Public Policy Sources #40: Is tobacco regulation useless?

[Previous] [Contents] [Next]

To understand the history of tobacco regulation it helps to know whether government is acting in the public interest or whether there are some special interests pushing government along its course. In the previous sections, I suggested that government regulation and taxation of tobacco were suspect because these efforts gather steam after tobacco consumption is in decline. Governments efforts also seem suspect when we examine their track records. Some case might be made for regulation, even if it arrives late on the scene, provided that regulation can be proved to reduce smoking and increase the health of smokers. Such proof is hard to come by. Let us look at a few of these regulations to see what effects they have had.

Advertising bans

In a 1998 review I made of dozens of studies of advertising bans, I found mixed and largely negative evidence that smoking bans reduce the amount of tobacco consumption (Palda 1998). In Canada, after the 1988 law banning tobacco advertising, tobacco consumption rose. Even with the best statistical techniques, event studies trying to find that advertising bans reduce smoking can show no results or false positive results.

False positives arise because bans often go hand-in-hand with grassroots movements to stop the proscribed activity. In a study of anti-alcohol policies and vehicle fatalities, Ruhm (1996) explains that statistical models of alcohol consumption fail to account for rising social stigmatization of excessive drinking. If the ban is the result of social stigma, then the ban itself will not be responsible for a fall in the targeted behaviour. The ban will be a stand-in for a crest in social disapproval of smoking. Put more technically, any study that finds a positive relationship between advertising bans and reductions in smoking may not be accounting for the endogeneity of advertising bans. Bans do not come into being by themselves. They are the result of pressure by broad coalitions of interest groups such as physicians, teachers, parents, and social activists and it may be the disapproval of these groups for smoking that reduces smoking. Karen Conrad and her colleagues found that young people decide to smoke after weighing the opinions and actions of their parents and peers (Conrad, Flay, and Hill 1992). An advertising ban may be simply the raising of a flag after a long battle. It is a symbol of victory, not the cause. Philip DeCicca and his colleagues from the Cornell University School of Human Ecology found that the most powerful variable explaining smoking onset was a variable for social disapproval (DeCicca, Kenkel, and Mathios 1998).

It is even possible that anti-smoking laws have slowed the downward trend in smoking. An advertising ban might endanger health by blocking information cigarette companies provide about the safety of their product and the dangers of rival cigarettes. Ever since the 1920s, cigarette companies have tried to give consumers information about what their product will do to health. Initially, cigarette advertising proposed that cigarettes in moderation could be good for health. This type of advertising did not last long. Cigarette companies discovered through extensive marketing efforts that smokers were worried about tar contents of cigarettes and did not like to be "winded" by smoke. In response, companies bred light forms of tobacco and invented and refined the filter tip. In a recent history of the cigarette, Richard Kluger (1996) describes the hundreds of millions of dollars that cigarette companies have sunk into researching the health effects of cigarettes and the further millions they spent to market cigarettes with safety features they believed were important to consumers. Cigarette advertisements from the 1930s to the 1970s reproduced in Viscusi (1992) emphasize a negative trait of smoking: the danger to health. As Business Week wrote in 1953: "Why has the industry persisted in this negative form of advertising even when, as tobacco growers and others complain, it hurts the trade by making people conscious that cigarettes may be harmful?" (Business Week 1953: 68). In his book Smoking: Making the Risky Decision, Kip Viscusi, a moving force behind the academic Journal of Risk and Uncertainty and a leading researcher in the way consumers perceive risk, writes:

Advertising the comparatively lower risk of a brand name may hurt that brand if individuals pay excessive attention to the presence of risk rather than its lower level. Moreover it may stigmatize the entire product group. Although safety appears to have become more marketable in . . . the 1980s and 1990s . . . the advertising environment of the previous decades in which these ads appeared was quite different. The willingness to mention health risks in an ad is an implicit recognition that the safety of the product was a salient consumer concern. (Viscusi 1992: 37-38)

It is possible that the health advertising wars of the cigarette companies reduced consumption of cigarettes by bringing to light new information about what cigarettes do to one's body. This information highlighted a new feature of cigarettes. As explained earlier, new features are what change demand. There is no definitive answer to the question of whether negative advertising wars reduced industry demand but we know that shortly after these wars the market shares of cigarettes yielding 15 milligrams of tar or less grew from 0.3 percent in 1964 to 27.5 percent in 1978 (Schneider, Klein and Murphy 1981). From 1955 to 1978, pounds of tobacco per cigarette in the United States fell from 0.00263 to 0.00173. There was also a major shift to filtered cigarettes. These were the major methods by which cigarette companies reduced tar and nicotine contents. Even if cigarette consumption did not fall due to the advertising wars, the consumption of features of cigarettes thought to be harmful fell.

Tax increases

There is evidence that taxes can discourage smoking. The evidence is mixed but stronger than the evidence for the effects of advertising bans on smoking. The problem with concluding that by reducing smoking, tax increases have increased the health of smokers, is that we do not know whether smokers have diverted their spending away from unhealthy smoking to some other unhealthy activity. Even if smoking bans and all anti-smoking measures were successful in blotting out smoking, it is not clear that smokers would live longer or more happily than before. Research by Sam Peltzman (1975) suggests that the introduction of safety belts and airbags in cars may not have reduced by much the risk of fatal accidents. Even though engineering studies prove these features will reduce the risk of death at any given speed, Peltzman and later researchers concluded that consumers were offsetting their increased safety by driving faster or more recklessly. The "law of offsetting behaviour" suggests that consumers are aware of risks and adjust the riskiness of their environment to a precise level that suits them. It is possible that smokers would offset any ban by substituting other high-risk substances such as alcohol. In the only study of its sort of which I am aware, Goel and Morey (1995) calculated for the United States that a 15 percent rise in the price of cigarettes would lower cigarette demand by 15 percent and raise alcohol demand by 20 percent among young people. While much work remains to be done to determine how tobacco prices turn smokers into drinkers, Goel and Morey's results force us to acknowledge that it may not be possible to control people bent on taking risks. As Viscusi (1992) has shown, smokers in every part of their lives--even the types of work-places they choose--are bigger risk-takers than non-smokers.

[Previous] [Contents] [Next]

info@fraserinstitute.ca You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.

If you know someone who would be interested in this web page, please enter their email address below, and we will forward this URL to them: Email Address:

Last Modified: Thursday, August 5, 1999.