Skip Navigation
Homeschooling in the United States: 2003

NCES 2006-042
February 2006

Executive Summary

This report represents the latest survey information from the National Center for Education Statistics on the prevalence of homeschooling in the United States. Homeschooling in the United States: 2003 uses the Parent and Family Involvement Survey of the 2003 National Household Education Surveys Program (NHES) to estimate the number and percentage of homeschooled students in the United States in 2003 and to describe the characteristics of these students and their families. It reports on the race and ethnicity, income level, and educational attainment of students’ parents; compares the characteristics of homeschoolers to those of public and private schooled students; examines how homeschooling rates have changed between 1999 and 2003 for different segments of the student population; and describes parents’ primary reasons for homeschooling their children, as well as the resources and curricular tools homeschooled students use in their education. Children were considered to be homeschooled if their parents reported them being schooled at home instead of at a public or private school, if their enrollment in public or private schools did not exceed 25 hours a week, and if they were not being homeschooled solely because of a temporary illness.

Interviews were conducted with the parents of 11,994 students ages 5 through 17 with a grade equivalent of kindergarten through 12th grade. Of these students, 239 were homeschooled. The NHES is designed to collect data on a wide range of educational indicators and types of students, including, but not limited to, homeschooling. Therefore, readers should note that the number of questions asked of homeschoolers and the number of homeschoolers represent only a small portion of the NHES collection. The overall response rates for the survey were 54 percent in 2003 and 65 percent in 1999. When the sample is weighted, it represents the approximately 50 million students ages 5 through 17 with a grade equivalent of kindergarten through 12th grade in the United States in 2003.

The results of the 2003 NHES survey reveal that the weighted estimate of the number of students being homeschooled in the United States in the spring of 2003 was 1,096,000, a figure which represents a 29 percent increase from the estimated 850,000 students who were being homeschooled in the spring of 1999 (table 1). In addition, the estimated homeschooling rate—the percentage of the student population being homeschooled—rose from 1.7 percent in 1999 to 2.2 percent in 2003. In this latest survey, parents were asked whether any of a set of reasons for homeschooling applied to them. Parents were then asked which one of the applicable reasons they considered to be their most important reason for homeschooling—31 percent of homeschooled children had parents who cited concern about the environment of other schools, such as safety, drugs, or negative peer pressure, as the most important reason for homeschooling and 30 percent had parents who said the most important reason was to provide religious or moral instruction (table 4). While these were the two most common responses, another 16 percent of homeschooled students had parents who said dissatisfaction with the academic instruction available at other schools was their most important reason for homeschooling.

Many of the 2003 survey findings concerning homeschooling rates by student and family characteristics paralleled those found in 1999. In 2003, as in 1999, the homeschooling rate for White students (2.7 percent) was higher than for Black students (1.3 percent) or Hispanic students (0.7 percent) (table 2). The homeschooling rate was also higher for students in families with three or more children in the household than for students in families with fewer children, higher for students in two-parent households than for students in one-parent households, and higher for students in two-parent households with only one parent participating in the labor force than for students with other parent labor force participation patterns. A multivariate analysis was conducted to determine whether these relationships held when controlling for all other demographic factors investigated in this report. In the multivariate analysis, most of the relationships held, but differences were not detectable between White and Black students and between students in two-parent households and students in one-parent households.

Between 1999 and 2003, homeschooling rates increased for a number of groups. Homeschooling rates increased from 0.9 to 1.7 percent among students with parents who have a high school diploma or less, from 2.0 to 2.7 percent among White students, from 1.6 to 2.4 percent among students in grades 6–8; and from 0.7 to 1.4 percent among students in single-parent households where the parent was in the labor force.

Finally, the 2003 report also investigates the sources homeschoolers used to obtain curricula or books for home education. A majority of homeschooled students had parents who used one or more of the following sources of curricula or books for their children’s home education: a public library (78 percent); a homeschooling catalog, publisher, or individual specialist (77 percent); a retail bookstore or other store (69 percent); and an education publisher that was not affiliated with homeschooling (60 percent) (table 5). Some students also used distance learning media. Forty-one percent of students who were being homeschooled in 2003 had engaged in some sort of distance learning (figure 3 and table 6).

As with results from any sample survey, the numbers and percentages discussed in this report are estimates of the actual numbers and percentages of homeschooled students in the population. NHES data, like all survey data, are subject to sampling error. Comparisons in the text are tested for statistical significance to ensure that the differences are larger than might be expected due to sampling error. All differences described are significant at the .05 level. Typically, NHES reports focus on statistically significant differences of at least 5 percentage points. However, this report presents estimates of a low-frequency event, the homeschooling rate. Because the homeschooling rate is less than 5 percent for almost all subpopulations, it is not possible to have a difference of greater than 5 percentage points, therefore, all statistically significant differences in the homeschooling rate between subpopulations are discussed.