The Cost-Coverage Trade-off: Title and subTitle Break“It's Health Care Costs, Stupid”

According to recent polls, many Americans consider health care reform the No. 1 domestic issue.¹ Presidential candidates, other politicians, health policy experts, labor leaders, business groups, and others have responded with numerous reform proposals. And somehow in the clamoring, health care reform has become equated exclusively with expanding coverage to the 47 million uninsured Americans.

This is a mistake. As serious as it is, the problems of the uninsured and lack of coverage are symptoms, not the underlying problem. Focusing on them is like treating a fever without addressing the causal infection. Instead, the diagnosis and treatment need to focus on health care costs. The fundamental problem arises because of a cost-coverage trade-off. Without controlling health care costs, any attempt at universal coverage will be transient. Sustainable expansion of coverage to all Americans requires credible changes in the rate of health care inflation—the slope of the health care cost curve. Fortunately, focusing on controlling costs may actually enhance prospects for health care reform. Those who count in the political process—voters, employers, governors, and others—are concerned about costs in a way they have not been genuinely concerned about the fate of the uninsured.

The number of uninsured Americans has been increasing, from 38.7 million in 2000 to 47 million in 2006,^{2 - 3} a 21.4% increase. During these years, health care costs in the United States have increased from $1.4 trillion to $2.1 trillion,⁴ and in real terms approximately 10%.⁵ Similarly, uninsured rates and health care costs differ markedly between states. For instance, in 2004, 9.2% of Iowa's population was uninsured whereas 19.4% of Florida's population was.⁶ Concomitantly, in 2004 the average Medicare spending per enrollee was $5767 in Iowa and $8462 in Florida.⁷ Similarly, average health insurance premiums for a family totaled $9422 in Iowa and $10 444 in Florida.⁸

For the last 30 years, comparisons over time and comparisons between states reveal a strong relationship between health care cost and coverage: higher state health care costs mean worse coverage, and as costs increase, the rate of uninsured individuals also increases (Figure). There are several reasons for this relationship. First, higher health costs drive up insurance premiums, which may induce employers and the self-insured to eliminate coverage. Moreover, as workers are forced to assume a higher fraction of their premiums, more of them may not choose health insurance even when offered. In addition, higher health care costs drive up the cost of Medicaid and other need-based government health programs, inducing states to constrict eligibility requirements.

Factors other than health care costs also affect coverage and rates of uninsurance among Americans. From the employer perspective, whether to provide health insurance depends not only on costs, but also on the competition for labor, unemployment rates, average age of workers, union contracts, average wage rates, tax incentives, mandates, and a company's sense of responsibility for its workers, a factor that is more difficult to quantify. Whether individuals are uninsured further depends on government policies regarding taxes, guarantee issue, mandates for specific services, and insurance rate reform and on the levels at which states define income eligibility and impose other eligibility requirements for Medicaid and the State Children's Health Insurance Program.

The finding of a health care cost-coverage trade-off is robust even when these factors are considered. Many researchers, using a variety of different data sets from a variety of different periods and controlling for many potential confounders, have documented that higher health care costs are associated with lower levels of coverage and higher levels of uninsurance.^{12 - 16} For instance, Sheils et al¹⁵ estimated that a 1% increase in health insurance premiums nationally is associated with 300 000 individuals losing employment-based coverage. Using state-level data, Glied and Jack¹³ reported that between 1981 and 2001, “a 10% increase in health care costs was associated with a 0.88% decrease in health insurance coverage overall.” Using data from 64 large metropolitan areas between 1989 and 2000, Chernew et al¹² reported that a “1% [health insurance] premium increase results in a net increase in uninsured of 164 000 people.” Indeed, in the 1990s, this same group calculated that increases in the cost of health insurance accounted for nearly two-thirds of the decline in coverage.¹² Analysis of the uninsurance rates and insurance costs in different states shows that every 10% increase in average family insurance premiums increases the rate of uninsured Americans younger than 65 years by 0.55%. No matter which data are used and what models are used for analysis, the empirical literature confirms the cost-coverage trade-off. The cost-coverage trade-off may not be an invariable law; it may not apply in countries with different health systems or at certain times in US history. However, since the late 1970s, it has been a predictable phenomenon in the United States.

This emphasis on controlling health care costs as an essential element in health care reform may induce skepticism. Warnings that costs are too high and cannot go higher are a perennial and recurrent theme dating back to at least the 1970s. But costs always increase, and the system has thus far accommodated them. This time, however, things may be different. Between 2000 and 2003, the number of Americans covered by employer-based coverage declined by more than 3.2 million, even as the number of working Americans increased by nearly 1 million.^{17 - 19} This was in part due to the recession after September 11, 2001. However, beginning in 2004, the economy improved significantly. Between 2004 and 2006, the gross domestic product increased 10.3%, the number of jobs in the economy increased by 6.7 million (4.9%), yet the number of individuals covered by employer-based health insurance increased only to its 2000 level.¹⁹ In other words, since 2000, the number of workers has increased by nearly 8 million with no change in the number of people covered by employer-based insurance. During the recovery, the median health insurance premiums for individuals and families have increased 21.9% and 24.1%, respectively.²⁰

This trend suggests that employers and workers are simply finding premiums too high. Providing family health care coverage to 1 worker is like hiring a second worker at minimum wage.²¹ A cost of approximately $12 000 per year for family health insurance—about a quarter of the median income—seems to constitute a cost “brick wall” that begins to make health insurance coverage unaffordable despite other positive economic factors.²²

The cost-coverage trade-off has important policy implications. To paraphrase James Carville's now-famous phrase from the 1992 presidential campaign: “It's health care costs, stupid.” Costs are the important determinant and underlying diagnosis. Consequently, health reform proposals by presidential candidates or others should be critically evaluated primarily on whether they establish a financing structure and incentives for the delivery system reform that really control costs. If they lack a serious plan, they are not credible reforms.

The cost-coverage trade-off does not mean that cost control and universal coverage should occur sequentially. Waiting to cover all Americans until costs are controlled is like blaming the victim. The uninsured are not driving health care cost increases. Moreover, lack of insurance adversely affects their health and economic well-being.²³ Cost control and universal coverage must occur simultaneously. Expanding coverage and then worrying about controlling costs, as was done in Massachusetts, is not a tenable policy. Without policies to restrain cost increases over time, universal coverage will not be sustainable.²⁴

What is a serious cost control plan? True cost control means reducing how much health care cost increases from year to year, to about 1% more than overall economic growth. Vague promises of savings from cutting waste, enhancing prevention and wellness, installing electronic medical records, and improving quality are merely “lipstick” cost control, more for show and public relations than for true change. Reducing the waste from insurance underwriting, sales, and marketing costs is valuable but constitutes a 1-time savings. Furthermore, because these costs are in large part a consequence of selling insurance individually to more than 6 million businesses, they can be achieved only by completely revamping employer-based insurance. Cost control will require comprehensive reform of both employer-based insurance and the dysfunctional health care delivery system that will take years of sustained effort.

This trade-off also undermines the simplistic statements that there is no worry if health care costs can increase to 20% or even 30% of gross domestic product.²⁵ True, the richer the country, the more it can afford to spend, and will spend, on health care. But the increase in costs does have a real effect in the United States: an increase in the uninsured. Greater spending on health care may be good for the well insured, but it also continuously constricts that demographic.

Finally, there are real political advantages from focusing on costs. The politically powerful constituencies whose support is integral to any health care reform really care about rising health care costs. Employers worrying about global competition, state governors handcuffed by rising Medicaid bills, and the 85% of insured Americans all care about rising premiums, deductibles, co-payments, and prescription drug prices. Because of self-interest, costs can motivate these groups in ways that covering the uninsured has not. In the strange calculus that is American politics, the more politically salient issue of costs may provide a better way to achieve the comprehensive reforms necessary to cover the uninsured than the hitherto futile direct moral appeal.