Small Area Health Insurance Estimates

Frequently Asked Questions about SAHIE

1. What is SAHIE?
2. How are the SAHIE program state and county estimates constructed?
3. What is the definition of uninsured?
4. What is the definition of income-to-poverty rates (IPRs) less than or equal to 138, 200, 250 and 400%?
5. Why don't you have data for ages 65 and over?
6. What does MOE stand for? What do they mean?
7. Why were two new IPRs added to the 2008 and 2009 data?
8. Why did SAHIE switch from using data from the Current Population Survey (CPS) to the American Community Survey (ACS)?
9. Why are there two different uninsured percents shown in the table? What's the difference between the two?
10. There are two different percents in your tables. How are these two percents calculated?
11. If I have further questions, whom should I contact?
12. How can one make statistical comparisons using SAHIE data?
13. Why were the SAHIE 2009 county-level data revised?

1. What is SAHIE?
   

   SAHIE is the Small Area Health Insurance Estimates program of the U.S. Census Bureau. SAHIE produces and disseminates model-based estimates of health insurance coverage for counties and states.

   The SAHIE program recently released 2010 COUNTY estimates of people with and without health insurance coverage by:

   • ages 0-64, 18-64, 40-64, and 50-64;
   • female, male, and both sexes;
   • all incomes and less than or equal to 138, 200, 250 percent or 400 percent of the poverty threshold;
   • a special estimate for the 0-18 age group for 'both sexes,' and 'all incomes,' less than or equal to 138, 200, 250, or 400 percent of the poverty threshold; and
   • measures of uncertainty of the estimates.

   Also recently released are the 2010 STATE estimates of people with and without health insurance coverage by:

   • ages 0-64, 18-64, 40-64 and 50-64;
   • female, male, and both sexes;
   • all Races, White alone (not Hispanic), Black alone (not Hispanic), and Hispanic (any race);
   • all incomes, less than or equal to 138, 200, 250 percent and 400 percent of the poverty threshold;
   • a special estimate for the 0-18 age group for ‘both sexes,’ and ‘all incomes,’ less than or equal to 138, 200, 250, 400 percent of the poverty threshold; and
   • measures of uncertainty of the estimates.

   For more information, please see our About SAHIE page.

   Return to top.

2. How are the SAHIE program state and county estimates constructed?
   

   The SAHIE program models health insurance coverage by combining survey data with population estimates and administrative records. SAHIE estimates are based on data from the following sources:

   • American Community Survey (ACS);
   • demographic population estimates;
   • aggregated federal tax returns;
   • participation records for the Supplemental Nutrition Assistance Program (SNAP), formerly known as the Food Stamp program;
   • County Business Patterns;
   • Medicaid and Children's Health Insurance Program (CHIP) participation records; and
   • Census 2010.

   More information is available at Methodology.

   For further information on these data sources, see information about data inputs.

   Return to top.

3. What is the definition of uninsured?
   

   The American Community Survey (ACS) insurance question asks "Is this person currently covered by specifically stated health insurance or health coverage plans?" In the SAHIE models, this question replaces the Current Population Survey (CPS) question that was used in prior releases which asked "At any time during the previous year" was the person covered? Note that coverage solely by the Indian Health Service (IHS) does not count as health insurance; i.e., people who were only covered by IHS in the previous year are counted as uninsured.

   Return to top.

4. What is the definition of income-to-poverty rates (IPRs) less than or equal to 138, 200, 250 and 400%?
   

   These categories are defined by the ratio of family income to the federal poverty threshold. A lower ratio indicates lower income. Less than or equal to 138% of poverty indicates people in families with total money income less than or equal to 138% of the federal poverty threshold applicable to that family. The same reasoning holds for the additional IPRs listed.

   Return to top.

5. Why don't you have data for ages 65 and over?
   

   Most people ages 65 and over are covered by Medicare or Supplemental Security Income (SSI). According to recent CPS ASEC data, less than 2 percent of the 65+ population were uninsured nation-wide.

   Return to top.

6. What does MOE stand for? What do they mean?
   

   MOE stands for "margin of error" or the difference between an estimate and its upper or lower confidence bounds. Confidence bounds can be created by adding the margin of error to the estimate (for an upper bound) and subtracting the margin of error from the estimate (for a lower bound). All published margins of error for the SAHIE program are based at the 90 percent significance level.

   Return to top.

7. Why were two new IPRs added to the 2008 and 2009 data?
   

   These newly added IPRs are reflective of the recent Health Care Reform initiative. The Patient Protection and Affordable Care Act helps families gain access to health care by allowing Medicaid to cover families with incomes less than or equal to 138 percent of the poverty line. Also, families with incomes above the level needed to qualify for Medicaid, but fall less than or equal to 400 percent of the poverty line can receive tax credits that will help them pay for health coverage in the new health insurance exchanges.

   Return to top.

8. Why did SAHIE switch from using data from the Current Population Survey (CPS) to the American Community Survey (ACS)?
   

   The full-production ACS has a sample size of roughly 3 million addresses, and the sample is selected from all counties and county-equivalents in the United States, and from all municipios in Puerto Rico (PR). Single-year direct survey estimates are published for counties and other places with a population size of 65,000 or larger, and three-year estimates are published for counties and other places with population sizes of 20,000 or larger.

   Return to top.

9. Why are there two different uninsured percents shown in the table? What's the difference between the two?
   

   The percent uninsured in the demographic group for all income levels is comparing the number of the low-income uninsured population relative to the population in that demographic group. The percent uninsured in the demographic group for the low-income category is comparing the low-income uninsured estimate relative to the low-income population in that demographic group.

   Return to top.

10. There are two different percents in your tables. How are these two percents calculated?
    

    The percent uninsured in demographic group for all income levels is:

    Number uninsured in demographic and income group

    ---

    Number in demographic group

    Example—
    

    Number of uninsured under age of 19 less than or equal to 138% of poverty

    ---

    Number of people under age 19

    The percent of the uninsured in demographic group for low-income is:

    Number uninsured in demographic and income group

    ---

    Number in demographic and income group

    Example—
    

    Number of uninsured under age 19 less than or equal to 138% of poverty

    ---

    Number of people under age 19 less than or equal to 138% of poverty

    Return to top.

11. If I have further questions, whom should I contact?
    

    You can ask questions or send feedback by e-mailing the Small Area Estimates Branch or calling the Demographic Call Center Staff at 301-763-2422 or 1-866-758-1060 (toll free).

    Return to top.

12. How can one make statistical comparisons using SAHIE data?
    

    If one would like to make comparisons between estimates of uninsured rates (or other SAHIE concepts) between different areas or demographic groups in a given year, or between estimates for the same area and demographic group for any two years after 2007, a reasonable approximation is available.

    First, construct an approximate upper bound for the margin of error (MOE) for the difference between the two estimates chosen for comparison. This approximate upper bound MOE is constructed as the square root of the sum-of-squares of the individual MOEs for each estimate (see example below).

    If this constructed MOE for the difference is smaller than the absolute value of the calculated difference between the two point estimates, then one can conclude that the two estimates are significantly different for at least a 90% statistical significance level. If the MOE for the difference is larger than the calculated difference between the two point estimates, then the comparison is inconclusive as to whether or not there is a statistically significant difference.

    For example, say the SAHIE estimate for percent uninsured, within the population ages 18-64 at all income levels, were 15.1% for county A, with a MOE of 1.4, and 18.2% in county B, with a MOE of 1.5%. Then the calculation is:

    90% MOE for the difference between the two estimates = square-root(1.4 x 1.4 + 1.5 x 1.5) = 2.0% Absolute value of the difference between the two estimates = 18.2 - 15.1 = 3.1%

    Since the MOE of the difference, 2.0%, is less than the difference between the two estimates, 3.1%, one can conclude that the two estimates are significantly different, with at least a 90% significance level. In contrast, if county A's uninsured rate were instead 16.3%, and all other values as above, then the difference between the two estimates, 1.9%, is less than the MOE of the difference, 2.0%, and the test would be inconclusive.

    Note this method produces a reasonable upper bound approximation to the 90% margin of error for the difference, and not an exact MOE. The reason for this caveat is that modeled estimates like SAHIE have correlations between the estimates, and the exact MOE would account for these correlations. For simpler modeling programs, such as Small Area Income and Poverty Estimates, we have examined the factors leading to these correlations and concluded that nearly all of them are positive, which means the exact MOE would be generally smaller than the approximation described above and the recommended test would have a confidence level better than 90%. We expect that there would be similar results for SAHIE and we are conducting research to verify this assumption, and in general, such research could provide more accurate MOEs.

    For year-to-year comparisons, we recommend only comparing any two years after 2007 due to a substantial methodological change implemented between the 2007 and 2008 estimates, as described in our methodology documentation, resulting in a difference in how the two concepts measure health insurance coverage.

    
    

    Return to top.

13. Why were the SAHIE 2009 county-level data revised?
    

    The SAHIE 2009 county-level estimates have been revised and re-released on November 1, 2012 due to a processing error present in the original October 2011 release.

    The poverty universe used during the original 2009 processing was incorrect. The appropriate parts of the group quarters population were not excluded from the universe. This revision most affected county estimates where a large portion of the county population resides in group quarters, such as prisons, barracks, and dorms.

    Return to top.