State Health Care Reform: Key Questions and Answers

Studies | Health

No. 311
Monday, April 21, 2008
by Linda Gorman and R. Allan Jensen


Does Universal Coverage Lower Health Care Costs?

The governing assumption of U.S. health care policy is that people who can afford health care should subsidize essential care for those who cannot.  Yet, those who must pay for people who cannot pay for themselves deserve an efficient system for providing subsidized care — one that minimizes costs.  It should be noted that by international standards, virtually everyone in the United States, regardless of ability to pay, does get health care.

Individual health insurance mandates are usually combined with large subsidies to people judged unable to afford health insurance.  This is likely to increase the cost of existing government programs designed to ensure that those unable to pay get essential health care — in part, by encouraging people who presently pay for their own health care to stop doing so.

Even if users of Veterans Health Administration hospitals are counted as uninsured, studies suggest that the uninsured pay for at least half of their own health care.8 Expanding public programs to cover people who are already paying for their care eliminates such payments.  As health insurance is an expensive way to buy health care, it is possible that it may actually be less expensive to provide care under the existing mixture of public subsidies and private charities than under a system created by mandatory health insurance.  There is no compelling case that mandated health insurance buttressed by a large new bureaucracy dedicated to the control of insurance markets and medical practice, and to extensive income redistribution, is the lowest-cost method of providing health care to those who need it but cannot pay.9

Contrary to popular belief, the uninsured use emergency rooms at about the same rate as the insured.10 Generous estimates of uncompensated care for the uninsured suggest that it is about 3 percent to 5 percent of private insurance premiums, which is probably less than the taxes proposed under most universal coverage proposals.11

"Government health care spending crowds out private spending."

It is not clear to what extent estimates of premium increases caused by “cost shifting” include uncompensated care for those nominally insured under such government programs as Medicare and Medicaid.12 In Washington state, Milliman, Inc. estimated that the cost shift from Medicare and Medicaid to private payers was 14.3 percent of commercial hospital costs, or about 4.8 percent of commercial premiums.  Of the typical commercial health insurance premium of $850 a month per family in 2004, the government program shifted costs of about $490 a year.  Physician underpayment by government programs was higher.13 Thus, expanding Medicaid and Medicaid-like programs runs the risk of expanding uncompensated care.  These expansions also increase utilization and encourage people to substitute government payments for health care for their own payments.  For example, TennCare, the Tennessee Medicaid expansion designed to insure everyone, promised to reduce uncompensated care.  Some years later, however, uncompensated care costs had increased.  Tennessee radically changed the program after its cost threatened to bankrupt the state.14

There are also disincentives to work created by the high marginal tax rates people who receive subsidies will face as their incomes rise.  People are free to make choices between leisure and labor, part-time work and full-time work, and high paying jobs and jobs that pay less but are more congenial.  Proposals for rich subsidies for health insurance may increase the number of people who choose lower incomes in order to qualify for taxpayer-supported programs.  This appears to be a particular problem with the State Children’s Health Insurance Program/Children’s Health Plan Plus (SCHIP/CHPP) in Colorado, where an estimated 6 out of 10 new enrollees dropped private insurance to participate in the subsidized public program.  The crowd-out rate is higher as more high-income families become eligible for coverage.  And, contrary to assertions that waiting periods can control crowd-out, economist Jonathan Gruber found that “the anti-crowd-out efforts that have accompanied the SCHIP program have probably raised crowd-out more than lowering it.”15

"Many of the uninsured see health insurance as a bad deal at current prices."

The Colorado Commission recommends subsidizing any household with an income between 300 percent and 400 percent of the federal poverty level that cannot buy employer group insurance and spends more than 9 percent of its income on health insurance.16 This means that any family of four with an income of $61,950 that spends more than $5,576 on health insurance, and any family of four with an income of $82,600 that pays more than $7,434, is eligible for subsidies.  According to the 2006 Consumer Expenditure Survey, families in this income bracket spent 5.3 percent of their household incomes on entertainment, 4.3 percent of their incomes on cash contributions, 3.8 percent of their incomes on household furnishings and equipment, and 5.5 percent of their incomes on food away from home.  In view of this spending pattern, meeting a 9 percent premium burden would not seem to be impossible for these households, and subsidizing them would seem to place an unfair burden on other taxpayers.

Focusing on payments for health insurance discriminates against people who substitute cash savings for insurance, and only purchase health insurance that covers very large expenses.  People who purchase a health insurance policy with a $10,000 deductible may never pay more than 9 percent of their incomes for health insurance but may occasionally have total health expenses that exceed 9 percent of income in one or two years.  Encouraging the purchase of health care using third-party payment and discriminating against cash payments raises costs by increasing administrative overhead. 
One reason given for imposing an individual mandate is to limit the need for subsidies by requiring that everyone spend money on government-defined health insurance.  The fact that people who have low medical expenditures are exceptionally resistant to purchasing standard insurance policies indicates that simply expanding the insurance model is a mistake unless regulators also act to lower economic costs.  A substantial number of the uninsured see health insurance as a bad deal at current prices.  Shifting that bad deal to taxpayers does little to change the cost/benefit tradeoff.

The estimated elasticity of demand for individual insurance — the percentage change in policies bought divided by the percentage change in price — ranges from -1.0 to -0.3.  This suggests that a 10 percent increase in insurance premiums results in a 3 percent to 10 percent decline in the number of policies purchased.  Poor families without access to group coverage who are not eligible for public plans are least likely to purchase individual insurance, regardless of the subsidy.  Married couples tend to be less affected by price increases, while single people are more sensitive.  One study found that even substantial subsidies for individual insurance would “have modest effects on the number of uninsured.”17

"Taxpayer subsidies make unwanted coverage attractive to the uninsured."

In Wisconsin, a 2004 evaluation of the BadgerCare program speculated, “the mere perception of the premium [3 percent of income above 150 percent of the federal poverty level] could be holding back applicants who would not be required to pay it.”18 Kate Bundorf and Mark Pauly found that the likelihood of purchasing health insurance increases with expected health expenditures, and that this effect is more likely to be observed in the large group market than in the individual market.19 They reported that in 2002 the average employee payment for single coverage was $450 per month — about the average expected health expenditure for a 25-to-29-year-old man.  Bundorf and Pauly concluded that if the wage difference between jobs with and without health insurance “reflects the average premium for coverage ($3,060 for single coverage in 2002), the reduction in wages associated with coverage may generate income effects for low-income workers that make jobs with coverage unattractive relative to those without coverage.”  If people with the largest expected health costs are already insured, estimates of the savings from insuring the uninsured may be overstated.

For employer-provided insurance, economists Jonathan Gruber and Ebonya Washington used results from the transition of federal employees to pretax health insurance premiums over the 1991-to-2002 period to estimate the effect of after-tax price on insurance takeup and plan choice.  They found that lower premium shares led people to choose more expensive plans but had little effect on overall plan choice.  They point out that targeting people who are already offered employer-subsidized insurance but refuse it is very costly.  The reason:  The fact that these people have already turned down a highly subsidized product means that they are exceptionally price sensitive or already have insurance from another source.  They estimated that the federal government spent $31,000 to $83,000 per newly insured person.20 This conclusion is roughly in line with the results in Maine, where the DirigoChoice program spends almost $16,000 to insure one additional uninsured person.21

Economic theory predicts that people with smaller medical expenditures will be more sensitive to the price of health insurance than those with larger, and less discretionary, medical expenditures.  Empirical support for this supposition suggests that individuals self-select into insured or uninsured status depending on their knowledge of their own health.  This self-selection means that the uninsured are not an isolated population subgroup, and to insure the majority of people it is necessary to change behavior at relatively high levels of the income distribution.

"Means-tested subsidies are often extremely unfair."

There is little evidence that insurers “cherry pick” and sort across plans, suggesting that worries about adverse selection in insurance markets are likely exaggerated.  This means that the regulatory schemes proposed to correct the problem — mainly guaranteed issue and community rating — are unnecessary, and likely do more harm than good.22

Another problem is that means-tested subsidies are potentially extremely unfair and create a disincentive to act responsibly.  For example, if the Colorado Commission recommendations are followed, the state could end up taxing a young married couple with employer-provided health insurance, a baby and an income of $25,000 to provide health insurance subsidies to an older married couple with substantial home equity and retirement savings, three children, an annual income of $68,000 and a business that does not provide health insurance.

In order to provide more health care for all at a lower cost, other options need to be explored and evaluated, including:  subsidized clinics, designated hospitals to which those who cannot pay can be transferred, removing the regulations that discourage physicians from participating in charitable activities and charitable organizations from operating such programs, insurance plans that provide small benefits for low cost, reducing unnecessary licensing barriers and scope of practice restrictions, and insurance plans that offer catastrophic benefits.  [See Appendix B for some specific recommendations.]


Read Article as PDF