• Background
• Estimated Number of Homeschooled Students in the United States
• Characteristics of Homeschooled Students and Their Families
• Parents' Reasons for Homeschooling
• Public School Support for Homeschooled Students
• Future Research Plans
• References

Past estimates of the number of homeschoolers vary by hundreds of thousands of children. Brian Ray, president of the National Home Education Research Institute, estimated the number of homeschoolers to be around 1.15 million during the 1996–97 school year, and predicted that the number would grow to at least 1.3 million during 1999–2000 (Ray 1997). Patricia M. Lines, through her research at the U.S. Department of Education, National Institute on Student Achievement, Curriculum, and Assessment, estimated the number of homeschoolers to be around 700,000 during 1995-96, possibly growing to 1 million by 1997–98 (Lines 1999). Both Ray and Lines grant that their estimates probably anchor the range within which the actual number of homeschoolers could fall.

The methods used by Ray and Lines in the development of their estimates varied. Ray derived his most recent estimate of the number of homeschoolers using his own 1995 survey of homeschoolers and their use of curricular packages as his base and sales of homeschooling curricular packages to adjust for growth over time. Ray applied the ratio of users of curricular packages and nonusers identified in the 1995 survey to more recent sales of homeschool curricular packages to obtain his 1999–2000 estimate. Lines collected data from all states that obtained records on homeschooling children in both the 1990–91 and 1995–96 school years (32 states and the District of Columbia). Using the 12 states with high record-collection rates for homeschoolers, based on Ray's estimates of the percentage of homeschoolers who reported filing in their state, Lines estimated the percentage of school-aged children who were homeschooling in those 12 states. Lines estimated the number of children homeschooled nationally by applying the percentage distribution of homeschoolers from the state sample to national totals of school-aged children.

The National Center for Education Statistics (NCES) was the first organization to attempt to estimate the number of homeschoolers in the United States using a rigorous sample survey of households. A household sampling frame circumvents problems inherent in the use of incomplete sample frames, such as customers of curricular providers and administrative records. Attempts to develop estimates of homeschoolers through household surveys, however, can also be problematic. The first two efforts to estimate homeschoolers at NCES—through the October supplement to the 1994 Current Population Survey (CPS:Oct94) and through the Parent and Family Involvement in Education/Civic Involvement Survey of the National Household Education Surveys Program, 1996 (PFI/CI-NHES:1996)—produced very different estimates. One problem that may have contributed to the varying estimates was the difference in how the two surveys identified students who were both homeschooled and enrolled in school part time. Neither survey collected precise data on these part-time home-schoolers. An NCES technical report, Issues Related to Estimating the Home-Schooled Population in the United States With National Household Survey Data, explores in detail the differences in survey design and execution that may have contributed to the disparity between the CPS:Oct94 and PFI/CI-NHES:1996 estimates (Henke et al. 2000).

In this report, the Parent Survey of the National Household Education Surveys Program, 1999 (Parent-NHES:1999) is used to estimate the number of homeschoolers in the United States, to describe the characteristics of home-schoolers, and to document parents' reasons for homeschooling and parents' reports of public school support for homeschoolers. Students 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. The unweighted number of homeschooled students used in this analysis is 275 and the unweighted number of nonhomeschooled students is 16,833. Students are defined in this report as children ages 5 to 17 with a grade equivalent of kindergarten through grade 12.

Approximately 850,000 students were being home-schooled during the spring of 1999, according to data from the Parent-NHES:1999 (table 1). Homeschoolers accounted for 1.7 percent of students nationwide, ages 5 to 17, with a grade equivalent of kindergarten through grade 12. The estimate includes students who were homeschooled while also enrolled in school for 25 hours or less per week, and excludes students who were homeschooled due to a temporary illness.

As with all sample surveys, the numbers and percentages in this report are estimates of the numbers and percentages in the population. Although 850,000 is the best population estimate available from this sample survey, another similar sample survey might produce a different estimate. A 95 per-cent confidence interval defines a range of values such that 95 percent of the estimates from other similar surveys will fall within the range of values. The 95 percent confidence interval for the number of homeschoolers is 709,000 to 992,000. The estimate provided here—850,000—is the midpoint of the range. Figure 1 illustrates the confidence interval.

Despite research limitations on documenting the number of homeschoolers, recent research on homeschooling helps suggest some characteristics of students and families who homeschool. An extensive 1998 study of homeschoolers, although based on a convenience sample, suggests that homeschoolers differ from the general population in parents' educational attainment, household income, parents' marital status, and family size (Rudner 1999).¹ Other research suggests that although homeschooling in the United States may once have been primarily a trend within a homogeneous subgroup of White, middle-class, Christian families, growth in homeschooling may be reaching a broader range of American families and values (McDowell, Sanchez, and Jones 2000; Lines 2000a; Welner and Welner 1999).

The Parent-NHES:1999 provides descriptive data about the characteristics of homeschoolers in the United States and their families. This report includes students who were homeschooled only and students who were homeschooled and enrolled in school for 25 hours or less per week. As shown in table 1, about four out of five homeschoolers were homeschooled only (82 percent) and one out of five homeschoolers were enrolled in public or private schools part time (18 percent).

Table 1.—Number and percentage of homeschooled students, ages 5-17, with a grade equivalent of kindergarten to grade 12, by school enrollment status: 1999

Figure 1.— Point estimate and 95 percent confidence interval for number of homeschooled students, ages 5–17, with a grade equivalent of kindergarten to grade 12: 1999

Table 2 shows the number of all students by selected characteristics, the number of homeschooled students by those same characteristics, and for each characteristic the percentage of students who are homeschooled. As shown in table 2, the percentage of students who were homeschooled in 1999 differed based on various characteristics of students and their families. Depending on these student and family characteristics, the percentage of homeschoolers among students ranged from 0.7 to 4.6 percent. Characteristics that distinguished high percentages of homeschooling were two-parent families, especially when only one parent participated in the labor force; large family size; and parents' high educational attainment. The percentage of students who were homeschooled was similar for both boys and girls; across elementary, middle, and high school grades; and across the four income ranges used in the analysis.

Table 3 further explores the characteristics that distinguish homeschoolers by comparing the characteristics of homeschoolers to those of all students and to the characteristics of nonhomeschoolers. The similarities and differences between homeschoolers and nonhomeschoolers are discussed in detail below.

Homeschoolers distribute over the grade groupings in much the same way as nonhomeschoolers (table 3). While it may appear that a higher percentage of homeschoolers were in kindergarten compared to nonhomeschoolers, the difference was not statistically significant.

A greater percentage of homeschoolers compared to nonhomeschoolers were White, non-Hispanic—75 percent compared to 65 percent. At the same time, a smaller percentage of homeschoolers were Black, non-Hispanic students and a smaller percentage were Hispanic students. Girls and boys were equally represented among homeschoolers and nonhomeschoolers.

A much greater percentage of homeschoolers than non-homeschoolers came from families with three or more children—62 percent of homeschooled students were part of families with three or more children compared to 44 percent of nonhomeschoolers. Homeschoolers were just as likely as nonhomeschoolers to be an only child and were less likely than nonhomeschoolers to have just one sibling.

In order to homeschool, parents may need to dedicate a significant amount of time to schooling their children. Because of the time required, homeschooling usually involves two parents—one who participates in the labor force and one who homeschools. Rudner (1999) found that 97 percent of homeschooling parents were married couples. The Parent-NHES:1999 shows the percentage of home-schooled students living in two-parent households was much higher than the percentage for nonhomeschoolers—80 percent of homeschooled students lived in two-parent families compared to 66 percent for nonhomeschoolers. In addition, 52 percent of homeschoolers came from two-parent families where only one parent was participating in the labor force compared to 19 percent for nonhomeschoolers.

Although Rudner found that the median household income of homeschooling families was higher than the median household income of families with children nationwide, the Parent-NHES:1999 indicates that the household income of homeschoolers, reported in ranges from less than $25,000 to over $75,000, is the same as the household income of nonhomeschoolers. The same percentage of homeschooled and nonhomeschooled students lived in households with annual incomes of $50,000 or less (64 percent).²

Parents' highest educational attainment, however, was clearly associated with homeschooling. Parents of home-schoolers had higher levels of educational attainment than did parents of nonhomeschoolers. Table 3 shows that 37 per-cent of parents of nonhomeschoolers did not complete any schooling beyond high school, compared to 19 percent of parents of homeschoolers. Conversely, 25 percent of parents of homeschoolers attained bachelor's degrees as their highest degree, compared to 16 percent of parents of nonhomeschoolers.

Urbanicity refers to the classification of households as urban or rural. There are two classifications of urban, which are referred to in this report as cities and towns. Places not classified as urban are rural. The percentage of home-schoolers living in a city was about 9 percentage points lower than the percentage for nonhomeschoolers (53 and 62 percent, respectively). There were no statistically significant differences between the percentages of homeschoolers and nonhomeschoolers living in towns or rural areas.

Table 2.—Number of students and number and percent of homeschooled students, ages 5–17, with a grade equivalent of kindergarten to grade 12, by selected characteristics: 1999

Table 3.—Distribution of all students, homeschooled students, and nonhomeschooled students, ages 5–17, with a grade equivalent of kindergarten to grade 12, by selected characteristics: 1999

Parents may homeschool their children for a number of reasons. Previous studies suggest that the most common reasons that parents give for homeschooling their children are moral or religious reasons, a desire for high educational achievement, dissatisfaction with public schools' instructional program, and concerns about school environment, including safety, drugs, and peer pressure (Lines 2000a; Grubb 1998; Mayberry 1991).

Parents gave a wide range of reasons for homeschooling in the Parent-NHES:1999.³ Parents were asked to list their reasons for homeschooling and could provide as many reasons as applied. The reasons parents gave were coded into 16 categories and included better education, religious reasons, and poor school environment. Figure 2 shows 10 reasons cited by at least 5 percent of students' parents. Additional reasons are listed in table 4.

Public schools or school districts sometimes offer support for homeschoolers by providing parents with a curriculum, books and materials, places to meet, and the opportunity for homeschooled children to attend classes and participate in extracurricular activities at the school. Previous research found that only a small percentage of homeschoolers enrolled in classes, used textbooks, or used libraries when they were made available by public schools and that many home-schoolers express antipathy toward using public school support (Lines 2000b; Yeager 1999; Mayberry et al. 1995).

Table 5 shows the different types of public school support for homeschoolers asked about in the Parent-NHES:1999. Parents of homeschoolers were asked whether their child's assigned school or district offered any of the eight pre-specified types of support shown in table 5. The estimates are based on parents' reports of public school support and use, not what schools or districts may actually offer. Between 15 and 38 percent of homeschoolers' parents did not know whether various types of support were offered.

The first two columns of estimates in table 5 show that, altogether, 28 percent of homeschoolers' parents reported that public schools or districts offered extracurricular activities, 21 percent reported curriculum support, and 23 percent reported books and materials. Between about 3 and 11 percent of homeschoolers' parents said that support was available and that they used the support, and between about 5 and 22 percent said that the support was available but they did not use it. For example, as table 5 shows, about 6 percent of homeschoolers' parents reported that they had the chance to attend extracurricular activities and used this type of support, and 11 percent reported that schools offered books and materials and that they used this type of public school support.

NCES plans to collect and report data about homeschoolers with future Parent and Family Involvement in Education Surveys (PFI), slated to occur on a 4-year cycle next scheduled for 2003 as part of the National Household Education Surveys Program (NHES). Future Parent and Family Involvement in Education Surveys will provide a comprehensive set of information that may be used to estimate the number and characteristics of homeschoolers in the United States. Future areas of inquiry might also include items on homeschoolers' use of distance learning and the Internet, information about families' past use of homeschooling, more information about homeschoolers who attend school for some classes or subjects, and specific information about homeschoolers' plans for postsecondary education.

Figure 2.— Ten reasons for homeschooling and the percentage of homeschooled students whose parents gave each reason: 1999

Table 4.—Number and percentage of homeschooled students, by reason for homeschooling: 1999

Table 5.—Percentage of homeschooled students whose parents reported availability and use of support from public schools or districts: 1999