When the National Assessment for Educational Progress (NAEP) was first administered in 1969, nearly nine out of 10 American children were born into married, two-parent homes. By 2023, this number had decreased to six out of 10 children on average, with wide variations across racial groups. Stunning disparities in married, two-parent households by race tightly correlate with disparities in child poverty, domestic violence, and father absence—all factors that adversely affect individual children growing up under these conditions. 

So to what degree do married two-parent households affect K-12 student academic outcomes? The answer may be intuitively obvious, but the data collected on it is nearly impossible to analyze.

2024 NAEP results were once again disappointing, revealing that the percentage of 8th graders reading below NAEP Basic was the worst and largest in the assessment’s history—33%. 49 out of 50 states saw a decrease in scores from 2019 to 2024 in 8th-grade math. Tennessee, the sole exception, showed no change at all.

To help analyze results, NAEP is legally required to collect demographic and other information from participating students, teachers, and school administrators through surveys that measure contextual factors related to student achievement. Thus, student results are disaggregated by categories like sex, race/ethnicity, socioeconomic level, disability, and English learner status. 

Breaking student outcomes out by these “usual suspect” groupings leads to the conclusion that discrimination within these groupings—racism, sexism, etc.—must be the primary sources of implied causality that explain achievement gaps. As so-called anti-racist author Ibram Kendi notoriously said, “when I see racial disparities, I see racism.” When we ask narrow questions, we shouldn’t be surprised by narrow answers.

The good news is that NAEP already collects data from students about household composition and family structure. One question on the 2024 student questionnaire (8th-grade math) asks which family members live with the student, allowing children to choose multiple answers from the universe of mother, father, stepmother, stepfather, foster mother or other female legal guardian, or foster father or other male legal guardian. 

“NAEP does collect family structure data and makes it available on the NAEP Data Explorer,” Mark Schneider, former director of the Institute of Education Sciences, told us. “As [CEO of the Center on Child and Family Policy] Katharine Stevens has demonstrated, family structure, unsurprisingly, interacts with other more widely used measures of race and class—so that the effects of family structure are far from one dimensional. The terrible NAEP Data Explorer makes the exploration of such interactions close to impossible.” 

This explains why there is such little data analysis on the relationship between a potentially transcendent factor like family structure and academic achievement. As a result, simplistic conclusions about the primacy of race, sex, and other immutable characteristics persist. As the late Henry Kissinger once said, “The absence of alternatives clears the mind marvelously.” 

If the new Secretary of Education prioritized reforming the Data Explorer so that academic achievement could be easily analyzed by family structure, it would be a game changer. According to the American Community Survey, 82% of Asian children and 62% of white children live with married birth parents. This compares to only 43% of Latino children and a dismal 23% of black children. We believe reporting student outcomes by family structure would reveal far more than the role of race or ethnicity. 

Put simply: Family structure matters, likely more than any other factor for a child’s development. As Melissa Kearney explains in her book, children who grow up in stable two-parent homes enjoy, on average, more quality adult time, greater economic resources, and increased emotional support from their parents.

More than a decade ago, the obvious relevance of family structure to child outcomes led healthcare leaders and analysts to make common sense changes to their methods of measurement. Analyses using seven distinct and mutually exclusive family structures (nuclear, single-parent, blended, unmarried biological or adoptive families, cohabiting, extended, and other) are yielding new explanations for entrenched problems and ushering in a new wave of family-focused prescriptions in the health arena. The same breakthroughs could occur in education. 

It is true that educators cannot control or change the structure of the families that students are from. But we can influence how our students think about the families they form and the series of life choices that will likely lead to their life success and that of their children. One way of doing this is teaching middle and high school students the success sequence: 97% of young people who graduate with at least a high school degree, enter into full-time employment, and then get married before having children are not poor as adults, and the vast majority enter the middle class or beyond. 

In order for such reforms to take hold nationally, we need more reliable data on the relationship between family structure and outcomes. The Trump Administration can take one large leap in this direction by making the NAEP Data Explorer easier to use, and by incorporating family structure as a prominent matrix of analysis.