This Article Abstract Submit a response to this article Services E-mail this article to a friend Alert me to new issues of the journal

P E R S P E C T I V E
F U T U R E  E L D E R L Y 

26 September 2005  
  

Accounting For 
Future Costs In Medicare

The commonsensical syllogism that better health equals
lower costs is not invariably true.

By Bruce C. Vladeck

ABSTRACT:

The relationship between the health of an insured population and the costs of insuring them is a lot more complicated than might first appear. The excellent paper by Geoffrey Joyce and colleagues helps explain why. The obvious interaction between health status and longevity, interacting with the issue of how long the insurer remains responsible for care of a particular person, is central. So is the cost of improving health status. And perhaps most importantly, the relationship between health care use and costs in the United States is not as straightforward as is generally assumed.

The relationship between the health of an insured population and the costs of insuring that population is a lot more complicated than might first appear. All other things being equal, it is both intuitively and empirically true that it is cheaper to insure healthy people than it is sick ones. But, as Geoffrey Joyce and colleagues so clearly demonstrate in their excellent paper, providing services to make a sick population healthy may or may not save the insurer money over the long run, depending in large part on whether making people healthier causes them to remain insured for a longer time.¹

Joyce and colleagues demonstrate that if you start with two sixty-five-year-old newly minted Medicare beneficiaries, only one of whom is diabetic, the per year costs to Medicare will be higher for the diabetic beneficiary than for the nondiabetic beneficiary. But the lifetime costs to Medicare may not be much greater if the diabetic patient dies much sooner than his counterpart. On the other hand, if the money Medicare spends on the diabetic improves his health, it may well extend his life expectancy, adding years during which treatment services will still be required. And a sick person who lives a long time is clearly more expensive to an insurer such as Medicare than a sick person who dies sooner.

(Parenthetically, but not at all unimportantly, Joyce and his colleagues apply a relatively modest discount rate of 3 percent to future expenditures, to provide an intellectually defensible framework for comparing lifetime costs when some people have relatively few expensive years, while others have many less-expensive years. From an analytic point of view, that is perfectly appropriate, but it runs counter to the way in which Medicare, and the federal government as a whole, accounts for future costs. In general, the federal budgeting and cost-estimating process attaches no importance to the time value of money. Worse, in standard Medicare cost estimation, the absence of a discount factor is compounded by built-in inflation adjusters, so that five- or ten- or seventy-five-year estimates are expressed in inflated dollars with no adjustment to real current dollars. This serves very well the interests of people trying to make future budget estimates look as scary as possible.² It also largely obliterates the cost savings achieved by reduced morbidity, since a dollar spent ten years from now is counted as more than a dollar today.)

The U.S. population is healthier than it has ever been, but health care costs continue to go up at a multiple of other economic indices; obviously, the simple, commonsensical syllogism that better health equals lower costs is not invariably true. Several other things must be going on. Let me briefly suggest four.

Good health is expensive. First, achieving better health isn’t always cheap. Certain disease management programs, for example, have proven themselves to have a fair degree of clinical efficacy, but whether or not they provide economic benefit to the insurer depends in large measure on how much they cost, and perhaps in larger measure on how long the person remains insured. If the patient dies or leaves the group before the completion of the “payback period” from the investment in disease management, the insurer has lost money. The insurer also loses money if the patient lives too long, and remains the insurer’s responsibility, after the payback period has been reached. Further, interventions designed to improve health—particularly those of relatively low cost and therefore great political popularity, such as screening exams or immunizations—do not work 100 percent of the time, so the cost per illness averted can be much higher. That is why inexpensive health education or lifestyle modification programs can still not be cost-effective, even if they are cheap.

The elderly get chronic illnesses. Second, just because people do not have a particular illness, or have one and are cured, does not mean that they won’t get another. Indeed, a central dynamic in the analysis that Joyce and colleagues present is generated by the increased incidence of various chronic illnesses as people age, particularly among people who have had other illnesses. Just because someone doesn’t have diabetes or cancer at age sixty-five doesn’t mean she won’t get it by the time she reaches age seventy—and the longer she lives, and the more other diseases she has had, whether or not those other diseases have been successfully treated, the more likely she is to eventually get another disease.³

Dying is expensive. Third, and perhaps most importantly in quantitative terms, everyone dies sooner or later, and dying is expensive. Indeed, terminal episodes account for a considerable fraction of the total lifetime costs incurred by Medicare for most of its beneficiaries. That is not because, as is widely believed, the health care system throws intensive resources at people known to be dying; rather, it is because it throws resources at people who might be dying, to prevent them from doing so.⁴ In modern society, a large proportion of deaths are preceded by periods of very serious, very costly illness. This is especially an economic problem for Medicare, the only U.S. health insurer that keeps essentially all of its beneficiaries until they die. Conversely, the life events that increasingly precede dying—retirement or extended periods of disability—tend to remove most people from being financial problems for other insurers before they become fatally ill.

Health care costs are driven by prices. Finally, as we all should have learned by now but seem to keep forgetting, changes in health care costs are only partially a function of changes in the health or illness status of the population being served. Members of Congress and the public have trouble getting past the idea that improving people’s health will reduce health care spending, but all of the empirical evidence suggests that in the aggregate, this is just not true.

Optimal care of a heart attack for a Medicare patient can cost twice as much in one part of the country as another; some of the difference may be attributable to different input costs in different places, some to different market conditions, and some remains completely unexplained, at least in the statistical sense. Moreover, the health care system is remarkably adaptable and creative: The near disappearance of tonsillectomies has not produced economic ruin for otolaryngologists. So is nature: Twenty-five years ago, before the epidemics of AIDS and multi-drug-resistant staph, infectious diseases was thought to be a dying specialty.

More basically, while health care costs are partly driven by utilization, they are also largely driven, especially in the United States, by prices. Both production and systems inefficiencies, along with rents received by producers, may explain far more of year-to-year cost changes than anything driven by the degree of health or wellness in the population. Or to put it somewhat differently, projections of future Medicare expenses are far more sensitive to exogenous inflation rates than to any changes in the health of the Medicare population.

The health of the population is not unimportant, and here Joyce and colleagues deserve special credit for recognizing that reducing the burden of chronic disease “is a goal worth pursuing” whether or not it affects Medicare spending.⁵ Making people healthier is important to do in and of itself. So is reducing health care costs. They just aren’t the same thing.

The views expressed by the author herein do not necessarily reflect the views of Ernst and Young LLP.

NOTES

1. G.F. Joyce et al., “The Lifetime Burden of Chronic Disease among the Elderly,” Health Affairs, 26 September 2005, content.healthaffairs.org/cgi/content/abstract/hlthaff.w5.r18.
2. J. White, Understanding Long-Term Medicare Cost Estimates (New York: Century Foundation, 1 January 1999).
3. Joyce et al., “The Lifetime Burden.”
4. E.J. Emanuel and L.L. Emanuel, “The Economics of Dying—The Illusion of Cost Savings at the End of Life,” New England Journal of Medicine 330, no. 8 (1994): 540–544.
5. Joyce et al., “The Lifetime Burden,” W5-R26.

Bruce Vladeck (bruce.vladeck{at}ey.com) is a principal at Ernst and Young LLP in New York City. He is a past administrator of HCFA (now the Centers for Medicare and Medicaid Services, or CMS).

Access the table of contents for this package

DOI: 10.1377/hlthaff.W5.R94
©2005 Project HOPE–The People-to-People Health Foundation, Inc.