Frontiers in Health Policy Research

NBER Reporter: Summer 2000

Frontiers in Health Policy Research

The NBER held its fourth annual conference on Frontiers in Health Policy Research in Arlington, Virginia, on June 14. NBER Health Care Program Director Alan M. Garber of Stanford University organized this program:

Michael J. Moore, NBER and Duke University, and James W. Hughes, Bates College, "The Health Care Consequences of Smoking and Its Regulation"

Gabriel A. Picone, University of South Florida, and Frank A. Sloan, NBER and Duke University, "How Costly Are Smokers to Other People? Longitudinal Evidence on the Near Elderly"

Darius Lakdawalla, RAND Corporation, and Tomas Philipson, NBER and University of Chicago, "Public Financing and the Market for Long-Term Care"

Amber E. Bernato and Christopher R. Kagay, Stanford University; Alan M. Garber; and Mark C. McClellan, NBER and Stanford University, "Trends in the Use of Intensive Procedures at the End of Life"

Ernst R. Berndt, NBER and MIT; Susan H. Busch, Harvard University; and Richard G. Frank, NBER and Harvard University, "Price Indexes for the Treatment of Depression"

The literature on the health economics of smoking presents two principal facts: that smoking increases health care costs and that restrictions on smoking lead to reductions in smoking prevalence and intensity. Moore and Hughes extend the literature and ask whether cigarette tax increases lead to lower health care costs. Using data from the 1991 and 1993 National Health Interview Surveys, they find that the health care benefits of smoking cessation are greater than previously believed. There is also weak evidence that tax increases lead to higher rates of smoking cessation. In combination, these results suggest that, in addition to providing a source for funding excess health care costs, tax increases may lower health care costs directly by inducing smokers to quit.

Picone and Sloan use data from four waves of the Health and Retirement Study to assess the impact of smoking on use of hospital and physicians' services and nursing home care. They limit their analysis to people aged 51 to 67 ("near elderly"); during this phase of the life cycle, many adverse effects of smoking, measured in terms of mortality and morbidity, begin to occur. In contrast to past studies, this one separates the health expenditure burden of smoking by the type of health insurer, since smokers are more likely to be on public insurance and to be uninsured (holding a large number of other determinants of coverage constant). Picone and Sloan find that the net effect of smoking on expenditures for health care services is positive for this age cohort. The relative burden of cost is highly dependent on age.

Lakdawalla and Philpson develop a framework for understanding how economic forces govern the demand for and supply of long-term care, and how they affect the prevalence of public subsidies for long-term care. The authors argue that aging may actually lower the demand for market care by increasing the supply of family care-givers (which substitute for market care). This idea appears to explain important trends in the output of long-term care over the past 30 years. In addition, the authors document the extent to which public financing has grown over the past several decades in response to increased demand for nursing home care. They argue that by raising the private price of nursing-home care, the growth in demand provides incentives for people to qualify for public assistance and thus expands the share of public output.

Drawing from a 20 percent sample of Medicare claims for decedents and a 5 percent sample of claims from survivors, Bernato, Garber, McClellan, and Kagay study the use of 45 intensive procedures among Medicare beneficiaries in 1985, 1990, and 1995. Each of the procedures was performed more frequently among decedents than survivors, and the relative growth in the utilization of all 45 procedures combined outpaced the growth in survivors. Expenditures associated with hospitalizations in which these 45 intensive procedures were performed grew faster than overall inpatient expenditures over time. This is consistent with the shift of less intensive procedures to outpatient settings. This trend in increasing intensity of treatment among Medicare recipients contributes to growth in expenditures that outpaces changes in demographics and the incidence of disease. Furthermore, the relative increases in intensity of care among decedents over survivors may explain why the increased use of less expensive services -- such as hospice and home health care for some decedents -- has failed to slow the growth rate of expenditures at the end of life below the growth rate for the rest of Medicare beneficiaries.

Berndt, Busch, and Frank compare three major approaches to constructing price indexes for the treatment of depression: the methods used by the Bureau of Labor Statistics (BLS); an approach based on episodes of treatment which uses adherence to treatment guidelines as an indicator of outcome; and an episodes-based approach that incorporates information on expected treatment outcomes from the literature and expert clinical opinion. Their findings highlight the differences in results produced using the episodes-and outcome-based methods versus the service-based approach of the BLS. The BLS method shows a rising price index for depression treatment between 1991 and 1996 while the episodes-based approaches show a declining price index for treatment of depression.

These papers and their discussion will be published by the MIT Press in an annual: Frontiers in Health Policy Research, Volume 4. In advance of publication, these papers will be available at Books in Progress.


National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us