps the single place where externalities are best internalized and where the Coase Theorem applies. Finally, the current excise taxes on cigarettes, even before all the recent increases, equal or exceed the medical care costs of smoking.

One can certainly quibble about the numbers. For example, the National Cancer Institute's claims of well over 400,000 smoking deaths per year compares death rates of smokers in the general population with death rates of nonsmokers who volunteered for an American Cancer Society Cancer Prevention Study. Adjustments were made only for age, sex, and race. (2) A more conservative estimate of the death rates could, ironically, lead to a higher estimate of the costs. (3) Adjustments for low-tar cigarettes are also controversial. But it is clear from Viscusi's work that the settlement of the states' litigation, and the states' claims of economic harm, bore no resemblance to actual damages.

Furthermore, the proposed allocation of payments among the companies--taxes in proportion to future sales and, in some incarnations of the deal, lump sums in proportion to stock market value--are unrelated to any reasonable theory of harm. For example, Philip Morris now has half the market but had less than 10% 40 years ago. (4) If the theory is that the company that caused the smoker to become addicted is particularly to blame, or at least that lifetime smoking should count, then Philip Morris's share of the damages should be significantly smaller than its share of the market while its actual contribution has been greater. Companies that have been declining should pay a bigger share.

3. The Settlement

With all this considered, Viscusi argues that the deals were a mistake for the tobacco companies; that they originally hoped for a broad settlement putting all litigation behind them, and when proposed legislation failed in Congress in June 1998 the companies were left with deals with the states that, if anything, may have encouraged more litigation.

But tobacco company stock prices have soared since the deal; perhaps it is worth looking more closely at the structure of the agreements to see whether they were profitable for the companies. The two major elements of the deals are marketing restrictions and damage payments that are structured to be almost equivalent to a tax per pack.

The marketing restrictions seem to be a natural part of any tobacco deal. Companies market to young smokers, for young smokers are the ones who have not settled on a brand. Given the limited amount of brand switching, acquiring a new long-term customer is worth a good deal of money. But the converse is that the battle for younger smokers should "compete away" most of the profits associated with these people. If the companies could collude on a marketing ban that eliminated this expense but reduced youth smoking by only one third, for example, it would clearly raise industry profits. (5) This is effectively what the deal did.

It is no accident that R. J. Reynolds (RJR) agreed to give up Joe Camel in the first half hour of the first bargaining session, and other companies followed suit. Say, for the sake of argument, that a ban on marketing eliminated one third of all youth smoking. The companies would then effectively be trading three customers for whom they had competed fiercely for two customers on whom they would earn monopoly profits. Constitutional issues aside, both sides should be commended for working out this deal that probably reduced smoking by younger people, while giving the companies more profits to bargain away in other parts of the deal, and all without tax increases on older smokers.

But the largest component of the settlements was the agreement of the major companies to pay approximately 40 cents per pack to the states for each pack of cigarettes sold in the United States. It is this arrangement that shows the corruption of illegality of the tobacco deal, but it is an area that Viscusi does not explore in any depth. Indeed, he advocates the tax-like structure of the deal as a way to reduce smoking.

The deal covered 46 states, four other states having settled earlier on similar terms. (6) A national tax (including the first four states) of about 40 cents per pack was imposed, for the benefit of the participating states. But the coercive agreement was very different from each state legislature voting to raise its cigarette taxes. If a state did not wish to sue the tobacco companies then it was given the following options: (i) participate in the settlement and have local consumers pay their share of the national tax, while the state government received its share of the revenue, or (ii) not participate, in which case local consumers would still pay their share of the national tax but the state would not receive its share of the revenue. Some choice!

But even if not technically a tax the payments are still probably illegal for antitrust reasons. Imagine the four firms that dominate an industry jointly putting their debts into a trust, with the trust to be financed by a fee per unit sold on each of the producers. After each settlement the marginal costs of every firm would rise by the same amount, leading immediately to a similar increase in prices and causing consumers to have to pay off the firms' liabilities. (7)

Deals like these have the potential for unbelievable mischief. For example, what if the oil companies were sued by environmentalists for causing global warming, and the companies agreed to pay 50 cents per gallon in damages but with each company getting an exemption for 80% of base year sales? Because the marginal cost of supplying the last consumer with an extra gallon would rise by 50 cents, price would go up by that much. But the companies' average costs would rise by only a dime. Everyone would fare better, except the consumer.

Of course, many people think an increase in the price of coal, as with cigarettes, would be a good thing, precisely because of global warming. I personally would like to see higher prices for sport utility vehicles and taxes on unsolicited telemarketing calls. But the right venue for these concerns is Congress, not a collusion-facilitating agreement between the companies and the trial lawyers.

Furthermore, the artifice of describing the tax collections as damage payments meant that on its face the smaller cigarette manufacturers would be exempt from the multistate agreement. But giving these companies a 40-cent price advantage would have been a deal breaker for Big Tobacco, given that cigarettes cost only 20 cents per pack to produce. Therefore, both carrots and sticks were used to force small companies to sign the Multistate Agreement (MSA), limiting the damage to the Big Four companies.

The small companies were bought off with a provision that said that if they agreed to become subject to the MSA they would be allowed to pocket the tax revenue produced on all their sales up to 125% of base year volume. Because the wholesale price of generics, net of federal excise taxes, was a bit less than the subsidy per pack, this means that a small mom-and-pop cigarette manufacturer, with gross sales of $100,000 per month, is now eligible for a $1.5 million annual tax subsidy for selling its high-tar and nicotine cigarettes. Liggett, which sold 322 million packs in 1997, is eligible for subsidies on up to 400 million packs. These provisions closely mirror the gasoline price example above.

As to why the small companies would settle for only 125%, a crucial part of the deal is that each state was required to pass a model statute that forces any nonparticipating manufacturer (NPM) to pay huge amounts into escrow, where the money stays for decades. Because Liggett had made an earlier deal with the states, there was a question of whether it could escape the escrow payments. Liggett ultimately agreed to sign the deal in place of its earlier agreement in return for $150 million from Philip Morris, plus the right to sell three of its dying brands for another $150 million.

To ensure legislative and judicial compliance, the deal specifies that any state that does not pass a model statute is at risk to lose all its tax revenue. If the model statute is passed but is then declared illegal in state court, then the state still may lose up to 65% of its tax revenue, with the money used to support big companies who lost market share to the NPMs. The terms are somewhat complicated, but a consequence is that under some circumstances RJR will make more money if a consumer buys a pack of NPM cigarettes than a pack of its own Camels. And NPMs have an incentive to limit their sales to a small number of states.

The tobacco settlement would be above legal reproach had the tobacco companies simply agreed not to oppose a 40-cent tax increase in any state that chose to pass one, with the states dismissing their lawsuits. The level of state cigarette taxes is clearly the responsibility of the state legislature. A simpler settlement along those lines would also have eliminated all the Byzantine ancillary provisions. So why do we have the MSA instead?

There are three main reasons. First, the lack of transparency is a key to paying the lawyers. The lawyers could not have claimed a contingency fee for a percentage of the tax increase they lobbied to get passed. It is well known that some of the trial lawyers were awarded fees of over $100,000 per hour, which even one trial lawyer described as "beyond human comprehension." Perhaps less well known is that the firms padded their hours by asking expert witnesses like Viscusi to read voluminous documents into the record and by repeat appearances by the same lawyers in several states retreading the same ground with the same witnesses in each venue. Under the agreement the fees were set with little or no opposition from the companies. Second, it was in the interests of everyone, including the producers, to claim that the companies had been hit hard by the settlement; an easier claim to make if the revenues were characterized as damage payments from the companies rather than tax payments from consumers. Third, once the originally proposed deal broke down in Congress, any tax increase would have had to be state-by-state, with the taxes paid by the consumers in a state going to that state's treasury. The MSA and the four other state deals not only allowed for some redistribution among states, but also created the coercion necessary to get everyone to go along.

Since the MSA, the large companies have predictably lost market share, as smaller companies have had an incentive to increase output at least to their exemption limit, to set up companies that sell mainly in the four states that are not part of the MSA, or to simply try to evade the escrow rules. The health consequence of this growth in small companies is not good; for example, their focus is hardly on low-tar cigarettes. But because of the way the MSA works, the competition effects are truly perverse. To make a long story short, if RJR or Brown & Williamson loses market share because of the growth of NPMs, then they may be eligible for tax rebates that on the margin are worth over a dollar for every pack of sales diverted to the NPMs. That is, if a smoker went into a store and purchased a pack of NPM cigarettes the consequences could be much better for RJR's bottom line than if the consumer bought a pack of Camels.

Whereas before the MSA RJR and Brown & Williamson, as well as Liggett, served as constraints on Philip Morris's price leadership, those constraints are substantially lessened. Liggett's market share is effectively fixed by its MSA participation, and at least under some circumstances the two other companies will find it unprofitable to gain market share, as suggested above.

The clearest results of the tobacco deal are that prices of cigarettes are way up, the operating profits of Philip Morris have risen slightly, and the small cigarette manufacturers have been the second biggest beneficiaries of the deal, right behind the trial lawyers. The public health impact has probably been positive but not enormous. Cigarette sales fell by 9% in 1999 as the deal was implemented, compared with a long-term rate of about 1.5%. Self-reported youth smoking, which rose during the 1990s, has begun to fall back. With higher prices there is also some evidence of consumers switching to higher-tar and nicotine cigarettes and to unfiltered roll-your-own cigarettes. There are also behavioral problems analogous to those typically raised in discussions of light cigarettes: with prices up, are smokers more likely to cover filters with their fingers, draw more deeply, and smoke down to the cigarette butt?

4. The Future

So where do we go from here? Viscusi regards an outright ban on smoking as unlikely to be effective and without broad support. He advocates the traditional economist's solution of better disclosure, not so much to warn of the well-known dangers of smoking but to provide consumers with better information on the comparative risks of tobacco products.

Viscusi also argues that the government should encourage technological advancement in producing less hazardous cigarettes. His argument is that more and better disclosure about the relative risks of smoking will both enable consumers to make better choices and encourage firms to produce safer cigarettes, citing the declines in tar and nicotine that occurred in the late 1950s until, in 1960, the Federal Trade Commission (FTC) negotiated an industry agreement to ban tar and nicotine advertising.

Viscusi is right, but for the foreseeable future politics will stand in the way of his prescriptions.

The current ethos about tobacco seems to be a Nancy Reagan--like "Just Say No" with no discussion of relative risks.

The Department of Health and Human Services is almost two years late in providing a promised revision of the federal tar and nicotine rating system administered by the FTC. There is some question of whether it will ever be done. Meanwhile, public health officials have tried to get the FTC to drop the existing system. The Centers of Disease Control have lobbied the FTC to prevent the chewing-tobacco companies from claiming that their products are safer than cigarettes, even though a consumer's chances of getting oral cancer from chewing tobacco is roughly one tenth the chance of getting lung cancer from cigarettes. The National Cancer Institute released a monograph claiming that low-tar and nicotine cigarettes were no better than regular ones; the scientific co-editor had been one of a group that just a couple of years ago advocated nicotine limits in cigarettes.

Except for nicotine chewing gum and patches, there has been outright opposition in the public health community for alternative means of nicotine delivery, such as low-tar high-nicotine cigarettes and nicotine lollipops. Although these health groups are undoubtedly well meaning, it appears that they have made a decision that the public should be presented with an all-or-nothing choice about tobacco.

I would like to see the federal government buy out all the domestic tobacco companies. By including protection against future lawsuits for the companies this could be done for no more than two years' tax revenue. Although the government generally does a poor job of managing businesses, it should excel in a situation like this where the mandate would be to deglamorize the product and make it as unappealing as possible. For example, the whole business could be turned over to the U.S. Postal Service.

Of course this won't happen, either, because of the politics. Such a deal would mean that the industry would no longer exist as a political whipping boy. Blame and responsibility for smoking would be transferred from the tobacco companies to the government.

Tobacco politics is exceptionally dirty. It is an area where many have apparently adopted an "end justifies the means" approach, whereas other players have focused on personal political and financial gain. The tobacco companies have done some awful things over the years; their behavior in the developing countries remains objectionable, to say the least. But they are not the only villains in this piece.

Viscusi's book fully addresses only a subset of the issues germane to the tobacco deal, essentially the issues most relevant to the litigation merits of the states' cases. But on this subset of issues, the book does a fine job.

(1.) One study cited favorably by the National Cancer Institute implied that the lung cancer risk of smoking for 30 years is about 6 times the risk of smoking for 20 years and 180 times the risk of smoking for 10 years. See Nationnal Cancer Institute Monograp 13, Risks associated with smoking with low machine-measured yields of tar and nicotine, p. 138.

(2.) For an attack on the death estimates, see Lies, damn lies, & 400,000 smoking related deaths by Robert A. Levy and Rosalind Marimont, Regulation 21 No. 4, 1998, pp. 24-29.

(3.) Of course, it is not in the tobacco companies' interests to do their own calculations. How could they conceivably benefit by a careful, conservative report under their sponsorship that showed that "only" 200,000 people a year died from smoking? It is better for the companies to allow the exaggerated claims to exist, with people ultimately becoming aware that they are overstated.

(4.) See Federal Trade Commission, Competition and the financial impact of the proposed tobacco industry settlement, 22 September 1997, p. 18.

(5.) Note that this argument holds even if all the marketing is geared toward smokers who are above the legal minimum age. But it seems inevitable that marketing geared toward appeal to the youngest legal smokers is also the sort that will have the greatest appeal to those who are underage.

(6.) Mississippi, Florida, Texas, and Minnesota each receive a fixed amount per pack sold anywhere in the country by the Big Four tobacco companies. Their shares amount to 12.45% of the overall deal. They are outside the escrow system that applies to the other 46 states and that is discussed below.

(7.) Indeed, the wholesale price of a pack of cigarettes rose by 56 cents from August 1998 until April 2002, more than enough to cover the settlement plus some incremental discounting, while still increasing operating profits. Source: U.S. Department of Agriculture Economic Research Service, table available at http://www.ers.usda.gov/briefing/tobacco/tables.htm. Accessed 30 June 2002.

Jeremy Bulow +

* Review of Smoke Filled Rooms: A Postmortem on the Tobacco Deal by W. Kip Viscusi. Chicago, IL: University of Chicago Press, 2002. Pp. viii, 263. $27.50.

+ Stanford Business School, Stanford University, Stanford, CA 94305-5015, USA; E-mail [e-mail].