Tobacco Documents Online

The Costs of Poor Health Habits Z0806SL80Z

The Costs of Poor Health Habits

The Costs of Poor Health Habits £0806S680Z

THE COSTS OF POOR HEALTH HABITS Willard G. Manning Emmett B. Keeler Joseph P. Newhouse Elizabeth M. Sloss Jeffrey Wasserman A RAND Study Harvard University Press Cambridge, Massachusetts London, England 1991 I V08061C~g0Z

Preface In 1988 the five of us completed a study on the external costs of three poor health habits: smoking, drinking heavily, and not exercising. By external costs we mean the costs imposed on others by people who have these habits. Such costs arise from various sources. The ones we considered were collectively financed programs; for exam- ple, health insurance, group life insurance, retirement pensions, and lifetime taxes on earnings. We also considered the value of property damaged and lives lost in traffic accidenis as the result of excessive drinking. What we found is that smoking, heavy drinking, and lack of exercise have high costs. And although some of this cost falls on the individuals who have these habits, a major portion is paid by others. In 1989 we presented our results in two articles: "The Taxes of Sin: Do Smokers and Drinkers Pay Their Way?" in the fournal of the Ameri- can Medical Assoriation, 261 (March 17, 1989) ond "The External Costs of a Sedentary Lifestyle" in the Americwr Journal of Public Heidth, 79 Qune 1989). These articles and subsequeni reports in the media have created considerable interest in our results and the analyses that gen- erated them-in part because they have implications for taxes on tobacco and alcoholic beverages, the so-called sin taxes. The purpose of this book is to present our analyses of the costs of all three habits in greater detail, describing the results and our ap- proach, data, and methods more fully than was possible in the jour- nal articles. Inquiries about the study indicate that the book may have interest for a very diverse readership: for example, federal, state, and local policymakers; researchers in health cciences, health poli<y, and other academic disciplines; insurance cornpanies; corporate benefits managers; health and consumer advocatcs; the producers of tob.icco and alcohol products; and taxpayers. 90806S480Z

Contents Preface vii Copyright (f) 1991 by the RAND Corporation All rights reserved Printed in the United States of America 10 9 8 7 6 5 4 3 2 1 This bonk is printed on acid-free paper, and its binding malerials have been chusen for slrenglh and durability. Library of Congrrsa Catalaxingin-PuMimtion Data The Costs of poor health habits f Willnrd G. 1. An Overview 2. Conceptual Framework 3. Data and Statistical Methods 4. The External Costs of Smoking 5. The External Costs of Heavy Drinking 6. The External Costs of Sedentary Life-Styles 7. Conclusions, Limitations, and Implications 1 26 46 62 86 107 127 .i. Manning . . . let al.l. p. cm. I l d bibli hk l f d i d kf. 01^ ograp a re erences an nc u es n ex. ISBN 6b74-17485-2 (alAL paper( 1. Health behavior-Economic aspects, I. Manning. Willard 6. O IDNLM: 1. Alcohol Drinking-economics- Appendixes 2- Costs and Cost Analvsis. 3. Health Behavior. A Lit R f h U 4. Health Services-emnumies. 5. Life Style. 6. Smoking-economics. . erature eview o t e Costs of Smuking and Drinking B. Survival t arameters from the HRA Model 137 143 W 74 Cg425/ RA776.9.C69 1991 C. HIE Habit Batteries 146 338.4'33621-dc20 D. Statistical Methods 157 DNLM/DLC for Library of Congress 91-7C143 ap E. Comparability of HIE and NHIS F. Excise Taxes and Demand 161 166 G. Parameters Used in the Cost MCdel 186 H. A Note on the Alcohol Tax 190 pA Notes 195 Bibliography 209 Index 219 7Q8 UV7YOUZ if a/;OPl'

viii Preface To serve the various interests and purposes these audiences would bring to the book, we have adopted the following strategy. Chapter 1 provides a nontechnical summary of the issues, our esti- mates of the external costs, what determines those costs, and what can be done about them. It is intended for those who are primarily interested in the magnitude of the problem and its implications for policies and programs aimed at improving people's health habits, lowering the external costs, and/or making people with poor habits pay at least the equivalent of those costs. The rest of the book is intended for researchers in health sciences, health policy, and other field,, who are interested in our concepts, assumptions, data, and methods, as well as results that are not re- ported in Chapter 1. We assume throughout this portion that readers are familiar with economic concepts, terms, and analytic techniques. We have tried, however, insofar as possible, to make the discussion comprehensible to other readers who might be concerned about the nature and rigor of our analyses and, thus, the scientific credibility of our results. At the end we include several technical appendixes. We would like to acknowledge the help, advice, and support that we received in our study of health habits and in producing this book. Our work was conducted at the RAND Corporation, supported by a grant from the National Center for Health Services Research and Health Care Technology Assessment (NCHSR/HCTA), now the Agency for Health Care Policy and Research (AHCPR). We are partic- ularly indebted to Jean Carmody, Stephen Marcus, and Selwyn Wain- grow, project officers for NCHSR/HCTA, for comments and advice. We are also indebted to Erik Farag of the same organization, and to RAND for its support in producing the book. We greatly appreciate the assistance of many colleagues: Thomas Vogt (Kaiser Health Services Research Center of Portland. Oregon) provided suggestions and guidance throughout the work. We bene- fited from discussions with liernard Friedman, George Keeler, and Robert Leu. Robert Amler helped with the Health Risk Appraisal model. Jerry Green, Lester Lave, Charles Phelps, and Kenneth War- ner, as well as our RAND colleagues James Kahan, Bridger Mitchell, and James P. Smith, gave careful reviews that markedly improved our report. Bernadette Benjamin and Janet I lanley provided meticu- lous programming and data management, and Martha Cooper cheer- fully converted our various inputs into typed, legible English. Joyce Preface ix Peterson deserves our special thanks for revising the material to make it accessible to a wider readership. Finally, the opinions and conclusions expressed in this book are ours alone and should not be construed as representing the policies or opinions of the Agency for Health Care Policy and Research or its predecessor, the RAND Corporation, or any of the people acknowl- edged above. 10806S680Z

~1, An Overview Many signs suggest that the nation's health consciousness has in- creased phenomenally over the last two decades. "Health-helli" is a major category of the self-help books that pour out of publishing houses. Many news programs and newspapers regularly feature health segments. And a veritable industry ot health newsletters and magazines has sprung tip, spreading the latest word on the nature and benefits of good nutrition and other health habits. Health news has become big news, and people seem to be re- sponding- Consider just a few indicators. Most neighburhuods are now served by far more health clubs tban hospitals. That fact and the strong sales of exercise gear indicate that exercise has become a major leisure-time activity. People evidently are also taking nutri- tional advice to heart, to judge by the advertising dollars the beef and pork industries are spending to counter claims that fish and chicken are "better for you." As for smoking, health consdousness has obviously disturbed the market: 1'hilip Morris now publishes, and widely distributes, a free magazinc extolling the "smoking life- style" and championing "smokers' rights." Despite all these signals that health awareness has improved, sta- tistics indicate that we are not yet on the high road to health. In 1986, the national tab was $24 billion for tobacco products and $18 billion for alcoholic drinks.' Between 1977 and 1983, the proportion of the population who smoked dropped by ltl percent, bul the fractiun of people who were "less active than lhoir contemporaries" and the fraction of heavy drinkers rose by 12 percent and 28 percent, respec- tively (Schoenborn and Cohen, 198G).z Given these statistics, it is little wonder that improving health hab- its has become a significant concern. The prevalence of unhealthy habits has prompted large-scale efforts to promote healthy habits, and has raised questions of how government can influence behavior 608065680Z

24 The Costs of Poor I iealth Habits I An Overview . 25 About a sixth of the population is sedentary but not physically limited. Whether educational efforts would warrant the expense, whether subsidies to encourage exercise would make enough nonex- ercisers more active to justify the subsidy, are issues we must leave to others. If our estimate of $1,650 in lifetime external costs per seden- tary person is correct, a relatively small additional percentage of indi- viduals exercising could justify some subsidy. Taxes and l.ife-Styles T he costs of smoking, heavy drinking, and lack of exercise are high for individuals, their families, and others. Health promotion pro- grams attempt to reduce these costs by publicizing them and by giv- ing individuals who wish to lead a healthier life-style information on how to do so. Even with active health promotion programs, however, some individuals will choose to indulge in poor health habits.'" These individuals, for whatever reason, find the satisfactions from smoking, drinking, and avoiding exercise sufficient to offset the risks. Some may be tempted to say, "So be it; that is their own business." Such a position assumes that there are no consequences for others. But there are. Some are financial (for example, higher health costs paid by taxes) and some are nonfinancLil (an innocent bystander killed by a drunk driver). Indulgence in these habits would be closer to the individual's "own business" if sin taxes approximated the costs imposed on others r9 Such an approach implies that the nonfinancial costs are converted to some kind of dollar equivalent in aggregating costs. For exercise, the analogy to taxes would be subsidies of public recreation facilities such as swimming pools. We have estimated the costs that smoking, drinking, and lack of exercise impose on others and compared these costs with current tax levels. Although our estimates are necessarily uncertain, they are not too uncertain to preclude useful conclusions. Taxes on cigarettes are at a level such that smokers pay approxi- mately the costs they impose on others. This situation does not mean that cigarette taxes should not be increased, but it does mean that other grounds for increases must be found. Here are two such ratio- nales: higher taxes will discourage some adolescents from starting to smoke, a benefit they will later appreciate; and cigarette taxes cause relatively less distortion and tax evasion behavior than other taxes. In contrast, the costs on others impo.ed by excess drinking greatly exceed current taxes on alcoholic beverages, especially those on beer and wine. Our analyses strongly support the recommendations of former Surgeon General Koop to increase alcohol taxes. WeThe costs imposed on others byea sedentary life-style are also high, need to discover cosf-effectiv methods to induce individuals to exercise some minimal amewnt, because one cannot tax inactivity. Our analysis supports a considerable increase in alcohol t.ixes, but changing any tax leads to changes in inefficiencies and inequities. Decisions on the right level of the tax and the associated inequities and inefficiencies is a task that falls to uur political institutions. We hope that our work will contribute to more informed decisions. 6Z806M0Z

4 The Costs of Poor Health Habits earlier, he will pay fewer taxes for collectively financed nursing home care, Medicare, and Social Security. In other words, he will pay lower "taxes on eamings" that support nursing home care and retirement pensions.4 Yet he may benefit as much as, or more than, Jane from those programs. If John kills himself in an automobile accident, how- ever, he may have paid into Social Security and not collected any benefits. Then, indirectlv, Jane gains financially from his premature death. Collectively financed programs and taxes on earnings are not the only sources of external costs for smoking and heavy drinking. Al- though there is considerable debate about the magnitude of effects for "passive smoking," Ihe Surgeon General has estimated that 2,400 deaths per year result from breathing air filled with tobacco smoke (USDHHS, 1986). These lost lives are another source of external costs. To the extent that passive smoking generates health care and other collectively financed costs, a portion of those costs is also paid by nonsmokers. For heavy drinking, other sources of external costs are loss of inno- cent lives and property damage caused by drunk-driving accidents, and other crimes committed "under the inFluence." Some of the property damage is the drinker's, and many drunk drivers are their own victims. We do not consider those costs (or the costs of passive smoking within families) as external' Although they are part of the total costs to society, by our definition they are internal costs to the drinker. To return to our example, if John's accident claims lives other than his own, it imposes very high external costs.F No one knows exactly how many auto fatalities are caused by drunk drivers, but research suggests the percentage is large. A study of 44,000 fatal accidents indicates that 42 percent of the drivers involved were intoxicated (USDHHS, 1987, pp. 8-9). Drinking is implicated in many other crimes that generate high external costs from properly loss and damage, as well as from injuries and death of victims. There is some controversy about the drinking- crime nexus, but statistics show that many crimes other than drunk driving are committed under the influence of alcohol (ibid.). These crimes too impose costs on the criminal justice system. Regardless of what is known or not known about how smoking, heavy drinking, and lack of exercise affect health and well-being, it is possible to measure external costs. The nature and dimensions of these costs provide a strong economic justification and political An Overview 5 rationale for government (and private) efforts to curb these habits. The costs also, indirectly, provide evidence about how these poor health habits affect health and longevity. Estimating the External Costs of Poor Health Habits An enormous amount of research has been done on the (variously defined) costs of smoking, less on the costs of drinking, and much less on the costs of exercise, which is a relatively new area of interest. Widely varying estimates of these costs have emerged. In the case of smoking, the estimated annual costs have ranged from $50 billion to $66 billion (in 1986 dollars).' Apart from case studies on the costs of alcohol abuse to Individual corporations and industries, only two comprehensive studies of drinking costs have appeared so far. One (Berry and Boland, 1977) estimated that costs in 1971 were just over $85 billion [(1986 dollars)]; the other (Harwood et al., 1984) estimated that in 1983 costs were $129 billion /(1986 dollars)]. No comparable figures are available for lack of exercise. Some differences in cost estimates arise because different studies use different data and make different assumptions. Other differences result from the fact that studies ask different questions. (These factors are discussed at length in Appendix A.) We address two questions here: (1) When an individual decides to smoke, drink heavily, or not exercise, what are the lifetime external .osts-that is, by how much does society subsidize these habits7fl (2) What drives these costs? To estimate the lifetime external costs of these bad habits, we con- sidered how they affect life expectancy. use of collectively financed programs, and taxes on earnings. For smoking, we also included costs of fires caused by smokers, and for drinking, the costs of drunk driving-lost lives and property damage-and crime. Table I-1 shows the components of the external costs. Our study utilized a number of data sources, primarily the RAND Health Insurance Experiment (HIE) and the National Health Inter- viewSurvey (N1fIS). The fflE, a randomir.ed trial of alternative health insurance arrangements, was our primary data source for people younger than 60. Families at six sites participated in the study from 1974 to 1982. The HIE data contain a wealth of information on habits and medical reasons for use of care for 5,800 people. The HIE did not enroll anyone aged 62 or older. Fur information on people over 60, we used data from the 1983 NHIS supplement, which was admin- 64806S480Z