Tobacco Documents Online

T11E WALL STREET JOURNAL EUROPE TUESD=IUNE 7, 1994 By RoaeaT J. BAaao I admit that I dislike cigarette smoke. I also do not understand why apparently rea- sonable people, Including a few famous economists, voluntarily expose themselves to the hazards and general unpleasantness of smoking. But I recognize these tastes as my own and would never propose to use the gov- ernment to get smokers to "do the right thing." Many people In the U.S., Including some government offlclals, are more eager for the government to Intervene to reduce cigarelte consumption. Flush, apparently, with Its bril- liant successes In the control of Illicit drugs-such as marijuana, cocaine and hero- In=1he head of the U.S. Food and Drug Administration would Ilke to classify ciga- rettes as drugs. This classification Is the first step on the slippery slope to prohibition. The experience with drug enforcement shows that prohibltlon of recreational drugs drives up prices, stimulates Illegal activity, : has only a moderate negative effect on con- i sumpllon, and Imposes unacceptable costs in terms of high crime, expansion of prison pop- ulations and deterioration of relations with ( lhe foreign countries that supply the out- lawed products. A better ldea would be to leave intacl the present regulatory structure for cigarettes-which Includes substantial but not outrageous tax rates and restrictions on sales to minors-and apply this apparatus to the currently Illegai drugs. Although smoking has been demonstrated to harm smokers, the regulation of this acliv- Ity slill runs Into the traditional libertarian objection that well-informed persons ought to be allowed to assume these risks If they wish. (The familiarity of this point does not make It less compelling.) Thus, advocates of Increased regulation have found It important to argue that smoking also Injures "innocent partles." notably through secondhand smoke. The U.S. Environmental Protection Agency (EPA), relying mainly on evidence from non- smoking spouses ot smokers, reported last year that secondhand smoke was a serious health hazard. An Incomplete Case Even 11 these hazards exisl-and the med- Ical evidence on this point is seriously In doubt-the case for government regulation would remain Incomplete. In a restaurant, for example, the owner can determine whether to make the premises nonsmoking or to have non- smoking sections. This decision will be guided -If the owner Is a greedy maximizer of profil= by a weighing of the benefits from a smoke- free environment to nonsmokers (expressed in terms of lheir willingness to pay and to patron- ize the restaurant) against the costs to smok- • Send Regulations Up in Smoke ers lexpressed again In terms of thely demand for the restaurant's services). The restaurant will be smoke•free it lhe gains to nonsmokers- Including the value of reduced medical rlsks- exceed the costs to smokers. One would also expect some restaurants to cater to smokers and some to nonsmokers. The same argument holds for airplanes, employers, commerclal•bulidings and so on. It also applies to the Interactions among spouses; even If a smoker damages the health of his or her nonsmoking spouse, the pair does not need the assistance of the govern- ment to decide whether smoking should occur tfl published in 1990 as "Environmental Tobacco Smoke." With respect to lung cancer, the con- clusion (page 111) was "The weak and Inconsislent associations seen In the epideml- ologlc studies of ETS lenvironmental tobacco smokel and lung cancer, the fact that bias and confounding cannot be ruled out, and questions about the reliability of the reported results, all indicale that these data do not support a judgment of a causal relationship between exposure to E1'S and lung cancer." The difficulties In establishing a link between ETS and lung cancer are illustrated by a November 1992 study In the American The statistical evidence f or health risks f rom second-hand smoke is extremely weak, even by the standards of an empirical economist. In the household (or, Indeed, whether they should be marrledl. The Interactions between smoking par- ents and nonsmoking children are more complex. Parents typically care about their children and take account of the effects of their actions on the children's welfare. A smoker has therefore determined that the benefits of llghting up are greater than the costs Imposed on the children. (The main cost Is probably not from secondhand smoke, but rather the Increase In the probability that the child will subsequently become a smoker.) Unlike a spouse, however, the child Is not a volunlary, well-Informed participant In a con- tract and would be unable to express an aversion to the parent's smoking In terms of some kind of future promise to pay. The government has, accordingly, often prevented parents from making unrestricted decisions on matters that affect their chil- dren's welfare, as in laws on compulsory schooling and work hours. Since most parents are better equipped and motivated than the government to consider their chlldren's well- being, il is unclear that this type of regulation has net social benefits. Hw, in any case, an extension of Ihis reasoning could be used to rationalize restrictions on parental smoking in the homo. I suppose that these restrictions could be administered by an EPA police force. - Despite the recent assertions by Ihe EPA, I the statistical evidence for health risks from secondhapd smoke Is extremely weak, even I by the standards of an empirical economisl. The evidence comes primarily from epideml- ologic studies, most of which rely on the experience of nonsmoking spouses of smok- ers. The research studies available through I 1989 were carefully evaluated In an interna• tional symposium at McGill University and Journal of Public Health that includes a rela- tively large sample of 618 nonsmoking or ex-smoking whlte Missouri women who were diagnosed with lung cancer between 1986 and 1991. The authors find that the spouse of someone who smokes at low or moderate lev- els has a 3090 lower probability of lung cancer than the spouse of a nonsmoker, whereas the spouse of someone who smokes at a high level has a31196 higher probability. The authors choose to emphasize the last result, but the findings as a whole make no sense In terms of a causal relation. Remarkably, the conclusion was that "comprehensive actions to limit smoking In public places and work sites are well' advised." Such Judgments, like those of the EPA, must come from prior beliefs, rather than from the sclenlific results. The difficulties with the 1992 Missouri study typify the problems in this field: a grealer tendency to publish results that show significant and harmful effects of ETS, a ten- dency for nonsmokers to misreport their status (combined with the likelihood that spouses of smokers will be smokers), and a failure to take account of socio•economlc characteristics that may skew the results. Large effects from secondhand smoke on health outcomes are Improbable, in any case, because the exposure levels are too low. Ac' cording to "Environmental Tobacco Smoke" (page 229), active smokers receive 100 to L1100 times the nicotine of persons exposed to ETS. Unlike cyanide, nicotine in small doses Is not lethal; even In active smokers, the adverse effects take many years to materialize. Thus. It is implausible that doses of t/IOOIh or 1/1,OOOth of those experienced by active smokers would have any noticeable health Impact, such as that claimed by the EPA. A connection between parental smoking and children's health has more foundation- for example, "Environmental Tobacco Smoke" Ipage 227) cites adverse consequences on children's "respiratory health." lnterest- ingly, however, the effect (s detectable only In young children; the link between health out- comes and parental smoking seems to disap- pear once children are older than age two. A possible explanation Is that smoking by the mother during pregnancy, rather than second- hand smoke, Is the main culprit. Another approach to regulating cigarettes would focus on the well-documenled harmful effects of smoking on smokers. The case for government Intervention could then come from paternalism; the government knows better than ignorant indivlduals about what makes them better off. Suspect Argument A superficially more compelling argu- ment is that a person's health is every- body's business because society pays for al least a portion of most people's health care. (One of the problems with socialized medi- cine is that this linkage becomes stronger.) This point has been used to justify manda- tory use of seat belts in cars and of he4 mets on motorcycles. The argument is dubl- ous in those areas, but il is particularly suspect for smokers, who tend to have rela- tively brief terminal Illnesses-at least from lung cancer-and often die before they col- lect much In retirement benefits from So- cial Security. From this perspectlve. Ihe usual economic analysis would argue for a subsidy to smoking. The final recourse Is to admit that the scl- enliflc evidence on the health hazards of secondhand smoke Is (llmsy, but to point oul-correctly-ahat this evidence also does not conclusively rule out a small effect. Why not then err on the side of safety and restrict people's exposure to secondhand smoke? One problem with this logic is that It applies to a virtually unlimited array of activ- Ities, starting with global warming and depletion of the ozone layer. The U.S. would be In serlous trouble If it spent liberally and took restrictive actions In every area in which a semirespectable theory of health risk has been advanced but has not been shown sciem lificaily to be significant. In no time at all, Imaginative environmentellsls would exhaust lhe entire gross national product on activities with low or negative social rates of return. Afr: Barro, a contributing editor of ihe IS'oll Street Journal, is a professor of ecouorn- ics at llarvard University and a fellow of Ihe llouver huhtution at Stan/ord Umnersau