Health Insurance: How Much Does It Matter?

Brief Analyses | Health

No. 416
Wednesday, August 21, 2002
by Greg Scandlen

Two recent reports associate lack of health insurance coverage with less access to health care services and worse health outcomes. One study is written by Jack Hadley of the Urban Institute and published by the Kaiser Commission on Medicaid and the Uninsured, the other by an Institute of Medicine (IOM) committee, with support from the Robert Wood Johnson Foundation.

The two have much in common. Neither adds original research, instead compiling the results of previous studies. Both acknowledge the limitations of this approach. Also, both identify a correlation between lack of insurance and poor health. But neither is able to determine whether one causes the other or each is caused by something else entirely. In particular, the studies fail to consider the degree to which poverty and lack of education may cause both poor health and non-insurance. Yet there is evidence that income and education are more important determinants of health than insurance.

Unwarranted Conclusions

Based on the "preponderance of evidence," the IOM committee concludes that:

  • The health of uninsured adults is worse than it would be if they were insured.
  • Providing health insurance to uninsured adults would improve their health and lengthen their lives.
  • Increased rates of health insurance coverage would be especially beneficial to those who have the poorest health and the least access to care, and such increases likely would reduce health disparities among racial and ethnic groups.

IOM guesstimates that 18,000 people a year die from lack of health insurance coverage. Such a number is handy for media sound bites, but has no factual basis. The Urban Institute's Hadley acknowledges that these "observational studies... cannot answer the question of whether health insurance directly affects health outcomes."

Income and Health: International Evidence

Neither study addresses the clear association of low income and poor health status and the likelihood that income is far more important than insurance in determining a person's use of health services and overall state of health.

A recent survey of five English-speaking countries with very different health care financing systems found that many more people with below-average incomes report themselves to be in poor health than do higher-income people. In the United States, 37 percent of below-average-income adults said they were in fair or poor health, while only 9 percent of above-average-income people said the same. Similar results were found in the other countries, regardless of the level of insurance coverage. In the United Kingdom the ratio of low income to higher income adults reporting themselves in fair or poor health was 31 percent to 10 percent; in New Zealand, 20 percent to 6 percent; in Canada, 23 percent to 9 percent; and in Australia, 30 percent to 10 percent. All four of the latter countries are considered to have some version of "universal" health insurance; so if having insurance coverage were the primary factor in determining health status, these disparities should disappear.

Insurance and Health

In the United States, countless studies have shown that the uninsured tend to be poorer than the insured. Only 26.9 percent of people under the poverty line have private insurance coverage and 33 percent are uninsured. Among people over 200 percent of poverty, 86.2 percent have private coverage and only 10.5 percent are uninsured. Thus there is a direct association between having a low income and being uninsured, yet these two reports made almost no effort to control for income. According to the authors' descriptions of the 164 separate studies examined, only 18 indicate that any adjustments were made for income.

The Medicaid Population

While it might have been impossible for the two reports to adjust for income in studies already completed, the reports' authors ignored an available surrogate for income. Nearly one-third of the studies break out the Medicaid population from the privately insured population. Medicaid could serve very well as a "control group" to test the difference between lack of insurance and low income in receiving health care services. People on Medicaid are both low income and well insured. Of people below the poverty line, 41.4 percent are on Medicaid (far more than are uninsured), and of the people above 200 percent of poverty, only 3.3 percent are on Medicaid (far fewer than are uninsured).

To the extent that having insurance coverage results in better health care, the Medicaid population should be doing very well. To the extent income is the primary determinant, people on Medicaid should be doing quite poorly. Which proves to be true? In 61 percent of the cases (31 studies) that identify the three populations (privately insured, uninsured and on Medicaid), Medicaid recipients appear to do about as badly or worse than the uninsured in receiving health care services or maintaining good health.

If insurance coverage is the primary determinant in getting adequate health care, people on Medicaid should be at the head of the line for good outcomes. But they are not. In fact, in many cases, they have worse outcomes than do people with no insurance at all. This is consistent with other available information that income is a better predictor of health than is insurance status. Medicare enrollees, for instance, are all covered by the same insurance program, but those with low incomes are twice as likely to report being in poor health as those with higher incomes. [See Figure I.]

The Education Factor

Another strong predictor of health status is education. One massive study of the U.S. adult population found a strong correlation between poor education and poor health for most conditions. For example, people with eight or fewer years of school have four times as much arthritis and hypertension as do high school graduates. [See Figure II.] Other recent research concludes that each year of additional education lowers the chance of dying within 10 years by as much as 3.6 percent and that poor education is directly related to a number of behaviors such as smoking, drug abuse and poor nutrition that contribute to poor health, regardless of insurance status.


In their rush to find inequalities in the American health care system, the authors of these reports miss the most obvious discrepancy - one that plagues countries with "universal" health care as much as it does the United States. Simply put, people with low incomes and limited education have worse health than the rest of the population.

In the United States, some of the people in poor health also may be uninsured. Many, though, are enrolled in or eligible for Medicaid. And while Americans who are uninsured may suffer worse health than those with insurance, they also are likely to be poor and poorly educated. Putting an insurance card in their wallets will not restore their health.

Greg Scandlen is a senior fellow in health policy with the National Center for Policy Analysis.

Read Article as PDF