WHO KNOWS
Peter Cunningham, Ph.D., Senior Health Researcher, Center for Health System Change (RWJ)
Paul Fronstin, Ph.D., Senior Research Associate, Employee Benefit Research Institute
Has who is uninsured changed?
PC: I wouldn't say that there have been any major shifts. Most of the uninsured are members of working families who are in that gap where they are not eligible for public coverage and are not offered coverage by their employer. I would say that there may be some smaller shifts-some increase [in uninsurance] in the upper income ranges, primarily resulting from self-employed people who no longer can afford purchasing a separate policy for themselves because of the expense. We saw for the first time in the past 10 years a decrease in the uninsurance rate for children, which I think is a combination of both public coverage expansions and the strong economy.
PF: I don't think it really has changed that much. The characteristics are in large part the same. It's mainly low-wage workers and their family members. The people have changed, because people's circumstances change over time. If you move from a job in a small firm to one in a large firm you're much more likely to get health benefits. As people leave the ranks of the uninsured, other people join the ranks of the uninsured and you have a constant flow of people in and out of the state of uninsurance.
Are there things that we now know and understand about the uninsured population that we didn't before?
PC: I think since probably the late 1970s or early 1980s there has been so much data on it-I don't know if there's a whole lot out there that's new. I think that things may be coming to light that weren't so commonly known. One is that the uninsured problem is especially severe among Hispanics, [who] make up about one quarter of the uninsured. Because the Hispanic population is increasing, unless something is done to address the problem among Hispanics, the uninsurance rate will go up merely because of these demographic changes. Another not so well known problem is that the uninsured rate is especially high among young adults. The policymakers tend to focus on the groups that are commonly perceived as vulnerable-children, near elderly-but the uninsurance problem actually is less severe for those groups than for young adults.
PF: I think one of the things we understand now is that cost is the big issue. People just can't afford it. There is a small group of people who can afford it and don't want it, but if that group is covered, you still have 40 million people without health insurance coverage. I think the other thing is understanding the dynamics of how long [people are] uninsured, who's uninsured the longest, and why they are uninsured for so long. We have a better grasp on that, but we still have a long way to go before we completely understand these dynamics.
Is a different strategy needed now to reduce uninsurance?
PC: Short of universal coverage, what we probably need are a combination of things. One is an expansion of public programs such as SCHIP and Medicaid to cover more low-income families. That, combined with perhaps some sort of tax incentives-subsidies or credits-to help people purchase health insurance. Even with all that, it's not clear how much of a dent we're going to make. One thing policymakers have to grapple with is that virtually all the policy proposals during the past five years have relied on strictly voluntary programs. With a program like SCHIP, you're not just automatically eligible. You have to go through an enrollment and application process and a lot of people are not going to sign up for that reason. I think one thing-without advocating anything in particular-that policymakers need to be aware of is the limits of proposals that rely on voluntary sign-up. There always will be people who for one reason or another are not going to enroll. As long as we rely on that, then we're still going to have a significant uninsured problem.
PF: I think the strategy that we haven't tried is education. And the education I'm talking about letting both workers and employers know why it's important and why they should spend the money. I think if evidence of the benefits of offering insurance go beyond the insurance component-benefits such as more productive workers, less turnover, the other costs of not offering coverage that are offset by offering it-I think more people might be covered.
Are we able to measure state variations in uninsurance trends?
PC: The Current Population Survey can generate estimates at the state level. There also are other surveys. In our Community Tracking Study, we don't provide state-level estimates, but we do provide estimates for individual communities. The new federalism study by the Urban Institute is looking at 13 states. Basically, all the state or other local area analyses show considerable variation across states or across localities in the uninsurance rates. So we know that there is tremendous variation across the country.
PF: Dealing with the state data is tricky because it depends upon the size of the sample you're using. In their published state estimates the Census Bureau provides a three-year moving average that smooths out some of the statistical anomalies in the data set. I don't think any survey exists that's designed to provide state estimates consistently over time across all states. There are various state efforts to come up with the number of uninsured in their own state, which would be state-representative, but there are no 50-state surveys. The difficulty is small samples in some states, where you may only have 100 uninsured people. And even when you use a three-year moving average, you have only 300 uninsured people, which is not a lot to be using for an analysis. You have to be careful if you use state numbers.
How do we get good state-level data?
PC: I think there is a need for state-level data because many states are taking the initiative ot address the uninsured problem in their states. Actually, many states conduct their own surveys, as well. The only problem is that if all the states conduct separate surveys and they're using somewhat different sampling methods or interviewing methods or somewhat different questionnaires, then it's difficult to compare the estimates across states. A consistent methodology is necessary so you can make comparisons.
PF: I think in-person surveys are the way to get the best data, but it's the most costly. It's much cheaper to conduct a telephone survey, but it's not cheap. When you use random digit dialing to identify households, one person often answers the questions, and they are less likely to give you accurate information over the telephone. When you interview in person you can ask to see an insurance card, you can show them a Medicare card or a Medicaid card--a sample- and ask if they have something like that. Potentially, you obtain much more accurate information.
TO NEXT SECTION (WHAT WORKS)
BACK TO MAIN
|
