sample="quota" bates="TIMN0018869" isource="ti" decade="19xx" class="ui" date="19000000" ECONOMIC ASPECTS OF SELECTIVE EXCISE TAXATION, WITH SPECIAL REFERENCE TO TOBACCO TAXATION Richard E. Wagner Florida State University The purpose of this essay is to describe the central economic aspects of excise taxation, particularly as this form of taxation pertains to the taxation of tobacco products. The first section provides some background information on the taxation of consumption in general and of tobacco in particular. The following four sections explore the main elements of the economic aspects of excise taxation. The first of these topics to be examined, in Section II, concerns who actually pays the revenues that are collected by government. Federal and state governments presently collect about $8 billion from the taxation of tobacco products. The economic analysis of incidence seeks to determine just how the burden of this tax, or, more generally, of any tax, is likely to be borne by people in their different capacities as consumers, as workers, and as investors. The second main topic in the economic analysis of excise taxation explores the possibility that the burdens imposed on people by a selective excise tax will exceed the amount of revenue collected by government. This possibility is described as excess burden, with the adjective used to indicate the possibility that the loss to taxpayers may exceed the revenues collected by government. Excess burden, then, represents a form of waste produced by a particular method of taxation, and Section III of this essay describes what is meant by excess burden and explains how the presence of excess burden is generally seen as an argument for the general as against the selective taxation of consumption. The one line of argument that lends support to the use of selective excise taxation is what is referred to as corrective taxation, and this topic is examined in Section IV. The argument for corrective taxation is based on the possibility that people may undertake some activities for which they do not bear the full cost, because some of that cost is borne by people instead. In such a setting, a properly chosen excise tax is one possible means of getting people to take into account the full cost of their activities, though it is not the only means of doing so, nor is it necessarily the best means. For example, if people's consumption of alcoholic beverages imposes costs on others, as through an increase in automobile accidents, an excise tax on those beverages might be seen as a means of placing the cost of those accidents on the consumers of those beverages. The general idea of corrective taxation is one thing, its applicability to any particular case is another. The fifth and final part of this essay will examine the applicability of the theory of corrective taxation to the consumption of tobacco products. In this regard, there have been a number of claims that cigarette smoking entails a social cost or less on the order of $75 billion annually, through such things as lost production and higher medical expenses. Section V will examine the applicability of the theory of corrective taxation to the consumption of tobacco products (A brief annotated bibliography will follow the essay.) I. General and Selective Consumption Taxation There are many particular ways in which governments tax consumption, but these various ways are commonly collapsed into two main categories: general and selective taxes. The general taxation of consumption has received increasing attention in recent years, mainly out of a growing recognition that the American system of income taxation taxes saving too heavily relative to consumption. Income that is saved is taxed twice, once when it is earned and again when it earns interest. By contrast, income that is consumed is taxed only once, when it is earned Such a tax penalty on saving reduces our rate of capital accumulation and, hence, retards our economic progress. Placing more emphasis on the taxation of consumption would represent an approach to taxation that would not impair incentives to save and to invest. Within the revenue system of the federal government, the taxation of consumption could be instituted either by allowing people to deduct their savings in arriving at their taxable income or by exempting interest from the base of the personal income tax. Alternatively, some people have looked to the value-added method of tax collection, which in the way it is commonly administered in many nations is essentially equivalent to a general tax on consumption. Relatedly, some people have proposed to replace the present federal tax on corporation income with a tax on value added, under which firms would pay a tax based on the difference between the value of the products a company sells and the amount it spends for its various inputs. This line of argument is illustrated further by Figure 2, which shows the symmetry with the excess burden argument developed around Figure 1. In the absence of tax, consumers would buy 900 million bottles of whiskey per year at a price of $5 each. The sacrifice of other output entailed with that much consumption of whiskey is $10, $5 of which is paid by consumers and represents the cost of such inputs as corn and glass, and $5 of which is borne by third parties through damages to person and property. By reducing the consumption of whiskey, alternative output valued at 410 could be secured by sacrificing something worth only $5. And it will be possible to secure such a gain until consumption is reduced such that its value to consumers equals its full cost. This happens, in the illustration at hand, at an output rate of 600 million bottles per year. In this case, consumers pay $6 billion ($10 x 600 million) per year for whiskey, of which $3 billion ($5 x 600 million) is collected as tax and $3 billion (45 x 600 million is received by producers. The gain that results from the removal of excess burden, and which is made possible by the corrective tax, is approximately $750 million, as can be computed by the formula formulaname Some Complexities in the Theory of Corrective Taxation. To argue that if producers are able to ignore some costs of their production, a properly chosen tax will lead them to make the same choices regarding production that they would make if they took those costs directly into account is, of course, a truism. However, it does not follow that the actual imposition of an excise tax for what might seem to be corrective purposes actually performs in the required manner. The theory of corrective taxation is generally much simpler than the reality such efforts at taxation address. There is the obvious question of how a correct, or even a reasonably informed, judgment of external cost might be formed. A market system produces such knowledge as a by-product of its internal workings. If a firm fails to cover its cost of production, this means that the value that consumers place on the firm's output is less than the value they place on the alternative output that was sacrificed to produce the product in question. And the converse conclusion can be reached in the case of a firm that makes a profit. When a direct market test is absent, as is necessarily the case with external costs, for the very concept implies that some aspects of resource usage are not reflected in market transactions, the problem of securing knowledge is much more difficult. Moreover, a market system provides a strong incentive for people to make choices knowledgeably, because poor choices will result in losses. A growing body of contemporary literature on political economy has explained why political incentives generally operate less strongly than market incentives. In politics, the costs and gains of good and bad choices are concentrated less strongly on those who make the choices, because more of those costs and gains are diffused generally throughout the citizenry. The loss from a governmental choice that is more costly than it need be is spread over all taxpayers, rather than being concentrated on those responsible for making that choice. Beyond this problem of knowledge, but yet related to it, a variety of complex issues arise about the particular form a corrective excise tax should take. For instance, suppose it were concluded that the external cost of whiskey consumption were $3 billion annually, in which case it could be argued that a tax of $5 per bottle is the right amount. But it is really not so simple as this. Different acts of consumption by different consumers entail different external costs. Each bottle of whiskey that is consumed does not entail an equal external cost, even if the average external cost is $5 per bottle. Some people may consume a bottle evenly over a month, and always at home in the evening before retiring. Others may consume a bottle entirely in one evening at a party, after which they drive home. The latter but not the former type of consumption has an external cost associated with it. The social or general gain, as distinguished from the transfer, comes about precisely because the increase in cost to drinkers reduces the amount of drinking, and hence, and more importantly, the amount of damage. The corrective tax reduces the amount of accident damage by $1.5 billion ($5 x 300 million). Besides this reduction in damage, the tax also transfers income from drinkers to victims. The amount of transfer is $3 billion ($5 x 600 million). This transfer is a by-product of the program that imposes a tax on drinking to compensate victims for damages suffered. Both effects are present in the same program in this particular illustration. There is a transfer from drinkers to victims of $3 billion per year; here, the gains to one set of people equals the loss to the other set. There is also a shift in the pattern of resource usage in society, with resources being shifted away from the production of whisky, automobiles, and hospital facilities into other activities that consumers value more highly. What a system of subsidized medical care basically amounts to is a transfer of income from people who make below-average claims. With respect to smoking, an external cost would arise from such a transfer program only to the extent that the subsidization of medial care would induce people to smoke more, and with the increased smoking leading to increased illness and higher medical expenses. The mere fact of subsidy does not represent an external cost, the removal of which will provide a general gain for the members of society. The mere fact of subsidy means only that its removal will provide a gain for those who were making a below-average use of medical resources, and this gain will come at the expense of those who were making an above-average use. It is highly unlikely that the presence of subsidized medical care induces people to smoke more, and thereby to become more sickly. Removal of subsidized medical care would surely have little if any effect on the amount of smoking and, hence, on the amount of smoking-induced illness. Removing the subsidization of medical care will, of course, reduce the claims people make upon medical resources. Smokers as well as everyone else will make more use of medical care when it is subsidized than when it is not. The subsidization of anything increases the amount of it people will choose to buy. This is precisely why medical costs have risen from about five percent of national income before the advent of signification subsidization in the mid-1960s to about ten percent now. But such subsidies are unlikely to increase the amount of smoking and, hence, the amount of smoking-induced illness. Medicare and Medicaid encourage people to make greater use of medical resources, but they do not encourage people to become more sickly than they would otherwise have been. Smoking has little if any external cost, even if medical expenses are subsidized. The increased use that smokers make of medical resources is identical to the increased use that others make as well, and this is an inevitable outcome of any program of subsidization. VI. Concluding Remarks There is no way in which the $75 billion figure bandied about so prominently as an external cost of smoking can be said to constitute a general, social loss. The elimination of smoking will not provide a general bonus of $350 to each person in society. For the most part, any such gain will accrue to smokers, though at the sacrifice of the benefits they receive from smoking. A selective excise tax on tobacco does not seem warranted by a principled application of the theory of corrective excise taxation. Rather, the selective taxation of tobacco seems more likely to represent the imposition of a punitive tax on smokers, who are a disfavored and increasingly pilloried minority in present American society. It is this weak position of smokers in conjunction with the taxing-and-spending biases of government, and not some effort to promote matters of common or general value, that seem best able to explain both the present level and what would appear likely to be the future prospects for the taxation of tobacco products. A Brief Note on References In this brief bibliographical note I shall describe a few sources that are pertinent to various of the topics I discussed in the essay. Vast amounts of information about taxation at all levels of government are provided in the biannual publication, Facts and Figures on Government Finance, which is published by the Tax Foundation, Inc. of Washington, D.C. Richard E. Wagner, Public Finance: Revenues and Expenditures in a Democratic Society (Boston: Little, Brown, 1983), is a textbook for advanced undergraduates that explores the economic and political aspects of taxing and spending. For one study of the ability of excise taxation to modify the particular characteristics of products, see Charles J. Goetz and Italo Magnani, "Automobile Taxation Based on Mechanical Characteristics: Evidence from the Italian Case," Public Finance, 24 (1969), 480-94. For an explanation of why the cost of producing one product is equivalent to the value that consumers place on the output that is thus sacrificed by those resources that could otherwise have produced that alternative product, see James M. Buchanan, Cost and Choice (Chicago: Markham, 1969). The reference to Luce and Schweitzer noted in the text is Bryan R. Luce and Stuart O. Schweitzer, "Smoking and Alcohol Abuse: A Comparison of their Economic Consequences," New England Journal of Medicine, March 9, 1978, 569-71. Greater detail is presented in Luce and Schweitzer's The Economic Costs of Smoking-Induced Illness, National Institute on Drug Abuse Research Monograph Series, No. 7, published by the then Department of Health, Education, and Welfare, in 1977. The reference to Burch's study cited in the text is P. R. J. Burch, "Smoking and Lung Cancer: The Problem of Inferring Cause," Journal of the Royal Statistical Society, Series A, 141 (1978), 437-58. Immediately following this study is a set of commentaries, some supportive and some not, and running from page 458 to page 477. On the study for Swiss women, smoking, and cardiovascular disease, see E. Guberman, "Surprising Decline of Cardiovascular Mortality in Switzerland, 1951-76," Journal of Epidemiology and Community Health 33 (1979), 114-20. For an application of the same framework commonly used for smoking to the consumption of alcoholic beverages, see Ralph E. Berry, Jr. and James P. Boland, The Economic Cost of Alcohol Abuse (New York: The Free Press, 1977).