June 2002, (Vol. 3, No. 4)
What Cost Health?
By: David A. Kindig, MD, Ph.D., and Robert Stone Newsom, Ph.D.

Absolute health expenditures in the United States are at the highest ever experienced in the world. After lying dormant for almost a decade, recent annual increases have again focused the nation's and Wisconsin's attention to this issue. Escalating health costs are a burden to employers, governments and many individuals and historically have led to increased numbers of uninsured or underinsured individuals. The problem is enormously complex and no solution will be without major economic and social dislocation. We can, however, better understand the problem and approach a solution through consideration of the following:

  1. Which health expenditures are of concern?
  2. Is the concern about absolute levels or rates of increase? For the entire state, or within the state? For individuals, employers, or government?
  3. What is responsible for variation in expenditures across regions?
  4. Do higher expenditures mean better health?
  5. What could be done to control health expenditures? Can reasonable goals be set for Wisconsin?

Although part of the national agenda for over fifteen years, few quantifiable answers to these questions are available. Consequently studies examining the reasons for variation in health expenditures remain critically needed. In the short run state government in collaboration with the private sector should consider establishing processes for determining the appropriate and possible level of health expenditures that Wisconsin wants and can afford.


Health expenditures include all services such as hospitalization, physician and dentist care, pharmaceuticals, nursing homes, home health, and other costs. It also includes all persons in the state, including those on Medicare, Medicaid, different forms of private insurance and managed care plans, and the uninsured. In 1998, the last year in which state specific data are available Wisconsin's total per-capita expenditures were $3845, or 2% higher than the national figure of $3759. Our growth rate from 1991 to 1998 was 5.7% per year compared to the national average of 4.9%. The most recent national estimate made by the federal CMS for 2002 is $4674 per person based on national growth rates of 6.4% in 1999, 9.3% in 2000, and 7.9% in 2001. If Wisconsin had the same growth rates over that period, its 2002 expenditures would be about $5100 per person.

Each service differs in its growth experience. For example, in 2002 hospital expenses are projected to increase by 6.7%, nursing homes by 4.6%, home health by 11.3% and prescription drugs by 13.5%. Private expenditures (employers and individuals) are projected to increase 9.4% in 2002 while public expenditures (Medicare at 5.6%, and Medicaid at 11.3%) will grow overall at 7.6%. State and federal governments and the private sector have taken a range of actions to curb these growth rates, with mixed and usually short acting results.


Over the years health spending has increased as a percentage of Gross Domestic Product because growth in health expenditures has outpaced growth in the economy as a whole. Although health expenditures are projected to fall to about 7% in the 2003-2007 period and to 6.6% from 2008-2011, health spending still will outpace the overall economy by 2.5% per year, resulting in growth from 13.2 % of GDP in 2000 to 17% in 2011. Unquestionably such projections include many assumptions and are difficult to make over such long periods. Nonetheless, they are the best evidence we have. Further such projections give us a clear opportunity to consider what target expenditure growth would be either desirable or achievable in Wisconsin. What are we really willing to spend to maintain, or increase the health we now possess?

Answering the willingness to pay question implies prior knowledge of what we currently spend. National expenditure data are produced annually by the federal government. State estimates, beyond hospital expenditures and Medicaid spending, are much more difficult to do since private expenditures are not routinely reported. If we are to understand sub-state or regional differences more studies such as that done by Mercer and requested of the GAO are critical. Since employers and governments are more aware of what they spend for health insurance, policy focus often is limited to employer premium increases, without tracking how much individuals are paying in co-payments and deductibles or for uncovered services.


To approach this complex question we need to recall that expenditures are made up of both volume and price, and increases in either may be responsible for overall increases. While we often think that more volume is the primary reason (cardiac procedures, MRI scans, number of specialists), in comparing the United States to other countries is has been found that prices of services rather than volume are primarily responsible for higher costs in the United States.

No study, of which we are aware, compares states in this regard, although the Mercer study is reported to show prices as a significant factor. Certainly some combination of new technology and advertising driven increased demand has led to increases in both prices and utilization. Ten years ago, we found that variation in expenditures across the states was in part due to underlying demographics like age and wage rates, and in part due to the supply of hospitals and physicians, and these findings are confirmed in the recent Health Affairs article cited above. They also note that states with higher managed care penetration also have lower expenditures. This is thought to be due to "both lower premium rates charged by HMOs and spillover effects of competition on non HMO premiums". Nationally, our private insurance-based system administrative costs and profit margins contribute to making costs much higher than in other countries or in Medicare. When comparing across small regions, variation in the illness burden of the population also needs examination.


Again, the answer is not certain. Of course a certain level of health care is necessary for good health, but it is not the only factor; social factors like education and income as well as individual behaviors like smoking and obesity are critical. We don't have good measures of population health on which to conduct such studies, but we recently examined the relationship in mortality rate reduction and health expenditure increases across the states from 1990-1998 and found no relationship. While Wisconsin is comparatively in the middle with average values of both health increase and health expenditure increase these data do raise questions, given current levels of spending, regarding the commonly held assumption that "more money buys more health". The research would be more robust with other outcomes than mortality, but similar lack of relationships have been seen when comparing cardiac surgery outcomes and hospital expenditures for these procedures.


There are no easy solutions to this problem, or they would be in place already. The underlying demographic, technologic, and professional forces are enormously strong, and, if indeed it is possible to modify them, substantial, difficult, and contentious interventions would be necessary. The Figure below from an article titled "The Sad History of Health Care Cost Containment As Told in One Chart" shows that the history of attempting to control such expenditures through either market forces or regulation has not been outstanding or permanent.

Housing with Supportive Services:
Some Definitons
Congregate Housing An apartment building providing hotel type services including meals, housekeeping and activities; often transportation and information are offered.
Residential Care Apartment Complex Wisconsin's term for assisted living see note in Issue Brief above.
Adult Family Home A residence serving 1 to 4 unrelated adults offers meals, supervision and personal care.
Community Based Residential Facilityg A group living arrangement served by staff and offering room and board, supervision and other supportive services.
Nursing Home A skilled-care facility providing room, board and 24hour care for residents who need 7 hours a week of nursing or personal care, often-rehabilitative services and providing both short-term and long term care.

The next grid is an incomplete portrayal of previous attempts to deal with this issue. It illustrates how both market forces and regulation have been used to influence both volume and price of health services.


Market Forces Co-payments


Defined Contribution
Managed Care
Regulation State Rate Setting
(Maryland All-payer)

Medicare Physician Fees

Generic Drugs
Certificate of Need

Medicare DRGS

Oregon Rationing

Additionally, although the Clinton health plan involved an all payer system that was and would be extremely difficult to accomplish, the principles deserve revisiting. The goal of the plan was to reduce health expenditure increases from 3.5% over inflation to 1% over general inflation, through aggressive competitive mechanisms as well as price regulation when and in which states market forces were inadequate. As experience has shown in the Medicare program regulation can hold down expenditures apart from the problem of maintaining such reductions in a political environment. Covering the uninsured to prevent cost shifting can be considered, although this has a price of its own. Pooling of high risk or small employer groups still remains a worthy policy to consider.

Lastly it is worth considering a plan where employers and government only pay for the catastrophic and preventive services shown to be most effective, leaving individuals to pay the increases for less cost effective or lifestyle / cosmetic treatments. This solution would be complicated and controversial largely because it requires a means and the will to determine value received.

Given historical precedent and the real complexities of any solution, maybe we should ask whether anything should be done at all? We have seen the difficulty in having, and sustaining, an impact on rising costs. We project that current increases may moderate somewhat in the coming years and have evidence of moderation, albeit short lived, during the nineties. Still our attention is refocused by the fact that increases averaging 2.5% over inflation leads to spending 17% of our resources on health care. If 30% of income is a reasonable percent to spend on housing is 17% too much to spend on health - or just right? With housing most of us can evaluate the value equation; with health we can measure the expenditure but have yet to measure or truly define the outcomes.


  1. Heffler, S et al. Health Spending Projections for 2001-2011: The Latest Outlook. Health Affairs 21 (2):207-218, 2002.
  2. Levit, K. et al. Inflation Spurs Health Spending in 2000. Health Affairs 21 (1): 172-181, 2002
  3. Martin, A et al. Health Care Spending During 1991-98: A Fifty State Review. Health Affairs 21(4): 112-126, 2002
  4. Altman, D and Levitt, L. The Sad History of Health Care Cost Containment in One Chart. Health Affairs January 23 2002 Web Exclusive
  5. Wisconsin Dept of Health and Family Services. Wisconsin Health Expenditures 1980-1998. Office of Strategic Finance 2000.
  6. Kindig, D. How Expensive is Health Care in Wisconsin? June 2002 Issue Brief from the Wisconsin Network for Health Policy Research.
  7. Kindig, D and Libby, D. Setting State Health Spending Limits. Health Affairs 13(2):288-89, 1994