Research Note: The Long-Term Uninsured

09/01/2004

Research Note: The Long-Term Uninsured

This report is available on the Internet at:http://aspe.hhs.gov/health/long-term-uninsured04/

Report in PDF format for printing

Contents

Researchers at the U.S. Department of Health and Human Services (HHS) performed an analysis of the Medical Expenditure Panel Survey (MEPS) to investigate the characteristics of those who are uninsured for 13-24 months over a two-year period.(1)  The MEPS follows the same people over a two-year period (1999-2000), permitting analysis of people uninsured for over a year.  For this analysis people uninsured for over one year are referred to as the long-term uninsured. According to the MEPS, there were 38.6 million people who were uninsured for over a year during the 1999 to 2000 period.(2)  These 38.6 million individuals represent half of the 77.3 million individuals who were uninsured for at least one month during that two year period.  What follows are some key findings from our analysis.

The Long-Term Uninsured by Age

Those uninsured for a year or longer are more likely to be under age 45. Among those under 45, there is fairly even distribution of long-term uninsured individuals.  However, certain age groups are disproportionately affected:

  • Although 18 to 24 year olds are only 10% of the U.S. population, they are 21% of the long-term uninsured;
  • By contrast, children under age 18 are 27% of the U.S. population, but only 22% of the long-term uninsured.

Figure 1: Age Distribution of the Long-Term Uninsured Compared to the Overall U.S. Population, 1999  2000

Figure 1: Age Distribution of the Long-Term Uninsured Compared to the Overall U.S. Population, 1999-2000.

Older individuals are much less likely to be among the long-term uninsured because Medicare covers close to 100 percent of Americans over age 65.  In addition, only 10% of the long-term uninsured are between 45 and 55, and only 8% of the long-term uninsured are between 55 and 64.  However, although the age distribution of the long-term uninsured among the age cohorts appears quite even, the age cohorts are different sizes.  Thus, those 18 to 24, who represent a 6 year age range, are disproportionately represented when compared to children, who represent an 18 year age range, and those 45 to 64, who represent a 19 year age range.

The Long-Term Uninsured by Poverty Status

The long-term uninsured are more likely to be middle income and poor, and much less likely to be higher income.

  • Those below 200% of poverty are 24% of the population, but 46% of the long-term uninsured;
  • Those between 200% and 399% of poverty are 31% of the U.S. population and 36% of the long-term uninsured;
  • By contrast, those 400% of poverty and above are 44% of the U.S. population, but only 18% of the long-term uninsured.(3)

Figure 2: Poverty Distribution of the Long-Term Uninsured Compared to the Overall U.S. Population, 1999  2000

Figure 2: Poverty Distribution of the Long-Term Uninsured.

Although it may appear surprising that those below 100% of poverty represent only 16% of the long-term uninsured, this result highlights the impact of Medicaid on the coverage of lower-income children, parents, and the disabled.(4)  In the absence of Medicaid, those below poverty would certainly constitute a higher proportion of the long-term uninsured.

The Long-Term Uninsured by Race/Ethnicity

Those uninsured for 13 to 24 months are most likely to be white, but a disproportionate percentage of the long-term uninsured are Hispanic.

  • Non-Hispanic whites are 71% of the U.S. population, but only 53% of the long-term uninsured;
  • Hispanics are 12% of the U.S. population, but 27% of the long-term uninsured.(5)

Figure 3: Race/Ethnicity Distribution of the Long-Term Uninsured Compared to the Overall U.S. Population, 1999  2000

Figure 3: Race/Ethnicity Distribution of the Long-Term Uninsured Compared to the Overall U.S. Population, 1999-2000.

Unlike whites and Hispanics, African-Americans and those of mixed or other races are represented among the long-term uninsured proportionately to their representation in the U.S. population.

The Long-Term Uninsured by Marital Status

When looking at the long-term uninsured by marital status, those who were never married comprise the largest group.  Although the next largest group is married individuals, they represent a far smaller percentage of the long-term uninsured than their percentage of the overall population.  The data below are limited to those over age 15, since MEPS only asks individuals over this age about their martial status.

  • Although married individuals are 53% of the U.S. population over age 15, they are 35% of the long-term uninsured over that age;
  • By contrast, the never married are only 28% of the over-15 U.S. population, but 45% of the long-term uninsured.

Figure 4: Marital Status Distribution of the Long-Term Uninsured Over Age 15 Compared to the Population Over Age 15, 1999  2000

Figure 4: Marital Status Distribution of the Long-Term Uninsured Over Age 15 Compared to the Population Over Age 15, 1999-2000.

Other marital status groups, such as those separated and widowed, comprise small portions of the long-term uninsured.  For this analysis parental status was also examined and while non-parents comprise 60% of the long-term uninsured, there were no other significant findings from a policy or statistical perspective.

The Long-Term Uninsured by Employment Status

The long-term uninsured are far more likely to have worked than to not have worked.  The data below are limited to the working-age population age 16 to 64.(6)

  • While those individuals that worked (either full-time or part-time) represent 83% of the U.S. population 16 to 64, they represent 80% of the long-term uninsured in that age range;
  • Those that did not work represent 17% of the population age 16 to 64, but 20% of the long-term uninsured in that age range.

Figure 5: Work Status Distribution of the Long-Term Uninsured Age 16-64 Compared to the Population Age 16-64, 1999  2000

Figure 5: Work Status Distribution of the Long-Term Uninsured Age 16-64 Compared to the Population Age 16-64, 1999-2000.

As can be seen in the chart above, those who work and those who did not work are roughly the same percentage of the long-term uninsured population as their percentage of the general working population age 16 to 64.

The Long-Term Uninsured by Insurance Offer Status

Not surprisingly, the long-term uninsured are far less likely to receive an offer of insurance through their employer than the general working population. The data below are limited to the working age population age 16 to 64.

  • Those not offered insurance through an employer are only 38% percent of the working population age 16 to 64, but 79% of the long-term uninsured working population in that age range;
  • By contrast, those offered insurance through an employer are 62% of the working population age 16 to 64, but only 21% of the long-term uninsured in that age range.

Figure 6: Insurance Offer Distribution of the Long-Term Uninsured Age 16 to 64 Compared to the Population Age 16 to 64, 1999  2000

Figure 6: Insurance Offer Distribution of the Long-Term Uninsured Age 16 to 64 Compared to the Population Age 16 to 64, 1999-2000.

As would be expected, those not offered insurance are disproportionately represented among the long term uninsured.(7)  It should be noted that the percentages in the above chart are based on whether respondents were personally offered insurance.  Thus, the results would likely be skewed even more dramatically if the data counted those who received no offer themselves, but who could have been covered under a spouses employer insurance as having been offered coverage.

Conclusion

The demographic profile of the long-term uninsured is similar to the profile of the total uninsured population as reported by other surveys such as the Census Bureaus Current Population Survey.(8)  The long-term uninsured are more likely to be younger, lower income, disproportionately Hispanic, not married, not parents, and working people not offered insurance through an employer than the general population.  However, as with the total uninsured population, the long-term uninsured are a diverse population with representation from most demographic groups, except seniors age 65 and over.

Endnotes

1.  This memo has been prepared by Rob Stewart of the Office of the Assistant Secretary for Planning and Evaluation (ASPE) and Jeffrey Rhoades of the Agency for Healthcare Research and Quality (AHRQ).

2.  Many readers of this memo may be familiar with the uninsured estimate of 43.6 million individuals.  That figure comes from the Census Bureaus Current Population Survey (CPS) for 2003.

3.  Poverty status was based on average annual income over the 1999 and 2000 panel years.

4.  Note that not all individuals below poverty are eligible for Medicaid.  Medicaid eligibility is based on a combination of income and population category. The population groups that qualify for Medicaid are children, parents of dependent children, pregnant women, the disabled, and the elderly.  The income levels at which these groups qualify differs from group to group, with coverage of children and pregnant women being available at higher income levels, followed by the disabled and elderly, then parents of dependent children last (though this varies by State). Childless adults who are not disabled or elderly rarely qualify for Medicaid, even at the very lowest income levels.

5.  The 1999-2000 MEPS panel bases its weighting on the 1990 Decennial Census, while more recent Current Population Survey (CPS) data bases its weighting on the 2000 Decennial Census.  Using the 1990 Decennial Census weights Hispanics to a lesser degree than would be the case using the 2000 Census.  The 1999-2000 MEPS reports about 12% of the U.S. population as Hispanic, while the 2003 CPS shows about 14% of the U.S. population as Hispanic.  However, the over-representation of Hispanics among the uninsured is consistent in both surveys.  The 2000-2001 MEPS will be available in the Fall of 2004 and will be weighted based on the 2000 Decennial Census.

6.  Employment status was based on employment at any time during 1999.  That is, if a respondent was working during any of the interview rounds in 1999, that person was coded as working for purposes of this analysis.  This means that there may be some mismatch between reported employment status and reported uninsured status, with some working individuals having been unemployed during the time they were uninsured.

7.  Insurance offer status, like employment status, was based on status during 1999.  If a respondent was offered insurance during any interview round in 1999, that person was coded as having an insurance offer for purposes of this analysis.  The potential for some mismatching between offer status and insured status is similar to employment status.

8.  Robert Mills and Shailesh Bhandari, Health Insurance Coverage in the United State:  2002, U.S. Census Bureau.

Tags