Issue Brief. Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 1999 Current Population Survey

Transcription

1 January 2000 Jan. Sources of Health Insurance and Characteristics of the Uninsured: Analysis of the March 1999 Current Population Survey by Paul Fronstin, EBRI EBRI EMPLOYEE BENEFIT RESEARCH INSTITUTE This Issue Brief provides summary data on the insured and uninsured populations in the nation and in each state. It discusses the characteristics most closely related to an individual s health insurance status. Based on EBRI estimates from the March 1999 Current Population Survey (CPS), it represents 1998 data the most recent data available. In 1998, million nonelderly Americans 81.6 percent had some form of health insurance. More than 64 percent had it through an employment-based health plan; 6.5 percent purchased it on their own; and 14.3 percent were covered by a public program, mostly through Medicaid (10.4 percent). Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec Issue Brief In 1998, 18.4 percent of the nonelderly population was uninsured (43.9 million people), compared with 14.8 percent in The percentage of uninsured Americans has generally been increasing since at least 1987, although the percentage uninsured in 1998 was not statistically different from the percentage uninsured in 1997 (18.3 percent). The increase in the uninsured prior to 1993 can be attributed to the erosion of employment-based health insurance. However, since 1993, the percentage of nonelderly Americans covered by an employment-based health plan has increased from 63.5 percent to 64.9 percent. The decline in public sources of health insurance would mostly explain the recent increase in the uninsured. For example, between 1994 and 1998 the percentage of nonelderly Americans covered by CHAMPUS/CHAMPVA declined from 3.8 percent to 2.9 percent, in large part due to downsizing in the military. Similarly, between 1993 and 1998, the percentage of nonelderly Americans covered by Medicaid declined from 12.7 percent to 10.4 percent as people left welfare. The increase in employment-based coverage since 1994 was due mainly to a higher likelihood that children were covered by an employment-based health plan. Between 1994 and 1998, the percentage of children covered by an employment-based health plan increased from 58.1 percent to 60.2 percent. For adults, it increased less than one percentage point, from 66.1 percent to 66.9 percent. Adults started to realize gains in employment-based health insurance between 1997 and Between 1994 and 1997, the percentage of working adults with employmentbased health insurance coverage held steady at roughly 72.3 percent. During this period, health care cost inflation was essentially nonexistent. However, between 1997 and 1998, the percentage of working adults with employment-based health insurance increased from 72.2 percent to 72.8 percent, despite the apparent return of health care cost inflation in It is likely that the changing composition of the labor force accounted for some of the increase in employment-based coverage. EBRI Issue Brief Number 217 January EBRI January 2000 EBRI Issue Brief 1

2 Table of Contents Text Introduction... 3 (table 1) Recent Trends... 4 (chart 1, chart 2, chart 3, chart 4, chart 5) Determinants of Coverage... 7 (table 2, chart 6, table 3, table 4, table 5, table 6) The Uninsured Location (table 7, chart 7, table 8) Citizenship (table 9) Employment Industry Firm Size Income (chart 8) Race and Origin Family Type (chart 9) Age (table 10) Children (table 11, chart 10, chart 11) Policy Implications Conclusion Appendix Duration of Coverage References Tables Table 1, Nonelderly Americans With Selected Sources of Health Insurance Coverage, Table 2, Nonelderly Population With Selected Sources of Health Insurance, by Age and Own Work Status, and Work Status of Family Head, Table 3, Workers Ages With Selected Sources of Health Insurance, by Industry of Primary Employment, Table 4, Workers Ages With Selected Sources of Health Insurance, by Firm Size, Table 5, Nonelderly Population With Selected Sources of Health Insurance, by Family Income, Table 6, Nonelderly Population With Selected Sources of Health Insurance, by Race and Poverty Status, Table 7, Nonelderly Population With Selected Sources of Health Insurance, by Region and State, Table 8, Nonelderly Population Living in Consolidated Metropolitan Statistical Areas (CMSAs) With Selected Sources of Health Insurance, by CMSA, Table 9, Nonelderly Population With and Without Health Insurance, by Region, State, and Citizenship, Table 10, Persons Ages With Selected Sources of Health Insurance, by Gender and Age, Table 11, Children With Selected Sources of Health Insurance, by Poverty Level and Age, Charts Chart 1, Percentage of American Children, Ages 0 17, With Employment-Based Health Benefits, Medicaid, and Without Health Insurance, Chart 2, Percentage of American Adults, Ages 18 64, With Employment-Based Health Benefits, Medicaid, and Without Health Insurance, Chart 3, Percentage of Women Ages Who Are in Families With Welfare Income or Who Are Employed, Chart 4, Percentage of Workers, Ages 18 64, With Employment-Based Health Benefits, Medicaid, and Without Health Insurance, Chart 5, Percentage of Workers Who Are Self-Employed, Working for Large Firms, or Part Time, Chart 6, Percentage of Nonworkers, Ages 18 64, With Employment-Based Health Benefits, Medicaid, and Without Health Insurance, Chart 7, Percentage Uninsured, by State, Chart 8, Percentage Uninsured Among Workers Ages 18 64, by Total Earnings, Chart 9, Percentage Uninsured Among the Nonelderly Population, by Family Type, Chart 10, Percentage Uninsured Among Children Under Age 18, by Work Status of the Family Head, Chart 11, Children Under Age 18 Without Health Insurance, by Work Status of the Family Head, January 2000 EBRI Issue Brief

3 Paul Fronstin of EBRI wrote this Issue Brief with assistance from the Institute s research and editorial staffs. Any views expressed in this report are those of the author and should not be ascribed to the officers, trustees, members, or other sponsors of EBRI, EBRI-ERF, or their staffs. Neither EBRI nor EBRI-EBRI lobbies or takes positions on specific policy proposals. EBRI invites comment on this research. In 1998, million nonelderly Introduction Americans were covered by some form of health insurance (calculated from table 1). Overall, this represents 81.6 percent of the nonelderly population. Almost 65 percent were covered by an employment-based health plan, while 6.5 percent purchased a health plan directly from an insurance company or, in some cases, through a group that is not employment-based, such as an association. More than 14 percent of the nonelderly had health insurance coverage from a public program, such as Medicaid. In 1998, 18.4 percent of the nonelderly population was uninsured, compared with 14.8 percent in Between 1997 and 1998, the percentage of uninsured Americans was statistically unchanged, but, in general, the percentage of uninsured Americans has been increasing since at least While the increase in the uninsured between 1987 and 1993 can be attributed to the erosion of employment-based health benefits, 1 the portion of Americans covered by employment-based health insurance increased between 1993 (63.5 percent) and 1998 (64.9 percent). 2 The decline in public sources of health insurance would mostly explain the recent increase in the uninsured population. For example, between 1994 and 1998 the percentage of nonelderly Americans covered by CHAMPUS/CHAMPVA 3 declined from 3.8 percent to 2.9 percent, in large part due to downsizing in the military. Similarly, between 1993 and 1998, the percentage of nonelderly Americans covered by Medicaid (the federal-state insurance program for the poor) declined from 12.7 percent to 10.4 percent as people left welfare for work. It may be that, while the percentage of individuals covered by employment-based coverage is rising, individuals previously covered by Medicaid and CHAMPUS/CHAMPVA are not being fully absorbed into the employment-based market for health insurance. For example, they may be taking jobs that offer health insurance coverage but declining the benefit. In fact, previous research has shown that take-up rates have fallen for all workers, and have fallen more for low-wage workers than for high-wage workers (Cooper and Schone, 1997). They may also be taking jobs that do not offer health benefits to them. 4 The purpose of this Issue Brief is to examine the status of health insurance coverage in the United States. The data are based on the March 1999 Current Population Survey (CPS). The report focuses primarily on the nonelderly population (under age 65) because this group receives health insurance coverage from a number of different sources, depending, for example, on income, employment status, and location. Medicare covers 96 percent of the elderly population, the least likely group to be employed. The next section discusses recent trends in health insurance coverage and some of the underlying factors affecting these trends. The following section discusses the determinants of having employment-based health insurance coverage and other sources of coverage. The section after that discusses the uninsured population and the factors associated with being uninsured, and is followed by a section examining policy implications. The final section presents conclusions. The data are discussed in the appendix. 1 See Fronstin and Snider (1996/97) for an analysis of the decline in employment-based health insurance between 1988 and While the year-to-year changes may not be statistically significant, the fiveyear trend is clearly upward. 3 CHAMPUS is the Civilian Health and Medical Program of the Uniformed Services, which covers dependents of active duty and retired members of the armed forces, and CHAMPVA is the Civilian Health and Medical Program of the Veterans Administration, which covers dependents of totally disabled veterans. 4 This does not imply that the percentage of workers offered health benefits has been declining. Fronstin (1999a) found that the percentage of workers offered health benefits was virtually unchanged between 1988 and In 1997, 75 percent of workers were offered health benefits. In 1988, 76 percent were offered health benefits. January 2000 EBRI Issue Brief 3

4 Table 1 Nonelderly Americans With Selected Sources of Health Insurance Coverage, b 1998 (millions) Total Population Employment-Based Coverage Own name Dependent coverage Individually Purchased Public Medicare Medicaid CHAMPUS/CHAMPVA a No Health Insurance (percentage) Total Population 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Employment-Based Coverage Own name Dependent coverage Individually Purchased Public Medicare Medicaid CHAMPUS/CHAMPVA a No Health Insurance Source: Employee Benefit Research Institute estimates from the March Current Population Surveys. Note: Details may not add to totals because individuals may receive coverage from more than one source. a Civilian Health and Medical Program for the Uniformed Services and the Civilian Health and Medical Program for the Department of Veterans Affairs. b Medicaid and uninsured data are not completely consistent with data from previous years. Starting with the March 1998 Current Population Survey (CPS), the Bureau of the Census modified its definition of the population with Medicaid and the population without health insurance coverage. Previously, individuals covered solely by the Indian Health Service were counted in the Medicaid population. Beginning with data from the March 1998 CPS, individuals covered solely by the Indian Health Service are counted as uninsured. This change decreased the Medicaid population and increased the uninsured population by 300,000, or 0.2 percent. Recent Trends Since 1994, the percentage of nonelderly Americans covered by an employmentbased health insurance plan has been increasing (table 1). Overall, the increase in coverage was due mainly to a higher likelihood that children were covered by an employment-based health plan. Between 1994 and 1998, the percentage of children covered by an employment-based health plan increased from 58.1 percent to 60.2 percent (chart 1). For adults, it increased less than one percentage point, from 66.1 percent to 66.9 percent, with most of the increase occurring before 1997 and 1998 (chart 2). Fronstin (1999b) has shown that the likelihood of a child being covered by an employment-based health plan increased for a number of reasons. The study found that the percentage of children with a working parent increased, the percentage of children in families with incomes below the poverty level decreased, and more children had a working parent employed in a large firm. The increase in employment-based coverage among children can in part be attributed to a combination of welfare reform and the strong economy, both of which resulted in fewer adult women on welfare and more adult women working. Chart 3 shows how the percentage of women ages in families receiving public assistance or welfare income declined, while the employment rate increased. Between 1994 and 1997, the percentage of working adults with employment-based health insurance coverage held steady at roughly 72.3 percent (chart 4). During this period, health care cost inflation was essentially nonexistent. Between 1997 and 1998, the percentage of working adults with employment-based health insurance increased from 72.2 percent to 72.8 percent, despite the apparent return of health care cost inflation in 1998, when health care costs increased 4 January 2000 EBRI Issue Brief

5 Chart 1 Percentage of American Children, Ages 0-17, With Employment-Based Health Benefits, Medicaid, and Without Health Insurance, % 70% 60% 66.7% 66.5% 65.8% 64.0% 62.7% 62.0% 59.5% 58.1% 58.6% 58.9% 59.7% 60.2% 50% 40% Employment-Based Medicaid Uninsured 30% 20% 15.5% 15.9% 16.0% 18.9% 20.8% 22.0% 23.9% 22.9% 23.2% 21.8% 20.5% 19.8% 10% 13.1% 13.3% 13.6% 13.2% 12.9% 12.7% 13.7% 14.2% 13.8% 14.8% 15.0% 15.4% 0% a Source: Employee Benefit Research Institute estimates from the March Current Population Surveys. a Medicaid and uninsured data are not completely consistent with data from previous years. Starting with the March 1998 Current Population Survey (CPS), the Bureau of the Census modified its definition of the population with Medicaid and the population without health insurance coverage. Previously, individuals covered solely by the Indian Health Service were counted in the Medicaid population. Beginning with data from the March 1998 CPS, individuals covered solely by the Indian Health Service are counted as uninsured. Chart 2 Percentage of American Adults, Ages 18 64, With Employment-Based Health Benefits, Medicaid, and Without Health Insurance, % 70% 60% 70.3% 70.0% 69.8% 68.3% 67.8% 65.9% 65.3% 66.1% 66.1% 66.3% 66.2% 66.9% 50% 40% 30% 20% 10% Employment-Based Medicaid Uninsured 15.6% 16.4% 16.6% 17.4% 17.8% 18.9% 18.8% 18.5% 19.0% 18.9% 19.7% 5.6% 5.7% 5.7% 6.4% 7.0% 7.3% 7.8% 7.9% 7.8% 7.8% 6.9% 19.7% 6.4% 0% a Source: Employee Benefit Research Institute estimates from the March Current Population Surveys. amedicaid and uninsured data are not completely consistent with data from previous years. Starting with the March 1998 Current Population Survey (CPS), the Bureau of the Census modified its definition of the population with Medicaid and the population without health insurance coverage. Previously, individuals covered solely by the Indian Health Service were counted in the Medicaid population. Beginning with data from the March 1998 CPS, individuals covered solely by the Indian Health Service are counted as uninsured. January 2000 EBRI Issue Brief 5

6 Chart 3 Percentage of Women Ages Who are in Families With Welfare Income or Who are Employed, % 76.9% 77.5% 77.8% 78.3% 78.2% 70% 60% 50% 40% Percentage With Welfare Income Percentage Employed 30% 20% 10% 0% 4.6% 4.4% 3.9% 3.0% 2.4% Source: Employee Benefit Research Institute estimates of the March Current Population Surveys. 80% Chart 4 Percentage of Workers, Ages 18 64, With Employment-Based Health Benefits, Medicaid, and Without Health Insurance, % 76.1% 75.5% 73.9% 72.2% 73.9% 72.2% 71.8% 72.4% 72.4% 72.3% 72.2% 72.8% 60% 50% 40% Employment-Based Medicaid Uninsured 30% 20% 14.6% 15.6% 16.8% 17.8% 16.8% 17.8% 17.8% 17.3% 17.6% 17.5% 18.2% 18.1% 10% 2.4% 2.4% 3.2% 3.4% 3.2% 3.4% 3.7% 4.0% 4.0% 4.2% 3.6% 3.5% 0% a Source: Employee Benefit Research Institute estimates from the March Current Population Surveys. a Medicaid and uninsured data are not completely consistent with data from previous years. Starting with the March 1998 Current Population Survey (CPS), the Bureau of the Census modified its definition of the population with Medicaid and the population without health insurance coverage. Previously, individuals covered solely by the Indian Health Service were counted in the Medicaid population. Beginning with data from the March 1998 CPS, individuals covered solely by the Indian Health Service are counted as uninsured. 6 January 2000 EBRI Issue Brief

7 Chart 5 Percentage of Workers Who Are Self-Employed, Working for Large Firms, or Part-Time, % 40% 30% 44% 44% 45% 45% 46% % 19% 18% 18% 18% 17% 10% 10% 9% 10% 10% 9% 0% Percentage Self-Employed Percentage Employed by Firms With 100 or More Workers Percentage Part-Time Source: Employee Benefit Research Institute estimates from the March Current Population Surveys. between 4 percent and 9 percent. 5 It is likely that the changing composition of the labor force accounted for some of the increase in employment-based coverage. For example, the percentage of workers who were selfemployed declined between 1997 and 1998, while the percentage of workers employed in private-sector firms with 100 or more workers increased (chart 5). Similarly, fewer workers were employed on a part-time basis in 1998, compared with 1997 (chart 5). In fact, the movement of workers from small firms to large firms and the decreased use of part-time employees date back to The increase in employment-based coverage between 1997 and 1998 is both surprising and not surprising. It is not surprising because the strong economy and low unemployment rates may have caused more employers to provide health benefits and also may have resulted in more workers being able to afford health insurance. It is surprising because 1998 saw the return of health care cost inflation. In the late 1980s and early 1990s, the percentage of Americans covered by an employment-based health plan declined in large part because of health care cost inflation. In the late 1980s, health care costs increased an average of between 15 percent and 20 percent. However, between 1994 and 1997, 5 Various sources have provided different estimates of health care cost inflation. The U.S. Office of Personnel Management reported that health insurance premiums increased approximately 7.2 percent in 1998 and 9.5 percent in Premiums are expected to increase 9.3 percent in 2000 ( William M. Mercer reported that health care costs increased 6.1 percent in 1998 and were expected to increase 9 percent in 1999 ( resource_news_topic40.html). The Kaiser Family Foundation reported that health insurance premiums had increased 3.7 percent between the Spring 1997 and Spring 1998 and 4.8 percent between the Spring 1998 and Spring 1999 ( health care costs barely changed. In 1998, they started to increase again, but the increase does not appear to have affected the percentage of Americans with employmentbased health benefits. More research needs to be conducted in this area. Determinants of Coverage Full-time workers, publicsector employees, workers employed in manufacturing, and individuals living in families with high levels of income are most likely to be covered by employment-based health insurance. Persons in families with income below the poverty level, especially children and single-parent families, are most likely to be covered by public health insurance such as Medicaid. Employment status is the most important determinant of health insurance coverage. Almost twothirds of the nonelderly population have employmentbased coverage. This coverage can be obtained either directly through one s employer/union or previous employer or indirectly through an employed person in one s family. In this report, individuals who receive coverage directly through their employer/union or a previous employer are categorized as having coverage in their own name. Individuals who receive employmentbased coverage indirectly are categorized as having dependent coverage. Large employers that provide access to group health insurance often are able to provide health benefits at lower cost than small employers, because they are subject to less adverse selection and their average January 2000 EBRI Issue Brief 7

8 Table 2 Nonelderly Population with Selected Sources of Health Insurance, by Age and Own Work Status, and Work Status of Family Head, 1998 Employment-Based Coverage Public Own Work Status and Individually Work Status of Family Head Total Total Own name Dependent Purchased Total Medicaid Uninsured (millions) Total Own Work Status Child Adult worker nonworker Work Status of Family Head Full-year, full time worker Other worker Nonworker (percentage within coverage category) Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Own Work Status Child Adult worker nonworker Work Status of Family Head Full-year, full-time worker Other worker Nonworker (percentage within work status categories) Total 100.0% 64.9% 33.1% 31.7% 6.5% 14.3% 10.4% 18.4% Own Work Status Child Adult worker nonworker Work Status of Family Head Full-year, full-time worker Other worker Nonworker Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. administrative costs and marketing costs are lower. However, examination of actual health benefit costs across firms usually shows that per-person costs are higher in larger firms than in smaller firms. This occurs because large firms typically offer more extensive health benefits than small firms. Furthermore, the nature of employment, the industry, and the firm s size often determine the cost and extent of coverage. Workers in large firms are more likely to be covered by health insurance than those in small firms. In 1998, 64.9 percent of the nonelderly were covered by employment-based health insurance (table 1). Workers were much more likely to be covered by employment-based health insurance than nonworkers (table 2). Seventy-three percent of workers were covered by an employment-based plan (table 2), compared with 40.5 percent of nonworkers (chart 6). In addition, 75.5 percent of individuals in families headed by fullyear, full-time workers were covered by employmentbased health insurance, compared with 40.3 percent of those in families headed by other workers, and 19.7 percent of individuals in families headed by nonworkers (table 2). With respect to industry, workers employed in the public sector and in manufacturing were more likely to have employment-based coverage in their own name than other workers (table 3). In addition, the larger the firm the more likely were workers to have employment- 8 January 2000 EBRI Issue Brief

9 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 46.4% Chart 6 Percentage of Nonworkers, Ages 18-64, With Employment-Based Health Benefits, Medicaid, and Without Health Insurance, % 45.1% 43.3% 19.7% 20.2% 21.1% 21.8% 19.0% 19.8% 19.9% 21.2% 42.0% 23.1% 39.7% 23.5% 38.7% 24.7% 38.7% 24.8% 22.1% 23.4% 23.0% 23.5% Employment-Based Uninsured 38.8% 39.1% 39.0% 25.0% 25.3% 26.2% a % Medicaid 24.0% 21.7% 40.5% 27.0% 19.5% Source: Employee Benefit Research Institute estimates from the March Current Population Surveys. amedicaid and uninsured data are not completely consistent with data from previous years. Starting with the March 1998 Current Population Survey (CPS), the Bureau of the Census modified its definition of the population with Medicaid and the population without health insurance coverage. Previously, individuals covered solely by the Indian Health Service were counted in the Medicaid population. Beginning with data from the March 1998 CPS, individuals covered solely by the Indian Health Service are counted as uninsured. Table 3 Workers Ages With Selected Sources of Health Insurance, by Industry of Primary Employment, 1998 Employment-Based Coverage Public Individually Industry Total Total Own name Dependent Purchased Total Medicaid Uninsured (millions) Total Agriculture, forestry, fishing, mining and construction Manufacturing Wholesale and retail trade Personal services Public sector (percentage within coverage category) Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Agriculture, forestry, fishing, mining and construction Manufacturing Wholesale and retail trade Personal services Public sector (percentage within industry categories) Total 100.0% 72.8% 55.5% 17.3% 5.7% 6.1% 3.5% 18.1% Agriculture, forestry, fishing, mining and construction Manufacturing Wholesale and retail trade Personal services Public sector Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. January 2000 EBRI Issue Brief 9

10 Table 4 Workers Aged With Selected Sources of Health Insurance, by Firm Size, 1998 Employment-Based Coverage Public Individually Firm Size Total Total Own name Dependent Purchased Total Medicaid Uninsured (millions) Total Self-Employed Total Wage and Salary Workers Public sector Private sector less than ,000 or more (percentage within coverage category) Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Self-Employed Total Wage and Salary Workers Public sector Private sector less than ,000 or more (percentage within firm size categories) Total 100.0% 72.8% 55.5% 17.3% 5.7% 6.1% 3.5% 18.1% Self-Employed Total Wage and Salary Workers Public sector Private sector less than ,000 or more Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. based coverage in their own name. Twenty-seven percent of self-employed workers and 27.4 percent of privatesector workers in firms with fewer than 10 employees were covered through a group health plan sponsored by their own employer/union or former employer in 1998, compared with 66.5 percent of private-sector workers in firms with 1,000 or more employees (table 4). Health insurance coverage is also related to income. In general, individuals with higher levels of income are more likely to be covered by employmentbased health insurance, while those with lower levels of income are more likely to be covered by a publicly sponsored plan. In 1998, 10.9 percent of individuals in families with annual income below $5,000 were covered by employment-based health insurance, compared with 84.9 percent of those in families with annual income of $50,000 or more (table 5). Although many individuals in poor families are covered by public health plans, that coverage is far from universal. In 1998, 44.8 percent of the nonelderly with family incomes below the poverty line were covered by a public plan 41.6 percent by Medicaid (table 6) although many low-income individuals may be eligible for Medicaid coverage even though they do not report 10 January 2000 EBRI Issue Brief

11 Table 5 Nonelderly Population With Selected Sources of Health Insurance, by Family income, 1998 Employment-Based Coverage Public Individually Family Income Total Total Own name Dependent Purchased Total Medicaid Uninsured (millions) Total Under $5, $5,000 $9, $10,000 $14, $15,000 $19, $20,000 $29, $30,000 $39, $40,000 $49, $50,000 and Over (percentage within coverage category) Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Under $5, $5,000 $9, $10,000 $14, $15,000 $19, $20,000 $29, $30,000 $39, $40,000 $49, $50,000 and Over (percentage within family income categories) Total 100.0% 64.9% 33.1% 31.7% 6.5% 14.3% 10.4% 18.4% Under $5, $5,000 $9, $10,000 $14, $15,000 $19, $20,000 $29, $30,000 $39, $40,000 $49, $50,000 and Over Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. coverage. Other sources of public health insurance include Medicare (which primarily covers the elderly but also covers qualified nonelderly disabled persons), CHAMPUS, CHAMPVA, and Veterans Administration (VA) health insurance. Many factors influence whether or not The Uninsured an individual has any type of health insurance coverage. This section presents data on the characteristics of the uninsured population. Location The proportion of the nonelderly population with and without health insurance varies by location. In 12 states, 20 percent or more of the population was uninsured in 1998 (table 7). These states are in large part concentrated in the south central and southwestern parts of the United States (chart 7). In many of these states a smaller proportion of the population was eligible for employment-based insurance and/or a larger proportion was eligible for publicly financed health programs than the national average. Lower average income and higher unemployment rates may both contribute to this difference. In addition, many of these states have a higher concentration of racial and ethnic groups that are less likely to be covered by health insurance. In contrast, states with a low percentage of uninsured individuals January 2000 EBRI Issue Brief 11

12 Table 6 Nonelderly Population With Selected Sources of Health Insurance, by Race and Poverty Status, 1998 Employment-Based Coverage Public Race and Individually Poverty Status Total Total Own name Dependent Purchased Total Medicaid Uninsured (millions) Total % % 149% % 199% % or more White % % 149% % 199% % or more Black % % 149% % 199% % or more Hispanic % % 149% % 199% % or more Other % % 149% % 199% % or more (percentage within race and poverty categories) Total 100.0% 64.9% 33.1% 31.7% 6.5% 14.3% 10.4% 18.4% 0 99% % 149% % 199% % or more White % % 149% % 199% % or more Black % % 149% % 199% % or more Hispanic % % 149% % 199% % or more Other % % 149% % 199% % or more Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. 12 January 2000 EBRI Issue Brief

13 Table 7 Nonelderly Population With Selected Sources of Health Insurance, by Region and State, 1998 Employment-Based Coverage Public Individually Region and State Total Total Own name Dependent Purchased Total Medicaid Uninsured (millions) Total New England Maine New Hampshire Vermont Massachusetts Rhode Island Connecticut Middle Atlantic New York New Jersey Pennsylvania East North Central Ohio Indiana Illinois Michigan Wisconsin West North Central Minnesota Iowa Missouri North Dakota South Dakota Nebraska Kansas South Atlantic Delaware Maryland District of Columbia Virginia West Virginia North Carolina South Carolina Georgia Florida East South Central Kentucky Tennessee Alabama Mississippi West South Central Arkansas Louisiana Oklahoma Texas Mountain Montana Idaho Wyoming Colorado New Mexico Arizona Utah Nevada Pacific Washington Oregon California Alaska Hawaii (continued) January 2000 EBRI Issue Brief 13

14 Table 7 (continued) Employment-Based Coverage Public Individually Region and State Total Total Own name Dependent Purchased Total Medicaid Uninsured (percentage) Total 100.0% 64.9% 33.1% 31.7% 6.5% 14.3% 10.4% 18.4% New England Maine New Hampshire Vermont Massachusetts Rhode Island Connecticut Middle Atlantic New York New Jersey Pennsylvania East North Central Ohio Indiana Illinois Michigan Wisconsin West North Central Minnesota Iowa Missouri North Dakota South Dakota Nebraska Kansas South Atlantic Delaware Maryland District of Columbia Virginia West Virginia North Carolina South Carolina Georgia Florida East South Central Kentucky Tennessee Alabama Mississippi West South Central Arkansas Louisiana Oklahoma Texas Mountain Montana Idaho Wyoming Colorado New Mexico Arizona Utah Nevada Pacific Washington Oregon California Alaska Hawaii Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. 14 January 2000 EBRI Issue Brief

15 Chart 7 Percentage Uninsured, by State, 1998 Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. 20.9% 27.2% 17.5% 20.9% 16.4% 17.5% 12.5% 16.4% 20. 9% t o 27. 2% (11) 17. 5% t o10.2% 12.5% 20. 9% (10) 16. 4% t o 17. 5% (6) 12. 5% t o 16. 4% (13) 10. 2% t o 12. 5% (11) include Nebraska, Minnesota, and Iowa. The percentage of the population without any form of health insurance coverage also varies by metropolitan region. For example, 19.8 percent of the population residing in Consolidated Metropolitan Statistical Areas (CMSAs) were uninsured in 1998 (table 8), compared with 18.4 percent overall. The Houston-Galveston-Brazoria, TX, CMSA had the highest percentage uninsured among CMSAs, at 30.5 percent, followed by Los Angeles-Riverside-Orange County, CA, at 29.3 percent and Miami-Fort Lauderdale, FL, at 25 percent. This compares with 10.2 percent uninsured in the Cincinnati-Hamilton, OH-KY-IN, CMSA, and 10.6 percent in the Milwaukee-Racine, WI, CMSA. Citizenship Citizenship is a primary factor in the likelihood of an individual having coverage and in the source of that coverage (table 9). In California, for example, 16.6 percent of nonelderly individuals reported that they were noncitizens, compared with 6.5 percent of the nation as a whole. Over 45 percent of nonelderly respondents indicating they were noncitizens were uninsured in 1997, compared with 16.5 percent of citizens. In Texas, more than 60 percent of the noncitizen population was uninsured. In Oregon and California, more than 50 percent was uninsured. High uninsured rates may be due in part to the fact that a higher proportion of noncitizens than citizens were in low-income families, were likely to be nonworkers, or were likely to work in small firms. Employment Eighty-three percent of the uninsured lived in families headed by workers in 1998, primarily because most people live in families headed by workers, including oneperson families (table 2). Seventeen percent of the uninsured were in families in which the family head did not work. Industry Uninsured workers were most likely to be employed in the wholesale and retail trade industry (table 3). This is not surprising, as workers in general are most likely to be employed in the wholesale and trade industry. For example, 33.7 percent of all workers are employed in the wholesale and retail trade industry, while 41 percent of uninsured workers are in this industry. This indicates that workers employed in the wholesale and retail trade industry are more likely to be uninsured than most other workers. These data are shown in the middle part of table 3. Workers employed in agriculture, forestry, fishing, mining, and construction were also disproportionately more likely to be uninsured, accounting for 17.4 percent of the uninsured population, while representing 9.3 percent of the working population. January 2000 EBRI Issue Brief 15

16 Table 8 Nonelderly Population Living in Consolidated Metropolitan Statistical Areas (CMSAs) With Selected Sources of Health Insurance, by CMSA, a 1998 Employment-Based Coverage Public Individually CMSA Total Total Own name Dependent Purchased Total Medicaid Uninsured (millions) Total Boston-Worcester-Lawrence, MA-NH-ME-CT Chicago-Gary-Kenosha, IL-IN-WI Cincinnati-Hamilton, OH-KY-IN Cleveland-Akron, OH Dallas-Fort Worth, TX Denver-Boulder-Greeley, CO Detroit-Ann Arbor-Flint, MI Houston-Galveston-Brazoria, TX Los Angeles-Riverside-Orange County, CA Miami-Fort Lauderdale, FL Milwaukee-Racine, WI New York-Northern New Jersey-Long Island, NY-NJ-CT-PA Philadelphia-Wilmington-Atlantic City, PA-NJ-DE-MD Portland-Salem, OR-WA Sacramento-Yolo, CA San Francisco-Oakland-San Jose, CA Seattle-Tacoma-Bremerton, WA Washington-Baltimore, DC-MD-VA-WV (percentage within CMSA category) Total 100.0% 64.7% 33.4% 31.3% 6.0% 12.5% 9.9% 19.8% Boston-Worcester-Lawrence, MA-NH-ME-CT Chicago-Gary-Kenosha, IL-IN-WI Cincinnati-Hamilton, OH-KY-IN Cleveland-Akron, OH Dallas-Fort Worth, TX Denver-Boulder-Greeley, CO Detroit-Ann Arbor-Flint, MI Houston-Galveston-Brazoria, TX Los Angeles-Riverside-Orange County, CA Miami-Fort Lauderdale, FL Milwaukee-Racine, WI New York-Northern New Jersey-Long Island, NY-NJ-CT-PA Philadelphia-Wilmington-Atlantic City, PA-NJ-DE-MD Portland-Salem, OR-WA Sacramento-Yolo, CA San Francisco-Oakland-San Jose, CA Seattle-Tacoma-Bremerton, WA Washington-Baltimore, DC-MD-VA-WV Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. a The specific metropolitan identifiers on this file are based on the Office of Management and Budget s June 30, 1993, definitions. Firm Size Almost 60 percent of all uninsured workers were either self-employed or working in private-sector firms with fewer than 100 employees in 1998 (table 4). Twenty-five percent of self-employed workers were uninsured, compared with 18.1 percent of all workers (table 4). Thirty-four percent of workers in private-sector firms with fewer than 10 employees were uninsured, compared with 12.7 percent of workers in private-sector firms with 1,000 or more employees. 16 January 2000 EBRI Issue Brief

17 Table 9 Nonelderly Population With and Without Health Insurance, by Region, State, and Citizenship, 1998 (In Regions and States Where the Percentage of Noncitizens is Greater Than 5 percent and With 75,000 or More Noncitizens) Insured Uninsured Uninsured Region Total and State Population Noncitizens Total Citizen Noncitizen Total Citizen Noncitizen Citizen Noncitizen (millions) (percentage) (millions) (millions) (percentage) Total % % 45.1% New England Massachusetts Middle Atlantic New York New Jersey East North Central Illinois South Atlantic Maryland Florida West South Central Texas Mountain Arizona Nevada Pacific Oregon California Hawaii Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. a Fewer than 50,000 respondents (weighted) in this category. Chart 8 Percentage Uninsured Among Workers Ages 18 64, by Total Earnings, % 30% 25% 30.6% 29.3% 20% 15% 18.1% 15.8% 10% 5% 9.1% 6.4% 5.2% 0% Total Less than $10,000 $10,000 $19,999 $20,000 $29,999 $30,000 $39,999 $40,000 $49,999 $50,000 or More Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. January 2000 EBRI Issue Brief 17

18 30% Chart 9 Percentage Uninsured Among the Nonelderly Population, by Family Type, % 25% 20% 18.4% 17.3% 22.0% 15% 13.5% 10% Income The uninsured are concentrated disproportionately in low-income families. In 1998, 41.2 percent of the uninsured were in families with annual incomes under $20,000 annually (table 5). More than 45 percent of individuals in families with incomes less than $5,000 were uninsured, compared with 8.6 percent of those in families with annual incomes of $50,000 or more. Generally, as income increases, the percentage of the population without health insurance decreases, the percentage covered by employment-based health insurance increases, and the percentage covered by publicly financed health insurance programs decreases (table 5). Workers with low earnings are more likely to be uninsured than those with high earnings. Thirty-one percent of workers with earnings under $10,000 were uninsured, compared with 5.2 percent of workers with earnings of $50,000 or more (chart 8). Low-income workers are employed generally in industries less likely to offer health insurance, may have a weaker (or temporary) attachment to the work force, and have less disposable income to allocate to the purchase of health insurance. Race and Origin 5% 0% Total Married Without Children Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. While 69 percent of the nonelderly population is white, whites comprised 52 percent of the uninsured 1998 (table 6). Individuals of Hispanic origin were more likely to be uninsured than other groups (37.1 percent). This may be due in part to the fact that 56 percent of the Hispanic population reported income of less than 200 percent of the federal poverty level. However, even at higher income levels, Hispanics were generally more likely to be uninsured than other racial groups and were less likely to be covered by employment-based health insurance. Married With Children Family Type Single individuals and individuals in single-parent families were more likely to be uninsured than married couples either with or without children (chart 9). Married couples and two-parent families may have higher income levels, and both adults may be employed, increasing their chances of receiving employment-based coverage; in addition, if not covered through an employer, they may be more able to afford individually purchased private health insurance. Age Single Without Children Individuals ages were less likely to be uninsured (13.7 percent), and individuals ages were more likely to be uninsured (34.4 percent), than those in all other age groups in 1998 (table 10). The high proportion of young adults without health insurance may occur because they are no longer covered by a family policy and may not have established themselves as permanent members of the work force. Some young adults may also have lost access to Medicaid, which covered them up through age 18 in some states. Many in this group may think that they do not need health insurance because their probability of encountering a high-cost medical event is very low. 6 In addition, young workers may be ineligible for an employment-based plan because of waiting periods imposed prior to eligibility. Children Single With Children Over 15 percent of all children or 11.1 million children were not covered by employment-based health 6 Both Fronstin (1999a) and Cooper and Schone (1997) found that young workers are less likely than older workers to be covered by employment-based health insurance even when a plan is offered to them. 18 January 2000 EBRI Issue Brief

19 Table 10 Persons Ages With Selected Sources of Health Insurance, by Gender and Age, 1998 Employment-Based Coverage Public Individually Gender and Age Total Total Own name Dependent Purchased Total Medicaid Uninsured (millions) Total Ages Ages Ages Ages Ages Ages Males Ages Ages Ages Ages Ages Ages Females Ages Ages Ages Ages Ages Ages (percentage within gender and age categories) Total 100.0% 66.9% 47.3% 19.6% 6.1% 10.7% 6.4% 19.7% Ages Ages Ages Ages Ages Ages Males Ages Ages Ages Ages Ages Ages Females Ages Ages Ages Ages Ages Ages Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. insurance or privately purchased health insurance and were either ineligible or did not receive publicly financed medical assistance in 1998 (table 11). Sixty-six percent of all uninsured children were in families with incomes below 200 percent of the poverty level. Twenty-eight percent of children whose family head did not work were uninsured (chart 10). Most uninsured children were in families whose head was employed year-round, either full-time or part-time, with no unemployment (68.5 percent) (chart 11). However, children in families headed by January 2000 EBRI Issue Brief 19

20 Table 11 Children With Selected Sources of Health Insurance, by Poverty Level and Age, 1998 Public Poverty Level Employment-Based Individually and Age Total Coverage Purchased Total Medicaid Uninsured (millions) Total Under age % of Poverty Under age % 149% of Poverty Under age % 199% of Poverty Under age % or More of Poverty Under age (percentage within age and poverty categories) Total 100.0% 60.2% 7.4% 22.8% 19.8% 15.4% Under age % of Poverty Under age % 149% of Poverty Under age % 199% of Poverty Under age % or More of Poverty Under age Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. Note: Details may not add to totals because individuals may receive coverage from more than one source. 20 January 2000 EBRI Issue Brief

21 Chart 10 Percentage Uninsured Among Children Under Age 18, by Work Status of the Family Head, % 28.2% 20.6% 20.9% 21.0% 20% 15.4% 12.7% 10% 0% Total Full Year, Full Time Full Year, Part Time Full Year, Some Unemployment Part Year Nonworker Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. full-year, full-time workers were much less likely to be uninsured than those whose family head worked parttime or experienced some unemployment (chart 10). Policy Implications Americans without health insurance are a concern for a number of reasons. First, individuals without health insurance are less likely to receive basic health care services than insured individuals. The uninsured report having fewer ambulatory visits than individuals with private or public health insurance, in part because of the greater difficulty in obtaining access to care, and, as a result, are more likely to seek care in a more costly emergency room setting. 7 Hence, the population s overall health status may be lower, and individuals overall productivity may be lower. 8 Second, providers of health care, especially hospitals but also physicians, are often uncompensated for the care that they provide to uninsured individuals and may seek to shift the cost of that care to other private and public payers. 9 However, the movement toward a more com- Chart 11 Children Under Age 18 Without Health Insurance, by Work Status of the Family Head, 1998 Part Year 6.4% Full Year, Part Time 5.6% Full Year, Some Unemployment 8.3% Nonworker 16.9% 11.1 Million Children Under Age 18 Without Health Insurance Full Year, Full Time 62.9% Source: Employee Benefit Research Institute estimates from the March 1999 Current Population Survey. 7 Krauss et al. (1999) found that 55.7 percent of the uninsured had at least one ambulatory medical care visit in 1996, compared with 76.2 percent of individuals with only public insurance and 77.2 percent of individuals with any private insurance. They also found that among persons with at least one visit, the uninsured had an average of 5.1 visits, compared with 8.7 visits by persons with only public insurance and 6.5 visits by those with any private insurance. Another study found that among persons visiting a health care provider, 17 percent of the uninsured received health care in an emergency room, compared with 9 percent of the privately insured (Cunningham and Whitmore (1998). Furthermore, Fronstin (1998) found that 22.1 percent of the uninsured were in a family where someone had difficulty obtaining needed care, compared with 10.2 percent of the insured population, mainly because they could not afford health care. 8 When comparing health status by insurance status, it is important to control for age differences in the insured and uninsured population. It appears that uninsured individuals ages 35 and older are more likely than the insured to report their health status as being fair or poor. Among individuals under age 35, the insured and uninsured are roughly equally likely to report their health as fair or poor, but the insured are more likely to be in excellent health, while the uninsured are more likely to be in good health. Unpublished EBRI estimates of this data are available upon request from the author. 9 Traditionally, cost shifting occurs when a health care provider raises its prices to one set of payers because it lowered them to another set (Morrisey, 1996). January 2000 EBRI Issue Brief 21

22 petitive health care market and the use of alternative forms of third-party reimbursement arrangements, such as capitation, fee schedules, and discounting, have made it more difficult for health care providers to shift these costs to other payers of health care (Morrisey, 1996). As a result, less health care may be available to the uninsured. For example, Cunningham et al. (1999) found that physicians involved with managed care plans and those who practice in areas of high managed care penetration tend to provide less uncompensated care to the uninsured. It appears that the combination of a growing economy and the lowest unemployment rates in over 25 years are both having an impact on the uninsured. Between 1997 and 1998, the uninsured increased from 18.3 percent to 18.4 percent of the nonelderly population, but the 1998 value is not statistically different from the 1997 value. Furthermore, the increase in the percentage of uninsured is generally lower than prior annual increases. The annual increase in the number of uninsured individuals also appears to have slowed down. The number of uninsured increased by 800,000 between 1997 and This compares with a 1.7 million 10 increase in the uninsured between 1996 and 1997, and a 1.1 million increase between 1995 and The number of uninsured Americans can increase for two reasons: the likelihood of being uninsured increases, and/or the population increases. Since the likelihood of being uninsured was not statistically different in 1998 than it was in 1997, the 800,000 increase in the number of uninsured was due mainly to population growth. It is important to recognize that the increase in the uninsured has slowed while population growth is increasing. While the percentage of Americans who are uninsured may have stopped growing, the uninsured will 10 The uninsured actually increased by 1.4 million after controlling for a methodological change. This change affects how persons enrolled in the Indian Health Service program are counted toward the uninsured. More detail is available in the appendix. continue to be an important public policy issue for a number of reasons. First, there are still 44 million Americans without health insurance coverage, and that number is growing. Second, rising health care costs are the primary reason for the decline in employment-based health benefits in the late 1980s and early 1990s. In the mid-1990s, health care costs barely increased. At the same time, the percentage of Americans with employment-based health benefits started to increase, while the growth in the uninsured slowed. Health care cost inflation returned in 1998, and is expected to continue. This will ultimately put pressure on employers to pass along this cost, but as long as unemployment rates remain low, it will be difficult for employers to make changes. Employers, for example, may reduce wage growth over time instead of directly increasing the employee s share of the cost of health benefits. Even when health care costs were increasing at an average of between 4 percent and 9 percent annually, the percentage of workers with employment-based health benefits increased in It is possible that the percentage of workers with employment-based health benefits might have increased more than it did if health care costs had increased less. However, additional research needs to be conducted to sort out all of the factors that affect health insurance coverage. Ultimately, the challenge is how to reduce the number and percentage of uninsured. The health reform bills passed by the 104 th and 105 th Congresses will likely have little impact on the uninsured population. For example, the Health Insurance Portability and Accountability Act of 1996 (HIPAA) provided for a medical savings account (MSA) demonstration project of up to 750,000 individuals. However, according to Internal Revenue Service Announcement 99-95, the number of MSAs set up under HIPAA is far below the cut-off threshold of 750,000 policies set by the legislation. During 1998, 42,477 tax returns were filed with an excludable or deductible MSA contribution, with 10,106 of these individuals reported as being previously uninsured. During the first half of 1999, 11,727 taxpayers 22 January 2000 EBRI Issue Brief

23 had established new MSAs, with 4,468 of that number reported as previously insured and therefore not counted toward the cut-off. The only recent policy change that may have a major impact on reducing the uninsured population is the state Children s Health Insurance Program (CHIP). This program is expected to provide health insurance coverage to 2.7 million children by September However, not all of these children would have been uninsured as of this date, so the full enrollment amount does not necessarily reflect a corresponding reduction in the uninsured. Furthermore, the program may take longer to enroll all of the children it was originally expected to cover as an increasing percentage of children are covered by employment-based health plans (Fronstin, 1999b). Three proposals to reduce the uninsured have been put forth recently. They include the Quality Care for the Uninsured Act (H.R. 2990), recently passed by the House of Representatives, 12 as well as the proposals of Vice President Al Gore and presidential candidate Bill Bradley. The Quality Care for the Uninsured Act contains a number of provisions that may make it easier for individuals to obtain health insurance coverage. The provisions include tax incentives and small-group purchasing arrangements. The tax provisions include full deductibility of health insurance premiums by 2007 for individuals not covered by employment-based health plans, and for workers who pay more than 50 percent of the cost of coverage. The full deduction of the cost of health insurance for the self-employed would be accelerated to In addition, medical savings accounts could be offered on a tax-preferred basis to just about anyone. The small-group purchasing arrangements would The bill passed in the House of Representatives essentially includes all of the provisions in the Senate Patients Bill of Rights Plus Act (S. 1344) that are intended to reduce the uninsured. 13 See Congressional Record, October 6, provide incentives for the creation of association health plans, multiple employer welfare arrangements, HealthMarts, and community health organizations. But the tax and small-group purchasing provisions may do little to reduce the uninsured: The Joint Committee on Taxation has estimated that only 200,000 individuals would gain health insurance coverage in 2002 if 25 percent of the cost of health insurance premiums were tax deductible. 13 However, that number would increase over time as full deductibility was phased in. Studies have shown that it will take subsidies close to 100 percent of the premium in order to have a substantial effect on the low-income uninsured. For example, Thorpe (1999) finds that 75 percent of the uninsured with incomes at 150 percent of the poverty level would purchase health insurance if the value of the subsidy were $2,800. When the tax code is changed to provide incentives for individuals to buy health insurance coverage, policymakers must also consider the fact that many people will benefit from the tax change who would have purchased health insurance in the absence of the new tax incentive. Federal spending would increase disproportionately with changes in insurance coverage rates. For example, the Joint Committee on Taxation estimates that 9.1 million taxpayers would claim the 25 percent deduction for health insurance premiums in 2002, but, as mentioned above, only 200,000 would be individuals who were previously uninsured. Vice President Gore and presidential candidate Bradley take a much different approach than the Quality Care for the Uninsured Act, and a much different approach from each other. Gore, for example, would ensure that all children have access to health insurance. CHIP would be expanded, and parents with children not eligible for CHIP would be able to buy into either CHIP or Medicaid. Gore also includes provisions for uninsured parents of Medicaid and CHIP eligible children, a Medicare buy-in provision for individuals ages 55 and older, small-employer purchasing arrangements, and tax credits. Bradley takes a more aggressive approach, January 2000 EBRI Issue Brief 23

24 especially toward children. Under his proposal, all children would be required to have health insurance. For children living in households with incomes under $32,800 (for a family of four), the government would completely subsidize the cost of health care. Partial premium subsidies would be available for children in households with incomes up to $49,200. All premiums paid by parents for their children would be excluded from taxable income. The Bradley proposal would have the biggest impact on the uninsured by far. Roughly 65 percent of all uninsured children (7.3 million children) would gain free coverage under the plan. Additional children and adults could gain coverage with partial premiums, and some adults would gain coverage through fully subsidized premiums. As with the House of Representatives plan, the Bradley plan would benefit individuals who are already covered by a health plan. In addition, the plan does not come without other costs. Employers would have an incentive to drop dependent coverage, or at a minimum, eliminate the subsidy they provide for dependent coverage. Lower-income workers would then qualify for the full or partial subsidy under the Bradley plan, while higher-income workers would likely pay for the dependent coverage out of their own pocket. Premiums paid by higher-income parents would continue to be excluded from taxable income. This Issue Brief has provided a summary of the characteristics Conclusion of people with and without health insurance, and the sources of the health insurance, from the March 1999 CPS. It finds that many factors affect the likelihood of an individual having health insurance and the source of that coverage. These factors include both demographics and employment characteristics, and often vary by location. For example, work status and income play a dominant role in deter- mining an individual s likelihood of having health insurance. In addition, age, gender, firm size, hours of work, and industry are all important determinants of an individual s likelihood of having coverage; however, these variables are also closely linked to employment status and income. Variations by race, ethnicity, and citizenship are also closely linked to employment status and income. Recent trends in coverage have also been presented. The data indicate that while the percentage of Americans who are uninsured may have stopped growing, there are still 44 million Americans without health insurance coverage, and that number is growing. While an increasing percentage of Americans are being covered by employment-based health plans, this trend may not continue because of the re-emergence of health care cost inflation, especially if the economy goes into a recession. Research illustrates the advantages to consumers of having health insurance. In general, the availability of health insurance allows consumers to avoid unnecessary pain and suffering and improves the quality of life. Ultimately, the challenge that remains is how to reduce the number and percentage of the uninsured. The data presented in this Issue Brief Appendix come from the March Current Population Survey (CPS), conducted annually by the U.S. Bureau of the Census. The March CPS provides answers to questions on labor-force participation, unemployment, work experience, family income, demographics, and health insurance coverage. The CPS is the official source of data on unemployment rates, poverty, and income in the United States. The CPS has undergone a number of changes over the years that affect the comparability of data from the time series. For example, in March 1988, the questionnaire was substantially changed. Among the changes 24 January 2000 EBRI Issue Brief

25 that were made, questions were added that inevitably picked up more people with health insurance coverage and reduced the number of uninsured in the survey. Starting with the March 1995 CPS, the questionnaire was revised again. The Census Bureau utilized a more detailed set of health insurance questions designed to take advantage of computer-assisted survey interviewing collection (CASIC) technology. The order of the questions was changed, and the wording in some of the questions was changed. In addition, the sampling frame was changed, potentially complicating comparability of the estimates prior to March 1995, with the estimates starting in March The new questions appear to have affected responses regarding the total number of respondents covered by individual types of employmentbased health insurance coverage, individually purchased coverage, and CHAMPUS/CHAMPVA. Questions on Medicare and Medicaid were also revised, but because estimates from the CPS do not vary much from year to year even when the survey is unchanged, it is difficult to know how much the estimates were affected by changes to the survey and how much represents true changes. 14 The longer-term trends in coverage are likely representative of the true change because the estimates do not change much from year to year. Swartz (1997) documents these data issues in greater detail. The Census Bureau made another change in the CPS that had a very small effect on the health insurance trend data. Starting with the March 1998 CPS, the U.S. Bureau of the Census modified its definition of the population with Medicaid coverage. Previously, an individual reporting coverage only from the Indian Health Service (IHS) was counted as part of the Medicaid population. Beginning with the data in the March 14 Medicaid and Medicare estimates are under-reported in the CPS, according to comparisons of these data with enrollment and participation data provided by the Health Care Financing Administration (HCFA). See Jennifer A. Campbell, Health Insurance Coverage: 1998, Current Population Reports, P (Washington, DC: U.S. Department of Commerce, Economics and Statistics Administration, September 1999). HCFA is the federal agency primarily responsible for administering the Medicaid and Medicare programs CPS, individuals covered solely by IHS are counted as uninsured. This methodological change affected roughly 300,000 individuals. If this change had not taken place, the Medicaid population would have fallen by 0.9 percentage points between 1996 and 1997, instead of by 1.1 percentage points, and the uninsured would have increased to only 18.1 percent instead of 18.3 percent. In order to compare the March 1995 CPS and later years with earlier years in this paper, data from the March 1988 CPS through March 1994 CPS have been adjusted to reflect two changes that occurred with the change between the March 1994 CPS and March 1995 CPS. First, the data analyzed prior to March 1995 have been re-weighted to reflect the revised sampling framework that occurred in the mid-1990s. Second, the data on employment-based health insurance coverage and individually purchased coverage have been adjusted in response to what appears to be a reallocation of coverage from individually purchased coverage to employmentbased coverage between the March 1994 CPS and the March 1995 CPS. Duration of Coverage Data from the March CPS do not allow researchers to determine the length of time that an individual is insured or uninsured. The Survey of Income and Program Participation (SIPP), another survey conducted by the Census Bureau, allows longitudinal analysis of the uninsured. Copeland (1998) found that 37 percent of the uninsured population were uninsured for one to four months, 22 percent were uninsured for five to eight months, 9 percent were uninsured for nine to 11 months, and 33 percent were uninsured for 12 months or longer. Similarly, Bennefield (1998) found that 29 percent of all uninsured spells lasted 5.3 months or longer. These data would seem to indicate that even though many individuals may lose health insurance during any given month, the majority remain uninsured for a short time, and may even be eligible for coverage under COBRA or various state continuation-of-coverage laws. January 2000 EBRI Issue Brief 25

26 Bennefield, Robert L. Dynamics of References Economic Well- Being: Health Insurance, 1993 to U.S. Bureau of the Census. Current Population Reports, P Washington, DC: U.S. Government Printing Office, Cooper, Philip F., and Barbara Steinberg Schone. More Offers, Fewer Takers for Employment-Based Health Insurance: 1987 And Health Affairs. Vol.16 (November/December 1997): Copeland, Craig. Characteristics of the Nonelderly with Selected Sources of Health Insurance and Lengths of Uninsured Spells. EBRI Issue Brief no. 198 (Employee Benefit Research Institute, June 1998). Cunningham, Peter J., and Heidi Whitmore. How Well Do Communities Perform on Access to Care for the Uninsured? Research Report 1. Washington, DC: Center for Studying Health System Change, September Cunningham, Peter J., Joy M. Grossman, Robert F. St. Peter, and Cara S. Lesser. Managed Care and Physicians Provision of Charity Care. Journal of the American Medical Association 281(12) (March 24/31, 1999). Fronstin, Paul. Access to Health Care and Satisfaction: Differences by Insurance Coverage and Insurance Type. EBRI Notes no. 4 (Employee Benefit Research Institute, April 1998): Employment-Based Health Benefits: Who Is Offered Coverage vs. Who Takes It. EBRI Issue Brief no. 213 (Employee Benefit Research Institute, September 1999a).. Employment-Based Health Insurance for Children: Why Did Coverage Increase in the Mid- 1990s? Health Affairs. Vol. 18 (September/October 1999b). Fronstin, Paul, and Sarah C. Snider. An Examination of the Decline in Employment-Based Health Insurance Between 1988 and Inquiry (Winter 1996/97): Krauss, N.A., S. Machlin, and B.L. Bass. Use of Health Care Services, MEPS Research Findings No. 7. AHCPR Pub. No Rockville, MD: Agency for Health Care Policy and Research, March Morrisey, Michael. Hospital Cost Shifting, a Continuing Debate. EBRI Issue Brief no. 180 (Employee Benefit Research Institute, December 1996). Swartz, Katherine. Changes in the 1995 Current Population Survey and Estimates of Health Insurance Coverage. Inquiry (Spring 1997): Thorpe, Kenneth E. Changing the Tax Treatment of Health Insurance: Impacts on the Insured and Uninsured. In Dallas L. Salisbury, ed., Severing the Link Between Health Insurance and Employment. Washington, DC: Employee Benefit Research Institute, January 2000 EBRI Issue Brief

27 Between the covers of this book are countless facts about health care in America today. But underlying them all is the larger, fundamental fact on which all others rest: Accurate data are essential to fully understand the trends that drive health care costs, health benefit plan design, and health policy in America accurate, comprehensive, unbiased, and timely data. The EBRI Health Benefits Databook provides an indispensable data resource for this kind of information. It builds on the Employee Benefit Research Institute s well-established record of objective research and education on economic security issues primarily health and retirement, but other benefit issues as well. EBRI was established in 1978, and our research and analysis now encompasses 21 years of change and evolution in the U.S. employee benefits sector. For many years, one of our most popular and heavily used resources has been the EBRI Databook on Employee Benefits, currently in its fourth edition, which combines essential data on both retirement and health benefits in the private and public sectors. But one of the challenges of pulling together such an exhaustive and comprehensive research tool is the well-recognized fact that government data on health and retirement move according to very different schedules: health data are updated much more often than are pension data. In recognition of that fact and to avoid unnecessary delays waiting for new pension data we are releasing this first edition of the EBRI Health Benefits Databook, focusing just on health issues. This will ensure that our health data are as timely and useful as they are comprehensive and unbiased. This first edition of the EBRI Health Benefits Databook has more than a hundred tables and charts, many with valuable time series data to help you see trends in health benefits and costs. It has been carefully organized and tightly focused on the range of health issues to make it as user-friendly and relevant as possible. Appendices have been updated to include the most recent health-related legislation and data sources used in the book. A comprehensive index also makes it easy to quickly find answers to your questions. Copies of the EBRI Health Databook are available by ordering online at or by calling (202) EBRI Health Databook, 1st ed., $59.95 EBRI Members receive a 55% discount on all EBRI publications. To receive the discount, you must indicate that you are a member when placing the order. January 2000 EBRI Issue Brief 27