About the Health Insurance Data

Information on this page:

What do these data tell us?

  • The percentage of people without health insurance in Minnesota by year, age group, sex, and race/ethnicity (on the "View Charts" page).
  • County-level data on people without health insurance (choose the "Explore Data" button to view the data query).
  • If a segment of a population is at higher risk for being without health insurance.
  • The geographic distribution of people without health insurance.
  • If a measure is increasing, decreasing, or not changing.

How can we use these data?

  • To inform the public about the number and percent of people without health insurance.
  • To explore trends in the percentage of people without health insurance.
  • For program planning and evaluation by state and local partners.
  • To inform actions to address disparities by race/ethnicity, age, sex, and geography.
  • To evaluate the relationships between heath care access, poverty/income, and health outcomes like asthma, cancer, childhood lead poisoning, and more.
  • For examples of how these data are used by the MDH Health Economics Program, see its most recent fact sheet: Health Insurance Coverage in Minnesota: Results from the 2015 Minnesota Health Access Survey (PDF).

What can these data not tell us?

The health insurance data cannot tell us why a person does not have health insurance.

What is the source of the data?

Health insurance data for the Facts & Figures webpage are extracted from the Minnesota Department of Health's Minnesota Health Access Survey, a large-scale telephone survey conducted jointly by the Minnesota Department of Health and the University of Minnesota. Cell phone users were added to the survey in 2009; in 2015, nearly three-quarters (72%) of interviews were completed on cell phones. Survey results are statistically weighted to be representative of the state's population based on age, race/ethnicity, education, region, home-ownership, nativity, household size and whether the household had a landline, cell phone, or both.

County-level health insurance data are extracted from U.S. Census Bureau's Small Area Health Insurance Estimates (SAHIE) program and are accessible through the Population Characteristics: Data Query. The SAHIE program models health insurance coverage estimates using data from several sources including the American Community Survey (ACS).

Because of differences in survey methodology, estimates from the Minnesota Health Access Survey should not be compared to other sources of data on health insurance, such as SAHIE county-level estimates. State- and regional-level estimates in the charts are not directly comparable to the data query's county-level estimates of insurance.

How do we determine who has health insurance?

  • Health insurance estimates described here are available for nonelderly people (under the age of 65). Most adults ages 65 and older are covered by Medicare or Supplemental Security Income (SSI), with less than 2% of this age group uninsured nationwide and less than 1% uninsured in MN in recent years.
  • In the Minnesota Health Access Survey, health insurance status is assessed by asking survey respondents a series of questions covering all types of health insurance in the state and country. State-specific terms, such as MinnesotaCare and Medical Assistance, are used. Health insurance is measured at the time of the survey, which is referred to as the "point in time" uninsurance rate.
  • For SAHIE estimates, the American Community Survey insurance question asks "Is this person currently covered by specifically stated health insurance or health coverage plans?" Respondents are considered insured if they are covered by any type of health insurance coverage, and they are considered uninsured if they are not covered by any type of health insurance. People with no coverage other than access to Indian Health Service are also considered uninsured. 

How can I tell if differences between groups are statistically significant?

  • Unless otherwise noted, differences between groups described on the "health insurance: facts & figures" webpage are statistically different. A difference, increase, or decrease is indicated as "statistically significant" when the t test was significant at the ±=0.05 level. 
  • Assessing the confidence interval for the percent or rate is one approach to determine whether there are differences over time. If they do not 'overlap' then they statistically differ. Although it is not a 'true' statistical test, it is a commonly accepted way to compare percentages or rates and tends to be more conservative than statistical testing.
  • A confidence interval for a rate is a measure of reliability. In this analysis, either 90% or 95% confidence intervals were calculated, depending on the data source. A 90% confidence interval, for example, is the interval within which the true value of the rate would be expected to fall 90 times out of 100.
  • Confidence intervals should not be used to compare states or counties using data from the U.S. Census Bureau's Small Area Estimates program. It is not appropriate to compare confidence intervals to compare geographies because the model-based estimates result in correlated percentages.

What are the limitations of the data?

  • Two data sources are available for health insurance data in Minnesota. Because of differences in survey methodology, estimates from the Minnesota Health Access Survey should not be directly compared to other sources of data on health insurance, such as from the Small Area Health Insurance Estimates (SAHIE) program, which are available at the county-level and accessible in the Population Characteristics: Data Query.
  • Health insurance estimates from the Minnesota Health Access Survey are based on the state's non-institutional population (excluding people in nursing homes and prisons, for example). The survey sample is designed to provide estimates for the 13 Minnesota Economic Development Regions, specific racial and ethnic groups, and age categories. Data are statistically weighted to be representative of the Minnesota population for the current year. Data are available for the state as a whole and by region, but not at the county level.
  • Health insurance data from SAHIE are model-based estimates and not direct counts from surveys. SAHIE data combines survey data (like the American Community Survey) with population estimates and administrative data.
  • In 2013, SAHIE methodology changed and began using Medicaid enrollment data from the same year rather than 2 years prior. The U.S. Census Bureau made this change to better reflect Medicaid enrolllment in its insurance estimates, as the Affordable Care Act included some Medicaid expansions. Accordingly, county-level estimates for 2013 have been updated on this portal and estimates for 2014 and onward will use the new methodology. Learn more under the heading "Why were the SAHIE 2013 data updated?" at Frequently Asked Questions about SAHIE.
  • SAHIE's county-level estimates from 2005-2007 should not be compared to 2008 and onward due to a substantial methodology change implemented between 2007 and 2008. Starting for SAHIE 2008 estimates, the American Community Survey (ACS) replaced the Current Population Survey Annual Social and Economic Supplement (CPS ASEC) as a primary data source. 
  • Because these data are for nonelderly people age 0-64, the numbers might differ slightly from data from the MDH Health Economics Program, which publishes data for people of all ages. For example, the percentage of people covered by public insurance (like Medicare) is very different for the population of Minnesota as a whole compared to the nonelderly population (0-64 years).

Where can I find more technical information about the data?

To learn more about Minnesota Health Access Survey, see the MDH Health Economics Program or contact Minnesota Public Health Data Access.

To learn more about SAHIE, go to Small Area Health Insurance Estimates (SAHIE) or contact Minnesota Public Health Data Access.