Tobacco Documents Online

Environmental Tobacco Smoke: A Guide to Workplace Smoking Policies: Comments Robert D. Tollison Duncan Black Professor of Economics George Mason University Fairfax, Virginia 22030

I. Introduction I am Dr. Robert D. Tollison, Duncan Black Professor of Economics and Director of the Center for Study of Public Choice at George Mason University in Fairfax, Virginia. I am a past president of the Southern Economic Association, and the author of a leading college textbook on economics and of over 200 articles and scholarly publications in professional economics. I have served in government twice -- once as a Senior Staff Economist on the Council of Economic Advisers and once as Director of the Bureau of Economics at the Federal Trade Commission. As noted, I come to the issue of environmental tobacco smoke (ETS) from a background as a professional economist. It is therefore the economic aspects of the Environmental Protection Agency's (EPA) proposed guide to workplace smoking policies which I will critically assess. As a practical matter, the primary economic content of the EPA guide is contained almost exclusively in Chapter 8, which considers the "Cost Savings Related to ETS Reduction." My comments thus center on Chapter 8. My general overall assessment is that the EPA analysis of cost saving is flawed beyond repair. It is based upon a failure to apply basic economic principles, the application of which lead to the conclusion that there are presently no uncompensated costs in the workplace arising from ETS; hence, there can be no cost savings from banning ETS in the workplace.

2 II. The Economics of Indoor Air Quality The draft policy guide is premised on the notion that smokers impose uncompensated costs on their employers and therefore the government must step in and provide guidance to private industry about smoking and indoor air quality. This premise is flawed for two reasons. First, there is no persuasive evidence that smokers impose special costs on their employers or that smoking employees are more costly overall than nonsmokers. Second, to the extent smoking is an issue, employers and employees already have sufficient incentive to negotiate efficient employment relationship. an The types of estimates of the so-called costs of . smoking contained in the Report of the Office of Smoking and Health have been extensively critiqued by economists and shown to be fallacious. I will not go into this issue extensively here, but I refer the interested reader to my book with Richard E. Wagner (Tollison and Wagner 1988). In a nutshell, these alleged costs are not social costs at all; they are private costs paid for by smokers. If smokers impose productivity costs on the economy, it is they who pay for such costs through lower wage rates. When smokers pay for their own health care, privately or through private insurance mechanisms, these are private costs borne by smokers. And, indeed, if anyone would simply do the arithmetic correctly, they would see that smokers do not "overuse" programs such as Medicare. In short, smokers pay their own way in society. The "cost" estimates for smoking, such as

3 those published by the Office of Smoking and Health, are predicated upon quite profound confusions and faulty economic analysis. Other "evidence" cited in the Guide (p. 38) is, by its own admission, "anecdotal." Second, the "social costs" of ETS in the workplace are zero. Consider the ETS problem in a privately-owned firm. The owner of the firm has an economic interest in providing the kind of environment that workers want. For example, firms hiring employees in a competitive labor market will provide certain workplace environments as part of the optimizing compensation package. This could involve smoker-nonsmoker segregation on the job, investment in smoke-removal devices, paying smokers or nonsmokers a wage premium to work in a given environment, and so on. The point is that the owner will internalize the costs of smoking in the workplace. Suppose that all workers prefer to smoke on the job, but that the owner of a firm objects strongly to tobacco smoke in the workplace. Clearly, the owner must bear the costs of indulging his preferences. If he requires that his employees not smoke on the job and only offers the going market wage, no one will be willing to work for him. To induce his employees not to smoke, the owner must pay, over and above the competitive wage, a premium that is sufficient to make employment in his firm as attractive as alternative jobs where there are no restrictions on smoking. On the other hand, the owner can offer the market wage, allow smoking on the job, and invest in smoke-removal devices

that bring the air quality up to his liking. In either case, the costs of imposing a given smoking policy are internal to the owner of the firm. Suppose instead that the owner is indifferent between smoke-filled and smoke-free environments but that some of the workers wish to smoke on the job and others prefer no tobacco smoke in the workplace. How does the owner reconcile these conflicting preferences? There are several alternatives. As before, the owner can ban smoking and pay a wage premium to smokers. Similarly, he can allow smoking on the job and compensate nonsmokers. Other options are to segregate smoking and nonsmoking employees, improve the ventilation or install other smoke-removal equipment. Which of these is chosen will depend on such factors as the mix of smokers and nonsmokers in the firm's workforce, the cost and effectiveness of ventilation or other air cleaners, and the nature of the firm's production process. This last consideration will involve questions of the extent to which co-workers can be separated without adversely affecting overall productivity. Market forces will lead the owner to select the smoking policy that achieves the desired result at minimum cost. In a competitive market, one would therefore expect to observe a variety of smoking policies adopted across firms, each of which is optimal for the given circum- stances.

- 5 - In other words, with private ownership arrangements in place, the social costs of smoking in the workplace are zero. The situation in the case of publicly owned facilities, managed not by owners but by government civil servants with little motive to provide workers a smoking-nonsmoking environment consistent with the workers' preferences, is admittedly not subject to the same salutary incentive effects. But even in publicly-owned buildings, market forces can be expected to operate in a manner that minimizes any theoretical social costs that may be associated with ETS. Government agencies must compete for workers with private firms, which have strong incentives to provide employees with a work environment consistent with their ` smoking preferences. Therefore, those agencies must either provide similar accommodations to their employees or compensate them with a wage premium identical to what they could receive in the private sector, or they will be unable to attract suitable workers. Thus, in private contexts, the relevant social costs of smoking are zero. This does not mean that firms will not have smoking policies; it simply means that there is no need for additional government guidance with respect to workplace smoking policies. By the agency's own admission, the public is widely aware of the ETS issue. Further government action is unlikely to have any beneficial impact. Third, the discussion on p. 38 under the heading of "Direct Cost Savings as a Result of ETS Reduction" is based on allegations that the presence of smokers in the workforce adds

- 6 - to business costs. These allegations fall into several categories. It is claimed, for example, that smokers damage property in such forms as burned carpets and higher costs for cleaning draperies. This accounting fiction is based on a presumption that nonsmokers do not do such things. But it is totally illegitimate to add up such smoking-attributable costs as cigarette burns and call this an excess cost of having smoking workers, without also trying to consider the costs that nonsmoking workers might incur. Many nonsmokers drink coffee or soft drinks that on occasion are spilled -- sometimes even onto, or into, expensive office equipment like computers and copying machines. Another nonsmoking employee may have failed to place a ladder securely in its position in a store room, with the result being that vibrations from a nearby fork lift caused the ladder to fall and crash into some chemicals stored on a shelf, that in turn fell and broke open. This in turn shorted out the electrical system in the building, causing the loss of much computer work. Employers, of course, have incentives to mitigate accidental damage to property, whether by smokers or nonsmokers. But the kinds of accounting statistics commonly reported in these matters give no basis for concluding that smokers are a source of excess property damage. And in any event, the issues created are ones of individual business policy and not of public policy, for

- 7 - it is the owners of the businesses who bear whatever costs may be incurred. The claim that smokers increase costs of ventilation and air conditioning is the same kind of claim as the preceding one that smokers are sources of property damage. Indeed, in many cases there may be no added ventilation cost that can be reasonably attributed to smoking. To be sure, the inadequacy of ventilation in modern buildings is coming increasingly to be recognized. But in a great many cases, and perhaps in most, there is no assignable cost of ventilation for particular items. This point is one more illustration of the essential arbitrariness of allocating joint costs. To illustrate, suppose the only two things to be removed from the air are cigarette smoke and formaldehyde. It is doubtful if a less expensive ventilation system would be installed if there were no cigarette smoke to remove. And once it is recognized that indoor air contains numerous other things, the relative significance of cigarette smoke diminishes even further, which in turn makes even less plausible the claim that smoking is a significant source of ventilation costs. In addition, these costs also are borne by the individual owners of business enterprises. It is at least imaginable that some owners could decide to prohibit smoking within their business, and at the same time choose to install weak ventilation equipment. These costs savings would accrue to ~ the business owners if they were correct in their judgment that ~ ~ C l1 w1 GD

8 without smoking much less ventilation would be necessary. But if they were wrong, the growing unpleasantness and unhealthiness of the work place would increase their costs of operation, despite the saving of ventilation costs. Case studies by ventilation experts such as Healthy Buildings International suggest that this is not uncommon. In either case, business owners bear the costs or gains, and have good incentives to seek efficient outcomes. The treatment of time spent on smoking breaks is analytically equivalent to accounting for time lost because of illness. Suppose smokers take four smoking breaks per day of 15 minutes each. If the standard working day were eight hours, smokers would be viewed as effectively working only seven hours.• This would make smokers 12 percent more costly to employ than nonsmokers -- assuming that this was the only difference between smokers and nonsmokers. This assumption, however, is surely without foundation. It would have us believe that smokers and nonsmokers work equally diligently for seven hours per day, and that nonsmokers continue in the same manner for the eighth hour while smokers sit back and smoke for that eighth hour. Yet experience in any business establishment will show all kinds of ways that nonsmokers can also take what might be called on-the-job leisure: drinking coffee, making shopping lists on a computer, talking about sports with a colleague, and daydreaming while appearing to be thinking, are a few of many possible illustrations.

- 9 - Perhaps the main difference between smoking and these other illustrations is that smoking is visible and explicit. This is probably particularly so in workplaces that allow smoking only in particular areas and times. Yet even time spent smoking might well be time spent productively. In white collar occupations, this might be time spent composing a letter to be dictated upon return to the desk. In blue collar occupations, a bit of rest time periodically may lead to higher levels of overall performance than would be possible if breaks were not allowed. The discussion of "Indirect Cost Savings as a Result of ETS Mitigation," which begins at the bottom of p. 38, presents basically the same fallacy encountered in the discussion of so-called direct costs. Such costs as fire, accident, and health insurance premiums are not different, in principle, from the costs of cleaning the office curtains. The firm considers these expected costs when it hires a new employee, and presumably if the employee is hired, his or her contributions to the firm exceed these costs. Hence, such costs have been internalized by the firm, and cannot be lowered by arbitrarily banning smoking on the job. Further, as the most recent evidence on absenteeism suggests (see below), smokers are not more costly workers to employ. They are not absent (for whatever reason) more often than nonsmokers, and, hence, there is no basis to claim that smokers are sick more often than nonsmokers and thereby drive up