In the continuing debate over health care, and especially the uninsured, two seemingly contradictory claims are in play. The first is that the uninsured are being denied access to care. The second is that the uninsured receive too much care, overcrowding emergency rooms and placing hospitals and businesses that cover employee health costs at financial risk. In reality, both statements contain some truth, and both can be partly addressed by one simple action: Make catastrophic care insurance more broadly accessible.

We know that the uninsured receive some care and that much of hospitals' bad debt stems from catastrophic care provided to the uninsured. We also know that the uninsured tend to get 25 to 40 percent less care than the insured. The trouble with these comparisons is that the insured differ from the uninsured in ways that can affect their use of health care resources: Patients with chronic conditions may make sure that they are covered, which leads to increased care among the insured. Or the insured may consume unnecessary care, since they tend to ignore the costs of treatment covered by insurance. To understand whether there are real health consequences to not having insurance, we need to consider health outcomes.

In an effort to do so, I examined all the severe automobile accidents in Wisconsin over a five-year period. Such crashes offer a particularly "clean" look at the effect of health insurance coverage on treatment and health outcomes. For example, mortality rates can be compared for insured and uninsured victims. And unlike those with chronic conditions, people are unlikely to purchase health insurance because they think they're going to have a car accident. Another feature of crashes is that treatment and health outcomes can be compared from the moment the health problem begins, as professional medical care arrives automatically. While analysis of one health problem may not translate to others, crashes produce a wide range of injuries and tend to involve young adults, who are more likely to be uninsured. (Automobile accidents are the leading cause of death among those under age 35.)

What my research revealed was that the uninsured received about 80 percent of the care received by the insured and that they had a 37 percent higher mortality rate. (The mortality rate for all victims of serious car crashes was 3.8 percent, rising to 5.2 percent for the uninsured.) These outcome differences continued even after controlling for severity of injury, socioeconomic status and hospital resources that can vary between the insured and uninsured. Even insured and uninsured patients who were in the same accident and were rushed to the same hospital had these differences in treatment and mortality.

Consider the example of patients with skull fractures. Uninsured patients were less likely to receive an operation on the nervous system, such as a craniotomy: 34 percent of the insured patients received such surgeries, compared with 25 percent of the uninsured. The uninsured with this diagnosis also had higher mortality (22 percent, compared with 13 percent for the insured). The lack of health insurance appeared to have a significant effect on treatment decisions across a wide range of health problems. These decisions may be the domain of the health care providers, who need only refer to patients' charts to determine their insurance coverage. They may also reflect decisions by crash victims themselves, concerned about large medical bills. Either way, lack of coverage is associated with a lower likelihood of potentially life-saving procedures. To give an idea of the magnitudes, if these estimates are applied to the nation as a whole, they indicate that about 2,000 lives a year are lost because of lack of treatment for uninsured people involved in auto accidents.

Health care reform proposals that focus on catastrophic care could go a long way toward closing this health outcome gap, while also relieving emergency rooms and other health care providers of an enormous financial burden. And catastrophic insurance premiums, with a high deductible, are relatively inexpensive when compared with more comprehensive plans. Coupled with health savings accounts to cover the costs of preventive care, a program to subsidize catastrophic insurance could give real substance to the political goal of restraining spiraling health care expenses as patients begin to consider the costs of care. Such insurance could also offer peace of mind that large bills will be covered and that life-saving treatments will be provided.

The writer is an assistant professor of applied economics at the MIT Sloan School of Management.