Number B-14 in Series, "New Federalism: National Survey of America's Families"
The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.
Note: Those wishing to print this report may find it easier to use the PDF Version.
The problem of the uninsured is receiving renewed attention in the media and in the political process. Recent release of new data from the Current Population Survey (CPS) showed that the number of uninsured had reached 44 million. Several proposals to extend coverage through modifying the tax system have been introduced in Congress, and several presidential candidates have discussed strategies for addressing the problem.
In recent years, most of the attention on the uninsured has been focused on low-income children. In 1997, Congress established the Children's Health Insurance Program (CHIP), which gave grants to states to expand coverage of children. All states responded by adopting programs, and enrollment has grown to over 1 million children and is expected to increase as the program matures.
Much less attention has been given to low-income adults, despite the fact that there is a higher rate of uninsured adults and a far higher number without insurance. According to the Urban Institute's 1997 National Survey of America's Families (NSAF), 17 percent of adults lacked health insurance in 1997, compared with 12 percent of children. There were nearly three times as many uninsured adults as children-27 million versus 9 million.1 Of those with incomes below the federal poverty level (FPL), 42 percent of adults versus 21 percent of children lacked health insurance. Of those with incomes between 100 and 199 percent of the FPL, 34 percent of adults lacked insurance versus 21 percent of children.
Recent Proposals to Expand Coverage
As part of this growing interest in the problem of uninsurance, several proposals to expand coverage for adults have been introduced in the 106th Congress. Some of these involve systems of tax credits that would be made available to individuals and families if they obtained health insurance, while others propose the introduction of health marts and association health plans for the purpose of reducing the cost of health insurance for small employers. Additionally, there have been proposals to expand access to tax-deductible medical savings accounts that would also potentially make insurance attractive to low-income adults.
More recently, as a part of his presidential platform, Vice President Al Gore has proposed extending coverage to uninsured parents of Medicaid- and CHIP-eligible children with incomes below state-determined thresholds. The vice president's proposal would also introduce tax credits equal to 25 percent of insurance premium costs for workers in small firms (less than 50 employees). Adults without access to employer-sponsored coverage would also receive a refundable tax credit equal to 25 percent of the premium for nongroup health insurance.
The most far-reaching proposal to extend coverage to adults (and part of his presidential platform) recently came from Senator Bill Bradley. The Bradley proposal would have extended subsidies for the purchase of employer-sponsored coverage or a plan offered by the Federal Employees Health Benefits Program (FEHBP). Full subsidies would have been available for each low-income adult living in families with incomes below the FPL. Subsidies also would have been available on a decreasing scale for adults in families with incomes between 100 and 200 percent of the FPL. The Bradley proposal would also have made all health insurance premiums tax deductible.
In addition, many states have begun to look at existing methods of covering uninsured adults. For example, as part of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA), Section 1931 of the Social Security Act gives states additional flexibility in the establishment of more liberal eligibility rules to extend Medicaid coverage to families. Certain provisions under CHIP also give states limited ability to extend coverage to adults (Krebs-Carter and Holahan 2000).
The National Survey of America's Families
The NSAF is a national household survey that provides information on over 100,000 children and nonelderly adults representing the civilian nonelderly population under age 65 (Dean Brick et al. 1999). The NSAF oversampled the low-income population (those with incomes below 200 percent of the FPL) and the population in 13 selected states2 in order to generate reliable state- and national-level estimates. Detailed information on the health insurance coverage of one sampled adult and up to two sampled children in each household was gathered. Specifically, interviewers asked respondents about their and their children's current and past year's health insurance coverage. Overall information on health insurance coverage, access, and utilization was collected on 56,278 adults.3
This brief provides information from the NSAF on the insurance status of the adult population, by income, and how uninsured low-income adults are distributed according to a variety of demographic characteristics. Table 1 presents data on the distribution of insurance coverage by income for all adults; in addition, data are provided separately for adults with and without children, given that some policy proposals treat the two groups separately.
Income and Family Type
Table 1 shows that uninsurance rates are 42 percent for all adults with incomes below 100 percent of the FPL and 34 percent for those with incomes between 100 and 199 percent of the poverty level; of adults with children, 42 percent are uninsured, as opposed to 41 percent without children. Compared with adults without children, those with children are more likely to be covered by Medicaid and less likely to have employer-sponsored or private nongroup coverage. As income rises, public coverage is replaced by private coverage and the likelihood of being uninsured declines. A noticeable shift in employer-sponsored coverage occurs at the poverty level, below which only 22 percent of all adults report employer-provided coverage, compared with 47 percent of those living in families with incomes between 100 and 199 percent of the poverty level.
Table 2 decomposes the uninsured adult population by age, sex, race/ethnicity, family structure, work status, and health status. Analysis is restricted to those with incomes less than 200 percent of the FPL, because they account for 64 percent of the total number of uninsured adults and because most proposals to extend insurance coverage focus on this group. The number of individuals in each group is shown along with the uninsurance rate, the percentage of the low-income uninsured population represented by the group, and the percentage of the uninsured population as a whole represented by the group.
Table 2 demonstrates that adults between the ages of 18 and 34, particularly males, have a higher uninsurance rate than older age groups. Forty-seven percent of males ages 18 to 34 are uninsured, while females have an uninsurance rate of 38 percent. The lower rate for young females is primarily due to the greater likelihood of Medicaid coverage. Despite the lower rate for young females, low-income young men and women make up similar shares of the adult uninsured population, since young women are more likely to have low incomes than young men. Low-income adults between the ages of 18 and 34 account for 56 percent of the low-income uninsured and 35 percent of the uninsured population as a whole.
Older adults ages 55 to 64 account for only a small segment of the uninsured; however, given that health problems tend to develop with age, the uninsurance rate in this group is still a cause for concern. Low-income adults ages 55 to 64 are uninsured at a rate of 23 percent but account for only 8 percent of the low-income uninsured population and 5 percent of the uninsured population as a whole.
Table 2 further shows the strong relationship between race/ethnicity and lack of health insurance among low-income adults and adults in general. Thirty-one percent of low-income, white non-Hispanics are uninsured, and they account for 50 percent of the low-income uninsured and 31 percent of the uninsured population as a whole. Black non-Hispanics are uninsured at a somewhat higher rate, 34 percent, and account for 16 percent of the low-income uninsured. The highest rate of uninsurance is among low-income Hispanics, among whom 53 percent lack health insurance. Low-income Hispanics account for 29 percent of the low-income uninsured and 19 percent of the uninsured population as a whole.
The likelihood of being uninsured does not vary greatly by family structure for low-income adults. Low-income single parents have the lowest uninsurance rate of the four groups, 33 percent, primarily because they have more public coverage. Low-income single adults without children have the highest uninsurance rate, about 40 percent, and account for 25 percent of all uninsured adults. Because of the size of the groups, married adults with families account for 30 percent of the low-income uninsured; low-income single adults without children, 40 percent. Viewed another way, parents account for 45 percent of the low-income uninsured, and nonparents the remaining 55 percent; therefore, the impact of any coverage extension will depend heavily on whether or not having children is a determinant of eligibility.
Table 2 shows that most of the uninsured are in families with at least one full-time worker. Among those with incomes below 200 percent of the FPL, 39 percent of families with a full-time worker are uninsured, and those with two or more full-time workers are uninsured at a rate of 37 percent. Low-income adults in families with at least one full-time worker account for 65 percent of the low-income uninsured and 41 percent of the uninsured population as a whole. By contrast, low-income adults in families with no workers account for only 21 percent of the low-income uninsured and 13 percent of the uninsured population as a whole. The fact that the majority of the low-income uninsured do live in families with at least one full-time worker is noteworthy, given the public perception that uninsurance predominantly affects people living in families with no workers (Blendon, Young, and DesRoches 1999).
Table 2 also shows that the uninsurance rate for low-income adults reporting both good and fair or poor health was 41 percent, while the rate for those reporting excellent or very good health was 32 percent. That the uninsurance rate among those in excellent or very good health is lower than that of those in fair or poor health is somewhat surprising, considering that the latter should be more willing to pay for coverage and vice versa. Because less than a quarter of low-income adults report fair or poor health, those adults account for only 26 percent of the low-income uninsured. Finally, low-income adults reporting a condition limiting their ability to work have an uninsurance rate of 30 percent and account for 18 percent of the low-income uninsured and 11 percent of the uninsured population as a whole.
Access to Care
In addition to describing the characteristics of uninsured low-income adults, the uninsured population is compared here with the insured population according to several widely accepted measures of access to care. Table 3 illustrates that, for low-income adults, even after controlling for a variety of other factors that might affect health care access and utilization, there is a strong relationship between the lack of health insurance and various measures of access. The uninsured are more than two times more likely than the insured to report not having a usual source of care (including a hospital emergency room) and to lack confidence in their ability to access health care services. In addition, the uninsured are significantly more likely to be in fair or poor health, to have unmet needs for medical care or surgery, not to have had a physician or other health professional visit, and to lack satisfaction in quality of care received. These results provide further evidence that good health or lack of a need to access health care services are not the reasons why low-income adults lack health insurance, and again run counter to the perception that the uninsured are able to obtain needed health care (Blendon et al. 1999).
Variation across States
The uninsurance rate among low-income adults varies considerably across states (figure 1). The proportion of uninsured adults varies from a low of 21 percent in Minnesota to highs of 44 percent in California and 50 percent in Texas. In general, there is an inverse relationship between employer-sponsored coverage and the uninsured rate in each state; coverage by Medicaid and state programs also clearly matters. For example, Minnesota has an above-average rate of employer-sponsored coverage, 44 percent, and an above-average rate of Medicaid/state coverage, 23 percent. This results in a low rate of uninsurance among low-income adults. In contrast, Texas has below-average rates of employer-sponsored and Medicaid/state coverage (32 and 11 percent respectively), resulting in its high uninsurance rate. California has an even lower rate of employer-sponsored coverage than Texas, 29 percent, but a higher rate of Medicaid/state coverage, 18 percent. However, this higher rate of Medicaid/state coverage does not offset the low rate of employer-sponsored coverage, leaving California with the second-highest rate of low-income uninsured adults in the 13-state sample.
As shown in figure 2, the uninsurance rates in Alabama, California, Colorado, Massachusetts, Michigan, Minnesota, New York, Washington, and Wisconsin are significantly higher for adults in households with one or more full-time workers than those for households with no workers. Three of the four states where there was no significant difference in the uninsurance rates between households with one or more full-time workers and households with no workers were Florida, Mississippi, and Texas. These states all had the highest uninsurance rates in the country among households with no workers (with the exception of California), most likely reflecting their low rates of Medicaid coverage of low-income adults.
Figure 3 shows the distribution across states in the uninsurance rates of low-income adults by income, that is, for those with incomes less than 100 percent and for those with incomes between 100 and 199 percent of the poverty level. The uninsurance rates for those below the poverty level were generally higher than for those between 100 and 199 percent of the FPL. However, the differences were statistically significant only in Alabama, Colorado, Florida, Mississippi, and Texas. While all states cover relatively small percentages of adults between 100 and 199 percent of the FPL, these five states all have more limited public coverage of those below the FPL (relative to the national average), resulting in higher uninsurance rates for these individuals relative to the near poor.
These data are intended to provide an overview of adults lacking health insurance coverage in the United States. Findings show that younger, low-income adults, particularly blacks and Hispanics, have the highest uninsurance rates. At the same time, half of low-income uninsured adults are white, and the majority of uninsured low-income adults reside in households with at least one full-time worker. There is considerable variation among states both in the percentage and the composition of the low-income uninsured. Finally, uninsured low-income adults are significantly worse off on several measures of health care access and utilization. Given the scale of the problem, recent proposals that seek to address problems of uninsured adults are a welcome development. It is clear, however, that the scope and structure of these proposals affect how many uninsured adults they reach.
1. These estimates of the uninsured differ somewhat from those presented by the CPS for two key reasons. First, in developing the NSAF we sought to minimize any interviewee recall problems by asking questions about current health insurance coverage, as opposed to insurance coverage over the preceding calendar year. Second, instead of calculating the uninsured as a residual, we incorporated a question that confirms for each family member whether, in fact, they currently lack health insurance coverage (Rajan, Zuckerman, and Brennan 1999).
2. The 13 selected states are Alabama, California, Colorado, Florida, Massachusetts, Michigan, Minnesota, Mississippi, New Jersey, New York, Texas, Washington, and Wisconsin.
3. The household response rate for the NSAF is 70 percent (Brick, Flores-Cervantes, and Cantor 1999). Responses to the interviews are weighted to reflect the design features of the sample, including the oversampling of low-income households in 13 states, and contain adjustments for nonresponse and undercoverage. Variance estimates were computed using a replication method that adjusts for the survey's complex design. Flores-Cervantes, Brick, and DiGaetano (1999) describe this method and its application to the NSAF in detail. Imputed data for health insurance, income, and other variables with missing values are used. Imputed values account for 1.3 percent or less of all observations for health insurance (Dipko et al. 1999).
Blendon, Robert, John Young, and Catherine DesRoches. 1999. "The Uninsured, the Working Uninsured, and the Public." Health Affairs 18 (6): 203-11.
Brick, J. Michael, Ismael Flores- Cervantes, and David Cantor. 1999. 1997 NSAF Response Rates and Methods Evaluation. Washington, D.C.: The Urban Institute. National Survey of America's Families Methodology Report No. 8.
Dean Brick, Pat, Genevieve Kenney, Robin McCullough-Harlin, Shruti Rajan, Fritz Scheuren, Kevin Wang, J. Michael Brick, and Pat Cunningham. 1999. 1997 NSAF Survey Methods and Data Reliability. Washington, D.C.: The Urban Institute. National Survey of America's Families Methodology Report No. 1.
Dipko, Sarah, Michael Skinner, Nancy Vaden-Kiernan, John Coder, Esther Engstrom, Shruti Rajan, and Fritz Scheuren. 1999. 1997 NSAF Data Editing and Imputation. Washington, D.C.: The Urban Institute. National Survey of America's Families Methodology Report No. 10.
Flores-Cervantes, Ismael, J. Michael Brick, and Ralph DiGaetano. 1999. 1997 NSAF Variance Estimation. Washington, D.C.: The Urban Institute. National Survey of America's Families Methodology Report No. 4.
Krebs-Carter, Melora, and Holahan, John. 2000. "State Strategies to Cover Uninsured Adults." Urban Institute Working Paper. Washington, D.C.: The Urban Institute.
Rajan, Shruti, Steve Zuckerman, and Niall Brennan. 1999. "Verifying Insurance Coverage: Impact on Measuring the Uninsured within the National Survey of America's Families." Urban Institute Working Paper. Washington, D.C.: The Urban Institute.
About the Authors
John Holahan is director of the Health Policy Research Center at the Urban Institute. He has authored several publications on the Medicaid program. He has also published research on the effects of expanding Medicaid on the number of uninsured and the cost to federal and state governments. Other research interests include health system reform, changes in health insurance coverage, physician payment, and hospital cost containment.
Niall Brennan is a research associate in the Health Policy Center of the Urban Institute. He currently focuses on issues related to health insurance coverage and long-term care. His recent research has examined state variations in health insurance coverage, the efficacy of programs for low-income Medicare beneficiaries, and the effect of Medicaid managed care programs on health care access and use.