Section V: (B)(3), supra (addiction)). Defendants' internal documents demonstrate their understanding that, in order to obtain an amount of nicotine sufficient to satisfy their addiction, smokers of low tar cigarettes modify their smoking behavior, or "compensate," for the reduced nicotine yields by taking more frequent puffs, inhaling smoke more deeply, holding smoke in their lungs longer, covering cigarette ventilation holes with fingers or lips, and/or smoking more cigarettes. See Section V(E)(2)(b), infra (smoker compensation). As a result of this nicotine-driven smoker behavior, smokers of light cigarettes boost their intake of tar, thus negating what Defendants have long promoted as the primary health-related benefit of light cigarettes: lower tar intake. 2027. Defendants did not disclose the full extent and depth of their knowledge and understanding of smoker compensation to the public health community or to government regulators. 2028. Defendants' conduct relating to low tar cigarettes was intended to further their overarching economic goal: to keep smokers smoking; to stop smokers from quitting; to encourage people, especially young people, to start smoking; and to maintain or increase corporate profits.

Low Tar/Light Cigarettes Offer No Clear Health Benefit over Regular Cigarettes

¶a. History of Health Claims

In the early 1950s, on the heels of a series of studies linking smoking and disease, the Defendant cigarette manufacturers began making health claims, often using models in doctors' white coats, in their advertising:

American Tobacco Co. made representations "that its cigarettes were less irritating to the throat than competing brands, offered one's throat protection, were easy on one's throat, and provided protection against throat irritation and coughing. . . ." Federal Trade Commission, In the Matter of American Tobacco Co., Complaint, Findings, and Order in Regard to the Alleged Violation of Sec. 5 of an Act of Congress Approved Sept. 26, 1914 (47 F.T.C. 1393), decided June 20, 1951.

R.J. Reynolds Tobacco Co. "represented to the public . . . that the smoking of such cigarettes . . . aided digestion"; "represented that the wind and physical condition of athletes would not be impaired by the smoking of as many Camel cigarettes as desired"; and "represented that the smoking of Camel cigarettes was soothing, restful, and comforting to the nerves, and protected one against becoming 'jittery' or 'unsure' when subjected to intense nerve strain …." Federal Trade Commission, In the Matter of R.J. Reynolds Tobacco Co., Complaint, Findings, and Order in Regard to the Alleged Violation of Sec. 5 of an Act of Congress Approved Sept. 26, 1914 (46 F.T.C. 706), decided March 31, 1950.

Liggett & Myers Tobacco Co. represented "'directly or by implication, that Chesterfield cigarettes can be smoked by an [sic] smoker without inducing any adverse affect upon the nose, throat and accessory organs of the smoker.'" FTC v. Liggett & Myers Tobacco Co., 108 F. Supp. 573 (S.D.N.Y. 1952).

Lorillard made comparisons between the tar and nicotine yields of its cigarettes and those of its competitors. For

773

example, Old Gold advertisements included statements that Old Gold was "lowest in nicotine and throat irritating tars and resins when compared with 6 other leading brands." Langenfeld WD at 42:5-10 (citing In the Matter of P. Lorillard Co., 46 F.T.C. 735 (1950)). 2030. As discussed in great detail in section V(E)(1)(a), infra, the FTC successfully

¶prosecuted the Defendant cigarette manufacturers' for some of these health claims. See In the Matter of American Tobacco Co., 47 F.T.C. 1393 (F.T.C. 1951); R.J. Reynolds Tobacco Co., 46 F.T.C. 706 (1950), modified, 192 F.2d 535 (7th Cir. 1951), on remand, 48 F.T.C. 682 (1952); Federal Trade Commission v. Liggett & Myers Tobacco Co., 108 F. Supp. 573 (S.D.N.Y. 1952), affirmed on opinion below, 203 F.2d 955 (2d Cir. 1953); P. Lorillard Co. v. Federal Trade Commission, 186 F.2d 52, 56-57 (4th Cir. 1950).

Those prosecutions led to public calls for a ban on advertisements containing such health claims. (no bates) (JD 043377) (Advertising Age (Dec. 21, 1953)) ("Cigarette advertisers were urged today by the National Better Business Bureau to adopt an eight point code to eliminate unfounded health claims in cigarette advertising.").
Given the public's concern over tar and lung cancer, the "White Coat" ads of the 1940s and early 1950s gradually disappeared, and a wave of ads featuring claims about filtration and tar reduction became the new basis for competition. See (no bates) (JD 000636 at 5) (1981 Surgeon General's Report) ("In the 1950s, cigarette manufacturers introduced cigarette filters as 'health protection' and advertised them widely."); Harris TT, 10/18/04, 2783:15-18.
Defendants competed through comparative filtration ads in the period following 1953. (no bates) (US 58700 at 199) (NCI Monograph 13) ("Companies initially responded to this health scare by introducing filtered products that were accompanied by advertisements with explicit health-related statements.").

774

In 1954, the FTC issued a letter to all tobacco companies announcing its intention to adopt uniform standards for cigarette advertising "to prevent the use of false or misleading claims." (no bates) (JD 000332 at 276) (False & Misleading Advertising (Filter-Tip Cigarettes), Hearings before the House Subcomm. of the Comm. on Gov't Operations, 85th Cong. (1957)) . During the ensuing negotiations with cigarette manufacturers, the FTC advised the industry to conform its advertising with FTC decisions, "including decisions finding that comparisons of tar and nicotine between brands were false and misleading." Id.
In 1955, the FTC adopted the "Cigarette Advertising Guides," proscribing any implicit or explicit health claims in cigarette advertising. The Guides did, however, provide a limited exception to this general rule, for what the FTC believed were implicit health claims. This exception allowed comparative ads claiming that a cigarette was "low in nicotine or tars," provided it has "been established by competent scientific proof applicable at the time of dissemination that the claim is true, and if true, that such difference or differences are significant." 50202 3956-57 (JD 003616 at 2) (FTC Press Release of 9/22/55 (Cigarette Advertising Guides)); (no bates) (JD 021949) (4 Trade Reg. (CCH) ¶ 39,012 (1995)). At the same time, some members of the public health community urged the development and adoption of cigarettes with reduced tar yields. Harris TT, 10/18/04, 2782:7-13; Burns TT, 2/15/05, 13357:8-16; Townsend WD, 80:22-81:10.
By the mid to late 1950s, the American cigarette manufacturers responded with a heated "tar derby" of competing claims about the effectiveness of various filters: "[C]igarette companies advertised that certain brands were lower in 'tar' and nicotine and, by implication, less dangerous." (no bates) (JD 001032 at 1-49, n.174) (FTC Staff Report on the Cigarette Advertising Investigation (May 1981)); see also (no bates) (JD 004344 at 8343) (Unfair or Deceptive Advertising & Labeling of Cigarettes in Relation to the Health Hazards of Smoking,²⁹²⁹ "Meta-analysis is an approach for combining data from different studies" or "averaging results from different studies." Samet WD 175:18-21. -1226- Fed. Reg. 8324 (June 22, 1964)).

775

Smokers responded too, switching in droves to filtered cigarettes. Krugman TT, 12/15/04, 8603:8-14; (no bates) (US 58700 Fig. 4-3, at 93).
However, Congress and the FTC perceived that the resulting competition was confusing since different cigarette manufacturers sought to substantiate their "low tar" claims based on different "scientific" testing methods. (no bates) (JD 000332 at 276) (False & Misleading Advertising (Filter-Tip Cigarettes), Hearings Before a Subcomm. of the House Comm. on Gov't Operations, 85th Cong. (1957)) (discussing lack of standardized test method).
Faced with the "tar derby" and perceived consumer confusion, the FTC concluded that, "[i]n the absence of uniform testing procedures, it was impossible to make claims about 'tar' and nicotine levels that could be substantiated. . . ." (no bates) (JD 001032 at 1-49, n.174) (FTC Staff Report on the Cigarette Advertising Investigation (May 1981)). Accordingly, in 1959, the FTC called a halt to the "tar derby" and "reiterated its view that tar and nicotine claims would be regarded as conveying the additional claim that lower levels of tar and nicotine reduced health risks." (no bates) (JD 000435 at 41) (J. Calfee, Cigarette Advertising, Health Information and Regulation Before 1970, Working Paper No. 134 (Dec. 1985)).
"[T]he position taken by the FTC at this time was that the simple fact of listing tar and nicotine deliveries . . . constituted an implied health claim," because the "implication was that these cigarettes would be less hazardous or less harmful." Harris TT, 10/19/04, 2902:19-2903:3.

776

¶On December 17, 1959, the FTC informed tobacco manufacturers that it henceforth would bar all health claims in advertising, including "all representations of low or reduced tar or nicotine, whether by filtration or otherwise." 1005150070 (JD 004534). The FTC further "inform[ed] the industry that in its opinion the evidence then available would support a complaint against any marketer who made any reference to tar or nicotine content, charging that such a reference was false and misleading." 670310575-588 (JD 040931 at 3).

A month later, the FTC requested that the cigarette manufacturers agree to make no references to tar and nicotine in their advertising, and the manufacturers agreed. 1005150056-57 (JD 004535); 1005150051-52 (JD 003617).

b. The FTC Method 2042. The FTC thought the 1964 Surgeon General's Report provided "an evidentiary

¶foundation to support a Rule requiring a positive statement [of average 'tar' and nicotine yields] in [cigarette] labeling and advertising." Letter of 4/11/66 from Paul Rand Dixon, FTC Chairman, to Sen. Warren Magnuson (670310575-588), JD 040931 (at 3).

Within a week of the issuance of the 1964 Surgeon General's Report, the FTC proposed a Trade Regulation Rule that, among other things, would permit the advertising of tar and nicotine yields, provided that such advertising was "verified in accordance with a uniform and reliable testing procedure approved by the Federal Trade Commission."²⁹²⁹ "Meta-analysis is an approach for combining data from different studies" or "averaging results from different studies." Samet WD 175:18-21. -1226- Fed. Reg. 530, (no bates) (JD 040184 at 532);²⁹²⁹ "Meta-analysis is an approach for combining data from different studies" or "averaging results from different studies." Samet WD 175:18-21. -1226- Fed. Reg. 8324 (1964), (no bates) (JD 004344 at 8355-56).
According to the FTC, "[c]onfusion can be obviated, and the ability of consumers to make an intelligent choice among competing brands protected, only if the measurement of cigarettesmoke ingredients accords with a uniform, fully reliable and approved testing procedure." 670310575-588 (JD 040931 at 4).

777

"[T]here was substantial support for the proposition that an accurate statement of tar and nicotine content would be in the public interest. . . ." 0002905512 (JD 004537 at 5512).
The National Interagency Council on Smoking and Health "hope[d] that [the FTC would] take the steps necessary to make it permissible for cigarette manufacturers to list tar and nicotine content on the labels of cigarette packages." 670310575-588 (JD 040931 at 8); 1002905514 (JD 004537).
The American Cancer Society, the American College Health Association, the Roswell Park Memorial Institute, and others expressed support for the proposed rule. 1002905512-5519 at 5514-5519 (JD-004537).
On March 24, 1966, the FTC notified cigarette manufacturers that they would be permitted to advertise tar and nicotine yields provided they used the Cambridge Filter Method, as published by Dr. C.L. Ogg of the U.S. Department of Agriculture, to substantiate any yield claims, and so long as "no collateral representations" were made as to the "reduction or elimination of health hazards" from lower yield cigarettes. (no bates) (JD 004538); see also 680236589 (JD 004612); (no bates) (JD 001032 at 4-3).
The FTC Cambridge Filter Method uses a machine to "smoke" the cigarette for a designated puff volume at a designated interval for a designated period of time. As the smoke is drawn into the machine, it passes over a filter known as a Cambridge pad, on which the particulate tar matter is collected. That accumulated matter is measured to calculate the tar and nicotine yields for the cigarette. The FTC Method was developed to provide consumers with a relative ranking of nicotine, tar, and carbon monoxide yields from any cigarettes that were tested. Henningfield WD, 47:11-48:2; Henningfield TT, 11/22/04, 6794:8-6796:6.

778

When the FTC gave manufacturers permission to make disclosures of tar and nicotine yields, it "recognized that the result would be that consumers would, in fact, believe that lowered delivery cigarettes were less hazardous and less harmful." Harris TT, 10/19/04, 2913:20-24.
The FTC's change in policy to permit these claims was designed to achieve two goals: provide consumers with an incentive to smoke the lower tar/nicotine cigarettes rather than the higher tar/conventional cigarettes and give manufacturers a competitive incentive to produce cigarettes with low levels of tar and nicotine. Harris TT, 10/19/04, 2909:1-25.
The federal government wanted to provide consumers with information that they could use to compare brands. See, e.g., FTCDOCS 0259-1751-1793 (JD 004353 at 1) ("The 'tar' and nicotine testing program was intended to provide smokers seeking to switch to lower 'tar' cigarettes with a single, standardized measurement with which to choose among the then-existing brands.").
In addition, given the premise of a dose-response relationship -- i.e., more tar equals more disease risk -- the FTC wanted to encourage competition among the cigarette manufacturers, thereby increasing the research, development, and production of cigarettes with lower FTC-measured tar yields. See, e.g., (no bates) (JD 043418 at 17) ("Based upon the proposition that lower yield cigarettes present a lessened hazard to the American public," the FTC has acted to "prompt cigarette manufacturers to develop less hazardous cigarettes."); see also (no bates) (JD 004615 at 1) (Sen. Warren G. Magnuson, News Release, Nov. 27, 1967) ("The results of the first government tests ranking cigarette brands by tar and nicotine levels were released today. . . . Hopefully, the wide dissemination of this information and the growing awareness of its significance among the smoking public will channel competition in the cigarette industry toward the marketing of cigarettes of progressively lower tar and nicotine content.").

779

The theory was that the public would shift its consumption away from higher tar products, and toward lower tar products, just as it had done with the advent of filter-tipped cigarettes. In this way, it was anticipated that the national sales-weighted-average tar yields of cigarettes sold in the United States would decline, and the public health would benefit. (no bates) (JD 053570 at 1).
On July 31, 1967, the FTC directed its staff to commence "formal test[ing]" of cigarettes using the Cambridge Filter Test. (no bates) (JD 002477 at 2064360211); (no bates) (JD 004348 at 1).
On August 8, 1970, the FTC proposed a rule to mandate disclosure of tar and nicotine ratings in all cigarette advertising. (no bates) (JD 004350).
The FTC also invited the cigarette manufacturers to submit a voluntary proposal, in lieu of the proposed rule, for such disclosures in cigarette advertising. 2023098316 (JD 040304) ("If the industry can devise a voluntary plan that is feasible and appropriate, the Commission is willing to consider it.").
The tobacco companies complied, (no bates) (JD 040305), and the FTC solicited public comment on the industry plan. 1005045883-84 (JD 041337); (no bates) (JD 002066).
Ultimately, the FTC agreed to allow tobacco companies make certain disclosures about tar and nicotine content in cigarette advertising instead of issuing a formal rule. On December 22, 1970, the FTC formally adopted a revised version of the cigarette manufacturers' proposal for displaying FTC tar and nicotine ratings in all advertising. The agreement was implemented by the FTC on January 13, 1971, as a substitute for its proposed trade regulation rule requiring such disclosure. 1005045883-84 (JD 041337); (no bates) (JD 003634).

780

The FTC concluded that the voluntary agreement to provide "tar and nicotine disclosure and the voluntary agreement . . . to put the Surgeon General's health warning on the side of the pack in advertisements . . . [were] highly responsible activit[ies] by" the cigarette manufacturers. Public Health Cigarette Amendments of 1971, Hearings on S. 1454 before the Consumer Subcomm. of the Senate Comm. on Commerce, 92nd Cong. (1972) (statement of Robert Pitofsky, Dir., Bureau of Consumer Protection); (no bates) (JD 042276 at 58).

c. The FTC Method Does Not Measure Actual Tar and Nicotine Delivery 2061. Within months after it notified cigarette manufacturers on March 24, 1996 of its

¶decision to allow advertising of tar and nicotine yields so long as the Cambridge Filter Method was used, the FTC invited cigarette manufacturers to comment in detail on the precise method to be used to measure tar and nicotine yields. See, e.g., (no bates) (JD 040780); (no bates) (JD 003620).25

Defendants "initially resisted imposition of the Cambridge testing method and claimed it would be inaccurate" because different smokers "smoke differently -- and even smoke differently at different times." Henningfield WD, 48:3-50:12. This is known as smoker variation. Early in the FTC process of developing a standard testing method, Defendants advised the Agency that, because of smoker variation, the Cambridge Filter Method would not measure the tar or nicotine that a human being would ingest from smoking any particular cigarette:

781

No two human smokers smoke in the same way. No individual smoker always smokes in the same fashion. The speed at which one smokes varies both among smokers, and usually also varies with the same individual under different circumstances even within the same day. Some take long puffs (or draws); some take short puffs. That variation affects the [tar and nicotine] quantity in the smoke generated.

¶(no bates) (JD 004362 at 2). See also United States' Obj. and Answers to Joint Defs' Modified

¶Eleventh Set of Req. for Admis. to the Pl. United States, RFA Resp. 5, 49-50, 54 (4/12/02) (The

¶Government admits that in 1966, during the comment period for discussion of the Cambridge method, certain tobacco companies stated that the FTC method would only be an effective measurement for certain conditions of smoking behavior).

The FTC also heard from, among others, Clyde L. Ogg, Ph.D., of USDA, who developed the method initially adopted by the FTC. He admitted that: "Since smokers vary so greatly in their smoking habits, the proposed ... method will not tell a smoker how much tar and nicotine he will get from any given cigarette. It will indicate, however, whether he will get more from one than from another cigarette if there is a significant difference between the two and if he smokes the two in the same manner." (no bates) (JD 004748 at 38).
The FTC's press release announcing its decision clearly described the limitations of the standardized test method it was adopting. (no bates) (JE 061264 at 1-2). The FTC stated:

No test can precisely duplicate conditions of actual human smoking and, within fairly wide limits, no one method can be said to be either "right" or "wrong." The Commission considers it most important that the test results be based on a reasonable standardized method and that they be capable of being presented to the public in a manner that is readily understandable. . . . [T]he public interest requires that all test results presented to the public be based on a uniform method used by all laboratories. Use of more than one testing method would produce

782

different results which would only serve to confuse or mislead the public. The Cambridge Filter Method does not and cannot measure these many variations in human smoking habits. . . . It does not measure all of the tar and nicotine in any cigarette, but only that in the smoke drawn in the standardized machine smoking according to the prescribed method. Thus, the purpose of testing is not to determine the amount of tar and nicotine inhaled by any human smoker, but rather to determine the amount of tar and nicotine generated when a cigarette is smoked by machine in accordance with the prescribed method.

¶(no bates) (JE 061264 at 1-2).

On that same day, the Tobacco Institute issued a press release stating that the FTC method was "unsound" and declaring that the "'tar' and nicotine results" produced by the FTC method "may be inaccurate [and] misleading" to consumers. Tobacco Institute Press Release, Tobacco Institute Says FTC Chose Unsound Test Methods: 'Tar' and Nicotine Results May Be Inaccurate, Misleading, Aug. 1, 1967, 500031952-1955 (JD 047658 at 1). Among other things, the press release pointed out that humans smoke cigarettes differently and that "per cigarette" tar and nicotine yields therefore would be "useless and misleading" to smokers who do not smoke within the FTC parameters. (no bates) (JD 047658 at 1).
Defendants did not, however, disclose their knowledge that smokers would ultimately ingest as much if not more nicotine and tar from low-delivery cigarettes as they would from full-flavor products. Defendants knew that the phenomenon of smoker compensation, discussed in greater detail infra, would cause smokers to smoke low-delivery products more intensely and more frequently in order to obtain their desired level of nicotine. To feed their addiction, therefore, these smokers would defeat the stated purpose of the lower-delivery products. Henningfield WD, 48:3- 50:12. Nor did Defendants disclose to the FTC that "a major reason that the method could yield misleading data was that nicotine addiction would drive smokers to achieve relatively stable nicotine intakes" and that smokers' "physiological need to obtain nicotine substantially lessens the accuracy of the FTC ratings." Henningfield WD, 48-14-49:7. According to Dr. Farone, Defendants did not inform the FTC in 1966 "that smokers alter their smoking behavior to get nicotine." Nor did Defendants tell the FTC that people's "smoking behavior was driven by the need to satisfy their nicotine addiction." Farone TT, 10/12/04, 2170:5-23. .

783

There is a dose-response relationship between smoking and lung cancer. That is, the less smoke to which smokers are exposed, the lower their lung cancer risk. Benowitz TT, 11/1/04, 4521:13-16; Townsend WD, 80:9-18. The predicate for the development and marketing of lower FTC-yield cigarettes was the expectation that, as a group, smokers of lower FTC-yield cigarettes would be exposed to less smoke. Townsend WD, 80:9-18.
Because of compensation and the need of smokers to obtain a desired dose of nicotine, they may offset the decrease in their cigarettes' FTC tar and nicotine yields, in whole or in part, by one of two means. First, smokers may engage in so-called "puff" or "within cigarette compensation." This is done by smoking individual, lower FTC-yield cigarettes more intensively by taking bigger puffs, taking more frequent puffs, smoking the cigarette closer to the butt, blocking ventilation holes placed in the filter that dilute the smoke, or other means. Second, they may simply smoke more cigarettes. Benowitz TT, 11/1/04, 4512:11-4513:1; Dixon WD, 16:13-21.
The issue of compensation has been discussed in the scientific literature since at least the 1940s. Benowitz TT, 11/1/04, 4526:19-4527:3. The early literature on nicotine, including compensation, was summarized in a well respected compendium of articles collected by Larson, P.S. et al., in Tobacco: Experimental and Clinical Studies: A Comprehensive Account of the World

784

¶Literature (Baltimore, Williams & Wilkins Co. 1961). (no bates) (JD 000500); Rowell WD, 17:4- 17. This useful research tool was cited repeatedly in the 1964 Surgeon General's Report. Rowell WD ,18:1-18. The Larson volume was funded by TIRC. McAllister WD, 102:9-22; Rowell WD , 17:18-20.

It was a "common concern" in the early 1960s that smokers who switched to filtered cigarettes might "compensate" by smoking more cigarettes each day. Burns TT, 2/16/05, 13580:18- 13581:7; see also Samet TT, 9/29/04, 1183:5-9.
"Compensatory smoking behavior is a manifestation of nicotine addiction," and "occurs primarily due to nicotine." Dr. Benowitz described "the consensus in the medical and scientific fields" regarding compensation: "The concept of smoking to obtain desired levels of nicotine and the concept of nicotine titration with associated compensation is widely accepted by the scientific and public health communities." Methods of smoker compensation include taking "bigger puffs," taking "more frequent puffs," "block[ing] the ventilation holes in the filter," and smoking more cigarettes. Benowitz WD, 55:11-22, 56:22-23, 57:5-9, 57:23-1; Benowitz TT, 11/2/04, 4762:23-24, 4763:14-16. Even as to ultra low tar cigarettes, where a smoker switching down may not be able to compensate fully "on a per cigarette basis[,] . . . that smoker could always smoke more cigarettes." Farone WD, 103:18-104:1; accord Farone TT, 10/12/04, 2169:18-19 (testifying that "the requirement for nicotine" drives smoker compensation).
Because each smoker smokes to obtain his or her own particular nicotine quota, smokers end up inhaling essentially the same amount of nicotine -- and tar -- from so-called "low tar and nicotine" cigarettes as they would inhale from regular, "full flavor" cigarettes. This is referred to as "complete" compensation. Virtually all smokers, over 95%, compensate for nicotine.

785

¶Benowitz WD, 59:6-17; 61:15-62:13; Benowitz TT, 11/2/04, 4769:25-4770:4; (no bates) (US 58700 at 10) (Monograph 13); accord Burns WD, 1:10-15, 12:10-11, 43:19-45:2; Burns TT, 2/15/05, 13311:9-15; Burns TT, 2/16/05, 13537:6-9.

The amount of nicotine that smokers need to sustain their nicotine addiction does not change over time. Therefore, compensation for reduced deliveries is permanent, and occurs for as long as the smoker smokes the low tar product. Benowitz WD, 70:25-71:10; see generally DXA0310399-0650 at 0452-0476 (US 58700) (Monograph 13) (indicating no evidence to warrant conclusion that there is reduction in compensation over time).
Because compensation is essentially complete, low tar cigarette smokers inhale essentially the same amount of tar and nicotine as they would from full flavor cigarettes, thereby eliminating any purported health benefit from low tar cigarettes. In short, "light and ultra-light cigarettes" do not, in actuality, "reduce the risks of smoking":

Considering the overall exposure data for individuals selecting their own brand, there is little reason to expect that smokers of cigarettes with low machine measured yields will have a lower risk of disease than those who smoke higher yield cigarettes.

¶Benowitz WD, 72:9-14; (no bates) (US 58700 at 60); see also Benowitz WD, 61:6-13 (explaining the conclusions of Chapter 3 of NCI Monograph 13).

As Dr. Benowitz pointed out, Compensation explains why smoking of light cigarettes has not been associated with a reduction of smoking-induced disease risks. One would think, looking at the FTC yield data, that toxic exposures would be substantially reduced if one switches to light cigarettes; however, because of compensation, resulting toxic exposures are similar for light and regular cigarettes.

786

¶Benowitz WD, 60:23-61:14.

Despite the fact that tar deliveries, as measured by the FTC Method, decreased by more than two-thirds between 1954 and 1994, lung cancer in smokers actually increased. (no bates)

¶(US 58700) (Monograph 13); (no bates) (US 76212) (1997 CDC MMWR article); see also (no bates)

¶(US 88626) (1995 Thun et al. article).

Compensation behavior is distinct from "individual smoker variation":

Individual smoker variation refers to the fact that one smoker may smoke cigarettes -- either regular or low tar -- differently than another smoker, and that the same person may smoke the same cigarette differently on different occasions. . . . Individual smoker variability relates to the fact that cigarettes are smoked differently by different individuals. This type of variability is separate and distinct from the issue of compensation, which relates to the phenomenon of smokers smoking purportedly low-delivery cigarettes more intensely in order to achieve their particular desired level of nicotine intake.

¶Benowitz WD, 56:6-23.

In its August 1, 1967 press release, the FTC set forth the Commission's understanding of smoker variation:

No two human smokers smoke in the same way. No individual smoker always smokes in the same fashion. The speed at which one smokes varies both among smokers, and usually also varies with the same individual under different circumstances even within the same day. Some take long puffs (or draws); some take short puffs. That variation affects the tar and nicotine quantity in the smoke generated.

Even with the same type of cigarette, individual smokers take a different number of puffs per cigarette depending upon the circumstances. When concentrating, or talking, the number of puffs is usually less. When listening, or required to listen to another person talking, the number of puffs per cigarette, as well as duration of each puff, usually increases. Smoking rates while reading a book may differ from smoking rates while viewing a television program. The number of puffs and puff duration (as well as butt length) will vary

787

according to emotional state. Some smokers customarily put their cigarettes down in an ashtray where they burn between puffs; other smokers constantly hold cigarettes in their mouths; others hold them between their fingers.

¶(no bates) (JD 040254 at 2); 03573029-3030 at 3029 (US 22244).

Significantly, the August 1, 1967 press release does not demonstrate a similar understanding of nicotine or addiction. It does not even mention nicotine and does not discuss the fact that nicotine addiction would lead smokers to obtain essentially the same amount of nicotine from so-called low tar cigarettes as they would from regular cigarettes. (no bates) (JD 040254); Farone TT, 10/12/04, 2170:5-23.
Public service announcements of the Office on Smoking and Health from the early 1980s relating to the potential for compensation reflected a similarly incomplete and ultimately incorrect view of compensation. Rather, they "implie[d] that the individual has within [his/her] ability automatically not to compensate; that is, that the compensation is not driven by the addictive process. That was the understanding we had in the early 1980s . . . . It was only after that [] that we understood with precision and specificity how the nicotine drives that smoking change." Burns TT, 2/16/05, 13565:8-13566:4.
Defendants suggested an analogy between the FTC tar and nicotine yields and automobile gas mileage estimates, intimating that they are both useful, albeit imperfect. As Dr. Henningfield explained, this comparison is not valid:

[W]e know through that [gas mileage] rating system that if you buy a car with a better gas mileage rating, virtually no matter how you drive it, you're going to get better mileage than a car with a worse rating. But in cigarettes, by just subtle changes in the way you smoke and things that most people don't even know about, the ventilation and the channels and the burn accelerants and all these different

788

tricks, makes those two cigarettes look the same. Thus, for example, when humans smoke Marlboro cigarettes . . . Marlboro Lights can yield approximately twice as much nicotine as the Regulars are claimed to deliver by the standard FTC method. Marlboro Ultra Lights can deliver three times their advertised rating and most of the Carlton brands can deliver seven or more times their advertised rating.

¶Henningfield WD, 83:14-84:10.

Light cigarette descriptors also "are totally different" from the information on food labels and drug labels, because "if you eat the listed serving size of [foods], you will receive the amount of [the constituents] listed on the label. . . . By contrast, . . . the advertised FTC tar and nicotine ratings for cigarettes bear very little relation to the actual dose a smoker can and, in most instances, does receive from smoking that cigarette. The inaccuracy in the FTC ratings is especially pronounced for cigarettes sold as 'light' or 'low tar' by the tobacco companies. This discrepancy is especially serious because it is in the direction of more toxins than advertised." Henningfield WD, 84:11-85:3.
Dr. Whidby, a scientist, former employee, and consultant for Philip Morris USA, agreed that: "[m]easurements of tar and nicotine yields using the FTC method do not offer smokers meaningful information on the amount of tar and nicotine they will receive from a cigarette . . . [or about] the relative amounts of tar and nicotine exposure likely to be received from smoking different brands of cigarettes." Whidby TT, 2/22/05, 13993:14-19; DXA0310399-0650 at 0423, 0452-0476 (US 58700).
Compensation has been documented by various scientific methods. Three different kinds of studies are generally used to conduct research on compensation: (1) spontaneous brand switching studies, (2) forced brand switching studies, and (3) cross-sectional studies. Benowitz WD,

789

First, spontaneous brand switching studies are longitudinal studies that follow the same group of smokers over a specific period of time. At the start, the study measures the smokers' daily smoke exposure and records the FTC-yield of the smokers' usual brand. Later, at follow-up, the same smokers are re-contacted, at which time the study observes any change in the FTC-yield of the smokers' usual cigarette and again measures the smokers' daily smoke exposure. Benowitz WD, 62:21-63:4; Benowitz TT, 11/1/04, 4513:2-4514:3, especially 4513:13-21; (no bates) (US 58700 at 45) (Monograph 13).
Such a study permits estimation of the changes in the daily smoke exposure of those smokers who, over the course of the study, spontaneously and voluntarily switched to cigarette brands with higher or lower FTC yields, as well as any changes in the daily smoke exposure of those smokers who did not change the FTC yield of their cigarette. Benowitz WD, 63:22-64:2; Benowitz TT, 11/1/04, 4513:2-4514:3, especially 4513:13-21.
Spontaneous brand switching studies, like randomized experiments, may be long term or short term. Dixon TT, 3/9/05, 15051:14-15052:5.
Spontaneous brand switching studies "are more informative of smokers' exposure in the real world when switching from higher to lower yield cigarettes," because "the brand of cigarette has been selected by the smoker and not by the researchers." Benowitz WD, 63:22-64:2. "[S]pontaneous brand switching studies generally show that there is no reduction in smoke intake [including nicotine and tar intake] per cigarette. . . . The per cigarette figure [is important because it] shows what an individual can take in from a particular cigarette. Thus, it provides information on the delivery characteristics of the product." Id. at 64:8-16, 66:3-5.

790

There is only one complete, peer-reviewed long-term spontaneous brand switching study, Lynch and Benowitz 1987, "Spontaneous Cigarette Brand Switching: Consequences for

¶Nicotine and Carbon Monoxide Exposure," J. Public Health, 78(9): 1191-1194, (no bates)

¶(JD 063010); Benowitz TT, 11/2/04, 4753:11-4755:5. The study found:

a. per cigarette nicotine intake was about the same, comparing smokers who switched to lower FTC-yield cigarettes during the course of the study to their own baseline per cigarette nicotine intake. (no bates) (JD 063010 at 1192).

b. per cigarette nicotine intake was lower, comparing smokers who switched to lower FTC-yield cigarettes during the course of the study to a similar "control group" of full-flavor smokers who did not switch. Benowitz TT, 11/2/04, 4758:25-4760:13; (no bates) (JD 063010 at 1192).

Only daily nicotine intake was actually measured in the study. The per cigarette nicotine values were calculated by dividing daily nicotine intake by the number of cigarettes the smokers reported they smoked per day. Benowitz TT, 11/2/04, 4758:12-18; Wecker WD, 9:9-10:3.
Dr. Benowitz drew this conclusion from his study:

For smokers who switched to lower yield cigarettes, the analysis of cotinine concentration or carbon monoxide per cigarette showed no change despite the reduction in nominal machine measured yield. Therefore, these smokers obtained the same dose of nicotine and carbon monoxide from each cigarette even though the machine measured yield was lower.

¶Benowitz WD, 63:22-64:2. The Benowitz study demonstrated that: "For spontaneous brand switchers, there is complete compensation for each cigarette smoked. As a result, for these smokers, switching from higher to lower yield cigarettes is not likely to reduce the risk of smoking." Id. at

791

¶64:14-65:6, 65:14-17. The evidence that there is no reduction per cigarette by switching to lower tar cigarettes is particularly compelling in light of Dr. Benowitz's testimony that "we do know that on average people who are smoking lower-yield cigarettes smoke the same or even slightly more than higher-yield cigarettes." Benowitz WD, 63:22-64:2; 64:14-65:6; 65:14-17; Benowitz TT, 11/2/04, 4762:23-24; 4763:14-16; see also Benowitz WD, 63:5-10 (explaining that "[c]otinine is a major breakdown product of nicotine" that "is metabolized . . . by the liver" in humans, and therefore "has become the accepted marker for looking at nicotine exposure measurement from tobacco products").

Based on this study, Monograph 13 concluded: For spontaneous brand switchers, there appears to be complete compensation for nicotine delivery, reflecting more intensive smoking of lower-yield cigarettes. *** Spontaneous brand-switching studies suggest that there is no reduction in smoke intake per cigarette. . . .

¶(no bates) (US 58700 at 10 and 60); see also Dixon TT, 3/9/05, 15046:7-15047:7.

There are also short-term spontaneous brand switching studies on compensation. Dixon TT, 3/9/05, 15051:14-15052:05; Dixon WD, 40:7-13.
The peer-reviewed literature contains a 1999 meta-analysis of the brand switching studies that employed nicotine biomarker data. The mean estimate of the extent of compensation in that article was about 50-60%. Dixon WD, 40:14-20; see also (no bates) (JD 000547 at 1).
Second, in experimental studies for forced brand switching, smokers are randomly assigned to smoke cigarettes with higher or lower FTC yields. The smokers' daily smoke exposure, as measured by various biologic markers, is compared to see if the smokers assigned to smoke lower

792

¶FTC-yield cigarettes have a lower daily smoke exposure than smokers assigned to smoke higher FTC-yield cigarettes. Benowitz WD, 68:6-9.

Daily smoke exposure takes into account both forms of potential compensation -- the tendency of smokers of lower FTC-yield cigarettes to smoke individual cigarettes more intensively, as well as the tendency of smokers of lower FTC-yield cigarettes to smoke more cigarettes. Benowitz TT, 11/2/04, 4739:13-4740:3; Wecker WD, 5:22-6:7.
In these long-term randomized experiments on forced brand switching, smokers randomly assigned to smoke lower tar cigarettes were exposed to less smoke each day than the smokers randomly assigned to smoke higher tar cigarettes. The estimated compensation is about 75- 80%, suggesting substantial but incomplete compensation. Benowitz WD, 70:19-21; Benowitz TT, 11/2/04, 4751:1-7; Wecker WD, 5:10-17.
Dr. Benowitz explained that the substantial but incomplete compensation shown in the forced switching compensation studies is likely due to the act of forcing participants to switch brands:

[S]mokers are switched only for the purpose of the research. Motivation and cigarette acceptability differ from the natural situation of brand switching. . . . The forced brand switching studies show on average about eighty percent compensation. . . . Presumably compensation is not complete because the smokers have been switched to cigarettes that were not of their own choosing.

¶Benowitz WD, 68:6-21; 70:19-24.

Third, in cross-sectional studies, the daily nicotine intake of smokers smoking their usual, voluntarily-selected brand is measured, typically using biologic markers for exposure to cigarette smoke, such as cotinine in the blood or saliva, and compared with the FTC-yield of the smokers' cigarettes. Benowitz TT, 11/2/04, 4742:25-4743:11.

793

Cross-sectional studies also lead to the conclusion that compensation is essentially complete:

Cross-sectional studies involve sampling smokers in the general population who are smoking their own chosen brand of cigarettes . . . show that there is very little difference in tobacco smoke exposure in people smoking cigarettes with different machine-determined yields. . . . There have been many cross-sectional studies performed, and overall they demonstrate that while there are some differences in nicotine exposure when high- and low-yield cigarette brands are compared, these differences are quite small. . . . Cross-sectional studies show nearly 100 percent compensation.

¶Benowitz WD, 66:9-68:5.

In cross-sectional studies where the participants themselves choose the tar level of their cigarettes, there is a "very shallow slope" or "very tiny slope across the range of tar and nicotine" comparing the nicotine intake of smokers of various tar levels, demonstrating that smokers who smoke cigarettes of widely varied FTC tar levels are ingesting similar amounts of nicotine. This data indicates that compensation is essentially complete. As Dr. Burns explained, the fact that lower tar smokers may show, on the whole, slightly lower levels of nicotine than higher tar smokers does not mean that the lower levels are the result of the type of cigarette, but rather that the nicotine quota of smokers able to smoke lower tar cigarettes is customarily lower:

The effect is one that one would expect to be present as a small slope, since one would expect that high yield smokers would be likely to have higher nicotine levels and that the very lowest yield cigarette smokers would be there because they don't need much nicotine. That's independent of the brand of cigarettes they smoke. That's why they chose those brands. It's not an effect of the brand that they smoke. And so you would expect to see a small slope. The fact that

794

the slope is as small as it is suggests that . . . there is full compensation on a population level when people in the natural setting move from one brand to another.

¶Dr. Benowitz's 1983 study, which showed this "very shallow slope," has been "replicated [in]numerable times by other scientists," and "those studies show basically the same picture, that there's a very shallow slope between the machine yield and cotinine levels." Burns TT, 2/16/05, 13541:8-13542:14; Benowitz WD, 67:10-18 (discussing his 1983 cross-sectional study); Benowitz TT, 11/1/04, 4564:13-15; Benowitz TT, 11/2/04, 4826:13-4827:5; see also Burns TT, 2/16/05, 13547:7-11 (indicating that the Gori and Lynch compensation study "was presented and examined in [Monograph 13] by Dr. Benowitz and was part of the data that was examined in that chapter that reached the conclusion that compensation was essentially complete").26

The cross-sectional studies as a whole suggested that compensation is not complete but substantial. Benowitz TT, 11/2/04, 4737:22-4738:9, 4738:15-18; Dixon WD, 44:23-45:18; Mulholland WD, 19:18-20:13.

795

Evaluating all the types of studies as a whole, the evidence demonstrates, at a minimum, that compensation for daily nicotine is substantial if not complete. Benowitz TT, 11/2/04, 4737:22-4738:9, 4738:15-18 (cross-sectional studies); 4751:1-7 (randomized experiments); 4757:4- 8, 4757:19-23, 4766:24-4767:3 (spontaneous brand switching, daily nicotine intake); Wecker WD, 15:19-16:5; Dixon TT, 3/9/05, 14942:5-14943:14; Mulholland WD, 57:12-22; Mulholland TT, 4/25/05, 19943:12-19945:19.

d. The Public Health Community Has Concluded that Low Tar Cigarettes Offer No Clear Health Benefit 2104. Low tar cigarettes have not reduced the risks of smoking relative to full-flavor

¶cigarettes. Burns WD, 1:10-15; 12:10-11; 30:5-12 ("I have concluded that the changes in cigarettes that resulted in a lowering of the FTC tar and nicotine yields over the past 50 years have not resulted in a reduction in the disease risks of smoking cigarettes for the smokers who use these cigarettes."); Burns TT, 2/15/05, 13311:9-15. Dr. Jonathan Samet, a Government expert with extraordinary qualifications, is a physician and epidemiologist with extensive experience treating patients with lung cancer and COPD. He is the Chair of the Department of Epidemiology at the Johns Hopkins University Bloomberg School of Public Health and has served as author and/or editor of several Surgeon Generals' Reports over more than 25 years, contributed to several National Cancer Institute Tobacco Control Monographs, and served as an author of Chapter 4 of Monograph 13. As an expert in the science of tobacco smoking and health, including epidemiology, pulmonary medicine, and internal medicine, he concluded that the use of lower tar and lower nicotine cigarettes "has had no clear benefit on the health risks of active smoking." Samet WD, 1:3-12, 2:20-3:3, 3:7-11, 3:19-23, 10:15-12:16, 14:1-15:13, 18:12-16, 168:17-19; Samet TT, 9/29/04, 10-18. Similarly, Dr. William

796

¶Farone, fact and expert witness and former Director of Applied Research at Philip Morris, concluded that, based on his training and experience, "'light' cigarettes -- because they generally permit easy compensation and employ levels of dilution that increase the mutagenicity of the tar – are not any less hazardous than their full flavor versions." Farone WD, 123:21-124:4. The Court credits the testimony of these three experts.

The 1981 Surgeon General's Report concluded, referring to the FTC Method of measuring tar and nicotine:

[T]he smoking-machine model is limited in accurately reproducing human smoking behavior.... Smokers, however, are able to take larger, more frequent, and higher velocity puffs than the machines do. It appears that such compensatory adjustments often turn nominally lower 'tar' and nicotine cigarettes into higher 'tar' and nicotine cigarettes. . . . Even if the compensations made in smoking a single cigarette are small or nonexistent, smokers can increase their intake of 'tar' and nicotine by smoking more cigarettes.

¶(no bates) (JD 000636 at 180) (1981 Surgeon General's Report).

The 1981 Report recognized that there are still "smokers who are unwilling or as yet unable to quit." As to them, the Report concluded that they "are well advised to switch to cigarettes yielding less 'tar' and nicotine, provided they do not increase their smoking or change their smoking in other ways." Id. at v. 2107. Dr. Burns, an editor of the 1981 Surgeon General's Report as well as "an author, editor or reviewer for each of the annual Reports of the U.S. Surgeon General on the Health Consequences of Smoking since 1975," concluded that, in his expert opinion,

had the information available to the tobacco industry been available to the scientists preparing the 1981 Surgeon General's Report, that Report would not have drawn the erroneous conclusion that lower tar cigarettes produced lower risk or have made the recommendation that

797

smokers who could not quit were "well advised to switch to cigarettes yielding less 'tar' and nicotine."

¶Burns WD, 1:10-15, 12:10-11, 36:3-37:12, 55:17-56:13, 56:21-57:17; Burns TT, 2/15/05, 13311:9- 15; Burns TT, 2/16/05, 13666:25-13667:24 ("Had that information been available to us, we would not have then offered the recommendation to the population of the United States that it would be a good idea to shift to these products."); see also Burns WD, 56:14-20 ("The Surgeon General clearly expressed a concern [in the 1981 Report] about reduced yield smoking leading to compensatory increases in smoking behaviors; but, at that time, the public health community was not aware of the role of nicotine addiction in altering puffing behavior, the elasticity of delivery designed into cigarettes then on the market which facilitated compensation on the part of the smoker, or the observations made by the industry that showed compensation was essentially complete for some 'light' cigarettes. Had we known that, the recommendation would not have been made."); Burns WD, 38:10-13 (indicating that "some of the same concerns [relating to the lack of a health benefit to lower tar cigarettes] were expressed in the 1989 Surgeon General's Report").

The 1981 Report did not fully take into consideration the phenomenon of compensation, and how smokers smoke to get a certain amount of nicotine, and will even adjust their smoking behavior to get the amount of nicotine they seek or are accustomed to . . . we didn't know in 1981 the extent to which smokers would compensate after switching to a 'low tar' and low nicotine yield product.

¶Samet WD, 164:15-165:5.

Dr. Burns provided the "three principal reasons" that "the traditional epidemiological approaches that were employed at the time of the 1981 Surgeon General's Report" yielded results erroneously suggesting that lower tar cigarettes provided less lung cancer risk:

798

(1) "[T]hat people who smoked low-tar and nicotine cigarettes" were smoking them largely "based on the understanding that these cigarettes . . . offered less risk." As a result, the people who choose these cigarettes "have different health behaviors" and often "different smoking characteristics" than smokers of higher tar cigarettes, leading to different expectations of health outcomes.

(2) That "very few people in the epidemiologic studies started out smoking low tar and nicotine or even filtered cigarettes." Most smokers who smoke the high-tar cigarettes very intensely are not able to switch down to lower tar cigarettes, whereas people who do not smoke the higher tar cigarettes very intensely, "when they switched to a low-tar cigarette . . . may be successful because they didn't have much nicotine intake that they needed to satisfy, and correspondingly they didn't have much tar intake. So the process of switching to low tar and nicotine starts to separate individuals who have different intensities of smoking, different amounts of tar that they are ingesting and, therefore, will have different risks."

(3) That "a substantial fraction of people who switch from high-tar and nicotine cigarettes to low-tar and nicotine cigarettes use increased numbers of cigarettes . . . as a mechanism of compensation. . . . In order to control for intensity of smoking, the epidemiologic studies used the number of cigarettes smoked per day as a measure of intensity with the mistaken assumption that people wouldn't change the number of cigarettes they smoked per day. That leads us to underestimate the actual number of cigarettes smoked per day as a measure of exposure in low-tar and nicotine cigarette smokers," because the epidemiological analysis compares people who smoke a higher number of cigarettes per day after switching to a lower tar cigarette to people who smoked this higher number of cigarettes per day of the higher tar cigarette. "That produces an erroneous, or incorrect, perception that switching to [the lower tar] cigarette lowered your lung cancer risk as an individual."

¶Burns TT, 2/15/05, 13327:21-13331:3; 13334:6-13337:5.

799

Moreover, epidemiological studies may underestimate the risks for lower tar smokers because these smokers may have other characteristics – such as healthier lifestyles – that contribute to a reduction in risk regardless of the type of cigarette smoked:

Epidemiological studies often assume that smokers who switch to low-yield products are similar to smokers who do not. There is evidence that this may not be true. For example, switchers may smoke their cigarettes differently, they may have started smoking later as teenagers, they may attempt to quit more often. Switchers may have smoked less intensely before they switched, when compared to their high-yield, non-switching counterparts. Switchers may have smoked less at younger ages. When these aspects of smoking behavior are not accounted for, study results may be misleading. In addition, switchers are [a] generally healthier group in terms of diet, exercise, and lifestyle in comparison to smokers who do not switch to a low yield product. The cumulative effect of these group differences is that any reduction in the risks among switchers may be the result of these differences, rather than the fact that they switched to a low yield product. This was also an observation in Monograph 13.

¶Samet WD, 167:13-168:2.

Recent studies, including the National Cancer Institute's Monograph 13 and the 2004

¶Surgeon General's Report, have confirmed that low tar and filtered cigarettes are no less harmful than conventional delivery and unfiltered cigarettes. The 2001 NCI Monograph 13, "Risks Associated

¶With Smoking Cigarettes With Low Machine Measured Yields of Tar and Nicotine" ("Monograph

¶13") concluded:

Epidemiological and other scientific evidence, including patterns of mortality from smoking-caused diseases, does not indicate a benefit to the public health from changes in cigarette design and manufacturing over the last fifty years. . . . Widespread adoption of lower yield cigarettes in the United States has not prevented the sustained increase in lung cancer among older smokers. . . . Considering the overall exposure data for individuals selecting their own brands, there is little reason to expect that smokers of low yield

800

cigarettes will have a lower risk of disease than those who smoked higher yield cigarettes.

¶DXA0310399-0650 at 0422-0423, 0473 (US 58700).

Dr. Samet testified that: "The evidence is clear. We have tracked the risk of lung cancer closely and not seen a fall in relative risks to smokers." Samet WD, 169:14-16.
The 2004 Surgeon General's Report reached the definitive conclusion: "[C]igarettes with lower machine-measured yields of tar and nicotine (i.e., low-tar/nicotine cigarettes) have not produced a lower risk of smoking-related diseases." In addition, the Report concluded that "[s]moking cigarettes with lower machine-measured yields of tar and nicotine provides no clear benefit to health" and that, "[a]lthough characteristics of cigarettes have changed during the last 50 years and yields of tar and nicotine have declined substantially, as assessed by the Federal Trade Commission's test protocol, the risk of lung cancer in smokers has not declined." TLT0930001-0949 at 0042, 0340, 0911 (US 88621). See also VXA100001-0604 (US 77217) (NCI Monograph 8); MTP0032477-2481 (US 76212) (11/7/97 CDC MMWR article); VXA1611681-1689 (US 77222) (1996 Samet et al. article); VXA1601456-1742 (US 64059) (1984 Surgeon General's Report).
Both the 2004 Surgeon General's Report and the NCI's Monograph 13 were based on evidence "derived from research on human behavior and exposures, cigarette design and yields, smoke chemistry, epidemiological [and other] population-based data on human disease risk." DXA0310399-0650 at 0422-0423 (US 58700).
Extensive research into the relationship between research of biomarkers of nicotine in humans and FTC tar and nicotine yields demonstrates that lower tar cigarettes do not provide a reduction in harm:

801

Generally speaking, research using these biomarkers has indicated little, if any, correlation between the FTC-yield of tar or nicotine, and the levels of the biomarkers measured in smokers. . . . These results suggest that there is little difference in the levels of biomarkers comparing smokers of higher yield tar/nicotine cigarettes and lower yield tar/nicotine cigarettes, as measured by the FTC method. This implies that doses of carcinogens or other toxic materials that smokers ingest have little relationship, if any, to the FTC tar yield. This, in turn, suggests that the gradual reduction in tar yield over the past several decades has not resulted in a reduction in smokers' exposure to carcinogens, and that the FTC test method is not informative with respect to lung cancer risk or to the risk of smoking-caused diseases generally. . . . In fact, evidence with respect to smoker compensation and biomarkers shows that those smokers who switch to "Low Tar" cigarettes modify their pattern of smoking to obtain the same or similar amounts of tar and nicotine as from the "High Tar" cigarettes they used to smoke. The bottom line is that a "Low Tar" label-based brand under the FTC protocol does not mean that a smoker is actually ingesting "Lower Tar" than from any other cigarette.

¶Samet WD, 147:11-148:21; 149:23-150:4.

The conclusions of Monograph 13 and the 2004 Surgeon General's Report -- that lower tar cigarettes do not provide a health benefit -- "represent[] the consensus view of the scientific community on this issue." Burns WD, 1:10-15; 31:6-9; 41:12-18; 58:20-61:13 (discussing

¶Monograph 13 (US 58700); IOM Report (US 20919); WHO Sactob Report (US 86658); and 2004

¶Surgeon General's Report (US 88621)); Burns TT, 2/15/05, 13311:9-15; 13668:1-8; see also

¶Benowitz WD, 72:21-24 ("Most authorities are now convinced that there is little if any benefit with respect to health risk to smoking low yield versus regular cigarettes"); (no bates) (US 86657)

¶(Canadian Expert Panel, Putting an End to Deception: Proceedings of the International Expert Panel on Cigarette Descriptors. A report to the Canadian Minister of Health from the Ministerial Advisory

¶Council on Tobacco Control 9) (2001) ("There is no convincing evidence of a meaningful health benefit to either individuals nor to the whole population resulting from cigarettes marketed as 'light' or mild'").

802

In 2001, the Institute of Medicine published "Clearing the Smoke: Assessing the

¶Science Base for Tobacco Harm Reduction," which concluded that "the public health of PREPs

¶[potential reduced-exposure products] is unknown." The IOM report went on:

The major concern for public health is that tobacco users who might otherwise quit will use PREPs instead, or others may initiate smoking, feeling that PREPs are safe. That will lead to less harm reduction for a population (as well as less risk reduction for that individual) than would occur without the PREP, and possibly to an adverse effect on the population.

¶(US 20919).

In 2001, the World Health Organization Scientific Advisory Committee on Tobacco

¶Product Regulation issued a report that concluded:

Tar, nicotine, and CO numerical ratings based upon current FTC methods and presented on cigarette packages and in advertising . . . are misleading and should not be displayed. . . . 4. Banned terms should include light, ultralight, mild and low-tar, and may be extended to other misleading terms. The ban should include not only misleading terms and claims but also, names, trademarks, imagery and other means of conveying the impression that the product provides a health benefit.

¶TLT1010692-0699, 0695 (US 86658).

In 2004, the World Health Organization International Agency for Research on Cancer

¶("IARC") released its Monograph 83, "Tobacco Smoke and Involuntary Smoking," which concluded that "changes in cigarettes since the 1950s have probably tended to reduce the risk for lung cancer associated with the smoking of particular numbers of cigarettes at particular ages." However, the

803

¶IARC Monograph attributed any reduced risk largely to the shift from unfiltered to filtered cigarettes which occurred in the 1950s, not the shift from high yield to low yield cigarettes which occurred in the last few decades. Moreover, the IARC's conclusion did not apply to individuals who increased the number of cigarettes they consumed as they shifted to low yield cigarettes. 1000861953-1953 (US 35484) (Wakeham 3/24/61) ("As we know, all too often the smoker who switches to a hi-fi cigarette winds up smoking more units in order to provide himself with the delivery which he had before."). The IARC Monograph went on: "the introduction of cigarettes that can be misperceived as 'safe' may well have adversely affected smoking uptake rates, cessation rates, and consumption per smoker." TLT0970001-1455, 0180 (US 86746)

All the major scientific bodies that have addressed this question in recent years have clearly concluded that lower tar cigarettes provide "no clear benefit" to health:

I think there's no evidence of clear benefit. . . . I think the state of the evidence has been well summarized in the reports of the Surgeon General, IARC, the Institute of Medicine, each group that's looked at the question of whether today's lower yield cigarettes are likely to produce -- are likely to produce lower risk of lung cancer, has said, you know, no clear benefit.

¶Samet TT, 11/29/04, 1168:12-16; 1169:2-18.

Echoing the conclusions of Monograph 13 and the 2004 Surgeon General's Report, a January 2004 article in the British Medical Journal reported on a study intended "to assess the risk of lung cancer in smokers of medium tar filter cigarettes compared with smokers of low tar and very low tar cigarettes":

There was no difference in risk among men who smoked brands rated as very low tar . . . or low tar . . . compared with those who smoked medium tar brands. The same was seen for women. . . . Men and women who smoked very low tar . . . and low tar . . . brands had risks

804

of lung cancer indistinguishable from those who smoked medium tar . . . brands. . . . Our finding that there was no difference in the risk of lung cancer between people who smoked medium tar filter, low tar filter, and very low tar filter cigarettes is consistent with evidence of compensatory smoking.

¶TLT1020160-0167 at 0160, 0164, 0166 (US 88622).

Two very large American Cancer Society's Cancer Prevention Studies ("CPS"), "conducted approximately 20 years apart," which show that lung cancer death rates have not gone down as a result of the introduction of low-tar cigarettes, provide powerful confirmatory evidence. CPS-I was conducted in the late 1950s and early 1960s, and CPS-II was conducted in the 1980s. Because of this 20 year gap, the smokers in the CPS-I study were smoking mostly high-tar, unfiltered cigarettes, and the "vast majority" of the smokers in the CPS-II study were smoking filtered cigarettes with much lower machine-measured tar and nicotine yields. CPS-I and CPS-II are "the two largest studies of smoking and disease risks;" they included "over a million men and women each," and "followed those individuals for"¹²¹² While the Court has not relied upon certain portions of Dr. Brandt's testimony, a careful review of all of Dr. Brandt's testimony, the documents he cited, and the documents which Defendants cited to rebut his testimony reveals that many of Defendants' characterizations of both the testimony and individual documents are out of context, unrepresentative, and unfair. -232- to 18 years. The results of these studies showed that, "[d]espite the substantive reduction in tar yield of the cigarettes smoked in CPS-II, lung cancer disease risks increased rather than decreased in comparison to CPS-I." Burns WD, 1:10-15, 12:10- 11, 33:18-34:6, 35:14-35:9; Burns TT, 2/15/05, 13311:9-15, 13313:9-13314:6, 13316:19-13317:15, 13322:9-10.
Despite the dramatic shift to filtered and "light" cigarettes in the last 50 years, the effect the public health community was expecting to see from a change in the type of cigarettes smoked in the U.S never materialized. To the contrary, health risks increased significantly. With reference to the CPS-I and CPS-II studies, Dr. Burns explained:

805

For males, when you look at the risk of smoking, you see that it just about doubled between CPS-I and CPS-II. . . . For females, it went up almost fourfold. . . . [E]ven after adjustment for differences in number of cigarettes smoked and the duration of smoking, the rates increased for males and increased for females between these two studies that were conducted over a period of time when there was approximately a 50 to 60 percent decline in the tar value of the cigarettes being smoked and a dramatic increase in the number of smokers who were smoking filtered cigarettes. So, instead of seeing a reduction in the risk of smoking with the introduction of these products, we have seen the risk of smoking actually increase over that interval. . . . we had watched and waited for the decline in lung cancer to occur. It did not.

¶Burns TT, 2/15/05, 13322:22-13323:22, 13325:6-13326:22; (no bates) (U.S. 17802) (depicting the

¶"sales-weighted tar and nicotine values for U.S. cigarettes" over time); (no bates) (U.S. 17803)

¶(depicting data from the CPS-I study); (no bates) U.S. 17804 (comparison of lung cancer risk for nonsmokers and smokers based on CPS-I and CPS-II data); (no bates) (U.S. 17806) (graph showing that risk of lung cancer from smoking has not declined, notwithstanding the drastic shift to filtered cigarettes and those with lower machine-measured tar and nicotine deliveries).

As Dr. Samet explained, "[i]f there were substantial benefits of the change in tar yield over the 20 years between [CPS-I in the 1960s and CPS-II in the 1980s], we would expect lower relative risks; instead they increased." Dr. Samet also explained that the British Physician's study, which was conducted over two 20-year periods, also "shows that the relative risk values [for lung cancer] have gone up comparing the first 20 years (1951-1971) to the second 20 years (1972-1991)."

¶Samet WD, 154:21-156:4, 156:5-12, 157:20-158:15; VXA1601833-1843 (US 64058) (published article on British Physician's Study, cited in Samet Written Direct).

Even if it were true that lower tar cigarettes result in some minor incremental reduction in tar, they do not provide any meaningful health benefit relative to higher tar cigarettes

806

¶"from the perspective of human exposure . . . or meaningful exposure." Henningfield TT, 11/3/04,

¶7295:2-7298:2. As Dr. Henningfield explained:

It would be a little bit like low fat cheese has 100 grams, lets say, of X, and then there is another type that gave you 98 grams, and you could say, yes, that's lower, and the next lower one is 97 grams, but that's a meaningless difference, even though it's accurate and reliable by a machine test and you can say it does go down, it's a meaningless difference. . . . In this case, what Benowitz['s] study, and then many other studies showed, is that if you looked at actual intake of people, there was virtually no difference at all in intake. And it wasn't just that the ranking was off a little bit, it was that, for example, the Marlboro Light, according to the Massachusetts data can get – give you about three times as much nicotine as it was rated, and more than twice as much as the Marlboro regular, so it's off by several orders of magnitude, but most importantly, it is just -- if [a] consumer says, okay, I want to get lower tar and nicotine and they pick a light versus a regular, they're not getting biologically meaningful lower tar and nicotine. . . . There is no meaningful difference in exposure to people. . . . what U.K. realizes now, what U.S. realizes, the World Health Organization realizes is that it is still a meaningless difference. And that's why . . . the resistance to even using the . . . label "light" cigarettes.

¶Id.; see also Henningfield TT, 12/1/04, 7535:12-7536:23 (explaining that the "light cigarette debacle" centers around the fact that the historical reduction in machine-measured tar "was biologically meaningless," citing NCI Monograph 13).

Dr. Samet echoed these conclusions, stating that "while some earlier studies suggested a modest benefit in terms of lung cancer risk, late, more recent evidence suggests otherwise, namely that there is no benefit." He pointed out that, even excluding the more recent evidence and postulating some reduction in risk, the "overall risk [of smoking these cigarettes] is so high that even a small reduction is of no public health or medical significance." Samet WD, 170:11-23.

807

Notwithstanding the widespread acceptance and independent validation in the scientific community of Monograph 13 and its conclusions, Defendants tried to challenge its conclusions and methodology, including the selection of its contributors. Defendants criticized the authors of Monograph 13 for not acknowledging the few scientists who expressed skepticism about its results. In addition, Defendants challenged NCI's selection of contributors to the Monograph, namely, that the scientists chosen were biased in favor of concluding that low tar cigarettes yield no clear benefit. However, as Dr. Burns noted, many of the contributors to Monograph 13 "represent some of the more distinguished scientists and experts on this issue in the country" and:

[T]he consensus statement of the organization, that is the NCI. . . . prevents individual opinion from being presented as the consensus of scientific thought. . . . The way that you know that [collective biases are not influencing the final document] is by taking it through a series of reviews by the governmental organization. . . . The organization that then produces the volume and puts its seal of consensus approval on it, that is the National Cancer Institute, that undergoes a series of reviews that it takes to ensure that the data contained in the volume are scientifically accurate and represent the consensus of scientific thought.

¶Burns TT, 2/15/05, 13383:1-13389:4.

Defendants' scientists were not involved with the production of Monograph 13 because:

At the time which this was undertaken, the tobacco industry's position was still that . . . there were no disease risks that were causally associated with cigarette smoking. . . . For that reason, the tobacco industry has not been included in the Surgeon General's Report process and various other processes because they weren't part of the consensus of scientific thought at that point in time. They were perceived as adopting positions that had so little scientific credibility that they could not be meaningfully utilized in the formation of a consensus.

808

¶Burns TT, 2/15/05, 13389:11-13391:21 (explaining also that Monograph 7, in which RJR scientists participated, "was not a consensus document, it was simply the results of the proceedings of a meeting. Those proceedings come out under the author's name, are understood to be the individual opinions of the authors. That was not what was being requested with Monograph 13").

The Court finds that the testimony of Dr. Burns is totally credible and persuasive on each of the issues which he discussed, including low tar cigarettes and their relative health effects. Dr. Burns, qualified by the Court without objection as an expert in "[t]he science of tobacco and health, including disease causation," is a medical doctor and professor of family and preventive medicine with 30 years' experience studying the health consequences of smoking. Dr. Burns has extensive experience, including as a teacher, in various medical areas, including smoking and health, and has studied epidemiology, addiction, nicotine, cigarette design and ETS in the context of smoking and health. He has over 200 publications, most of which are peer-reviewed, in the area of smoking and health, including chapters in two of the leading medical textbooks. Furthermore, Dr. Burns has studied and taught extensively in the area of lung diseases, including lung cancer, and has personally participated in the treatment of thousands of patients with lung disease. In addition, he has served as "an author, editor or reviewer for each of the annual Reports of the U.S. Surgeon General on the Health Consequences of Smoking since 1975," and has served as contributing author and Senior Scientific Editor for several of the National Cancer Institute Tobacco Control Monographs. Dr. Burns has also received numerous award and honors for his work in the area of smoking and health, including the Surgeon General's Medallion. Finally, Dr. Burns's demeanor during his testimony, which spanned nearly two full days and consisted mostly of cross-examination, further demonstrated his credibility. He fully answered the questions posed to him by well-prepared counsel, he was totally versed in the complex scientific areas about which he was questioned, he was neither evasive nor combative, and demonstrated an enormous familiarity with the science of smoking and lung cancer, the history of the issue in this country as scientists learned more and more about the subject, and the manner in which preparation of the Surgeon General's Reports represented the most up-to-date scientific consensus on the topic being studied. Burns WD, 1:3-5; 1:10-15;

809

¶12:10-11; (no bates) (US 78526).

For the reasons set forth at length in previous Sections, the Court also finds that the testimony of Drs. Samet, Henningfield, Benowitz, and Farone is also highly credible and persuasive.
To rebut the compelling testimony of Drs. Burns, Samet, Henningfield, Benowitz, and Farone, Defendants called a statistician, William Wecker, who, by his own admission, has

¶"never been qualified by a Court as an expert in the subject of smoking and health."27 Wecker TT,

¶Dr. Jerry Whidby and Dr. Joseph Mulholland also offered testimony criticizing Monograph 13. Their testimony was neither reliable nor persuasive. Called to testify by the United States as an adverse witness, Dr. Whidby admitted that he was a paid consultant for Philip Morris and would earn $2,800 per day (total amount $14,000 for five days) for preparing and delivering his fact testimony; and that his agreement with Philip Morris obligates him to testify on its behalf in litigation, and does not leave him the option of not testifying.

¶Dr. Whidby was repeatedly impeached with prior inconsistent statements during his live examination. In some instances, portions of Dr. Whidby's Corrected Written Direct Testimony were in conflict with each other. Whidby WD, 2:5-3:15 (addressing consultancy agreement); Whidby TT, 2/22/05, 13942:23-13943:17 (same); Whidby TT, 2/22/05, 13945:23-13947:5 (impeached with prior inconsistent testimony relating to whether increased ventilation has an impact on biological activity); Whidby TT, 2/22/05, 13947:6-13948:10 (impeached with prior inconsistent testimony relating to whether witness had any empirical evidence that Marlboro Reds sold today are any less harmful than those sold in 1960); Whidby TT, 2/22/05, 13948:8-13949:18 (impeached with prior inconsistent testimony on whether witness had any evidence that Philip Morris's general reduction and selective reduction techniques have led to one less case of lung cancer in the United States); Whidby TT, 2/22/05, 13952:23-13954:3 (impeached with prior inconsistent testimony on whether witness had any empirical evidence that Philip Morris's general and selective reduction techniques have in fact made its cigarettes less harmful).

810

¶3/15/05, 15650:5-7. Because he is a statistician, and neither an epidemiologist nor a medical doctor, Dr. Wecker is "not able to offer opinions as to causation" relating to the relative health effects of low tar cigarettes. Wecker TT, 3/15/05, 15666:6-15667:2, 15670:20-15671:8. Moreover, Dr. Wecker's statistical analyses are unconvincing because they are flawed in several ways.

First, Dr. Wecker's core opinion -- that smokers who switch to lower delivery cigarettes do not increase the number of cigarettes per day that they smoke -- is flatly contradicted by Defendants' voluminous research reports and other documents, spanning decades, which demonstrate that smokers who switch to lower deliveries do smoke more cigarettes per day. 1000861953-1953 (US 35484) (1961 PM); 2022244449-4450 at 4449-4450 (US 36855) (1970 PM); 1000350158-0188 at 0161-0162 (US 20176) (1971 PM); 1003285403-5416 at 5403 (US 20159) (1972 PM); 1003285403-5416 at 5403 (US 20159) (same); 1003293476-3493 at 3480 (US 85073) (1974 PM); 2040066740-6766 at 6750-6751, 6754-6755 (US 20435) (1979 PM); 2071376833-6834 at 6833 (US 27273) (1992 PM); 501525355-5366 at 5360-5361 (US 29531) (1974-1976 RJR); 679009843-9867 at 9843 (US 85055) (1977 B&W); 775036039-6067 at 6050 (US 21053) (same);

811

¶536000000-0090 at 0050 (US 22338) (1984 B&W); 760008596-8803 at 8760 (US 54588) (Confidential) (1998 BAT); 00044522-4523 at 4522 (US 22012) (1976 Lorillard).

Second, Dr. Wecker testified that, in forming his opinion on this issue: "I don't reach my opinion by weighing all the evidence, but mainly on my own statistical work replicating and correcting figure 4-5" of NCI Monograph 13. Wecker TT, 3/15/05, 15656:22-15657:14. Figure 4-5 illustrates one of the new statistical analyses of CPS-I data that were performed as part of the production of NCI Monograph 13 relating to change in number of cigarettes smoked per day by smokers switching down to obtain lower tar yield. However, the Conclusions of Chapter 4 of Monograph 13 did not rely on the new analyses. Burns WD, 30:13-18, 58:4-19; Burns TT, 2/16/05, 13498:1-9; DXA0310399-0650 at 0509-0510 (US 58700).
Dr. Burns explained that the issue of "increases in number of cigarettes per day smoked by those who switch to lower tar brands of cigarettes" is "not the only [reason and], for that matter, it's not the princip[al] one" for Monograph 13's conclusion that lower tar cigarettes provide no reduction in harm relative to higher tar cigarettes.
The first sentence in the Monograph under the heading for these new analyses of CPS-I data clearly states that the new analyses of cigarettes per day were inconclusive and, as such, were not the basis for Monograph 13's conclusions:

A reexamination of the CPS-I data set was inconclusive as to whether compensatory changes in the number of cigarettes smoked per day when smokers switched to a lower nicotine cigarette introduce a bias sufficient to explain the observed increased lung cancer risk among smokers of high yield cigarettes.

812

¶DXA0310399-0650 at 0509 (US 58700); Wecker TT, 3/15/05, 15683:5-15687:14. Consequently, Dr. Wecker's testimony, which relied on his critique of the new CPS-I analyses, was basically irrelevant since Monograph 13 did not rely on that data in reaching its conclusion.

Third, Dr. Wecker's analysis failed to refute, or even consider, the conclusions reached by several other prominent scientific bodies -- including the 2004 Surgeon General's Report, the 2002 publication of the Scientific Advisory Committee on Tobacco Regulation to the World Health Organization, and the 2001 report to the Canadian Minister of Health from the Ministerial Advisory Counsel on Tobacco Control in Canada -- which all reached the same conclusion as did the NCI in Monograph 13, namely that lower tar cigarettes provide no significant reduction in lung cancer or other health benefit -- in direct contrast to Dr. Wecker's conclusions. Wecker TT, 3/15/05, 15661:11-15664:23. Since Dr. Wecker was only a statistician and not qualified to address the subject of smoking and health, he could not have addressed the substance of NCI's scientific conclusions in Monograph 13.
Fourth, Dr. Wecker made several changes to the analyses in Monograph 13 that reflected his lack of understanding of the relevant subject matter. For instance, he included women in his analysis of Figure 4-5 and was unaware that the contributors to Monograph 13 excluded women because the later increase in smoking prevalence among women resulted in the lack of a valid baseline dose-response relationship. Wecker TT, 3/15/05, 15691:18-15695:11; Burns TT, 2/15/05, 13323:23-13325:4 (explaining the temporal differences in the rise of prevalence of male and female smoking, which led to exclusion of women in graph); 13342:8-20 (explaining the gender differences in prevalence in the United States and France).

813

Fifth, Dr. Wecker acknowledged that many of his reanalyses of Figure 4-5 showed no statistically significant difference in risk for smokers at the various tar levels, which is entirely consistent with the conclusion of the authors of Monograph 13, quoted above, that the results on this issue were "inconclusive." Burns TT, 2/16/05, 13526:4-6; Wecker TT, 3/15/05, 15687:15-15691:17.
Sixth, Dr. Burns explained that studies done by Garfinkel et al. reported in the 1980 Banbury Product Liability and Health Risks report, which Dr. Wecker claimed were "consistent with [his] conclusions," were not "consistent with Dr. Wecker's findings . . . because it's a different analysis." Dr. Burns also noted: "We cited both of the [Garfinkel] studies that examined cigarettes per day in Monograph 13. We looked at them carefully." Burns TT, 2/16/05, 13517:11-13518:23; Wecker WD, 31:10-14.
Dr. Samet explained that comparing Dr. Garfinkel's calculations, published more than 20 years ago, to those performed for Monograph 13, is a case of comparing apples to oranges, as the analyses sought to answer markedly different questions:

They're very different. . . . Dr. Garfinkel said: If people changed their brand, did they smoke more, the same or less? Just those three bins [categories], if you will. What [the Monograph 13] analysis says is: Let's look at whether there's a relationship between the reported numbers of cigarettes smoked on the first brand and on the second brand and the difference in nicotine yield on the first brand and the second brand. So [the Monograph 13 analysis] is a more quantitative analysis and Dr. Garfinkel's was more sort of: Did people change from smoking more than they used to [or] to less than they used to?

¶Samet TT, 9/29/04, 1185:11-1186:19; 1191:12-16; 1193:8-1195:11.

Dr. Wecker maintained that examining lower tar and higher tar smokers' lung cancer rates, without controlling for cigarettes per day, provides "other empirical support" for his claim that epidemiological studies of the relative health effects of low tar cigarettes are not biased by controlling for cigarettes per day. Wecker WD, 49:5-11. However, Dr. Burns explained that neglecting to control for cigarettes per day fails to resolve "the core issue" necessary to have a reliable analysis, because it "introduce[s] other biases that are equally important":

814

The core issue is having comparable groups, that is, having groups [where] you can adjust for those characteristics other than the type of cigarettes that they smoke. If you include numbers of cigarettes for controlling, you over control. If you don't, then you're left with the two populations likely to have differences in intensity of smoking and no method by which you can control for that difference. Then you can't tell whether the difference you're seeing is due to a difference in intensity of smoking, i.e. completely independent of the type of cigarette or is a characteristic of the cigarette that you've chosen to smoke.

¶Burns TT, 2/16/05, 13661:14-13662:6; see also Burns TT, 2/16/05, 13484:18-13485:10; 13486:8-21 ("If you don't control for cigarettes per day, you have two different groups of individuals who have different intensities of smoking and, therefore, you can't compare their exposures without looking at intensity.").

Finally, Dr. Wecker's analysis of actual versus predicted death rates was not done in the manner utilized in Monograph 13. He admitted on cross-examination that there were several discrepancies between his analysis and that described in Monograph 13. The discrepancies included the fact that Dr. Wecker used mortality data, taken from the National Health Interview Survey, from 1993 to 2000, that was not contained in the calculations in Monograph 13. Again, Dr. Wecker was unaware of a critical fact, namely that the contributors to Monograph 13 specifically excluded data after 1992 because the National Health Interview Survey "changed the definition of smoker in 1992," making the post-1992 data inconsistent with the pre-1992 data. Wecker WD, 51:3-21; Wecker TT, 3/15/05, 15698:14-15701:18.

815

Dr. Wecker also acknowledged that, while Monograph 13 discussed a chart examining lung cancer death rates at ages under 50, he attempted to create the chart described in the Monograph but only included people aged 40-50, excluding death rates for all ages under 40. Wecker TT, 3/15/05,15702:5-15706:2.
Dr. Burns testified, upon review of a chart based on Dr. Wecker's calculations of actual versus predicted death rates, purportedly based on Monograph 13: "The data presented in this graph have essentially no meaning whatsoever. This is not the analysis that Sir Richard Doll was suggesting be done and it is not an analysis that has any valid, scientific or technical meaning." Burns TT, 2/16/05, 13666:20-23; see also Burns TT, 2/16/05, 13664:10-24 (testifying that, to his knowledge, the chart has never "appeared in published peer reviewed literature").
In sum, there is an overwhelming consensus in the public health and scientific community, both here and abroad, that low tar cigarettes offer no clear health benefit to smokers, have not reduced the risk of lung cancer and heart disease for smokers using them, and have not produced any decrease in the incidence of lung cancer. Moreover, because of the misleading nature of the advertising for low tar cigarettes, smokers who might have quit have refrained from doing so in the belief that such cigarettes reduced their health risks.

816

Based on Their Sophisticated Understanding of Compensation, Defendants Internally Recognized that Low Tar/Light Cigarettes Offer No Clear Health Benefit28

a. Defendants Internally Recognized that Low Tar Cigarettes Are Not Less Harmful Than Full-Flavor Cigarettes

¶(1) Philip Morris

A March 1, 1977 Philip Morris memorandum by industry-funded scientist Stanley

¶Schachter to Thomas Osdene, Director of Research, concluded that low tar/low nicotine cigarettes are not less harmful:

[I]t would certainly seem that the campaign for low nicotine cigarettes is misguided and rests on a set of fallacious premises. . . . The question is crucial and particularly so in light of . . . Ross's evidence that carbon monoxide, hydrogen cyanide, and nitrogen oxide delivery is considerably greater in most of the popular brands of low nicotine filter, [sic] cigarettes than in high nicotine, non-filter cigarettes. . . . It is . . . clear . . . that the major body of data that has been used to justify the campaign for low nicotine cigarettes does nothing of the sort.

¶1000046626-6661 at 6655, 6660 (US 20074).

Dr. Farone stated that Philip Morris's Marlboro full-flavor and Marlboro Lights cigarettes are "essentially identical except for dilution" -- i.e., that Marlboro Lights have more dilution, dilution referring to ventilation that dilutes the smoke, particularly when machine-smoked by the FTC method, with ambient air. "[A]s you increase dilution, the toxicity in [the Ames] test increases, which is more likely than not associated with a toxicity increase in smokers." Farone TT,

¶10/7/04, 1888:2-1889:5; 1891:17-19.

817

In fact, Dr. Farone explained that the very Ames mutagenicity testing that Philip Morris has conducted for the past 25 years, and that "Philip Morris has concluded . . . predicts carcinogenicity" has indicated that Philip Morris's Marlboro Lights cigarettes are, as designed, more mutagenic than Marlboro full-flavor cigarettes:

[I]n the case of Marlboro Lights, the Philip Morris test data that I have reviewed on that level of dilution for equivalent blends indicated that the product design for their Light cigarettes was more mutagenic than the full flavor Marlboro, Marlboro Reds, and therefore predictive of more potential cancer risk. These studies were repeated multiple times over the past 20 years and continue to be repeated to this day. The Philip Morris data, as was used by Philip Morris, was a strong warning that their product design change between a Marlboro Red and a Marlboro Light -- increased ventilation -- resulted in a potentially more dangerous product.

¶Farone WD, 119:7-120:15; Farone TT, 10/7/04, 1866:2-17. Philip Morris has not "changed the design of 'Light' cigarettes in response to its studies and knowledge concerning mutagenicity." Farone WD, 121:3-9.

Philip Morris consultant and former employee Dr. Whidby agreed with Dr. Farone's basic analysis and acknowledged that "increased filter dilution," one of the techniques Philip Morris uses to lower the FTC tar yield of its cigarettes, "is associated with increased biological activity." Dr. Whidby explained that biological activity in the context of Philip Morris's biochemical testing reports generally refers to biological reactions such as tumor growth, cell mutations, and toxic reactions, and that it was "a bad thing" that should be reduced. Whidby TT, 2/22/05, 13964:18-25; 13967:22-13968:19. Helmut Wakeham, of Philip Morris, acknowledged the same phenomenon as far back as March 1, 1974. 1003293476-3493 at 3492 (US 85073).

818

A November 1977 Philip Morris memorandum to Dr. Robert Pages from J. Booker and S. Drew about Ames testing stated: "The take home lesson from this experiment is that dilution of a cigarette appears to increase the activity of the WSC [whole smoke condensate] (more dramatically for some cigarettes than for others)." 1002978361-8363 at 8362 (U.S. 35635).
By 1978, Philip Morris had substantial evidence that "filter dilution [which Philip Morris used to reduce FTC tar and nicotine yields] was somehow acting to increase" the "activity" of the whole smoke condensate ("WSC") collected from its cigarettes. Further experiments confirmed that the tar from ventilated low tar reference cigarettes, i.e., cigarettes used for research purposes and not actually sold in stores, measured higher on mutagenicity tests than non-ventilated products. Additional research conducted in 1979 yielded the same result. 2001243600-3673 at 3610-11 (US 20298); accord 2022180219 (US 21479).
A May 11, 1982 Philip Morris document from INBIFO (Philip Morris's overseas research facility in Switzerland) revealed that Philip Morris learned from its testing of low tar reference laboratory cigarettes in Europe that these cigarettes registered higher in standard biological tests than the full-flavor delivery reference cigarettes -- i.e., were "more active" -- and thus were more likely to cause cancer: "Low tar reference cigarette . . . [m]ay be slightly more active than [the regular delivery reference cigarette] as a complete carcinogen." 1003121638-1643 at 1638 (US 20153).
A January 28, 1994 report from INBIFO to Philip Morris in Richmond, Virginia stated that increased cigarette filtration, porosity, and ventilation (primary methods used by Philip Morris to reduce the FTC Method tar and nicotine yields in its cigarettes) would result in an increase in the degree to which cigarette smoke was toxic to living cells (i.e., cytotoxicity), the irritation it caused to smokers, and the likelihood that the smoke would generate mutations such as tumors and/or cancer (i.e., mutagenicity). The document stated: "Increased filtration will result in a relative enrichment of gas phase constituents, leading to increased cytotoxicity and irritancy. . . . Increased porosity and ventilation will . . . increase the specific mutagenicity." 2024005509-5512 at 5509-5510 (US 20399); Farone WD, 122:1-14 (citing to, and agreeing with, INBIFO conclusions).

819

In this case, A. Clifton Lilly, Senior Vice President of Technology, confirmed that data from tests run at Philip Morris's INBIFO facility showed that the Ames test for mutagenicity from Marlboro Lights produces significantly higher results than the tar from Marlboro full flavor products. 2001243600-3673 at 3610-11 (US 20298); 2022180219-0219 (US 21479); 1000135419- 5439 (US 20078); Lilly PD, United States v. Philip Morris, 5/14/02, 229:17-231:21.
A 2001 document about Ames mutagenicity testing from Philip Morris's INBIFO laboratory in Germany demonstrated that, in every case, the mutagenicity of Marlboro Lights is higher than the mutagenicity of Marlboro full-flavor. 2505913831-3836 (US 46079).
James Morgan, former President and CEO of Philip Morris, conceded in 2002 that, in his opinion, lower tar cigarettes are not any safer than higher tar cigarettes. Morgan PD, Price v. Philip Morris, Inc., 6/5/02, 75:3-15.
According to Nancy Brennan-Lund, Philip Morris Senior Vice President of Marketing, "what we say on our web site we believe to be true." Philip Morris's position is that low tar cigarettes are no less harmful than full-flavor cigarettes, "based on what the Monograph 13 came out with." Lund later qualified her statement: it has "not been proven" that light cigarettes are less harmful, so one cannot assume they are less harmful. Brennan-Lund PD, Price v. Philip Morris, Inc., 9/20/02, 107:22-108:14; 109:16-110:22; 114:9-114:10.

820

Ellen Merlo, then Philip Morris USA Senior Vice President of Corporate Affairs, agreed that in 2002 that Philip Morris's policy at the time was that lights or low tar cigarettes are not safe or safer than any other cigarettes. Merlo PD, Price v. Philip Morris, Inc., 10/2/02, 80:3-80:15.

(2) R.J. Reynolds 2159. RJR's internal documents show that it, like the other Defendants, has long known that

¶it has evidence that low tar cigarettes are no safer than regular cigarettes.

In May 1980, RJR scientist C.T. Mansfield performed the Ames test for mutagenicity "on the tars from twenty-four domestic brands of cigarettes with various [FTC] 'tar' deliveries," and found "a trend for low 'tar' cigarettes to show higher numbers per mg [of] 'tar,'" indicating that the low 'tar' cigarettes caused more mutations. 514903578-3610 at 3579 (US 20863).
A September 29, 1992 RJR internal presentation reported that lower tar cigarettes were more likely to cause mutations such as tumors and cancer than higher tar cigarettes. The presentation stated: "Higher tar cigarettes tend to have lower Ames activity . . . than lower tar cigarettes." 509643825-3832 at 3825 (US 20830).
In 2003, Arnold Mosberg, an RJR scientist, and other RJR scientists (Doolittle and Morgan) reviewed "data [they] have had for decades" (some for more than two decades) to conduct a comparison of the relative harmfulness of lights and full flavor cigarettes, using various tests, including animal skin painting tumorgenicity, rodent inhalation, and Ames mutagenicity studies. The results of these studies indicated that low tar cigarettes do not reduce risk relative to full-flavor cigarettes. Mosberg PD, Turner v. R.J. Reynolds, 8/19/03, 2:12-16, 6:7-12:4, 14:21-15:19, 15:24- 16:1, 16:13-24, 19:3-5, 21:20-22:10, 22:19-26:8, 55:1-11, 57:15-58:16, 97:15-100:12, 103:23-104:1 (discussing in part Deposition Exhibit 2).

821

A February 4, 1976 memorandum from Ernest Pepples, B&W Senior Vice President, titled "Industry Response to Cigarette/Health Controversy," reveals Defendants' knowledge that the low tar and filter cigarettes they were marketing as less harmful were not producing less tar and less nicotine to the smoker and were not likely to actually be less harmful:

The industry has moved strongly toward filtered cigarettes, which have increased from 0.6% in 1950 to 87% in 1975. . . . This became known as the 'tar derby' of the late 1950's. It was characterized by sharply intensified advertising competition. . . . The new filter brands vying for a piece of the growing filter market made extraordinary claims. . . . It was important to have the most filter traps. Some claimed to possess the least tars. In most cases, however, the smoker of a filter cigarette was getting as much or more nicotine and tar as he would have gotten from a regular cigarette. He had abandoned the regular cigarette, however, on the ground of reduced risk to health. . . . The manufacturers' marketing strategy has been to overcome and even to make marketing use of the smoking/health connection. . . . Thus the 'tar derby' in the United States resulted from industry efforts to cater to the public's concern and to attract consumers to the new filtered brands. . . . The current duel between True and Vantage and between Carlton and Now are other examples of competitive efforts to capitalize on the smoking/health controversy.

¶170042567-2574 at 2568, 2574 (US 20292); Smith WD, 79:5-22.

An August 5, 1980 B&W document signed by J. Kendrick Wells III, B&W Assistant

¶General Counsel, acknowledged that "[t]here was question about the degree of support . . . at the present time" for the "scientific opinion that certain low levels of 'tar' consumption are relatively safe to the smoker," and that "for the longer term the support may be quickly eroding." 680050983-1001

822

An October 31, 1989 B&W internal memorandum, titled "Objections to Product Innovation Strategy," from Wells to RJ Pritchard, B&W executive and member of the Tobacco Institute's Executive Committee, conceded that "it is not established that the reduction or removal of specific smoke constituents or of smoke constituents across the board, such as in low tar cigarettes, is significant for smoking and health." 680701034-1038 at 1035 (US 21010); Wells WD, 60:3-61:10.
Sharon Blackie Boyse, Director of Scientific Communications and a spokesperson for B&W on scientific issues as late as 1998, acknowledged that, "based on [her] experience as an employee within the cigarette industry, [she has] been aware for some time that some smokers believe that low tar cigarettes are less hazardous to their health [and that] some smokers believe that by switching to low tar cigarettes, they will achieve a health benefit." From at least as early as 1998, B&W acknowledged that the company "did not know whether low tar cigarettes were, in fact, less hazardous," and "was not confident that the science showed any health benefit from low tar cigarettes." Blackie WD, 184:22-185:5.
As of 2005, B&W's website admits that low tar cigarettes are not safer than regular cigarettes. It states that "despite a dramatic lessening of tar yields, the hoped-for reduction of smoking-related illnesses has not been conclusively demonstrated." Furthermore, the website directs the reader to the National Cancer Institute's Monograph 13, citing its conclusions that "[e]pidemiological and other scientific evidence, including patterns of mortality from smoking-caused diseases, does not indicate a benefit to public health from changes in cigarette design and manufacturing over the last fifty years," and that "[w]idespread adoption of lower-yield cigarettes in the United States has not prevented the sustained increase in lung cancer among older smokers." The website further states: "[W]e continue to believe that smokers should rely on the public health authorities' views on low tar cigarettes and other smoking issues." Ivey WD, 63:9-16; (no bates) (US 86656).

823

(4) BATCo 2168. A 1976 BATCo document from S.J. Green to P.L. Short and P. Sheehy revealed both

¶that BATCo planned to market low tar cigarettes as safer and that BATCo did not have a sufficient basis to believe that low tar cigarettes were safer, stating: "Before we do work aimed to sell low delivery cigarettes, unless we are already satisfied, we should do some work to establish that in fact they are safer." 110076428-6432 at 6430 (US 34957).

A June 9, 1982 BATCo document, "Technical Exchange Meeting," noted that Ames testing revealed that "[t]he specific activity [a measure of mutagenicity] of a plain cigarette was found to be lower than that of a ventilated filter cigarette." 109883189-3192 at 3191 (US 20265).
A February 18, 1988 BATCo study of cigarette mutagenicity from the B&W Research & Development Library's E.D. Massey found that the "lighter" the purported delivery of the cigarette, the higher the mutagenicity. Using Philip Morris cigarettes as an example, Merit cigarettes had higher mutagenicity than Marlboro Lights, which in turn had more mutagenicity than regular Marlboro cigarettes. 620000021-0032 at 0027,³⁰³⁰ In response to the overwhelming weight of scientific authority finding that (continued...) -1234- (US 20944).

(5) Lorillard 2171. Asked whether low tar/low nicotine cigarettes are any safer than conventional,

¶full-flavor cigarettes, Christopher Coggins, Senior Vice President of Science and Technology at Lorillard, stated: "[O]ur policy is that cigarettes can cause cancer and that goes for all cigarettes." Coggins PD, United States v. Philip Morris, 8/16/01, 115:22-116:4.

824

¶(6) Liggett

Comments by Liggett scientists on a "Memorandum of June 13, 1966; C.F. Woodward and C.L. Ogg to P.A. Wells" acknowledged that there was no basis to conclude that reductions in tar and nicotine and/or the use of filters reduced the harmfulness of cigarettes:

Although the public does have a right to know what it is buying, extreme care must be exercised to avoid leading a non-technically oriented public to erroneous conclusions regarding the relative merits of one brand versus another based on minimal differences in "tar" or nicotine -- to neither of which can be attached any quantitative measure of health hazard. . . . We would question if any differences between any filter brands would show correlation with tumorgenicity. We know of no correlation of tar delivery among filter brands with tumor production in mice -- or for that matter, even among non-filter cigarettes. The level of uncertainty in current biological testing is so great that distinction between cigarettes is not possible on this basis.

¶LWDOJ00068944-8949 at 8944-45 (US 21214) (emphasis in original).

b. Internally, Defendants Had an Extensive and Sophisticated Understanding of Smoker Compensation

Defendants have known since at least the 1950s that the central component that drives the smoking habit is nicotine, an addictive substance. Accordingly, Defendants also have long been aware that the reason people smoke cigarettes is to obtain a sufficient "dose" of nicotine to sustain their addiction. 1003287880-7890 at 7884 (US 20163); accord 500380562-0564 (US 20630);

¶100515899-5910 (US 20230); 1003285403-5416 (US 20159); 500917468-7476 at 7474-76 (US

¶20660); 105553905-3914 (US 34799).

Defendants also have known since the 1960s and 1970s that, because smokers smoke to obtain the desired effects of nicotine, smokers of lower-yield cigarettes tend to adjust their smoking behavior to titrate (i.e., control) their nicotine intake of nicotine to achieve the necessary levels of nicotine. That adjustment or titration of nicotine levels is called compensation. Defendants' internal understanding of compensation was decades ahead of that of employees and scientists of the Government and the scientific community. See Section V(B)(2)(b), supra. According to Dr. William Farone, Philip Morris employee from 1976 to 1984, who served as Director of Applied Research, and was accepted as an expert in "the chemistry and biochemistry of alkaloids and addictive drugs, the chemistry and physics of cigarette smoke, cigarette design and technology, and the chemistry and biochemistry of toxic substances and how they interact with living systems," during his employment at Philip Morris, the company had "a greater understanding of compensation than the outside scientific community," and, in his expert opinion, "the same is true for the other tobacco company Defendants." In 1966, when the FTC was considering the FTC Method, Defendants knew "that smokers smoked for nicotine" and "that smokers alter their smoking behavior to get nicotine." Farone WD, 2:2-8, 2:15-19, 117:15-118:8; Farone TT, 10/12/04, 2169:18-22, 2170:5-11, 2171:25-2172:8, 2182:11-2190:7; Wigand WD, 8:11-17; 120:5-17.

825

When Dr. Farone was Director of Applied Research, Philip Morris's own research found that "if we adjusted the design to reduce the nicotine delivery, or if people were given a cigarette of lower nicotine delivery than their usual brand, smokers would 'compensate' -- change how they smoked -- to get the amount of nicotine they need." Farone WD, 102:2-14; see also Farone WD, 104:7-15 (testifying that he knows Philip Morris was aware of compensation for nicotine "[f]rom conversations that I had with many of my colleagues at Philip Morris while I was working there, including people working under Dr. Dunn in his behavioral research group," and that this knowledge "is evident from the company's own documents").

826

In a March 24, 1961 Philip Morris memorandum from Wakeham to Hugh Cullman,

¶"Trends of Tar and Nicotine Deliveries over the last 5 Years," Wakeham stated: "As we know, all too often the smoker who switches to a hi-fi cigarette winds up smoking more units in order to provide himself with the delivery which he had before." 1000861953-1953 (US 35484); see also

¶Farone WD, 104:16-105:9 ("[T]his research into and understanding of compensation influence[d] how Philip Morris designed cigarettes").

As Philip Morris marketing researcher Myron E. Johnston noted in a June 1966 Philip

¶Morris report titled "Market Potential for a Health Cigarette":

[A]ny health cigarette must compromise between health implications on the one hand and flavor and nicotine on the other. It seems clear from the performance of existing health cigarette entries that flavor and nicotine are both necessary to sell a cigarette. A cigarette that does not deliver nicotine cannot satisfy the habituated smoker and cannot lead to habituation, and would therefore almost certainly fail.

¶1001913853-3878 at 3860 (US 20123).

A July 28, 1967 Philip Morris USA memorandum from W.L. Dunn, Jr., then

¶Associate Principal Scientist, to R.B. Seligman, Director of Development, discussed ventilation holes and compensation:

An earlier study (Memo of June 27, 1967) established that lip contact with the tipping paper extended to 9.96 mm from the outer end of the tipping paper for the average smokers. Since the air dilution holes are located in a band from 8.0 to 9.7 mm from the outer end of the tipping paper, it follows that some of these holes are likely to be occluded under normal smoking conditions, whereas no occlusion is likely to occur when the cigarettes are machine smoked for analysis.

827

¶The memorandum also documents that "[s]mokers adjust puff intake in order to maintain TPM

¶[total particulate matter] and/or nicotine constancy." 1003295500-5502 at 5500, 5502 (US 88627).

An August 11, 1967 Philip Morris USA document from Helmut Wakeham, Vice

¶President of Corporate Research and Development, to Paul D. Smith, Vice President and General

¶Counsel, stated that human smokers increased their smoke intake when switching from non-filter to filter cigarettes:

Two tests conducted at Product Opinion Laboratories demonstrate that in smoking a dilution filter cigaret [sic], the smoker adjusts his puff to receive about the same amount of "undiluted" smoke in each case. . . . In the smoking machine the puff volume is constant so that with dilution the quantity of "equivalent undiluted smoke" delivered to the Cambridge filter is reduced. Not so with the human smoker who appears to adjust to the diluted smoke by taking a larger puff so that he still gets about the same amount of equivalent undiluted smoke. . . . The smoker is, thus, apparently defeating the purpose of dilution to give him less "smoke" per puff. He is certainly not performing like the standard smoking machine; and to this extent the smoking machine data appear to be erroneous and misleading. It has probably always been so for diluted smoke cigarettes, whether dilution is obtained by porous paper or holes in the filter.

¶1000322554-2555 (US 35224) (emphasis in original); see also Dr. Jerry Whidby WD, 17:16-19:11

¶(testifying that "Product Opinion Laboratories was a facility established by Philip Morris to evaluate smokers' reaction to the cigarette brands Philip Morris was selling, as well as to Philip Morris's prototype cigarettes," and that he is not "aware of any instance, at any time between when Dr. Wakeham wrote this document in 1967 and when [Dr. Whidby] left the company in 1998, in which

¶Philip Morris informed the American public directly of Wakeham's conclusions that the FTC tar and nicotine yields are apparently 'erroneous and misleading,'" and "dilution filter cigarettes generated lower FTC yields than non-dilution cigarettes, but delivered about the same amount of smoke to smokers").

828

Dr. Farone explained the extraordinary significance of Wakeham's statements in this document:

It shows that Philip Morris understood the puff compensation phenomenon. This document shows that by 1967, Philip Morris recognized that when you have dilution or ventilation, the mechanism for compensation is puff adjustment . . . this document [also] shows that Philip Morris knew in 1967 that human smokers compensated by increasing their smoke intake when switching from non-filter to filter cigarettes, and in doing so, smokers received the same amount of tar and nicotine from their filter cigarettes as from non-filter cigarettes. It also shows Wakeham's understanding that the FTC tar and nicotine yields for low tar cigarettes are erroneous and misleading.

¶Farone WD, 111:5-112:15.

Since roughly the mid-1970s, the "vast majority of the low tar and ultra-low tar cigarettes sold by Philip Morris in the United States . . . are dilution cigarettes." Whidby WD, 19:16- 18.
In an August 25, 1967 Report on Project 1600, William Dunn, Senior Scientist, outlined an additional study performed by Philip Morris on puffing, with findings that provided "further support to the postulate that smokers adjust puff intake in order to maintain constant smoke intake." 1003288337-8338 at 8337 (US 85049).
In a Fall 1969 speech, William Dunn reported to the Board of Philip Morris: "It would appear that smokers do modify their smoking habits in order to obtain a preferred [nicotine] intake level." 1003287880-7890 at 7884 (US 20163).

829

Helmut Wakeham presented a November 26, 1969 internal industry paper, titled "Smoker Psychology Research," to the Philip Morris Board of Directors stating:

This great variability among smokers results from the fact that a smoker tends to seek his own level of intake. Even while smoking a single cigaret [sic], he adjusts the volume of his puff as he goes down to the rod, compensating for the change in the density of the available smoke. . . . A smoker's intake level is determined by the smoker himself, not by the manufacturers of the cigarettes.

¶1000273741-3771 at 3748 (US 26080).

A September 2, 1970 Philip Morris memorandum from Ray Fagan to Wakeham confirmed Philip Morris's understanding that smokers compensated for lower deliveries by smoking more cigarettes:

In the last 15 years particulates in cigarette smoke have declined by 33%; however, the number of cigarettes per smoker has increased. Furthermore, experimental studies have shown that a smoker will increase the number of cigarets he smokes if the cigaret he is offered contains less particulates and less nicotine.

¶2022244449-4450 at 4449-4450 (US 36855); Farone WD, 110:6-14; 118:9-23.

A November 1971 Philip Morris Special Research Report written by Tom Schori also addressed compensation by increasing the number of cigarettes smoked. The last sentence of the abstract of this report stated: "These findings support the hypothesis that the smoker does have daily intake quotas for tar and/or nicotine and that he titrates his smoke intake to meet these quotas." 1000350158-0188 at 0161-0162 (US 20176).
A January 1972 document from Philip Morris Research Center, written by Tom Schori and William Dunn, reviewed Philip Morris's evidence indicating that smokers compensated when smoking brands supplying less nicotine in order to receive their "daily nicotine intake quota," stating:

830

Cigarette consumption rate, i.e., number of cigarettes smoked per day, was found to vary as a function of the nicotine delivery of these cigarettes. Specifically, as nicotine increased, cigarette consumption decreased. These findings support the notion that smokers develop a daily nicotine intake quota and that when smoking cigarettes differing in nicotine delivery from that which they are accustomed they tend to modify their consumption rate in order to maintain their normal quota. No support was found for the analogous notion of a daily tar intake quota, however.

¶1003285403-5416 at 5403 (US 20159); 2062951266-1279 at 1266 (US 39723); Farone WD, 111:1-4

¶(referring to US 20159); Henningfield WD, 94:14-21 (referring to US 39723 as "significant in terms of Defendants' knowledge and understanding of the addictiveness of nicotine from the 1970s").

A May 14, 1975 Philip Morris memorandum from William Dunn to Robert Seligman,

¶Vice President for Tobacco Science and Research, stated:

Underlying all of our work in this area is the conviction what the smoker gets in the way of smoke is independent of smoke concentration levels as delivered within the range of commercially available cigarettes. He has a variety of regulatory maneuvers at his disposal for accommodating supply to a fairly constant need [for nicotine]. To monitor all of these maneuvers simultaneously is a major objective of our behavioral research program.

¶1000024914-4920 at 4915 (US 26072).

An August 19, 1977 Special Report from the Philip Morris USA Research Center written by Barbara Goodman, a Research Scientist, described Philip Morris's "Human Smoker

¶Simulator," a mechanism the company used to replicate human smoking behavior. The Simulator recorded how smokers smoked particular cigarettes by measuring their puffing behavior, then played back the recording into a smoking machine so the machine could replicate -- and then measure -- the amount of smoke constituents and the chemical composition obtained from a cigarette when smoked the same way the human had smoked it. The document states: "The Smoker Simulator program has the instrumentation to measure those smoker variations that constitute a smoker's puffing profile and a programmable smoking machine to measure the resulting tar, nicotine, and water deliveries." 1003728025-8039 at 8027 (US 20179); Whidby WD, 21:19-28:9; Whidby TT, 2/22/05, 13989:19- 13991:6; see also 1003293476-3493 at 3484 (US 85073) ("The cassette tape impulses an electronic smoking machine which duplicates exactly the smoking behavior of a given individual with a given cigarette. Delivery of tar, nicotine, and other smoke components can then be determined at the same conditions.").

831

A Philip Morris Report dated July 20, 1981, written by Frank Gullotta and J.A. Jones and sent to William Dunn and 40 others, including Dr. Whidby, described Philip Morris's Human Smoker Simulator program as "a system . . . which permits relatively unobtrusive monitoring of a smoker's inhalation patterns outside the laboratory setting," and indicated that "the system's accuracy was highly satisfactory throughout the experiment" and had "a mean accuracy reading of 96% . . . for 70 experimental sessions." The Report also indicated:

A major barrier to investigations on smoke-laden inhalation patterns has been the lack of instrumentation which would accurately measure inhalation parameters, and yet be unobtrusive to the smoker. The system we have acquired breaks this barrier by permitting accurate and relatively unobtrusive monitoring of inhalation patterns under natural smoking conditions.

¶2025986350-6401 at 6352-6353, 6382 (US 87080).

The July 20, 1981 Report includes a number of statements reflecting Philip Morris's understanding of smoker compensation:

832

That roughly 20% or less of smokers take puffs equal to or smaller to those taken by the FTC Method machine smoking protocol.

"The varying puff sizes in turn also give increased deliveries above those of" the FTC Method.

"Smoker profile characteristics have been found to be affected by the cigarette design parameters to varying degrees."

"A cigarette designed such that it causes the smoker to take larger puffs than the compared model could easily be perceived as having more impact (desirable or undesirable as the case may be)"

Lists low resistance to draw and "high filter dilution" as the top two "physical cigarette designs that have the effect of increasing a smoker's puff volumes."

"In conclusion, when a smoker is presented with a cigarette other than his normal brand, it is possible to estimate the maximum flow rate to a certain degree. The puff duration will increase with increasing RTD [resistance to draw] and/or filter dilution. Since the volume is based on both flow and duration, the puff volume will change accordingly."

Test results showing that smokers of full-flavor, light and ultra light cigarettes all took larger puffs than in the FTC Method, and that the smokers' puffs were increasingly large for lower tar cigarettes, so that, for full-flavor smokers, "average tar deliveries were 45% higher than" the FTC yields, low tar smokers' "showed a higher rate of increase, 81%," and, for ultra light cigarettes, "[t]he average tar delivery" was "more than three times that of" the FTC yield.

¶1003728025-8039 at 8027, 8028, 8032, 8034, 8036, 8037 (US 20179); Whidby WD, 22:18-23:9,

¶23:17-21, 24:6-13, 25:14-18, 26:21-27:5, 27:15-28:9; see also 2025986350-6401 at 6353, 6384 (US

¶87080); Whidby WD, 53:10-55:21 ("Preliminary results suggest that inhalation patterns are modified in response to changes in the available nicotine in the cigarette smoked. . . . The results from studies which analyze blood plasma and urine nicotine concentrations . . . suggest that nicotine compensation is fairly complete.").

833

A September 17, 1975 Philip Morris document from Goodman to Leo F. Meyer,

¶Philip Morris Director of Research, reflecting results of Philip Morris's studies with its Human

¶Smoker Simulator, reported that, due to compensation, smokers got as much tar and nicotine from

¶Marlboro Lights as from full-flavor Marlboros:

Marlboro Lights cigarettes were not smoked like regular Marlboros. There were differences in the size and frequency of the puffs, with larger volumes taken on Marlboro Lights by both regular Marlboro Smokers and Marlboro Lights smokers. . . . The panelists smoked the cigarettes according to physical properties; i.e., the dilution and the lower RTD of Marlboro Lights caused the smokers to take larger puffs on that cigarette than on Marlboro 85's. The larger puffs, in turn, increased the delivery of Marlboro lights proportionally. In effect, the Marlboro 85 smokers in this study did not achieve any reduction in smoke intake by smoking a cigarette (Marlboro Lights) normally considered lower in delivery.

¶The report's "Conclusions" section noted that "[t]he smoker data collected in this study are in agreement with results found in other project studies." 2021544486-4496 at 4486-4488 (US 20348);

¶see also Whidby WD, 45:11-12 (noting, in the context of this exhibit, that "Marlboro 85's" refers to

¶Marlboro Reds, a full-flavor cigarette brand).

As Dr. Burns explained, "there are three things that are powerfully significant in this document":

(1) It "very clearly demonstrates that, in contrast to what we believed six years later when we wrote the 1981 Surgeon General's Report, smokers who smoked brands of cigarettes on the market in 1975 were not getting different yields when they smoked those products. We [in the public health community] believed they were."

(2) "[T]his is dated 1975, six years prior to the time the [1981] Surgeon General's Report reached its conclusion. And we did not have access to this information or comparable information."

834

(3) "[T]his study was done on a machine that mimicked actual smoking behaviors, that actually matched the behavior of the individual when the machine smoked the cigarette. In 1981, one of the recommendations that we made . . . was that this type of machine should be developed so that we could develop a better understanding of the relationship between delivery of tar and nicotine of these cigarettes when they were actually smoked. So . . . six years prior to the time we were reviewing that evidence for the Surgeon General, this information was available to Philip Morris."

¶Burns WD, 52:15-53:12.

One other Philip Morris Human Smoker Simulator report compared the deliveries of the full flavor and light versions of brands Philip Morris was selling. That report, dated September 23, 1976, measured the smoking behavior -- and resulting tar and nicotine deliveries -- of 150 full-flavor cigarette smokers for several years. This Human Smoker study's results revealed that, while Marlboro 100s are both full-flavor cigarettes and longer in length than Marlboro Lights, "the tar and nicotine yields for these two brands are basically identical." Whidby TT, 2/22/05, 14099:12- 14106:6, 14106:15-14107:1; 1003727277-7298 at 7280, 7281, 7290 (JD 040539).
Philip Morris stopped all testing with the Human Smoker Simulator in 1981. Whidby TT, 2/22/05,13991:15-13992:13.
A March 1, 1977 Philip Morris memorandum from Stanley Schachter to Thomas Osdene, Director of Research, concluded: "Serious smokers smoke to prevent withdrawal. Smokers regulate nicotine intake . . . . The smoker who fails to regulate suffers withdrawal." 1000046626- 6661 at 6654-6655 (US 35105).
An October 16, 1981 memorandum from Jan Jones to William Dunn, titled "Nicotine Retention Research Proposal," stated:

835

Research on smoke-laden inhalation patterns, using the ambulatory monitoring instrumentation, has provided preliminary evidence that inhalation behavior is modifiable, and is altered as a function of changes in the nicotine delivery of the cigarette. We are observing changes in inhalation parameters in the direction which would suggest compensation for increases or decreases in nicotine relative to the subject's usual brand.

¶1000136372-6373 at 6372 (US 35162).

In a November 29, 1982 report, "The Effect of Cigarette Nicotine Content on Smoker Puff Parameters and Deliveries," Philip Morris scientists reported the results of their study of "puff number, puff volumes, puff durations, flow rates and puff intervals," which showed smokers "generally tended to decrease puff duration, puff number and puff volume and increase puff interval as the nicotine level of the cigarette increased." 1000408760-8809 at 8762, 8764, 8771 (US 35272).
Carolyn Levy, a research scientist for Philip Morris in its Behavioral Research Group from 1975-1980, testified that she "worked on the issue of whether individuals regulated the amount of nicotine they obtained from smoke," and as part of that work she "monitored how smokers inhaled smoke from cigarettes with varying tar and nicotine deliveries." Levy further testified that, as a result of this research, she was able to "gather evidence that some people change their smoking behavior in response to cigarettes with differing tar and nicotine deliveries." When Levy requested publication, she "was told not to publish or was not given approval to publish by the manuscript review board." Levy WD, 10:16-11:15.
A November 1999 presentation, titled "PM USA Discount Brands," given to Geoffrey Bible, Chairman of the Board and CEO of Philip Morris Companies, noted in a Product Comparison chart that Ultra Light products have a higher puff count than Full Flavor products.

836

¶(2) R.J. Reynolds

In a March 28, 1972 memorandum marked "RJR SECRET" from Claude Teague to

¶E.A. Vassallo and Murray Senkus, titled "A Gap in Present Cigarette Product Lines and an

¶Opportunity to Market a New Type of Product," Teague stated: "I believe that for the typical smoker nicotine satisfaction is the dominant desire, as opposed to flavor and other satisfactions." The document went on to state:

Given a cigarette that delivers less nicotine than he desires, the smoker will subconsciously adjust his puff volume and frequency, and smoking frequency, so as to obtain and maintain his per hour and per day requirement for nicotine (or, more likely, will change to a brand delivering his desired per cigarette level of nicotine).

¶Teague further stated:

[R]egardless of which cigarette the smoker chooses, in obtaining his daily nicotine requirement he will receive about the same daily amount of tar. If, as claimed by some anti-tobacco critics, the alleged health hazard of smoking is directly related to the amount of tar to which the smoker is exposed per day, and the smoker bases his consumption on nicotine, then a present "low tar, low nicotine" cigarette offers zero advantage to the smoker over a regular filter cigarette, but simply costs him more money and exposes him to substantially increased amounts of allegedly harmful gas phase components in obtaining his desired daily amount of nicotine.

¶The document ends with the statement that "[t]he thoughts and philosophies expressed above come from many sources and certainly are not solely those of the writer." 500790776-0784 at 0778,

¶0782-0784 (US 29473).

A document titled "Smoking Satisfaction" and labeled as a "[t]alk delivered to RJR

¶Tobacco Company management, June 23, 1974 and RJR Tobacco International management, August

¶4, 1976" by Murray Senkus, Director of Scientific Affairs for RJR until 1979, stated

837

[T]he amount of nicotine that one can get in the lungs from low tar cigarettes is much less. So the smoker then resorts to other means to get the nicotine he needs in the blood from low tar cigarettes by longer puffs, by bigger puffs, by more frequent puffs, and also by smoking more cigarettes each day. It has been observed that as one switches from a non-filter to a filter, one smokes more cigarettes per day. But eventually one can change his style of smoking so one can get enough nicotine in the blood during the inhaling step by changing the smoking style; i.e. longer puffs, bigger puffs, and more frequent puffs. Surveys have shown that in switching to lower tar cigarettes, smokers have not necessarily increased the number of cigarettes per day.

¶501525355-5366 at 5360-5361 (US 29531).

Senkus, in a speech he gave at RJR in both late 1976 and early 1977, "Some Effects of Smoking," demonstrated RJR's knowledge that smokers compensate for lower delivery cigarettes to obtain their required nicotine level, and confirmed that the other Defendants also knew this:

[T]here are ways to increase or decrease the amount of nicotine one can obtain by smoking a single cigarette: One can take a deeper puff or shallower puff. . . . One can puff more frequently or less frequently. . . . One can take a deeper puff and hold the smoke in the lungs longer before exhaling to assure complete transfer of nicotine into the body fluids. Without any question, the desire to smoke is based on the effect of nicotine on the body. . . . [T]he amount of nicotine that one can get in the lungs from low tar cigarettes is much less. So the smoker then resorts to other means to get the nicotine he needs in the blood from low tar cigarettes, by longer puffs, by larger puffs, by more frequent puffs, and also by smoking more cigarettes each day. One can get enough nicotine into the blood during the inhaling step by changing the smoking style; i.e., longer puffs, bigger puffs, and more frequent puffs. . . . It is worth noting that our competitors are aware of the significance of the quality and quantity attributes of nicotine. Moreover, they are fully aware of the advisability of maintaining a low tar value and also maintaining the nicotine as high as possible [referring to Philip Morris's Marlboro and Merit, and Lorillard's True brand].

¶500251711-1722 at 1714,1718, 1720 (US 48076).

838

In an April 5, 1982 RJR report from J.H. Robinson and J.H. Reynolds to Dr. D. Werner, the authors admitted that the nicotine delivered under human smoking conditions was "more than 200%" of that advertised, stating that "the smoker can adjust his puffing characteristics to obtain the same level of nicotine from different cigarettes. This represents the first concrete evidence that smokers compensate to obtain a consistent amount of nicotine. Relevant to this, it should be noted that all cigarettes experienced a marked reduction in nicotine filter efficiency under human smoking conditions compared to the nicotine filter efficiencies obtained under standard FTC conditions."

¶508028982-8984 at 8983 (US 85053).

John Robinson, RJR's Principal Scientist in psychopharmacology, wrote a July 25,

¶1983 memorandum to Alan Rodgman, titled "Critique of Smokers of Low-Yield Cigarettes Do Not

¶Consume Less Nicotine," which essentially agreed with the conclusions in an article written by Dr. Neal Benowitz showing that smokers compensate to obtain a stable nicotine dose in their bloodstream. The article stated:

The paper itself expresses what we, in behavioral, have "felt" for quite some time. That is, smokers smoke differently than the FTC machine and may very well smoke to obtain a certain level of nicotine in their bloodstream. If a given level of nicotine in the blood is the final goal of a smoker, one would predict that he would smoke an FFT [full-flavor tar] and ULT [ultra low tar] cigarette differently. If the smoker could obtain the same nicotine in his bloodstream from an FFT and ULT cigarette by modifying his puffing/inhaling pattern, it would be expected that the blood cotinine level would be the same after smoking either cigarette on a regular basis . . . the data reported in this paper remind us of the HMSM experiment done with the German Camel and Marlboro cigarettes. While there were certain imperfections in this experiment, you may recall that the smokers apparently obtained almost exactly the same amount of nicotine no matter which of the four cigarettes they smoked. This was one of the first indications that smokers may, in fact, smoke to obtain a certain

839

level of nicotine in their bloodstream. Data like these made me feel that the data reported in this current publication are probably correct.

¶502680871-0871 (US 49198); see also 508978013-8025 at 8014 (US 20819) (acknowledging that smokers who switched to low-tar products typically "compensated," and indicating that a smoker "has his or her own nicotine requirement from each cigarette" and "adjusts [his/her] smoking maneuver" to obtain the desired level of nicotine").

(3) Brown & Williamson 2206. Minutes from a January 12-18, 1974 B&W/BATCo conference stated: "[W]hatever

¶the characteristics of cigarettes as determined by smoking machines, the smoker adjusts his pattern to deliver his own nicotine requirements." 109882674-2679 at 2675 (US 21507).

B&W's scientific research on compensation was confirmed by its consumer research on low tar cigarettes, in which smokers reported compensating for the reduced deliveries. For instance, a February 23, 1977 consumer research report, "Consumer Discussions of Low Delivery Cigarettes," written by R.F. Brotzge and W.H. Deines, demonstrates B&W's awareness that low tar smokers were compensating by smoking more cigarettes, stating:

Findings, in order of importance, to participants in the five focus groups were: 1. Health -- Participants expressed general fears about cancer, emphysema, other lung diseases, etc. Despite these fears, they stated their determination to continue smoking. . . . 5. Compensation -- Participants noted they smoked more low tar cigarettes and received less satisfaction.

¶679009843-9867 at 9843 (US 85055).

Similarly, a July 25, 1977 B&W Internal Marketing Study, titled "Low 'Tar' Satisfaction, Step 1 Identification of Perceived and Underperceived Consumer Needs," analyzed smokers' satisfaction with low tar cigarettes with regard to switching behavior, and stated: "It was noted earlier that new arrivals to the Hi-Fi category realize that they are smoking more cigarettes [quoting a study participant]: 'You can also go down to the lower tar, but increase your smoking. So you're right back where you were.'" The study further noted that "Cigarette consumption, as reported in a 1976 Consumption Study, increases as nicotine (satisfaction per cigarette) decreases." 775036039-6067 at 6050 (US 21053).

840

A report bearing the stamp "Brown & Williamson June 24, 1980 R&D Library" prepared by BATCo on April 23, 1980, titled "Compensation: A Review [of] the Relationship Between Compensation and Changes in Cigarette Design," stated that studies indicated that compensation was a permanent phenomenon: "On the basis of the German studies, compensation would therefore be seen as a long-term tendency to permanently adjust towards some preferred (or minimum) level [of nicotine]." 650032329-2385 at 2356 (US 53429); Ivey WD, 68:1-10.
A July 1983 B&W report by W. Wiethaup and W. Schneider, "Filter effects on smoke and smoke effects," stated:

One factor, which may be responsible for a relatively intensive "strength" impression within a given tar segment, is the "smoke elasticity." The smoke elasticity describes the potential of a cigarette, to provide the smoker with more smoke, if he draws harder. This becomes relevant at least for the first few puffs of a low tar cigarette, as all recent investigations show.

¶512107109-7120 at 7110 (US 85056).

A March 26, 1999 e-mail from Hugh Honeycutt to Mike Dixon and numerous other BATCo employees referenced B&W's "Atlanta Study," which was conducted by scientist Kelly St. Charles, to examine smoking behavior and puffing profiles for low tar cigarettes. Honeycutt's email expressed concern that "B&W had just made a big splash in the US touting Carlton as the '1' for you," -- a marketing slogan indicating that Carlton delivered only one milligram of tar -- in light of research finding that "smokers of ultra low tar brands like our Carlton 1 mg appeared to actually get 3 mg." of tar. 321155579-5580 at 5579 (US 46683); 2073168412-8414 (US 22024). As of June, 2004, B&W admits that smoker compensation causes Carlton 1 mg smokers, on average, to inhale 5 to 6 times the amount of tar that B&W has advertised. (no bates) (US 88628) (stating "a Carlton 1 mg. smoker will, on average, get 5 to 6 milligrams of tar").

841

(4) BATCo 2212. A BATCo document from the late 1970s, "Why do People Smoke?," reflects

¶BATCo's understanding of compensation, stating that "[n]icotine sustains smoking behaviour," that "smoking behaviour is highly responsive to cigarette design," and that "[a] key determinant of product preference will be the design 'Effort-Reward Gradient,'" i.e., elasticity of delivery. The document added: "Increase in Cigarette Consumption [is] Related to Change in Nicotine Yields," noting that "[m]ost compensation must occur at the individual cigarette level." 403626692-6802 at 6729, 6731, 6734, 6762, 6768 (US 85018*).

An undated BATCo document by Dr. S.J. Green, a Senior Scientist for BATCo Research and Development, titled "Ranking Cigarette Brands on Smoke Deliveries," discussed compensation to equalize nicotine intake in several contexts, including smokers "increas[ing] puff volume to receive the same nicotine" when smoking a lower tar cigarette and smokers "adjust[ing] their smoking behaviour on the basis of nicotine intake." 110077247-7268, at 7247-7250 (US

842

A March 11, 1971 report written by D. Creighton, BATCo R&D Research Scientist, and L.M. McGillivray, titled "The Effect of Changed Deliveries at Constant Pressure Drop on

¶Human Smoking Pattern," stated:

It was found that there is indeed a degree of compensation for the reduced delivery. The panel as a whole took larger puffs from the lower delivery cigarette, inhaled the smoke more deeply and held the smoke in the lungs for a longer time. The six individual panel members compensated for changed delivery in different ways, some by increased volumes, and others by increased number of puffs on the lower delivery cigarette.

¶The report also stated: "An increase in puff volume, but not puff duration, means that subjects must have drawn harder on the lower delivery cigarettes, and that the flow rate was greater during puffing.

¶This implies that smokers are willing to work harder to achieve an optimum delivery from a lower delivery cigarette." The report further stated:

The fact that the panel compensated for the lower delivery by increasing the depth of inhalation, the depth of exhalation and the total time for which the smoke was held within the body is of particular interest in the light of the finding that more is retained from a puff of smoke that is inhaled deeper, and held within the lungs longer.

¶757001173-1185 at 1174, 1180 (US 85058); see also 102793967-3980 at 3969 (US 34698).

A June 17, 1975 BATCo document, "Compensation for Changed Delivery," sent to several BATCo employees, including David Geoff Felton, Senior Scientist for BATCo Ltd.'s

¶Research and Development Department, acknowledged that smokers compensate to achieve a stable dose of nicotine:

A number of experiments . . . have been interpreted as showing that compensation for changed delivery does occur. . . . Our own results showed that when the [tar] to nicotine ratio was changed more smoke was taken from the lower delivery cigarette. . . . My own view is that

843

compensation for changed delivery of nicotine does occur. . . . The weight of evidence at present available is for nicotine compensation [referring to several studies].

¶105658168-8179, 8168, 8178 (US 85418).

A June 1, 1976 BATCo report, also titled "Compensation for Changed Delivery," written by D. Creighton, R&D Research Scientist, concluded that

the evidence is strongly in support of the hypothesis that many smokers do change the way they smoke in response to cigarette design changes that affect nicotine delivery. . . . The tendency amongst the majority of established smokers is to attempt to equalise nicotine delivery if the cigarette design allows them to do so.

¶The study further concluded that "[d]ue to the differences in the delivery of individual cigarettes from the same brand . . . and the differences between subjects and within a subject" "[e]qualisation [of nicotine levels] within the range + 20%" was expected. The report stated that "there are eight suggested methods by which a subject may regulate his nicotine intake; any number of which may be used simultaneously or at different times," namely by varying puff volume, puff number, puff distribution, cigarette butt length, puff interval, puff profile, inhalation pattern, and number of cigarettes smoked. 650008449-8480 at 8470, 8460-8462, 8469, 8464 (US 76192).

A June 27, 1978 BATCo memorandum, also titled "Compensation for Changed Delivery," written by Creighton, confirmed BATCo's knowledge that compensation is not temporary, and occurs as a result of a need for nicotine, and contradicted "the advice of Health Authorities" that smokers who would not or could not quit should switch to a lower delivery cigarette:

It is difficult to ignore the advice of Health Authorities who advise smokers to give up smoking or change to a lower delivery brand but there is now sufficient evidence to challenge the advice to change to a lower delivery brand, at least in the short-term. Numerous experiments have been carried out in Hamburg, Montreal, and

844

Southampton within the company as well as many other experiments by research workers in independent organizations, that show that generally smokers do change their smoking patterns in response to changes in the machine smoked deliveries of cigarettes. . . . Further findings from these results were that the modified smoking patterns used to smoke the changed delivery brands were maintained for the month during which they were smoked. This shows that there was no adaptation during that time. . . . In general, a majority of habitual smokers compensate for changed delivery, if they change to a lower delivery brand than their usual brand. . . . If they choose lower delivery brand which has a higher tar to nicotine ratio than their usual brand (which is often the case with lower delivery products) the smokers will in fact increase the amounts of tar and gas phase that they take in, in order to take the same amount of nicotine.

¶105553905-3914 at 3905, 3907, 3913 (US 34799); Ivey WD, 69:14-70:1.

A June 29, 1979 study written by Creighton, "A Comparison of Smoking Surveys

¶Separated by Four Years," compared the smoking behavior of a group of smokers in 1974 of a cigarette with 1.7 mg nicotine and 27 mg TPM to their behavior in 1978 when smoking a cigarette with a slight reduction in nicotine (by 17% to 1.4 mg) and virtually identical TPM delivery (26 mg).

¶The study found that the group smoked the reduced nicotine cigarette "more intensely" (i.e., increased their puff volume), which likely "equalised" the nicotine delivery of the two cigarettes.

¶The study concluded that "it is probable that as a result of the changes in smoking behaviour observed in this study, subjects took about the same amount of nicotine from the two different cigarettes but, because of changes in TPM to nicotine ratios, received more TPM from the [reduced nicotine cigarette]." The study further found that

[t]he fact that smokers have changed their smoking patterns to take more smoke from a cigarette with lower nicotine delivery but similar TPM delivery adds support to the contention that nicotine is a major determinant of smoking behavior, and that TPM, as long as it is delivered in sufficient quantity, plays a lesser role.

845

¶650008946-8960 at 8948, 8953-8955 (US 85059).

An April 27, 1981 memorandum by Martin Oldman, an employee in BATCo's Research & Development department, to L.C.F. Blackman, Director of BATCo, stated that "[s]ome people appear to smoke for nicotine, others don't . . . nicotine dependent smokers. . . are more likely to compensate for nicotine than others." 105399692-9693 at 9693 (US 85060).
A July 9, 1984 document by Imperial Tobacco Limited, sister company to B&W, was distributed to various B&W and BATCo employees, including: Blackman; A.M. Heath, BATCo Executive Director of Marketing; Erhard Koehn, BATCo Manager of Product Development; Rainier Wernitz, BATCo Manager, Market Research; Tilford Riehl, B&W Division Head of Product Development; A. Mellman, B&W Director of Marketing Research; T. Wilson; Brennan; C.I. Ayers; and G.O. Brooks, BATCo scientist. It acknowledged that smokers who switch to low tar brands increase the number and intensity of puffs taken and number of cigarettes smoked to achieve a higher dose of nicotine: "BRANDS SWITCHING DOWN DELIVERY: increase in puffing parameters -- increase in numbers of cigs. smoked -- more puffs taken means to achieve a higher dose." 536000000-0090 at 0050 (US 22338).

A January 24, 1985 BATCo letter from Charles H. Keith to Lance Reynolds stated: [H]uman smokers, even though they ingest much more Nicotine and Tar than is indicated by the FTC values, get about the same amount of Tar and one and a half times the Nicotine from Barclay, Carlton and Cambridge. . . . [I]t is clearly apparent that the human smokers are ingesting much more nicotine and tar than the nominal values obtained by FTC tests. The human levels are six to eight times higher than the normal values.

846

A BATCo document dated October 12, 1987 sent by M.L. Reynolds to H.F. Dymond

¶(researcher) and H. Ibig, under the heading "Easily Achieved Tar Deliveries from Low Tar

¶Cigarettes," displays a chart depicting the increased tar deliveries of RJR's Now cigarette, and Philip

¶Morris's Merit and Merit Ultra cigarettes, caused by the compensatory behaviors of vent hole blocking, taking more puffs, and taking bigger puffs. This chart shows that these behaviors in combination caused a thirteen-fold increase of tar inhaled for Now cigarettes, a nearly three-fold increase for Merit, and a four-fold increase for Merit Ultima. 400015695-5696 at 5696 (US 85063).

A December 21, 1987 BATCo document with subject heading "Notes on Meeting

¶With Dr. Eicher" written by "HFD" (H.F. Dymond, researcher) and sent to Nick Cannar and several

¶B&W and BATCo personnel, stated: "BAT acknowledged the discussion on compensation and described how channel ventilation was an alternative form of ventilation, both systems could be manipulated by a consumer. Other companies were also beginning to acknowledge compensation, but they were reluctant to debate the issue in public." 400015634-5635 at 5634 (US 85064).

A May 6, 1992 BATCo report, titled "Topics In Smoking and Health 'Bible'" stated:

[T]he expression of product smoke deliveries in the form of a league table, while understandable, can be misleading. There can be no guarantee that a smoker who switches from one product to another delivering a lower "tar" value will thereby reduce his intake of "tar." He may well alter the way he smokes the second product in some subtle fashion and so adjust his intake of smoke to fit his needs. In this way, he may inadvertently increase his intake of other substances in the smoke. League tables and delivery data on products may, therefore, be misleading to the consumer, who will be unaware of the sub-conscious ways in which he manipulates his own behaviour . . . smokers of higher delivery cigarettes may find that they need to smoke more low delivery cigarettes to achieve the same satisfaction. . . . Increasingly smokers will accept the alleged harmfulness of smoking, and while wishing to continue will look for health reassurance brands. . . . Smoking behaviour is also of

847

importance. For example research into the effects of low tar and nicotine cigarettes on ease of quitting smoking will be undertaken.

¶500887584-7709 at 7606-7607, 7614, 7679, 7704 (US 20656).

(5) American Tobacco 2225. A November 11, 1976 report prepared by Fay Ennis Creative Research Services for

¶F. William Free & Company, an advertising agency used by American, demonstrates American's awareness that smokers of low tar cigarettes employed several different methods of smoker compensation. The report summarized focus group sessions relating to low tar cigarettes. When asked about Now and Carlton cigarettes, the panelists "concluded that you would smoke twice as much of this type of cigarette in order to get any satisfaction. One man said that he didn't like the draw on these cigarettes because he had to puff so hard, his throat tickled." When asked to define a low tar cigarette, some panelists stated: "You have to drag on the cigarette 'real' hard to get any satisfaction out of it." ATC0137310-7324 at 7319-7320 (US 87916).

(6) Lorillard 2226. A July 16, 1976 Lorillard memorandum from M.S. Ireland to H.J. Minnemeyer on

¶the subject of "Research Proposal -- Development of Assay for Free Nicotine" revealed Lorillard's awareness of smoker compensation:

Cigarette sales are made for one reason. The customer is satisfied with the product either from the taste or the physiological satisfaction derived from the smoke. The consensus of opinion derived from a review of the literature on the subject indicates the most probable reason for the addictive properties of the smoke is the nicotine. Indications are that the smoker adjusts his smoking habits to satisfy the desire for nicotine either by frequent or large puffs on the cigarette, or smoking a large number of cigarettes.

¶00044522-4523 at 4522 (US 22012); 94937037-7038 at 7037 (US 56775).

848

A December 10, 1976 document by H.S. Tong, which included a review of the scientific literature, reached several conclusions confirming Lorillard's understanding that smokers compensate to receive their desired level of nicotine:

It seems that, within limits, smokers can and do control their nicotine intake from smoke by varying their smoking techniques. . . . Smokers were known to smoke more when offered low nicotine cigarettes. . . . It would seem desirable to have a low tar cigarette with a nicotine content between the threshold and optimum doses level.

¶00045061-5071 at 5061-5063, 5068 (US 34210).

A July 30, 1980 Lorillard memorandum, "A Review of Behavioral and Psychopharmacological Factors in Smoking," from S.T. Jones (Product Design), included conclusions by Lorillard personnel based on review of scientific articles in the literature, including the following: "The evidence to date clearly indicates that smokers titrate or regulate their intake of nicotine, e.g. smokers of cigarettes which deliver large amounts of nicotine will adjust -- when given low nicotine cigarettes -- their smoking to get a larger nicotine dose than the machine determined values indicate." Lorillard also independently acknowledged that, in the 1980s, it knew that smokers of low tar/low nicotine cigarettes would compensate by altering their smoking habits in order to obtain a higher level of nicotine. 01105000-5021 at 5010 (US 20030); Spears PD, State of Minnesota v. Philip Morris Inc., 9/23/97, 62:29-65:11.
An August 21, 1984 Lorillard memorandum from E-Chung Wu to W.R. Deaton, reporting on a "puff profile study of 15 brands of cigarettes . . . with 5 smoking panel members" acknowledged smoker compensation: "The general trend shows that puff volume increases with the decrease of both TPM or nicotine. . . . Obviously, the higher dilution of smoke needs larger volume to compensate for the decrease of the flavor or nicotine." 89213491-3501 at 3491 (US 56466).

849