Methodology: Latino Children’s Coverage Health Fact Sheets

Fact Sheets are additional analyses building upon the following brief report:

Whitener, K., Lopez, S., Roygardner, L. & Snider, M. (2020). Decade of Success for Latino Children’s Health Now in Jeopardy (March 10, 2020).

Unless otherwise noted, the data cited is based on Georgetown University Center for Children and Families analysis of the U.S. Census 2016-2018 American Community Survey (ACS) and Puerto Rico Community Survey (PRCS) using Integrated Public Use Microdata (IPUMS) via usa.ipums.org/usa. Significance testing for IPUMS data was computed using Stata/SE 16.0 statistical software with survey design and replicate weights, unless otherwise noted.


Number of Latino Children by State              

  • Children were defined as those individuals under age 19 (0 to 18 years). The Census Bureau recognizes and reports race and Hispanic origin/Latino (i.e., ethnicity) as separate and distinct concepts and variables. We report “Hispanic or Latino,” as “Latino.” “Latino” refers to a person’s ethnicity, Latino and non-Latino individuals may be of any race. “Latinx” may also be used to respect various gender identities and expressions.
  • The reported number of Latino children are estimates and should be interpreted as approximations of the population size rather than precise population counts.

Trends Over Time: Annual Estimates of Number of Uninsured Latino Children and Uninsured Rates for Latino Children (National, by State, and Puerto Rico)

  • * Change is significant at the 90% confidence level and relative to the prior year indicated.
  • ^ Change is significant at the 90% confidence level and relative from 2016 to 2018.
  • The reported number of uninsured Latino children are estimates and should be interpreted as approximations of the population size rather than precise population counts.

Uninsured Rates for Latino and Non-Latino Children

The disparity scores (i.e. the number of times one group is more likely to be uninsured than another) was computed and then interpreted into language so that a disparity of 1.49 would be translated into “almost 1.5 times more likely to be uninsured.” 

Sources of Children’s Coverage

  • In the ACS, data on sources of health insurance coverage are point-in-time estimates that convey whether a person has coverage at the time of the survey. Individuals can report more than one source of coverage, so such totals may add to more than 100 percent. Furthermore, due to rounding percentages may not total to 100 percent. Additionally, the estimates are not adjusted to address the Medicaid “undercount” when comparing federal and state administrative data, which, for example, may be accentuated by the absence of state-specific health insurance program names in the ACS.
  • Coverage groups: Children covered by Medicaid/CHIP exclusively or Medicaid/CHIP in combination with another coverage type were reported here as “Medicaid/CHIP.” Children covered by Medicare, TRICARE/military, Veteran coverage, or two or more types of health coverage—with the exception of Medicaid in combination with another type— were reported as being covered through “Other source of health coverage.” The health coverage categories discussed herein include: uninsured, employer sponsored insurance (ESI), purchased directly from an insurance company/direct purchase, Medicaid/CHIP alone or in combination, and other source of health coverage. People who indicate Indian Health Service (IHS) as their only source of health coverage do not have comprehensive coverage according to ACS survey definitions and are therefore considered to be uninsured.

Covering Parents Helps Children

Uninsured Rates for Latino and Non-Latino Children by State In Comparison to All Children National Average

Uninsured Rates for Latino Children by Age

  • Age groups here are divided roughly by pre-kindergarten entry (ages 0-5), elementary school aged (ages 6-12), and teens (ages 13-18).  

Uninsured Rates for Latino Children by Income

  • Data on poverty levels include only those individuals for whom the poverty status can be determined for the past year. Therefore, this population is slightly smaller than the total non-institutionalized population of the U.S. (the universe used to calculate all other data in the brief). The Census Bureau determines an individual’s poverty status by comparing that person’s income in the past 12 months to poverty thresholds that account for family size and composition, as well as various types of income.
  • The Census definition of income may vary considerably from how state Medicaid and CHIP programs measure income for purposes of determining eligibility due to differences in how income is counted and household size is determined and other factors.
  • Income is reported here according to 2018 poverty thresholds for a family of three (1 adult and 2 children): Retrieved from https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-poverty-thresholds.html.
  • Disproportionate Job Losses among Latinos during COVID-19 pandemic: Latino Jobs Report, May 2020, “Latino Unemployment Rate Spiked to 18.9%.”

Uninsured Rates for Latino Children by Child’s Citizenship Status

  • Citizenship status was derived from a variable named “citizen” in IPUMS which reports on citizenship status of respondents, distinguishing only between a) citizens born abroad to American parents, b) naturalized citizens and c) non-citizens.
  • Note: The non-citizen variable does not differentiate between documented or undocumented immigrant statuses.

Uninsured Rates for Latino Children by Parent’s Citizenship Status

  • Parent’s citizenship status was derived in IPUMS via attached characteristics by linking parent variables (i.e. mom and pop) with a variable named “citizen” in IPUMS which reports on citizenship status of respondents, distinguishing only between a) citizens born abroad to American parents, b) naturalized citizens and c) non-citizens. The attached characteristics produces parent citizenship variables “citizen_mom, citizen_mom2, citizen_pop, and citizen_pop 2” which were analyzed here in connection with Latino children to report of child’s uninsured rate by the parent’s citizenship status. Children were coded as having citizen parents when their sole parent or both parents were citizens. Children with one or two non-citizen parents were coded as children with non-citizen parents.
  • The non-citizen variable does not differentiate between documented or undocumented immigrant statuses.

Uninsured Rates for Latino Children by Child Race and Ethnicity

  • When we report “all” children, we mean all U.S. children regardless of race and ethnicity, including Latino children. We report “Hispanic or Latino,” as “Latino.” “Latino” refers to a person’s ethnicity, Latino and non-Latino individuals may be of any race. Latino ethnicity includes all children who report Latino ethnicity of any race. Analysis here shows the corresponding race category endorsed for all children who identified their ethnicity as Latino. For more detail on how the ACS defines racial and ethnic groups, see “American Community Survey and Puerto Rico Community Survey 2018 Subject Definitions.”
  • Arizona: The uninsured rate for Latino children identifying as Asian/Pacific Islander is suppressed due to poor reliability. In 2018, according the ACS the population size of Latino Asian/Pacific Islander children in Arizona is estimated to be about 2,228, or 0.3% of the Latino child population.

Uninsured Rates for Latino Children by County

  • Georgetown University Center for Children and Families analysis of the U.S. Census 2018 American Community Survey (ACS) and Puerto Rico Community Survey (PRCS) data using 5-year estimates from Data.Census.Gov.
  • Data is sorted and presented here in ascending order by number of uninsured Latino children by county.
  • Health indicators can be presented at the most aggregated level (i.e., national means, population estimates or rates) and may also be computed for smaller subgroups, like in this report looking at uninsured rates of subgroups by looking at both ethnic group (in this case Latino children ages 0-18) and also geographic regions (within each of the 50 states in the U.S. plus the District of Columbia by smaller geographies including Counties).
  • Data suppression rules were applied to portions of this analysis for the purpose of ensuring reliability. This is due to ACS data quality issues related to some estimates that have a large standard error resulting in states with number and percent estimates with low reliability. The U.S. Census encourages the computation of standard errors and coefficients of variation to calculate the reliability of estimates in the ACS. Therefore, to avoid potentially unreliable estimates which can likely lead to misleading findings and inaccurate conclusions, we have applied the following data suppression rule in the Congressional District and County-level analyses. In cases where estimates and/or percent estimates yield large coefficients of variation (CV) (counties with CVs greater than 25 percent) have been suppressed. See methodology from full 2020 report here for further details.

Uninsured Rates for Latino Children by Congressional District

  • Georgetown University Center for Children and Families analysis of the U.S. Census 2018 American Community Survey (ACS) and Puerto Rico Community Survey (PRCS) data using 1-year estimates from Data.Census.Gov.
  • Data is sorted and presented here in ascending order by number of uninsured Latino children by congressional district.
  • Health indicators can be presented at the most aggregated level (i.e., national means, population estimates or rates) and may also be computed for smaller subgroups, like in this report looking at uninsured rates of subgroups by looking at both ethnic group (in this case Latino children ages 0-18) and also geographic regions (within each of the 50 states in the U.S. plus the District of Columbia by smaller geographies including Congressional Districts).
  • Data suppression rules were applied to portions of this analysis for the purpose of ensuring reliability. This is due to ACS data quality issues related to some estimates that have a large standard error resulting in states with number and percent estimates with low reliability. The U.S. Census encourages the computation of standard errors and coefficients of variation to calculate the reliability of estimates in the ACS. Therefore, to avoid potentially unreliable estimates which can likely lead to misleading findings and inaccurate conclusions, we have applied the following data suppression rule in the Congressional District and County-level analyses. In cases where estimates and/or percent estimates yield large coefficients of variation (CV) (congressional districts with CVs greater than 25 percent) have been suppressed. See methodology from full 2020 report here for further details.

Latest