The level of health care spending continues to make headlines,

Transcription

1 DataWatch The Concentration Of Health Expenditures: An Update by Marc L. Berk and Alan C. Monheit Abstract: An earlier study tracing trends in health spending from 1928 to 1980 found health expenditures concentrated among the top 1 percent of those spending money for health care. In an update to that study, the authors found that that trend toward concentration has increased. In the top 1 percent of spenders accounted for 30 percent of health spending, up from 26 percent in 1970 and 29 percent in In contrast, the bottom half of the population in terms of spending accounted for only 3 percent of total spending in. The average expenditure for the top 1 percent of spenders in was $47,331. Nearly half of those top spenders in were elderly. Also, more than 16 percent were black, up from nearly 10 percent in The level of health care spending continues to make headlines, amidst continued talk of health system reform to curb rising health inflation. Of equal interest is how those funds are spent. In an earlier paper in Health Affairs, we published an analysis of the distribution of health expenditures for the U.S. population from 1928 to That analysis indicated that the distribution of spending for health among the noninstitutionalized segment of the population was heavily skewed, with the top 5 percent of spenders accounting for close to half of health expenditures, while almost half of all Americans combined spent less than 5 percent of the total amount. In general, this distribution held steady over the five decades studied. Between 1963 and 1970, however, the distribution among the top percentile shifted dramatically. This shift was associated with two important structural changes in the U.S. health care system: growth in private and public health insurance coverage (especially major medical coverage), and diffusion of medical technologies in the hospital sector. We cautioned in our earlier work that continued increases in the level of resources devoted to catastrophic illness could jeopardize the health care system's ability to achieve other desirable reforms in health care, such as increasing primary care services for the disadvantaged. That study concluded that "[c]ost containment efforts that try to get 'average' people to Marc Berk is director of the Project HOPE Center for Health Affairs in Chevy Chase, Maryland. Alan Monheit is an economist with the Agency for Health Care Policy and Research in Rockville, Maryland.

2 146 HEALTH AFFAIRS Winter 1992 use services more prudently will not have a large impact on total national expenditures." 2 These findings take on new significance in view of recent debates over health care rationing. As Henry Aaron has noted, the concentration of health care expenditures creates "enormous incentives" for private insurers to identify high-cost users of care and to keep their numbers to a minimum in order to ensure profitability and to control premium costs. 3 For example, a recent study of 1989 claims experience of Blue Cross/ Blue Shield of Ohio found that 1 percent of beneficiaries accounted for 24 percent of total benefits paid, and 10 percent of beneficiaries accounted for almost two-thirds of benefits. 4 Aaron has suggested that from the government's perspective as well, [t]he concentration of expenditures on a few patients underscores the fact that any measures that successfully reduce the rate of growth of spending on health care must eventually affect outlays on high-cost episodes.... In short, successful cost control will require rationing of services to the very ill. Update Of Spending Concentration In this DataWatch we present additional data that allow further examination of changes in the concentration of health expenditures. Exhibit 1 presents our original tabulations of the distribution of health care expenditures, with an additional column of data from the Exhibit 1 Distribution Of Health Expenditures For The U.S. Population, By Magnitude Of Expenditures, Selected Years, Percent of U.S. population ranked by expenditures Top 1 percent Top 2 percent - 17% 26% 35 27% 38 29% 39 30% 41 Top 5 percent Top 10 percent 52% Top 30 percent Top 50 percent Bottom 50 percent Sources: Data for 1928 are from I.S. Falk, M.C. Klem, and N. Sinai, The Incidence of Illness and Receipt of Medical Care among Representative Families (Chicago: The University of Chicago Press, 1933); data for 1963 are from R. Andersen, J. Lion, and O.W. Anderson, Two Decades of Health Service: Social Survey Trends in Use and Expenditures (Cambridge, Mass.: Ballinger, 1976). Data for 1970 are from National Center for Health Services Research tabulations of the 1970 CHAS/NORC survey; for 1977, from the 1977 National Medical Care Expenditure Survey (NMCES); for 1980, from the 1980 National Medical Care Utilization and Expenditures Survey (NMCUES); and for, from the National Medical Expenditure Survey (NMES).

3 DATAWATCH 147 National Medical Expenditure Survey (NMES). These new data similarly reveal a highly skewed expenditure distribution; in fact, there is some evidence to indicate a small shift in the aggregate expenditure distribution toward increased concentration. The top 1 percent of the population, which accounted for 26 percent of health expenditures in 1970 and 29 percent in 1980, accounted for 30 percent of health expenditures in. The average expenditure for persons in the top 1 percent of spenders in was $47,331. The top 2 percent of the population accounted for 35 percent of health care expenditures in 1970 and 41 percent in. Conversely, the bottom half of the population as ranked by health expenditures accounted for only 3 percent of total health expenditures in. The analysis shows that during the seven-year period between 1963 and 1970, the percentage of health expenditures accounted for by the top 1 percent of the population increased nine percentage points. During the seventeen subsequent years, that percentage increased only another four percentage points. Nevertheless, while the growth in concentration is slow, each of the three subsequent health expenditure surveys conducted since 1970 shows increased concentration of spending. In our earlier work we noted that the shift in concentration over the study period was associated with the growth of both private and public health insurance coverage, particularly major medical coverage, during the 1960s and 1970s. We also observed that a number of expensive "prestige" technologies (for example, open-heart surgery, coronary care units, intensive care units, and long-term renal dialysis, among others) were widely adopted in the hospital sector at the same time. Between 1977 and the number of uninsured Americans of all ages increased from 12.3 percent of the population to 15.5 percent (based on tabulations from first-quarter data in the 1977 and expenditure surveys), and it is unlikely that increased access to costly medical technologies by the general population (especially in the hospital sector) during this period matched the gains in coverage of the mid-1960s and early 1970s by groups previously uninsured. The current shift, therefore, may be influenced more by changing demographic patterns than by changes in insurance coverage and increased access to technology. Exhibit 2 presents a comparison of the demographic characteristics of persons with high health expenditures from 1970 to. The top health care spenders in were an older population than was true a decade earlier. More than 43 percent of persons ranked in the top 1 percent of health care spenders in 1980 were elderly, up from 40 percent in 1977 and 32 percent in Data from the NMES show that

4 148 HEALTH AFFAIRS Winter 1992 Exhibit 2 Characteristics Of Persons With High Expenditures, Top 1 percent of persons ranked by expenditures Characteristics Black 8.2% 8.8% 9.6% Total U.S. population % 11.5% 9.2% 11.7% 11.9% Over age Fair or poor perceived health Sources: Data for 1970 are from National Center for Health Services Research tabulations of the 1970 CHAS/NORC survey; for 1977, from the 1977 National Medical Care Expenditure Survey (NMCES); for 1980, from the 1980 National Medical Care Utilization and Expenditures Survey (NMCUES); and for, from the National Medical Expenditure Survey (NMES). 48 percent of persons with the highest health expenditures in that year were elderly. Although the elderly still predominate among persons incurring the top 1 percent of health care expenditures, the most dramatic change observed is in the percentage of persons with high health expenditures who are black. In 1970, 8 percent of persons with high expenditures were black; this rose to nearly 9 percent in 1977 and nearly 10 percent in Data from the survey indicate that 16 percent of persons in the highest expense category were black. In the three earlier surveys (1970, 1977, and 1980) blacks were underrepresented in the highexpenditure category compared with their representation in the general population. In the survey blacks were much more likely to incur a level of health care expenditures that ranked them in the top 1 percent than would be expected given their representation in the general population. Discussion Since the publication of our original findings, the debate over health care rationing has intensified. Much of that debate now centers on the Oregon Medicaid initiative, which would expand the number of Medicaid beneficiaries while restricting coverage for selected high-cost procedures. At this writing, the administration has rejected the Oregon proposal but has invited a revised application after additional study. As our analysis indicates, health care expenditures have become somewhat more concentrated among the top spenders even as public and private health insurance coverage has declined. An important consideration for public policy is whether new initiatives to expand health insurance to the 15 percent of the U.S. population that is currently uninsured will

5 DATAWATCH 149 cause the concentration of expenditures to accelerate as it did during the expansion of coverage in the 1960s, If insurance expansions include unlimited coverage of higlvcost or experimental technologies, the concentration of health expenditures may continue to increase. Conversely, a minimum benefit package that encourages nonusers to obtain some services may decrease concentration. The debate over health care reform must focus, therefore, not only on expanding coverage to the uninsured but also on the consequences of alternative levels of benefits. A particularly difficult choice is deciding to what extent persons should be insulated from the expense of high-cost illness. The increasing concentration of health expenditures suggests that the unpleasant choices available to policymakers in the current rationing debate are unlikely to be any less painful tomorrow. The views expressed in this paper are those of the authors. No official endorsement of the Project HOPE Center for Health Affairs, the Agency for Health Care Policy and Research, or the Department of Health and Human Services is intended or should be inferred. NOTES 1. M.L. Berk, A.C. Monheit, and M.N. Hagan, "How the U.S. Spent Its Health Care Dollar, ," Health Affairs (Fall 1988): Ibid., H. Aaron, Serious and Unstable Condition: Financing America 1 s Health Care (Washington: The Brookings Institution, 1991). 4. T. Wicker, "Code Blue on Insurance," The New York Times, 21 July 1991, Aaron, Serious and Unstable Condition, 53.