Market Forces Information and Reduced Flammability Cigarettes

Gary T. Ford, The American University
John E. Calfee, University of Maryland and Federal Trade Commission
ABSTRACT - Fires started by cigarettes are a significant problem that tends to be concentrated in certain socio-economic groups. The federal government may require that cigarettes be changed to reduce their tendency to start fires. The net effect would depend partly on how the market would adjust. Among the possibilities are increased tar and nicotine yield decreased care in preventing fires and possibly, decreased consumption of cigarettes
[ to cite ]:
Gary T. Ford and John E. Calfee (1987) ,"Market Forces Information and Reduced Flammability Cigarettes", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 274-278.

Advances in Consumer Research Volume 14, 1987      Pages 274-278


Gary T. Ford, The American University

John E. Calfee, University of Maryland and Federal Trade Commission

[Ford is Professor of Marketing in the Kogod College of Business Administration. Calfee is Visiting Lecturer in the College of Business and Management and Assistant to the Director Bureau of Economics Federal Trade Commission. The views expressed here are the author's and not necessarily those of the Federal Trade Commissioners or other members of the FTC staff. Results reported in this paper were partially supported by federal government contract CPSC86-122501.]


Fires started by cigarettes are a significant problem that tends to be concentrated in certain socio-economic groups. The federal government may require that cigarettes be changed to reduce their tendency to start fires. The net effect would depend partly on how the market would adjust. Among the possibilities are increased tar and nicotine yield decreased care in preventing fires and possibly, decreased consumption of cigarettes


Government intervention often requires that products possess certain attributes (such as seatbelts) or meet certain standards (such as with bumper standards). When intervention is concerned with only one or a few of the characteristics of the product, other product attributes remain free to be determined by the market. The manner in which manufacturers and consumers adjust to the new requirements may dominate the total changes from the intervention and result in consequences not intended by the regulators Government planning for intervention should therefore include an assessment of how markets are likely to adjust to the intervention

In this paper we describe an interesting case of a possible intervention that is still in the early planning stages Our purpose is to illustrate how market factors especially information could play a role in the outcome of the intervention The potential intervention is to require reduced flammability cigarettes i.e. cigarettes that are less likely to cause fires. This possible intervention arose because fire-starting is a major but easily overlooked outcome of smoking. Recent estimates are that smoking caused fires result in approximately 1 500 deaths each year (FEMA 1984 p. A18.) The federal government is now investigating the possibility that mandating new standards for cigarettes will significantly reduce this cost [Cigarette Safety Act of 1984 (Pub. L. 98-567 98 Stat. 2925 Oct. 30, 1984.)]

The cigarette market is particularly appropriate for analyzing the effects of public policy because information in this market has been highly regulated while the product itself has remained virtually unregulated (cigarettes and other tobacco products are exempt from FDA regulation.) The cigarette market is also unusually well documented and experience has shown that this market is capable of adjusting rapidly when consumers are exposed to new and important information about smoking -- or at least the market tends to react quickly when the flow of information is not impeded by government (Calfee 1985, 1986). Moreover some of the specific attributes of cigarettes especially health-related ingredients such as tar nicotine and carbon monoxide could be affected by chances to reduce fires.


A natural question is whether the potential effects of reduced flammability cigarettes will vary considerably among socio-economic groups. We begin by estimating the distributions by age race and sex of those who smoke cigarettes and of those who die in fires started by cigarettes. This allows us to determine whether fires are being started disproportionately by a few types of smokers. [Complete details concerning the methodology used to estimate the distribution of fire-starters by age race and sex are contained in Ford Brown and Calfee (1986).]

The most thorough analysis of the demographic characteristics of those who die in fires started by cigarettes is contained in a United States Fire Administration report (Hall and Helzer 1983.) That report used data for 1978. 1979 for six states with high fire rates, and six with low rates and calculated the death rate per million for population groups jointly categorized by age race and sex. [Hall and Helzer warn that the age sex and race cross-tabulations should be treated with caution since the incidence of death calculations were based on small numbers of cases. This note of caution is worth bearing in mind since the calculations we perform rely on this data. In our defense we aggregate the data from the high- and low-fire rate states by calculating weighted averages based on 1980 census population count of each state Thus the small sample size problem occurring in some cells is diminished somewhat.] We used population data to convert these rates into estimates of the actual number of deaths. The separate estimates for the two groups of states were then aggregated by calculating a weighted average based on the total population in the high- and low-fire rate states; this was necessary because the high-fire rate states had twice the relative death rates but only half the population of the low-fire rate states.

Estimating the distribution of smokers was more straight forward since the National Center for Health Statistics (NCHS) supplies such data for smokers aged twenty and above by age race and sex. The only complicating factor was the fact that the incidence of teenage smoking is not estimated in the NCHS data and had to be estimated from other data sources. [For purposes of this report (and as is explained in more detail in Ford Brown and Calfee (1986), we assumed that the participation rates of 17-19 year-old smokers were three percent less than comparable percentages for 20-24 year-olds. While admittedly arbitrary this approach results in estimates that are consistent with the smoking patterns for even younger smokers and with the patterns observed for white-males black-males white-females and black-females.]

Table 1 shows the distribution of deaths from fires caused by smoking, the distribution of smokers by age race and sex and the ratio of smokers to deaths. The marginal percentages show that white males account for slightly over one-half (i.e. 53.7 percent) of all deaths from smoking fires and that white females represent almost 30 percent of the victims. Less than eight percent of the victims are sixteen years or younger (and although not shown here only three percent are less than five years old). The data indicate quite clearly that most victims of cigarette caused fires are adults, and in fact 63 percent are 45 years or older.



The lower section of Table 1 shows for each category how the ratio of deaths to smokers compares to the ratio for the average category. An entry greater than one indicates a category of smokers with a greater than average chance of dying in a smoking caused fire while a ratio of less than one indicates the opposite. We see disproportionately high entries for smokers 45 years and older; this age group accounts for 35 percent of the smokers and 63 percent of the deaths from smoking caused fires. Black males black females and white males over 64 also seem particularly likely to benefit from a reduction in ignition propensity since these three groups represent approximately four percent of the smokers but account for nineteen percent of the victims of smoking caused fires. In contrast are the smokers under age 45 especially those between 25 and 44. They are the heaviest smokers (e.g approximately 30 percent of this group smokes 25 or more cigarettes per day; NCHS 19861) but have the lowest death to smoking ratio (.45). Thus they will benefit the least (at present) from reduced flammability cigarettes.


If the government is to require reduced flammability cigarettes there are quite different methods for doing so. We concentrate on the interactions between the manner in which changes are implemented and the incentives of producers. It is clear that the more regulation is in accord with market incentives, the more likely the regulations will be followed. Less obvious is the fact that some kinds of regulation are more compatible with incentives than others and that taking this into account can improve the degree to which regulatory goals are achieved (Rubin and Cohen 1985.)

In the present context there would be a choice between two approaches. One is to require that cigarettes be made a certain way (paper thickness tobacco weight and so on.) This may be referred to as a "design" or "input" standard. The second approach is to require cigarettes to perform in a certain manner (for example extinguish within forty seconds of being placed on a standard textile surface.) Such a requirement may be called a "performance" standard

Performance standards offer considerable advantages Chief among these are that this method more thoroughly harnesses competitive market forces in support of the purposes of introducing the new cigarettes. Firms would be free to achieve the same effects using new methods that could reduce costs improve the trade-offs between flammability and other features such as tar content or even improve upon flammability itself. We shall see below that these trade-offs may be quite different for different varieties of reduced flammability cigarettes. Thus performance standards would tend to minimize adverse effects while achieving regulatory goals

A disadvantage of the performance standard approach is the obverse of its main advantage. The market may eventually compete so exclusively on the exact attributes of the performance standards that the results could be perverse For example product development on reduced flammability could concentrate on techniques that work well with the particular textiles used in tests but work poorly with other surfaces The experience with the Federal Trade Commission's tar and nicotine measurement service has shown that this could happen although the adverse effects may be slow to emerge (Calfee 1985 and 1987.) If new cigarettes are required and performance standards are used the problem of competing for the favors of the testing machines should be kept in mind


The availability of reduced flammability cigarettes in the market will alter consumer behavior in two basic respects: precautions against starting fires and how cigarettes are smoked. Fire precautions depend mainly on what consumers know of the changes in flammability whereas changes in smoking behavior depend upon the precise ways in which the new cigarettes differ in taste, smoking characteristics, tar and nicotine content and other traits.

Potential Effects on Precautions Against Fires

Smokers are presumably aware that fires can be started by careless smoking especially smoking in bed. Thus we may think of smokers as choosing a level of precaution by balancing the risks of fire against inconvenience and other costs of prevention. The availability of reduced flammability cigarettes will alter the trade-off between inconvenience and fire danger. We would expect a lower level of precautions against fires: more smoking in bed for example. Similarly holding all else constant safer cigarettes should result in somewhat more smoking since presumably some people refrain from smoking at all in certain situations where fire danger is high such as when about to fall asleep in an easy chair

The magnitude of these effects is unknown A few points are clear however. The degree to which smokers change precautions in response to changed cigarettes will depend on what they know of the new cigarettes. If reduced flammability cigarettes are mandated but smokers are uninformed of the new properties no adjustments would take place until knowledge of reduced flammability becomes disseminated (as it eventually will of course.) This almost sounds like a good thing, but as we shall see in a moment smokers of reduced flammability cigarettes may be tempted to ingest substantially more tar and nicotine and unless they are aware of the advantages of the new cigarettes they may not make the best adjustments from an overall health standpoint.

Adjustments to reduced flammability cigarettes may vary substantially according to the kinds of people involved. Socio-economic groups that frequently encounter fires started by cigarettes face different trade-offs from those faced by other groups. There is nothing inherently wrong in this, of course. Trading off life-style and safety is a necessary part of life, such as when one decides to go to work on a day when roads are icy. The point here is simply that when one takes into account the fact that the new cigarettes are less likely to cause fires in any given situation, one also has to remember that consumer adjustments will tend to off-set this effect. Presumably, the net effect is still to reduce fire incidence, although it is hard to demonstrate this with certainty.

Potential Effects on Smoking Behavior.

Reduced flammability cigarettes may be smoked differently. Whether this happens, and how, depends on exactly how the new cigarettes differ from the old ones. Policy-makers are considering two types of reduced flammability cigarettes: "self-extinguishing" and "reduce" ignition propensity." Self-extinguishing cigarettes will go out if not puffed within a certain period, so as to reduce free burning to a period less than is required to start a fire under most conditions. "Reduced ignition propensity" cigarettes, on the other hand, will have a lower propensity to start a fire when the cigarette is burning. The two kinds have somewhat different implications for consumer behavior and market changes.

Self-Extinguishing Cigarettes. The effect of self-extinguishing cigarettes on smoker behavior depends primarily on four factors: price, taste, smoking characteristics (especially, the time allowed between puffs) and changes in the ingestion of health-related ingredients such as tar, nicotine and carbon monoxide. We deal with price changes elsewhere. We have reason to believe that taste will not be appreciably worsened in these cigarettes, perhaps because of the addition of new flavor enhances. Two important aspects of self-extinguishing cigarettes remain to be considered: smoking characteristics and health-related ingredients. These are interrelated, because the manner in which smokers change smoking habits to counteract the tendency to self-extinguish may affect how much tar, nicotine and carbon monoxide are ingested.

The data on puff frequency, duration and volume of which we are aware indicate that smokers puff approximately every 40 to 60 seconds, that there is wide variation within and across smokers, that the average puff duration is between two and three seconds and that smokers take between nine and twelve puffs on average per cigarette. (Rawbone, Murphy, Tate and Kane 1978, p. 187 and 190; Creighton and Lewis 1978, p. 292; Guillerm and Radziszewski 1978, p. 365) (One study found that mean duration between puffs ranged from 23 to 115 seconds for different smokers (Guillerm and Radziszewski 1978, p. 365)). If new cigarettes self-extinguish within, say, 60 seconds, many smokers will have to puff more frequently or relight occasionally.

Smokers often increase puff frequency to compensate for reduced tar and nicotine, and thus it seems likely they will do the same in order to avoid the annoyance of a cigarette that self-extinguishes. If so, more smoke would go into the lungs of the smoker, and less into the ambient environment. Thus self-extinguishing cigarettes could have the unintended consequence of dramatically increasing tar and nicotine consumption. For example, Kozlowski (1981) estimated that increasing puff frequency increases tar delivery by 58 percent or more. [Altering puff frequency is one of a number of ways in which smokers "compensate" for changes in cigarette content and construction. The literature on compensation is extensive and somewhat mixed in its assessment of overall effects on what actually reaches smokers' lungs. Some of this research is summarized in Calfee (1985) at n. 119.]

The amount of tar consumed by smokers could also increase because of changes in cigarette construction or ingredients. We understand that in order to reduce burn rates, self-extinguishing cigarettes are likely to have reduced paper porosity and/or increased tobacco density. Less porous paper will cause less dilution of the smoke with air, and will therefore increase the extent to which smokers inhale undiluted smoke. The effect could be substantial. Kozlowski estimates that in an ultra low tar cigarette (i.e., 1 mg of tar), as designed, as much as 80 percent of the smoke in each puff is diluting air (Kozlowski, n.d., p. 8). In addition, denser cigarettes exhibit a slower burn rate and this, too, may result in increased tar levels.

To some extent, changes in cigarette characteristics will be reflected in FTC tar and nicotine ratings. But the effects of changes in smoking behavior such as puff frequency and/or duration are unlikely to be captured by the FTC rating system. The FTC's system has not been altered in the twenty years since its inception, despite the well-documented existence of compensatory smoking behavior that sometimes makes FTC ratings inaccurate, and there seems little reason to expect changes soon (Calfee 1987). These problems with the FTC system could be compounded by reduced flammability cigarettes.

Mandated self-extinguishing cigarettes could even render ultra-low tar cigarettes (those yielding roughly 1 to 5 mg of tar) obsolete. These cigarettes use extremely porous paper (and filters), and contain "expanded" tobacco that is puffed so that it is less dense. Requiring that the tobacco in cigarettes be denser and the wrapping paper less porous may therefore remove the least harmful type of cigarette from the market.

Reduced Ignition Propensity Cigarettes. Reduced ignition propensity cigarettes will probably be thinner, faster burning, longer, covered with thicker and more porous paper, and made of expanded tobacco. The new cigarettes may also be coated with a neutral tasting silica gel designed to reduce ignition propensity, and if so, this may curtail the need for the other modifications (Ruegg, Weber and Lippiatt 1986). There seems little reason to expect that smokers will be required to substantially modify puff frequency or other aspects of their smoking behavior. Therefore, the only factor of concern (other than increased costs which are discussed below) involves changes in cigarette construction and/or ingredients, especially the increased use of expanded tobacco.

Expanded tobacco is used in a large proportion of cigarettes manufactured today, including most low tar brands which account for well over fifty percent of the total cigarette market (Maxwell Report 1981). Increasing its use is likely to decrease average tar and nicotine content -- just the opposite effect from what is expected from self-extinguishing cigarettes. The effect on taste will presumably be about the same as would be associated with the reduction in tar, offset by enhanced use of flavorings. We understand that taste is also influenced by the amount of paper being burned relative to the amount of tobacco. If flavorings can be used to mask the paper taste, which apparently is possible, then cigarette taste will be relatively unaffected (Lago 1986) Cost may go down, because less tobacco is required.

The choice between self-extinguishing and reduced ignition cigarettes is an important example of the potential advantages of performance standards over design standards. If cigarettes met whatever reduced ignition standards were mandated, the market could decide whether reduced ignition propensity or self-extinguishing cigarettes would dominate or whether each would find a substantial market segment. Further, manufacturers would be free to bundle certain attributes that consumers may desire (ultra-low tar and reduced flammability, for example) in the same product as long as the product met the flammability standards.


Consumption of Cigarettes

Overall consumption of cigarettes will be affected by several factors that tend to off-set one another. One is price. Reduced flammability cigarettes will probably cost more, and if so, this will tend to reduce consumption. The price change will almost certainly be small; actual construction of cigarettes accounts for only a small part of the retail price. We assume that the maximum price increase will be a few cents per pack, which is about a two or three percent increase. A number of studies have estimated the price elasticity of demand for cigarettes. These suggest an overall elasticity of, very roughly, one-half for adults, with a substantially higher elasticity for youths. Moreover, the changes in demand tend to derive substantially from the few smokers who quit altogether, rather than from all smokers reducing consumption equally (Porter 1986, Ippolito and Ippolito 1984, Lewit and Coate 1982). Thus a two percent price increase, for example, would cause roughly one percent decrease in overall consumption, with a larger decrease among youths, and might cause a significant number of persons to quit smoking or refrain from taking it up.

Consumption will also be affected by several factors that have already been mentioned. These include the fact that the cigarettes will be safer to smoke in bed (thus increasing consumption), taste worse (decreasing consumption), have less desirable smoking characteristics (decreasing consumption) and may change in rated tar and nicotine (decreasing or increasing consumption). The net effect of these factors, along with price, seems likely to be negative, but the magnitude is impossible to predict. We speculate, however, that if either taste or smoking characteristics change substantially, this could dominate other factors.


The net effect of reduced flammability cigarettes on fires is a combination of partially off-setting forces: fewer precautions against fires, a probable net decrease in consumption, and, of course, a substantially decreased likelihood that a burning cigar will cause a fire. It seems reasonable to assume that the overall effect will be to decrease the incidence of fires started by smoking. Whatever happens, it is clear from demographic data that the effects will be concentrated on certain socio-economic groups, those described earlier as exhibiting a large ratio of smoking caused fires to magnitude of smoking.

Health Effects of Smoking

The net effect of reduced flammability cigarettes on the health effects of smoking is, again, a matter off-setting forces. Overall consumption may be down slightly, but cigarettes may yield more tar, nicotine and other dangerous ingredients. Clearly the outcome depends on what kind of reduced flammability cigarettes are introduced. We noted earlier that self-extinguishing cigarettes are almost incompatible with ultra-low tar cigarettes. Epidemiological data suggest, moreover, that a substantial increase in tar yield would have a major effect on lung cancer incidence (see the survey in Participants 1985.)


Reduced flammability cigarettes will probably became a serious topic for intervention only if they show promise of yielding substantial net benefits. The government could then face a choice between requiring changed cigarettes or simply encouraging these presumably superior products to find their way into the market through more or less natural means. We refer to the latter option as "voluntary standards." The decision between mandated and voluntary standards will depend partly on factors beyond the scope of this study (for example, the extent to which smokers do not bear the full costs of the fires they start).

We think it useful, however, to point out that the "voluntary" approach merits serious consideration. This approach would have some strong advantages. To a large extent, these advantages follow from perhaps the greatest disadvantage of changing the market by means of regulation: if reduced flammability cigarettes were required, it would be nearly impossible for market (or other) processes to reveal consumer valuation of the changes. In an unregulated environment, on the other hand, competing cigarettes would offer a variety of levels of reduced flammability, and consumers would better be able to trade this off against tar and nicotine yield (if there were such a trade-off). This opportunity for choice could be important in view of the strong differences in socioeconomic groups in their vulnerability to smoking-induced fires. The information required for such choices would probably be available through advertising. Finally, "voluntary" standards would be less subject to manipulation by vested interests, which often acquire so strong a stake in required product attributes that requirements can be difficult to dislodge even when improvements are possible.

This is not to say that all considerations are on the side of voluntary methods. The free market often works slowly and imperfectly, at least from the standpoint of regulators and legislators. Moreover, requiring that all cigarettes be changed could, as we noted earlier, partly obviate the possibility that smokers will compensate for safer cigarettes by being more careless.


Requiring reduced flammability cigarettes could have unintended consequences, mainly because of how market participants would adjust to the new cigarettes. Depending upon the kinds of cigarettes required, levels of tar, nicotine and other health-related ingredients could increase substantially, and do so in a manner not fully detected by FTC ratings. Smokers will probably be less careful about causing fires. Implementing reduced flammability cigars by means of performance standards instead of design standards could minimize these adverse effects. Whatever happens, the effects will tend to be concentrated in certain socio-economic groups in which fires caused by smoking occur with much greater than average frequency.


Calfee, J. E. (1985) "Cigarette Advertising, Health Information and Regulation before 1970," Bureau of Economics, Federal Trade Commission, Working Paper #134 (December 1985.)

Calfee, J. E. (1986) "The Ghosts of Cigarette Advertising Past," Regulation (November-December).

Calfee, J. E. (1987) "Cigarette Advertising Regulation Today: Unintended Consequences and Missed Opportunities," forthcoming in Advances in Consumer Research, 14 (this issue).

Creighton D. E. and P. H. Lewis (1978) "The Effect of Different Cigarettes on Human Smoking Patterns," in Smoking Behavior Physiological and Psychological Influences, Raymond E Thornton, editor, Churchill Livingston: London, 289-300.

Federal Emergency Management Agency (FEMA) (1984), Fire in the United States Fifth Edition, (FA 68) 60 pages.

Ford, Gary T., John P. Brown and John E. Calfee (1986) "The Effects of Reduced Fla lability Cigarettes on Smoker Behavior," report prepared for the National Bureau of Standards, October.

Guillerm R. and E. Radziszewski (1978) "Analysis of Smoking Patterns Including Intake of Carbon monoxide and Influences of Changes in Cigarette Design," in Smoking Behavior. Physiological and Psychological Influences, Raymond E Thornton, editor, Churchill Livingston: London, 361-370.

Hall, John R. and Susan G. Helzer (1983) "Civilian Residential Fire Fatality Rates: Six High-Rate States Versus Six Low-Rate States," Federal Emergency Management Agency, NBSIR 83-2754, 23 pages.

Ippolito, P. K., and R. A. Ippolito (1984) "Measuring the Value of Life Saving from Consumer Reactions to New Information," J. of Public Economics 25, 53-81.

Kozlowski, Lynn T. (1981) "Tar and Nicotine Delivery of Cigarettes: What a Difference a Puff Makes,," Journal of the American Medical Association, 245, (Jan. 9), 158-59.

Kozlowski, Lynn T. (no date) "Less-Hazardous Tobacco Use as a Treatment for the Smoking and Health Problem," unpublished manuscript, 50 pages.

Lago, Armando (1986), "Notes on April 14 Meeting with Dr. John Krasny," (June 13), mimeo, 5 pages.

Lewit, Eugene M. and Douglas Coate (1982) "The Potential for Using Excise Taxes to Reduce Smoking," J. of Health Economics, 1, 121-145.

Maxwell, jr., John C. (1981), "The Maxwell Report," Advertising Age, (December 7.)

National Center for Health Statistics (1986a) "Trends in Smoking, Alcohol Consumption, and Other Health Practices Among U.S. Adults, 1977 and 1983," Advancedata, No. 118, (June 30).

National Center for Health Statistics (1986b) Health 86, preliminary data.

Participants of the Fourth Scarborough Conference on Preventive Medicine, "Is There a Future for Lower-Tar-Yield Cigarettes?" The Lancet, November 16, 1985, pp. 1111-1114

Porter, R. (1986) "The impact of government policy on U. S. cigarette industry," in Ippolito, P. and D. Scheffman, ed's. Empirical Approaches to Consumer Protection Economics, Bureau of Economics Report, Federal Trade Commission, 447-482.

Rawbone, Roger, G., K. Murphy, M. E. Tata and S. J. Kane (1978), "The Analysis of Smoking Parameters, Inhalation and Absorption of Tobacco Smoke in Studies of Human Smoking Behavior," in Smoking Behavior. Physiological and Psychological Influences, Raymond E Thornton, editor, Churchill Livingston: London, 171-194.

Rubin, Paul, and Mark Cohen (1985) "Private Enforcement of Government Regulation," Yale Journal of Regulation, 2.

Ruegg, Rosalie, Steven Weber and Barbara Lippiatt (1986) "Cigarette Ignition Propensity: A Benefit-Cost Study, Second Progress Report," Nation l Bureau of Standards, July 31, mimeo.

United States Public Health Service (1981), The Health Consequences of Smoking: The Changing Cigarette, U. S. Department of Health and Human Services.