Family Structure


The number of parents in the household plays an important role in the growth and development of children. Previous studies found that two-parent households were generally associated with better health outcomes than single-parent households. For example, children in two-parent, married households were less likely than children in single-parent households to be in fair or poor health. Children in single-parent households were more likely to have a learning disability or attention deficit hyperactivity disorder (ADHD) and certain chronic health conditions than children in two-parent, biological households.1

In 2013, more than two-thirds of all U.S. children less than 18 years of age lived in households with two parents (includes biological, adopted, or stepparents), nearly a quarter lived in a mother-only household, and 4.1 percent lived in a father-only household (Figure 1). A small proportion of children (1.9 percent) lived with a grandparent. Between 2000 and 2013, the percentages of children under 18 years of age living in two-parent and single-parent households remained relatively stable.

Family Structure of Children

Figure 1 Source

Family structure differs with race and ethnicity. In 2013, less than half of non-Hispanic Black and non-Hispanic American Indian/Alaska Native children lived in two-parent households, compared to 77.4 percent of non-Hispanic White children (Figure 2). The majority of non-Hispanic Asian (86.0 percent), non-Hispanic Native Hawaiian/other Pacific Islander (70.0 percent), non-Hispanic White (77.4 percent), and Hispanic (65.0 percent) children lived in two-parent households.

Family Structure of Children by Race

Figure 2 Source

Household income as a percent of poverty is also related to family structure. In 2013, children in single-parent households were most likely to live in poverty, with 41.2 percent living in households with incomes below 100 percent of poverty ($23,834 for a family of four in 2013), compared to 13.2 percent of two-parent households.

Data Sources

Figure 1. U.S. Census Bureau and Bureau of Labor Statistics, Current Population Survey, Annual Social and Economic Supplement. Analyses conducted by the Maternal and Child Health Epidemiology and Statistics Program.

Figure 2. U.S. Census Bureau and Bureau of Labor Statistics, Current Population Survey, Annual Social and Economic Supplement. Analyses conducted by the Maternal and Child Health Epidemiology and Statistics Program.


1 Blackwell DL. Family structure and children’s health in the United States: findings from the National Health Interview Survey, 2001&ndash2007. National Center for Health Statistics. Vital Health Statistics. 2010;10(246).


Statistical Significance Test

Calculate the difference between two estimates:

Calculated Z-Test Result 0.9567433 Not statistically significant

We follow statistical conventions in defining a significant difference by a p-value less than 0.05 where there is a less than 5% probability of observing a difference of that magnitude or greater by chance alone if there were really no difference between estimates. The 95% confidence interval includes a plausible range of values for the observed difference; 95% of random samples would include the true difference with fewer than 5% of random samples failing to capture the true difference.

This website allows comparisons between two estimates using the independent z-test for differences in rates or proportions. This test is appropriate for comparing independent populations across years (e.g., 2011 versus 2012) or subgroups (e.g., Male versus Female) on corresponding measures. To the extent possible, the functionality of this application has limited estimate comparisons based on appropriate use of the independent z-test. However, some tables present subgroup categories within broader categories that will allow comparisons between non-independent populations (e.g., low birth weight and very low birth weight). Users should exercise caution when interpreting these test results, which will frequently overstate statistical significance.

For some tables, the website does not allow for comparisons between two estimates, even though the data represent independent populations. Generally, this is because the standard errors were not publicly available at the time this website was created.