Minding the Outcome Gaps for American Indian Children in the Race for Results Report

An index of child development shows that while American Indian outcomes are below average, they vary widely across the U.S., suggesting opportunities to learn from more successful areas.

Richard M. Todd | Vice President and Advisor to the CICD

Published January 17, 2018

The Annie E. Casey Foundation’s Race for Results report aims to “bring a fresh perspective and new data analysis to the national conversation about how we make sure that all children realize their potential.”1 Introduced in 2014 and updated every three years, the report boils 12 separate indicators down to a single one-thousand-point index that “measures how children are progressing on key milestones by race and ethnicity at the national and state levels.”2 The update for 2017, released in October, aligns with other studies in finding that American Indian children “face some of the biggest obstacles on the pathway to opportunity.”3 However, the report also shows a wide range of results across states, including some states with relatively strong results for American Indian children. To stimulate conversation about how we make sure that all American Indian children realize their potential, I use data from the 2017 Race for Results report to highlight between-state and within-state outcome gaps pertinent to the future of American Indian children and families.

For example, looking at the American Indian index values across state lines reveals some potentially informative gaps. Figure 1, taken from the 2017 Race for Results report, shows that the index values for American Indian children vary widely over 26 states.4 In fact, the report notes (p. 36) that, “[t]he range of scores for American Indian children — 220 in South Dakota to 615 in Alabama — is the widest [among racial/ethnic groups] in the index.” Although almost 100 points below the national average of 713 for White children, Alabama’s American Indian score is higher than the value for White children in 5 states, including Alabama itself. By contrast, “[t]he score for American Indian children in South Dakota is the lowest of any group in any state.”

In some cases, index values differ sharply between adjacent states. For example, American Indian children in Texas have an index value of 609, almost 300 points higher than the value of 316 for their peers in neighboring New Mexico. Other large differences among neighboring states include gaps of 219 points between California (505) and Arizona (286), 214 points between Idaho (481) and Montana (267), 157 points between Wisconsin (520) and Minnesota (363), and 143 points between Minnesota and South Dakota (220). Although some of these gaps may be linked to parallel large differences in social and demographic circumstances, large differences in outcomes over relatively short distances are inherently thought-provoking and merit further study. For example, it is not readily apparent why Minnesota’s outcomes lag so far behind Wisconsin’s, especially since Minnesota ranks on par with or higher than Wisconsin on children’s outcomes for every other racial/ethnic group in the report.

Figure 1: 2017 Race-for-Results Milestones Index for American Indian Children/Youth, by State

The inter-racial gaps within states may also merit attention. Figure 2 shows the White-American Indian score gap. Nationally the gap is 300 points. This wide divergence is driven by a few areas with very big gaps and relatively high American Indian populations--a block of 5 Northern Plains states (Minnesota, Montana, Nebraska, North Dakota, and South Dakota), a pair of Southwestern states (Arizona and New Mexico), and North Carolina. At the other end of the spectrum, American Indian children in Alabama score slightly ahead of their White peers, and 8 more states have gaps below 200 points. Why do American Indian children fare as well as or not too far behind White children in some states but face profound outcome gaps elsewhere?

Figure 2: White vs. American Indian Gaps in the Race-for-Results Child/Youth Milestones Index, by State

Again, some of these disparate state results occur in adjacent states. For example, the score for American Indian children in Idaho is only 156 points less than for White children, compared to a gap of 404 points in Montana. These and other gaps in the Race for Results report deserve further study and discussion, with the goal of making sure that all American Indian and other children reach their full potential.


1 See page 2 of the 2014 “Race for Results” report at

2 For the full report, see the website “2017 Race for Results” at For details on how the index is created, see The index combines statistics from the American Community Survey (3- to 5-year olds in nursery school, pre-school, or kindergarten; 19- to 26-year-olds in school or working; 25- to 29-year olds with at least an associate’s degree; children living with a householder with at least a high school degree; children in two-parent families; children in families with income at or above 200 percent of the poverty level; children living in low-poverty areas), the U.S. Department of Education (4th graders proficient in reading; 8th graders proficient in math; young adults graduating high school on time), and the Centers for Disease Control and Prevention (babies born at normal birthweight; 15- to 19-year-old females who delay childbearing to adulthood).

3 The report’s American Indian group basically corresponds to non-Hispanic individuals who identify racially as American Indian only (no other race), based on the following system. The report first classifies as Latino those who identify as Hispanic, Latino, or Spanish origin, according to definitions from the U.S. Office of Management and Budget (OMB). Non-Latinos are then placed in racial groups, including “two or more races” for all multi-race individuals and “American Indian” for those “who identified as belonging to an American Indian or Alaska Native tribal group.”

4 In the remaining 24 states (the white areas on the U.S. maps below), the data on American Indian children are too sparse to compute a valid index.