Tobacco Documents Online

Although a careful analysis may suggest that the regulation will not have counter-productive effects on mortality, what these estimates do suggest is that the assessment of the benefits and the costs of regulation must be more thorough and accurate than those that have been prepared. It is essential to obtain unbiased assessment of the true expected risks, the value of the risk reduction to society, and the associated costs of the regulation. The current analysis provides only a sketchy treatment of these components. In effect, this analysis takes as a given that society should proceed with the regulation provided that the agency can show that there is some significant risk reduction. What such a calculus neglects is that there may be other risk effects of the regulation that are adverse and may in fact outweigh the purported beneficial impacts. 5. The cost estimates defy economic logic and are not credible. In its assessment of the indoor air quality regulation, 03HA breaks out three different classes of effects. First, there will be health benefits to workers, which presumably should have some value to them and would be reflected in compensating wage differentials. Second, the regulation will have direct costs, which the agency estimates to be $8.1 billion annually. Third, the regulation will have cost savings, which the agency estimates to be $15 billion annually (page 16002). For the purposes of this discussion, let us neglect the health effects and focus on the cost savings. What this analysis 10 . TIBR 0007367

suggests is that the regulation would cost businesses $8.1 billion, but would save them $15.1 billion annually. Firms' failure to take advantage of this opportunity to make almost $2 for every dollar expended cannot be attributed to the market imperfections alleged by OSHA. These market imperfections, such as an assumed lack of worker information about the risks of ETS, pertain to the value of the health effects and their transmittal through the compensating differential mechanism. Instead, what we have is an assumed market failure on the part of firms, which fail to realize that they could make $15.1 billion more every year in return for an annual expenditure of $8.1 billion. There is no reason why firms would not take advantage of such an opportunity, if it in fact existed. The major implication of these numbers is not that there is a large market failure that merits government intervention. Rather, it is that the components of the OSAA regulatory analysis simply do not make economic sense. This lack of internal consistency calls into question the overall legitimacy of all the cost estimates that have been put forth. 6. Market forces will lead to the adoption of smoking restrictions and smoking policies that are desirable. Individuals who smoke in the workplace will generate a perceived environmental risk to their nonsmoking workers. If nonsmokers believe that there is a risk associated with ETS, they will demand a compensating differential to work in such environments. Firms will have an incentive to institute policies " TIBR 0007368

to limit workplace smoking or to provide smoking areas in much the same manner as they would to control other public aspects of the workplace environment, such as the level of noise and noxious fumes. In effect, the compensating differential mechanism gives firms a financial incentive to carry out the bargains that are appropriate given the perceived risks that are being generated. The OSHA proposal recognizes that these mechanisms are pertinent (page 16008) but discusses various market forces that could impede these mechanisms. Much of this discussion of market imperfections draws upon analyses that I first advanced in the economic literature. For example, the discussion of marginal workers and worker mobility on page 16008 first appeared in my book, Emnlovment Hazards: An Investigation of Market Performance (Cambridge: Harvard University Press, 1979), and in my article, "Unions, Labor Market Structure, and the Welfare Implications of the Quality of Work," Journal of Labor Research, Vol. 1, No. 1, Spring 1980, pp. 175-192. However, this discussion of market imperfections is much more pertinent to dimly understood risks than to smoking. The critical link in all these arguments is information. The discussion of the proposed rule correctly notes that in general workers do not always have "perfect knowledge of job risks." In this case, the risk involved is that of smoking. My past research on the perception of smoking risks indicates that throughout society there is not an accurate understanding of these risks. However, the nature of the bias is the opposite of 12 TI$g 0007369

the direction that would be needed to have a market failure that warrants government intervention. As I document in my book, Smoking: Makina the Riskv Decision (New York: Oxford University Press, 1992), in 1985 the average adult in the population assessed the risk of lung cancer from smoking to be 0.43, whereas estimates based on reports of the Surgeon General at that time placed the risk between .05 and .10. Even more recent official estimates of the lung cancer mortality risk are only as high as .06-.13, so that the perceived risk of lung cancer greatly exceeds the risk value that is assessed by the U.S. Surgeon General. To test the sensitivity of this result, I have explored other measures of the risk as well. In particular, I have examined the overall total smoking- mortality risk perc=ption in society, which I found to be 0.54. In contrast, the overall mortality risk to the smoker based on estimates by the U.S. Surgeon General ranges from 0.18 to 0.36. Thus, even in the case of the total mortality risks of smoking, there is substantial overestimation of these risks on the part of the population. A third measure of the risk perception pertains to the loss of life expectancy involved. In particular, do people fully understand the extent of the loss of life that the Surgeon General assesses based on smoking behavior? In the case of public perception regarding loss of life, one finds a result that parallels attitudes about the alleged risks of smoking: the overall assessed life expectancy loss is 11.5 years, whereas the 13 TIBR 0007370

estimated life expectancy loss based on reports by the u.s. Surgeon General range from 3.6 to 7.2 years.' What distinguishes the potential risks of smoking from other health hazards in the workplace is that they have been highly publicized. Indeed, the potential hazards of smoking are among. the most widely publicized risks in society. I have documented the substantial publicity given to smoking in Viscusi (1992). The current public debate over smoking and ETS involves prominent officials from the U.S. EPA, the U.S. Occupational Safety and Health Administration, and the FDA. Most workers would certainly be acutely aware of this debate. Indeed, OSHA cites evidence indicating that "88 percent of nonsmokers are aware of the negative health consequences of ETS" (p. 16007). These results are inconsistent with OSHA's view that the ETS risks are ignored. The opposite problem may in fact be the case. Unfortu- nately, the practical effect of the substantial publicity given to ETS is not necessarily the fostering of accurate risk perceptions. Literature on the economics and psychology of risk perception clearly documents that highly publicized risks tend to be overestimated. The potential hazards of smoking are among the most highly publicized and widely discussed risks in our society; in consequence, it is not surprising that the net result is that these risks are currently overassessed. What this overassessment implies from the standpoint of compensating differentials is that the market response to the 'See page 80 of W. Kip Viscusi, Smoking. 14 TIBR 0007371

risks of smoking will in fact be excessive. Rather than leading to too little market accommodation of the preferences of nonsmokers, this over-perception of the risk will lead to excessive restrictions on smoking in the workplace. Thus, the actions that have been undertaken may already be too stringent from the standpoint of their overall social desirability. As is indicated in the discussion of the proposed rule, the private market agreements have already been quite extensive in this area. OSHA summarizes the nature of these actions (page 16007): "A 1991 survey of company smoking policies shows that of the 85 percent of firms with smoking policies, 34 percent have complete bans and another 34 percent prohibit smoking in all open work areas. Over 90 percent of non-manufacturing establishment have smoking policies [H-030EX.77]. "Workplace smoking policies are more common in larger [than in smaller] businesses. In a survey of personnel managers, 63 percent of those with 1,000 or more employees report having a smoking policy compared with 52 percent of companies with fewer employees." Thus, we have not only a substantial role of smoking policies but also substantial heterogeneity of smoking policies across the labor market. One would expect, for example, that smoking policies would be more prevalent for large workplaces. There should be economies of scale in providing smoking areas in such work environments and also a greater need to standardize smoking policies as opposed to letting the voluntary discussions of small worker groups address the appropriate smoking policy on a decentralized basis. 15 TIBR 0007372

Another class of market imperfection that is discussed in the OSHA proposal (page 16008) pertains to the lack of "perfect employee mobility between jobs." This limitation pertains not only to potential ETS risks but also to every workplace hazard. If there were in fact worker immobility, one could use this rationale to justify the regulation of any type of workplace risk in any industrial situation. However, even though worker mobility is not perfect it is indeed substantial. There is little lifetime commitment to jobs in the U.S. economy. Moreover, I have documented in my past work that job hazards alone account for 1/3 of all manufacturing quit rates.e The textbook discussions of a coal mine worker trapped at a single firm in a company town are apocryphal and do not pertain to the 1990s. Moreover, even if there were a problem of worker mobility in which a worker began work on the job, subsequently discovered that there were hazards associated with it, and felt unable to quit because of the transactions cost involved, this stylized scenario would not fit the potential risks associated with ETS. The presence of ETS and the ventilation conditions of the workplace are readily observable characteristics that can be monitored by the worker. They are not the kind of dimly understood health risks for which there is associated the usual type of market failure. BSee W. Kip Viscusi, Risk By Choice: Regulating Health and Safety in the Worknlace (Cambridge: Harvard University Press, 1983). 16 TIBR 0007373

7. The discussion of societal externalities of smoking claims that smokers on balance impose a burden on societv, whereas in fact the opoosite is the case. Smokers pay substantial taxes specifically linked to their smoking behavior. For the fiscal year ending in June 30, 1993, federal taxes on cigarettes were $5.5 billion and state taxes were $6.2 billion. Municipal taxes were a total of $187 million, so that the total tax amount was $11.9 billion. Potentially offsetting these taxes are the effects claimed by OSHA on Social Security, welfare, Medicare, and related programs (page 16008). However, even taking the estimated hazards of smoking at face value, on balance the savings to society from the estimated early mortality of smokers offset the added cost to society in terms of higher health costs. At reasonable rates of interest, such as a 3 percent real rate of return, on balance smokers more than pay their own way. Indeed, the tax payments they provide is an additional bonus since even without taxes smokers on balance save society money. This result has been documented in a recent study by the Congressional Research Service as well as in a recent book published by the RAND Corporation.' Much of this discussion of societal externalities pertains not to the risks of ETS but rather to the risk to smokers who will be associated with the dominant external costs. This class of issues involves a much larger debate over the societal effects 9See Jane Gravelle and Dennis Zimmerman (1994) and Willard Manning, et. al., (1991). 17 TIBR 0007374

and, in particular, the role of smoking outside the workplace. Because the OSHA rule is not primarily directed toward that issue but rather toward workplace smoking, this aspect of OSHA's analysis is quite appropriately not well-developed. 8. The most important cost component, the decreased welfare of smokers and the societal loss from smoking restrictions, has been completely ignored in this analysis. As in other consumer contexts, individuals who smoke spend money on cigarettes and other tobacco products because the economic value they derive from these products exceeds their cost. In particular, they reap a welfare gain from smoking behavior. Similarly, there is a benefit to companies as well, who earn profits from producing tobacco products. There are also widespread benefits to society in terms of increased employment and tax revenues. None of these effects on either cigarette smokers or the tobacco industry has been recognized in the analysis. If smoking is banned in the workplace, for eight hours of the day, or approximately half of their waking hours, smokers will be unable to smoke. If smokers are relegated to a specific smoking area, their welfare will also be decreased and their productivity may be affected as well. The decreased consumption of cigarettes not only will impose a welfare loss on smokers, but would also lead to a reduction in the consumption of tobacco products overall. From the standpoint of a comprehensive economic analysis, it is important to 18 TIBR 0007375

calculate the lost producers' surplus (loss to companies) as weli as the loss in consumer surplus. The first matter that will be considered is the magnitude of the lost consumers' surplus from a decrease in the demand for cigarettes stemming from limitations on smoking in the workplace. From an economic standpoint, the consumer surplus that an individual consumer reaps from consuming a commodity is the difference between the price the consumer would be willing to pay for the good and the price that the consumer is actually charged. To estimate the amount of the consumers' surplus for the market as a whole, the main critical component is the shape of the consumer demand curve, or more specifically the demand elasticity, which gives the percentage change in the quantity of the good purchased that will result from a unit percentage change in its price. Table 1 presents estimates of the lost consumer surplus for five different demand elasticities ranging from -0.2 to -1.4. This range of estimates spans the estimated range in the literature, which I have surveyed in my book, Smoking: Making the Risky Decision (see especially pages 102-105). Most of the demand elasticities tend to be clustered in the range from -0.4 to -1.0 so that the elasticity of -0.7 can be viewed as the midpoint of the estimated range. The 1994 Congressional Research Service study (p. CRS-27) focused on a population-wide elasticity of -0.31. Teenagers appear to be most responsive to cigarette prices, as they exhibit demand elasticities of -1.4. What is 19 TIBR 0007376