NEWS RELEASE, 09/09/98

UC economists find the total health care bill for smoking comes to $72.7 billion per year or 11.8 percent of all medical payments

By Patricia McBroom, Public Affairs

BERKELEY -- The total cost of caring for people with health problems due to cigarette smoking - counting all sources of medical payments - is estimated at $72.7 billion per year, according to health economists at the University of California.

The figure is almost six times higher than the cost per year of smoking-related Medicaid payments alone, reported last spring by the same economists from the University of California, Berkeley and the University of California, San Francisco.

The new total estimate "translates the adverse health effects (of smoking) into dollar terms, the universal language of decision makers," said the analysis published today in Public Health Reports.

All payments made in 1993 by Medicare, Medicaid, Veterans Administration medical programs, military medical programs, private health insurance companies and out-of-pocket payments were analyzed for that fraction of illness caused by cigarette smoking. By comparison, the March report counted only the cost of smoking-related Medicaid payments - $12.9 billion that year.

"I am not surprised by these costs," said Leonard Miller, an economist and professor in the School of Social Welfare at UC Berkeley, who is first author on the report.

"You expect a figure of this magnitude for the impact of smoking on health care, when you consider that one in five deaths per year is due to cigarette use," said Miller.

Of total medical expenditures in the U.S. in 1993, the health effects of smoking accounted for 11.8 percent of the costs, according to the analysis.

The report was co-authored by Dorothy Rice, professor emeritus of health economics at the UCSF Institute for Health and Aging and former director of the National Center for Health Statistics. Other authors are Xiulan Zhang, UC Berkeley graduate student, and Wendy Max, associate professor at UCSF.

"We can now see the tremendous burden of smoking on society," said Rice. "These are very high costs."

She pointed out that the 1993 bill for California alone amounted to $8.7 billion, the highest total in the nation, followed by New York, with $6.6 billion in smoking related disease costs. Wyoming at $80 million in 1993 had the lowest expenditure for illness caused by cigarette smoking.

Rice also said that if a proposed financial settlement between the U.S. Congress and tobacco companies had been reached last spring, it would not have come close to compensating private and public insurance payers for the cost of smoking-related illnesses.

That settlement, now off the boards, would have cost tobacco companies $368.5 billion, paid out over 25 years, in exchange for a cap on future liability for any lawsuits against the tobacco industry. The tobacco industry pulled out of the negotiations when the proposed settlement rose above $500 billion over 25 years.

But according to the Miller-Rice report, the actual cost of medical care for smoking related disease in the next 25 years will be an astronomical $1.8 trillion in 1993 dollar values.

"The amount being considered was clearly well below the actual amount that will be spent for the health care of smokers whose health has been damaged by cigarettes," said Rice.

Estimates in the report of "smoking-attributable expenditures" (SAEs) are derived from 11 equations that link smoking history with health in two ways: (1) the likelihood of a prior treatment for a tobacco-related disease (lung cancer, chronic obstructive pulmonary disease, coronary heart disease, stroke and arteriosclerosis); and (2) self-reported poor health. Other calculations then analyzed the impact of self-reported poor health on direct medical expenditures.

The analysis took particular care to isolate direct smoking health effects from other confounding factors that affect health such as poverty and race. Miller accounted for the impact of twelve such factors: age, gender, region of the country, education, income, body mass, seat belt use, smoking history, self-reported health status, previous treatment for tobacco-related diseases and type of medical expenditure.

Each of these factors helped define the independent effect of smoking on expenditures. For instance, seat belt use provided a measure of risk-taking behavior that sorted out the contributions of smoking from dangerous behaviors in general.

The data for constructing the SAEs and a related measure called "smoking attributable fractions" (SAFs) came from three large federal and state surveys:

(1) NMES (National Medical Expenditure Survey, 1987) a face-to-face survey of 35,000 people that linked population characteristics with use of and payments for medical resources.

(2) CPS (Current Population Survey, 1993), conducted by the Census Bureau for employment information.

(3) BRFSS (Behavioral Risk Factor Surveillance System, 1993), an annual telephone survey of health risks and medical care, conducted by state health departments in collaboration with the CDC.

In the final stage, the SAFs were applied to state medical expenditures for persons over age 18, using data published by the Health Care Finance Administration.

"These are the best federal and collective state data. It's the best we can do at this point to estimate the impact of smoking on health costs in all 50 states," said Miller.

A refinement of the model described in this paper was used recently in a suit against tobacco companies by the State of Minnesota that resulted in an out-of-court settlement for the state that came to more than $6 billion, covering Medicaid, Blue Shield and general assistance medical payments.

The full text of the Public Health Reports article can be found at

This server has been established by the University of California at Berkeley Public Information Office. Copyright for all items on this server held by The Regents of the University of California. Thanks for your interest in UC Berkeley.
More Press Releases | More Campus News and Events | UC Berkeley Home Page


Send comments to: