CARDUS

Home | Research & Policy Library | Diverse Outcomes for a Diverse Population

Diverse Outcomes for a Diverse Population

Findings About Homeschooled Adults from the Cardus Education Survey

February 3, 2025

Albert Cheng

Angela R. Watson

Education

Research Report

Cardus Education Survey Education Outcomes Educational Pluralism

No single, overarching profile accurately describes Americans who were homeschooled.

Key Points

  • The Cardus Education Survey (CES), conducted in the United States in 2011, 2014, 2018, and again in 2023, examines a range of outcomes for a nationally representative sample of adults aged 24 to 39. Presented in this report are the economic, mental-health, civic, spiritual, and family-formation outcomes for the group of CES respondents who were homeschooled for a portion of their primary and secondary education.
  • Adults who were homeschooled are disaggregated into short-term homeschoolers (who were homeschooled one to two years), medium-term homeschoolers (three to seven years), and long-term homeschoolers (eight years or more) and compared to adults who were never homeschooled, controlling for a range of demographic characteristics.
  • Top-line findings:
    • Diversity in terms of time spent in homeschooling: Adults who were homeschooled spent a wide range of time in that learning environment. Among them, 20 percent were homeschooled for only one year of their primary and secondary education, and 17 percent were homeschooled for all K–12 years.
    • Educational attainment and employment: Bachelor’s and graduate degree attainment rates were similar between short-term homeschoolers and non-homeschoolers. But medium- and long-term homeschoolers were more likely than short-term and non-homeschoolers to have ended their education with a high school diploma or an associate’s degree. For all three groups of homeschooled adults, the likelihood of being employed full-time and having a household income above the median was lower relative to adults who were never homeschooled.
    • Mental health and well-being: Compared to the other respondent groups, long-term homeschoolers exhibited the highest levels of optimism, gratitude, and life satisfaction. Long-term homeschoolers were also the least likely to “feel helpless dealing with life’s problems” and to report symptoms associated with depression and anxiety.
    • Civic behaviors: Short- and long-term homeschoolers were more likely to have volunteered and given to charity in the past twelve months compared to medium-term homeschoolers and adults who were never homeschooled.
    • Religious engagement: Homeschooled adults were much more likely to report that they believed in God and life after death and regularly engaged in religious practices. The prevalence of religious belief and practice increased with the number of years spent in the homeschool sector.
    • Family formation: Compared to the other respondent groups, long-term homeschoolers were more likely to be married, had the lowest divorce rate, and had more children, on average.

Introduction

Past Research on the Outcomes of Homeschooled Children

Questions about the effectiveness of homeschooling, debates over its legitimacy, and concerns about how it is practiced have risen since the 1970s and persist today. Throughout this time, educational theorists, policymakers, legal experts, and homeschooling parents themselves have debated the legitimacy of homeschooling, the role of the family in education, whether the child has a right to a different kind of education or other protections, and practical issues such as child safety. Meanwhile, academic researchers have conducted numerous studies to evaluate homeschoolers’ academic achievement, socialization, and other outcomes. According to that research, homeschoolers do not conclusively appear different from their non-homeschooling counterparts in terms of those outcomes. Some studies demonstrate differences between the two groups, yet it is unclear from them whether those differences are due to homeschooling or other factors. Nor do many of the studies generalize to the entire population of homeschoolers, which has become increasingly diverse over the past several decades. 1 1 R. Kunzman and M. Gaither, “Homeschooling: An Updated Comprehensive Survey of the Research,” Other Education: The Journal of Educational Alternatives 9, no. 1 (2020): 253–336, https://www.researchgate.net/publication/374195331_Homeschooling_An_Updated_Comprehensive_Survey_of_the_Research; C. Valiente, T.L. Spinrad, B. Ray, N. Eisenberg, and A. Ruof, “Homeschooling: What Do We Know and What Do We Need to Learn?” Child Development Perspectives 16, no. 1 (2022): 48–53, http://dx.doi.org/10.1111/cdep.12441.

All of these issues surrounding homeschooling are growing increasingly salient. The onset of the COVID-19 pandemic catalyzed a surge of parents who decided to homeschool their children. When schools closed, students were educated from home, but as schools re-opened and the pandemic abated, some parents continued the practice. According to one estimate, enrollment in traditional public schools throughout the United States declined by 2.1 million students from the fall of 2019 to the fall of 2021. Over the same period, homeschooling enrollments increased by 30 percent. 2 2 T. Dee, “Where the Kids Went: Nonpublic Schooling and Demographic Change During the Pandemic Exodus from Public Schools,” Urban Institute, 2023, https://www.urban.org/research/publication/where-kids-went-nonpublic-schooling-and-demographic-change-during-pandemic/.

Since 2011, the Cardus Education Survey (CES) has contributed to the scholarship and public discourse about a range of educational topics in the United States, including homeschooling. The CES has been regularly administered to a nationally representative sample of high school graduates aged 24 to 39 who attended traditional public schools, Catholic schools, Protestant schools, nonreligious independent schools, or homeschool settings. The CES queries respondents about various life outcomes, including educational attainment, employment, civic engagement, social connectedness, and religious belief and practice.

Past reports about homeschooling based on analysis of the CES data examined only respondents who were religious homeschoolers—defined for that purpose as those who were homeschooled for a majority of their high school years and whose parents regularly attended religious services. The findings from the CES described in those reports provide a picture of how adults who were homeschooled as children for religious reasons fare today. The 2018 iteration of the CES revealed evidence that religious homeschoolers were less likely than other respondent groups to have earned a postsecondary degree and were more likely to have lower incomes. It also revealed that religious homeschoolers were less socially connected; they had fewer people they could confide in overall but were more likely to list a family member as someone they were relationally closest to. The 2011 CES also demonstrated that religious homeschoolers were more likely to agree that they lacked a clear sense of direction or purpose and felt helpless dealing with life’s problems. The 2018 CES, by contrast, demonstrated that religious homeschoolers were less likely to agree that their lives lacked clear goals or sense of direction. The 2018 data also revealed that religious homeschoolers were more distrustful of politicians and the media, and that they spent more time volunteering, especially through religious organizations. Religious homeschoolers, like Protestant school respondents, were most likely to remain religiously engaged in adulthood. 3 3 Pennings, R., and K. Wiens, “Cardus Education Survey: Do the Motivations for Private Religious Catholic and Protestant Schooling in North America Align with Graduate Outcomes?”,  Cardus, 2011, https://www.cardus.ca/wp-content/uploads/2023/05/Cardus-Education-Survey-Phase-I-report.pdf; “Cardus Education Survey 2018: Spiritual Strength, Faithful Formation,” Cardus, 2019, https://cardus.ca/research/education/research-report/2018-us-cardus-education-survey-spiritual-strength-faithful-formation/.

Purpose of Report

This report uses 2023 CES data to provide an updated nationally representative description of US respondents who were homeschooled for at least one year of their primary and secondary education. Unlike prior iterations of the CES, the sample includes all respondents who were homeschooled for at least one year during their primary and secondary education, not just those who were homeschooled for a majority of their high school years and whose parents regularly attended religious services. Homeschooled respondents in the 2023 CES could have been homeschooled from as early as 1989 to as recently as 2017. Notably, the 2023 CES administration does not address homeschooling during the COVID-19 pandemic, though such a study would be possible in a few years once young people who spent part of their primary and secondary schooling during the pandemic come of age.

Drawing upon past research that has identified distinct groups of homeschoolers based on the amount of time they were homeschooled, this report additionally disaggregates the population of respondents who have ever been homeschooled into short-term, medium-term, and long-term homeschoolers. 4 4 D. Hamlin and A. Cheng, “Homeschooling, Perceived Social Isolation, and Life Trajectories: An Analysis of Formerly Homeschooled Adults,” Journal of School Choice 16, no. 2 (2022): 332–59, https://doi.org/10.1080/15582159.2022.2028338. The rationale for this disaggregation is to examine whether the outcomes of homeschoolers relate to the amount of time spent in the homeschool sector. For the purpose of this study, short-term is defined as one to two years, medium-term as three to seven years, and long-term as eight to thirteen years. These three groupings, furthermore, roughly divide the sample of homeschoolers into three groups of equal size, maximizing the statistical power needed for making comparisons among the different groups of homeschoolers.

Disaggregating results by the amount of time that a respondent was homeschooled—something prior research has not done—is important for providing a clearer picture of homeschooling outcomes. Almost all existing quantitative research of homeschoolers treats them as a uniform group, overlooking the diversity of their experiences. As a result, policy debates about homeschooling and popular conceptions of homeschoolers typically assume that homeschoolers are homeschooled for the entirety of their primary and secondary education. However, recent estimates from some studies suggest that such a characterization applies to only 10 percent of homeschoolers. Most are homeschooled for only one or two years and spend the rest of their years of education enrolled in conventional schools. 5 5 Hamlin and Cheng, “Homeschooling, Perceived Social Isolation, and Life Trajectories.”

The following sections in the report describe how respondents who are short-, medium-, or long-term homeschoolers are faring in life by comparing them across a range of outcomes, including employment, mental health, civic engagement, faith, and family structure. Each section is devoted to one of these outcome areas. Alongside the results for the three groups of homeschooled respondents, results are presented for the respondents who were never homeschooled, to provide another point of comparison. In this report, these respondents are referred to as non-homeschoolers. Readers interested in comparing homeschoolers to non-homeschoolers from specific school sectors such as traditional public or other independent schools should refer to the Cardus report “School-Sector Influence on Graduate Outcomes and Flourishing.” 6 6 L.E. Swaner, A. Cheng, and J. Eckert, “School-Sector Influence on Graduate Outcomes and Flourishing: Findings from the 2023 Cardus Education Survey,” Cardus, 2024, https://www.cardus.ca/research/education/reports/school-sector-influence-on-graduate-outcomes-and-flourishing/.

It is important to note that this report does not make a causal argument about the effects of being homeschooled or the effects of being homeschooled for particular durations of time. On the one hand, all the analyses control for important demographic characteristics, including childhood poverty, whether the respondent grew up with both biological parents, and whether the respondent grew up in a religious household, as well as the respondent’s gender, race, mother’s educational attainment, age, and urbanicity of residence. Therefore, the possibility that any outcome differences between short-, medium-, and long-term homeschoolers and non-homeschoolers are attributable to these factors can be ruled out.

However, the data and analysis cannot account for unmeasured systematic differences between those four groups of respondents. For example, some families decide to homeschool due to bullying or other unsafe experiences at school. 7 7 S. Green-Hennessy and E.C. Mariotti, “The Decision to Homeschool: Potential Factors Influencing Reactive Homeschooling Practice,” Educational Review 75, no. 4 (2023): 617–36, https://doi.org/10.1080/00131911.2021.1947196. Those adverse experiences leading to the transition to homeschooling may have ongoing effects on the child’s educational outcomes. As a result, one cannot clearly discern whether it is the experience with an unsafe school environment or the length of time someone is homeschooled that causes any differences in educational outcomes. Unmeasured factors such as this, which may influence both length of homeschooling and educational outcomes, preclude the ability to use the 2023 CES data to make causal claims about the effects of homeschooling. Despite these methodological limitations, there is much to learn about how homeschoolers fare in adulthood in the following pages.

Study Sample

Between October 10 and 22, 2023, the polling firm Ipsos Public Affairs administered the CES to members of KnowledgePanel, one of the largest representative panels of US adults. The sample comprised 2,350 adults aged 24 to 39 who completed at least high school. About 8 percent (or 181) of these respondents were homeschooled for at least one year between kindergarten and twelfth grade. As discussed above, these 181 respondents were grouped into three approximately equally sized groups: (1) short-term homeschoolers, who were homeschooled for one to two years, (2) medium-term homeschoolers, who were homeschooled for three to seven years, and (3) long-term homeschoolers, who were homeschooled more than seven years.

Information about how long each respondent was homeschooled is presented in table 1.1. Based on the CES data, the number of years that someone spent in a homeschool setting is quite varied. About 56 percent of adults who were ever homeschooled were homeschooled for six years or less, that is, for less than half of their thirteen years of primary and secondary education. In fact, one in five adults who were ever homeschooled were homeschooled for only one year. On the other hand, a mere 17 percent of adults who were ever homeschooled were homeschooled for every year of their K–12 schooling.

Many respondents who were homeschooled also spent significant time in conventional school settings. Short-term homeschoolers spent about eight years in traditional public schools and three years in charter or independent schools, on average. Meanwhile, medium-term homeschoolers spent about five years in traditional public schools and almost six years in charter or independent schools, on average. Even the average long-term homeschooler spent about one year in traditional public schools and half of a year in charter or independent schools during their primary and secondary schooling years.

Educational Attainment and Employment Outcomes

Perhaps one of education’s most valued outcomes is equipping children for a career and employment in adulthood. While readiness for college and career has been a guiding educational mantra for schools and policymakers for several decades, the American understanding of the purpose of education has changed over time. Once focused more on becoming a productive member of society, goals for education have shifted to a more holistic and humanistic view. Nevertheless, preparation for work and completing a sufficient amount of formal schooling contributes to human flourishing. Thus, concerns about educational attainment and employment outcomes are central and are a key part of the CES. Americans traditionally value four-year postsecondary degrees—and for good reason, given the connection between obtaining such a credential and employment outcomes. Although the value of a four-year degree has been increasingly questioned in recent years, formal postsecondary training and education plays some role in preparing young people for employment. 8 8 B. Caplan, The Case Against Education (Princeton University Press, 2019); C. Goldin and L.F. Katz, The Race Between Education and Technology (Harvard University Press, 2008).

How do homeschooled respondents fare regarding educational attainment and employment outcomes? The 2018 CES found that respondents who were homeschooled for religious reasons had completed one or two fewer years of education, and were over twice as likely to be unemployed, as respondents who attended public and independent schools. Religious homeschoolers were about four times less likely to pursue any postsecondary education, and those who pursued postsecondary education were one and a half times more likely to have pursued a two-year degree than a four-year degree. 9 9 “Cardus Education Survey 2018: From the Classroom to the Workplace,” Cardus, 2019, https://cardus.ca/research/education/research-report/cardus-education-survey-2018-from-the-classroom-to-the-workplace/.

However, as described in the introduction, analysis of the 2018 CES data did not disaggregate these homeschoolers by years spent in the sector. Nor did it consider nonreligious homeschoolers, masking important variation within the total homeschooling sample. Educational attainment and employment outcomes are more fully explored in this analysis by disaggregating homeschooled respondents into short-, medium-, and long-term homeschoolers. Moreover, the entire population of homeschoolers was considered, not just those from religious families.

The 2023 CES asked respondents about their academic attainment and employment. Respondents from the medium- and long-term homeschooler groups were less likely to have obtained at least a bachelor’s degree. Forty-five percent of short-term homeschoolers had at least a bachelor’s degree, comparable to the 46 percent of non-homeschoolers who had at least a bachelor’s degree. In contrast, only 29 percent and 27 percent of respondents from the medium- and long-term homeschooler groups, respectively, had at least a bachelor’s degree.

The results regarding four-year and graduate degree attainment raise the question of whether medium- and long-term homeschoolers were more likely to get two-year associate’s degrees or to simply not pursue any postsecondary education. Analysis of the 2023 CES data suggests the latter explanation. Rates of associate’s degree attainment, or of having enrolled in any postsecondary education without completing a program, were similar between medium- and long-term homeschoolers, compared to non-homeschoolers.

On the other hand, 47 percent and 41 percent of medium-term and long-term homeschoolers, respectively, had at most a high school diploma or equivalent credential, compared to 27 percent of non-homeschoolers. At 36 percent, short-term homeschoolers also had a somewhat higher rate of having no more than a high school diploma, compared to non-homeschoolers.

When comparing the employment status of the different homeschooling groups, long- and short-term homeschoolers were more similar to each other than to medium-term homeschoolers. Nearly two-thirds of long- and short-term homeschoolers reported being employed full-time, compared to 48 percent of the medium-term homeschooled respondents. These rates for all homeschoolers were lower than the nearly three-quarters of non-homeschoolers who were employed full-time. The gap in full-time employment rate between non-homeschoolers and those who were homeschooled for any length of time appears driven more by differences in unemployment rather than part-time employment. Part-time employment rates were comparable across all groups, except for the higher likelihood of part-time employment among medium-term homeschoolers. However, the proportion of homeschooled respondents who were not working ranged from 26 percent of long-term homeschoolers to 33 percent of medium-term homeschoolers, compared to 17 percent of non-homeschoolers.

When asked about income, respondents again varied by group. Results are shown in figure 2.3 for the proportion of respondents who reported household incomes above the median income level, which was about $81,000 according to the US Census Bureau. 10 10 G. Guzman and M. Kollar, “Income in the United States: 2023,” United States Census Bureau (Sept. 2024), https://www2.census.gov/library/publications/2024/demo/p60-282.pdf. The sample size was not large enough to disaggregate income levels into finer categories. Homeschooled respondents, regardless of years spent in the sector, were less likely to report incomes above the median income level, compared to non-homeschoolers. A similar percentage of medium- and long-term homeschooled respondents reported household earnings above the median—about one-third of them. Meanwhile, 41 percent of short-term homeschoolers reported having household incomes above the median. These findings about earnings appear consistent with those about educational attainment and employment rates. That is, earnings were generally higher for groups that completed more education and lower for groups where unemployment rates were higher.

To better understand the underlying reasons behind these employment outcomes, the survey asked respondents to state how important money and hard work were as personal values.

One-third of respondents who were never homeschooled indicated that money was a very important personal value. Short- and medium-term homeschoolers were slightly less likely to hold this view, at about 29 and 28 percent of them, respectively. However, only 19 percent of long-term homeschoolers said money was a very important personal value.

As for the value of hard work, over half of medium- and long-term homeschoolers considered it a very important personal value. Those rates were significantly higher than the 39 percent of short-term homeschoolers and the 42 percent of non-homeschoolers who held this view.

In other words, long-term homeschoolers may value other things in life more than money, which might partially explain their lower educational attainment and income levels. However, long-term homeschoolers also placed more value on hard work, which might be a partial explanation for why their likelihood of employment was similar to that of more highly educated short-term homeschoolers.

Prior administrations of the CES found that homeschooled respondents were less likely to have obtained four-year degrees and less likely to have higher incomes. The findings presented in this section suggest that some additional nuance is necessary to characterize the educational attainment and employment outcomes of homeschooled respondents. Data presented in this section provide evidence for differences not only between respondents who were homeschooled and respondents who were not but also within the homeschooled cohort, depending upon the number of years spent in the sector. Distinct patterns emerged after disaggregating the results by the years spent being homeschooled. For instance, bachelor’s degree attainment rates were similar between respondents who were homeschooled for only one or two years and non-homeschoolers. Moreover, the likelihood that a respondent had not pursued an education after high school was highest for medium-term homeschoolers. The likelihood of being unemployed was highest for medium-term homeschoolers as well. Likewise, household incomes were lower overall among respondents who were ever homeschooled, compared to non-homeschoolers, but the income gap was smallest for short-term homeschoolers.

Mental Health Outcomes

Concerns about a child’s mental health are a significant driver of families’ decisions to homeschool. The National Household Education Survey, administered every few years by the National Center for Education Statistics, asks parents about their reasons for homeschooling. Concerns with safety, defined broadly and likely related to physical and psychological safety, as well as the desire to provide moral or religious instruction, have ranked at the top of the list of reasons. In the 2019 National Household Education Survey, 80 percent of homeschooling parents reported that concerns about the school environment were a reason they chose to homeschool. One-quarter of homeschooling parents said that a concern about the school environment was the top reason for homeschooling. 11 11 L. Hudson et al., “2019 Homeschooling and Full-Time Virtual Education Rates,” National Center for Education Statistics, US Department of Education (2023), https://nces.ed.gov/pubs2023//2023101.pdf.

Further, concerns about mental health reached a new level of visibility during the pandemic. In some cases, mental health professionals were concerned about the mental health of children who were isolated at home, citing increased suicide rates. 12 12 National Institute of Mental Health, Youth Suicide Rates Increased During the COVID-19 Pandemic, Research Highlight (2023), https://www.nimh.nih.gov/news/science-news/2023/youth-suicide-rates-increased-during-the-covid-19-pandemic. In other cases, students who were not attending brick-and-mortar schools during school closures may have flourished, leading to decreased suicide rates among this age group, raising questions about the conditions of school environments. 13 13 M. Roeloffs, “Teen Suicide Plummeted During Covid-19 School Closure, New Study Finds,” Forbes, July 19, 2023, https://forbes.com/sites/maryroeloffs/2023/07/19/teen-suicide-plummeted-during-covid-19-school-closures-new-study-finds/. Some attribute teen suicide and other mental health challenges to stressors from school, such as bullying, academic pressure, and peer pressure; suicide is the most extreme mental health outcome, but different manifestations of poor mental health such as anxiety, depression, and hopelessness are also of concern and likely more widespread. According to data from the Centers for Disease Control, mental health has worsened over the last decade among high school students, particularly among females. 14 14 US Centers for Disease Control and Prevention, Youth Risk Behavior Survey Data Summary and Trends Report: 2011–2021 (2022), https://www.cdc.gov/yrbs/dstr/pdf/YRBS_Data-Summary-Trends_Report2023_508.pdf. The general well-being of young adults has also worsened, particularly among males. 15 15 R.V. Reeves, Of Boys and Men: Why the Modern Male Is Struggling, Why It Matters, and What to Do About It (Brookings Institution Press, 2022).

The 2011 and 2018 administrations of the CES have drawn some attention to mental health concerns among religious homeschoolers, that is, respondents who were primarily homeschooled as children and came from religious family backgrounds. Those data indicate that religious homeschoolers, when compared to other respondents, were less socially connected, having fewer people in whom they could confide. Religious homeschoolers were also more likely to say that they lacked a clear sense of direction in life and felt helpless dealing with life’s problems. What is the mental health status of homeschooled respondents now, and is it related to how long they were homeschooled? The 2023 CES data provide some insight into these questions.

When asked about the state of their mental health, respondents expressed some differences depending on the length of their homeschool experience. For example, when asked how frequently they had feelings of helplessness, respondents who were homeschooled for short and medium terms responded similarly to respondents who were never homeschooled. About one out of three respondents in these three groups agreed that they often “feel helpless in dealing with the problems of life.” However, only one out of five long-term homeschoolers felt the same.

Moreover, short- and long-term homeschoolers and respondents who were not homeschooled had similar responses to the question about their perception of their life’s goals and purpose. When asked to provide their level of agreement or disagreement with the statement “My life often seems to lack any clear goals or sense of direction,” 38 percent and 39 percent of short- and long-term homeschoolers agreed, respectively. Non-homeschoolers felt the same as short- and long-term homeschoolers, with 37 percent agreeing that their lives lacked direction. However, 59 percent of medium-term homeschoolers reported a lack of direction in their lives.

When asked about feelings of gratitude, almost all respondents agreed that they “have so much in life to be thankful for.” Even so, there is some evidence of differences among homeschoolers and non-homeschoolers. About nine out of ten non-homeschoolers agreed with the statement, and a similar proportion of medium-term homeschoolers did so. Levels of gratitude were highest for short-term and long-term homeschoolers. Among long-term homeschoolers, 97 percent agreed with the statement, as did 95 percent of short-term homeschoolers.

The CES asked respondents to indicate their level of agreement or disagreement with the statement “When I think about the future, I am positive.” Agreement was highest among long-term homeschoolers, compared to short- and medium-term homeschoolers and non-homeschoolers. Almost all respondents in the long-term homeschool group agreed with the statement, compared to just over three-quarters of other respondents.

In addition to the survey items discussed thus far, the CES used four items from the Patient Health Questionnaire, asking respondents to indicate how frequently over the last seven days they had been bothered by (1) little interest or pleasure in doing things, (2) feeling down, depressed, or hopeless, (3) feeling nervous, anxious, or on edge, and (4) not being able to stop or control worrying. 16 16 K. Kroenke, R.L. Spitzer, and J.B.W. Williams, “The PHQ-9: Validity of a Brief Depression Severity Measure,” Journal of General Internal Medicine 16, no. 9 (2001): 606–13, https://doi.org/10.1046/j.1525-1497.2001.016009606.x Although these scores, averaged for each group, were quite low and do not indicate depression, the likelihood of being free from depressive symptoms was greatest for long-term homeschoolers (aggregate scores of these items, with higher scores indicating a greater frequency of depressed moods, are displayed below).

When asked to respond to items from the Satisfaction with Life Scale, 17 17 E. Diener et al., “The Satisfaction with Life Scale,” Journal of Personality Assessment 49 (2010): 71–75, https://doi.org/10.1207/s15327752jpa4901_13. medium-term homeschoolers showed the least favorable responses, averaging 3.62 out of 6 points on this scale. On the other hand, long-term homeschoolers were noticeably more satisfied with life, averaging 4.38 scale points.

The CES asked respondents questions from the Close Social Relationships domain of the Flourishing Measure developed by the Human Flourishing Program at Harvard University, along with other items designed to measure sense of belonging. 18 18 T.J. VanderWeele, “On the Promotion of Human Flourishing,” Proceedings of the National Academy of Sciences 114, no. 31 (2017): 8148–56, https://doi.org/10.1073/pnas.1702996114. The results were similar to the findings regarding life satisfaction. Medium-term homeschoolers expressed the lowest levels of belonging, while long-term homeschoolers expressed the highest levels. In between were short-term homeschoolers and non-homeschoolers, who expressed similar levels of sense of belonging.

These findings suggest a few important points to consider. First, many of the mental health findings varied depending on years of homeschooling, underscoring the need to refrain from considering homeschoolers as a monolithic group with a single set of outcomes. On many of the indicators, long-term homeschoolers appeared to have the most favorable levels of mental health, compared to all other respondents. Short-term homeschoolers were quite similar to respondents who were never homeschooled. Medium-term homeschoolers exhibited the least favorable levels of mental health of all respondent groups. These findings bring additional nuance to the lower levels of mental health of religious homeschoolers found in the 2011 and 2018 CES administrations. Although it may be true that respondents who were homeschooled for religious reasons experienced adverse effects due to their education—and certainly, individual cases of harm should not be dismissed—one must also be careful not to overgeneralize. Hopefully, a better understanding of the relationship between homeschooling and mental health will enable the further development of homeschooling policies and practices that promote positive mental health.

Although the CES data on their own do not explain the reasons behind all of these patterns, it is possible that more stable educational experiences and possibly, by extension, more stable upbringings are conducive to promoting longer-term mental health. Switching schools may affect students negatively in the short run as students make the adjustment to a new educational setting. 19 19 E. Hanushek, J.F. Kain, and S.G. Rivkin, “Disruption Versus Tiebout Improvement: The Costs and Benefits of Switching Schools,” Journal of Public Economics 88, no. 9–10 (2004): 1721–46, https://www.sciencedirect.com/science/article/abs/pii/S004727270300063X?via%3Dihub. It is possible that switching in and out of homeschooling too often, as medium-term homeschoolers may have done, has some later-life impact on mental health. Equally likely is that medium-term homeschoolers may have switched in and out of homeschooling frequently for reasons that are connected to adverse mental health. In contrast, long-term homeschoolers, short-term homeschoolers, and non-homeschoolers may have had less disruption. Testing these explanations is beyond the scope of the CES, but they are offered for consideration.

Civic Engagement

Another key purpose of education is the formation of civic virtue and preparation for citizenship. As such, a common critique of homeschooling has been that isolating students prevents them from practicing democratic norms and developing dispositions that motivate them to engage with the broader community and fellow citizens in adulthood. 20 20 R. Reich, “The Civic Perils of Homeschooling,” Educational Leadership 59, no. 7 (2002): 56–59, https://ascd.org/el/articles/the-civic-perils-of-homeschooling. The CES examines respondents’ participation in civic behaviors, particularly volunteering and charitable giving.

In general, homeschooled respondents had volunteered at least as much as, if not more than, non-homeschooled respondents, with nearly half of the long-term homeschoolers reporting that they had volunteered within the past twelve months. Just over one-third of short-term homeschoolers reported having done so. Meanwhile, 27 percent of non-homeschoolers and 23 percent of medium-term homeschoolers reported having done so.

Medium- and long-term homeschoolers were most likely to have volunteered regularly. Of respondents who reported that they had volunteered, over half of medium- and long-term homeschoolers reported volunteering at least monthly. Just under half of the non-homeschoolers who had volunteered reported doing so at least monthly, while only one third of short-term homeschoolers who had volunteered reported doing so at least monthly.

Medium-term homeschoolers were most likely to have held a leadership position with a volunteer organization. The difference between medium-term homeschoolers and other respondents was quite large. While only 28 percent of non-homeschoolers and short-term homeschoolers had held a leadership position in volunteer organizations, 54 percent of medium-term homeschoolers had done so. Long-term homeschoolers were least likely to have held a leadership position in a volunteer organization.

As in prior iterations of the CES, respondents were asked to indicate what kinds of organizations they had volunteered with. Respondents who volunteered were presented with a list of organizations, including a religious organization other than their own church or congregation; a social-service organization that is non-religious; a political organization; an environmental organization; an arts, cultural, or sports organization; and a neighborhood association. For each type of organization, respondents said “yes” or “no” to having volunteered in that organization.

Homeschooled respondents were more likely to have volunteered with religious organizations, especially those in the medium- and long-term homeschooler groups. In contrast to the 19 percent of non-homeschoolers who had volunteered with a religious organization, 37 percent and 42 percent of medium- and long-term homeschoolers, respectively, had done so (figure 4.4). In contrast, volunteering in nonreligious organizations was much more common among non-homeschoolers. As shown in figure 4.5, 38 percent of those respondents who had volunteered did so in a non-religious setting, compared to 12 and 24 percent of medium- and long-term homeschoolers, respectively. Long-term homeschoolers were least likely to have volunteered with environmental organizations and sports, arts, or cultural organizations. Medium-term homeschoolers tended not to volunteer with neighborhood and political organizations.

Finally, the CES measured respondents’ charitable giving. Despite having lower incomes, on average, than non-homeschoolers, short- and long-term homeschooled respondents were most likely to have given money to charitable organizations in the past twelve months, with half of short-term homeschoolers and 58 percent of long-term homeschoolers having done so. As a point of comparison, 42 percent of non-homeschoolers had given charitably within the past twelve months. Meanwhile, only 31 percent of medium-term homeschoolers had done so.

When it comes to volunteering and charitable giving, there was clear variation among homeschooled respondents depending on the length of time spent in the sector. Long-term homeschoolers, in particular, appeared to have the strongest civic dispositions, based on the indicators considered by the CES. They were most likely to have engaged in charitable giving and to volunteer, and quite likely to have volunteered at least monthly. The higher likelihood of charitable giving among long-term homeschoolers was also consistent with the lower degree of importance they placed on money as a personal value (figure 2.4). The 2023 CES data do provide some evidence contrary to the claim that homeschooling does not cultivate civic virtue. That said, there are other important forms of civic engagement that the 2023 CES did not examine, such as voting, learning about how government institutions work, and specific civic virtues such as tolerance and appreciation for democratic norms. Bringing empirical evidence to bear on how homeschooled respondents fare on these measures would be of great value.

Faith

Families are motivated to homeschool their children for various reasons. One of the more common reasons among religious parents is a desire to provide their children with moral and religious education. 21 21 J. Murphy, Homeschooling in America: Capturing and Assessing the Movement (Corwin, 2012). Many of these parents also hold the conviction that they, instead of a school, should be the primary directors of their children’s education. 22 22 R. Kunzman, “Homeschooling and Religious Fundamentalism,” International Electronic Journal of Elementary Education 3, no. 1 (2010): 17–28, https://www.iejee.com/index.php/IEJEE/article/view/236. Although not all homeschooling parents decide to homeschool for religious reasons, the question of how homeschoolers fare when it comes to faith during adulthood remains. This section of the report considers aspects of religious belief and experience as well as religious practice among respondents. Measures of religious belief, practice, and values as indicators of faith were adapted from the Spectrum of Spirituality Index developed by Cardus and the Angus Reid Institute. 23 23 A.P.W. Bennett, “Still Christian(?): What Canadian Christians Actually Believe,” Cardus, 2024, https://www.cardus.ca/research/faith-communities/reports/still-christian/.

Religious Belief and Experience

Medium- and long-term homeschoolers were more likely to indicate that they definitely believe in God or a higher power than were short-term homeschoolers, and they were even more likely to do so than non-homeschoolers. Forty-one percent of non-homeschoolers stated that they definitely believe in God or a higher power. The proportion of short-term homeschoolers who shared this view was 56 percent. About 61 percent and 72 percent of medium-term and long-term homeschoolers, respectively, indicated belief in God or a higher power. In other words, the proportion of respondents who believed in God or a higher power increased with the number of years spent in the homeschool sector.

Respondents were also asked about their belief in an afterlife. One out of every three non-homeschoolers expressed this belief, whereas 47 percent of short-term and 46 percent of medium-term homeschoolers did so. Meanwhile, nearly two-thirds of long-term homeschoolers indicated that they definitely believed in an afterlife—almost double the proportion of non-homeschoolers who shared this view.

Consistent with frequency of belief in God and an afterlife, frequency of religious experience tended to increase as duration of homeschooling increased. Only 31 percent of non-homeschoolers reported experiencing God’s presence at least weekly. While 40 percent of short-term homeschoolers indicated having this experience at least weekly, 55 percent of medium-term homeschoolers indicated the same. Long-term homeschoolers were most likely to say they felt God’s presence at least weekly, at about 60 percent.

Religious Practice

Findings about religious practice were similar to those about religious belief. Medium-term homeschoolers appeared to engage in practices such as prayer, reading a sacred text, or attending a religious service almost as frequently as long-term homeschoolers did. Both medium- and long-term homeschoolers engaged in these practices more often than short-term homeschoolers and non-homeschoolers. Almost two out of every three medium- and long-term homeschoolers prayed at least weekly. Slightly less than half of short-term homeschoolers prayed as often, and a still smaller percentage (37 percent) of non-homeschoolers prayed at least weekly.

In general, all respondents were less likely to read a sacred text and to attend religious services than to pray. However, the decrease in the likelihood of reading a sacred text was less pronounced among medium- and long-term homeschoolers. Almost 60 percent of medium- and long-term homeschoolers did so, compared to 27 percent of short-term homeschoolers and 18 percent of non-homeschoolers. Just less than one-quarter of short-term homeschoolers attended religious services weekly, as did about one-fifth of non-homeschoolers. In contrast, 45 percent of medium-term homeschoolers attended religious services at least weekly, and a still higher percentage of long-term homeschoolers (54 percent) did so.

The frequency with which respondents engage in religious practices reflected their self-reports of how important the value of religion was to them personally. Thirty-seven percent of short-term homeschoolers and 21 percent of non-homeschoolers reported that religion was very important to them. These proportions contrast sharply with the over half of medium- and long-term homeschoolers who reported that religion was very important to them.

Similarly, medium- and long-term homeschoolers were more likely to agree that it is important for parents to teach their children about religion. Seventy-three percent of long-term homeschoolers agreed with that statement, as did 68 percent of medium-term homeschoolers. Meanwhile, 60 percent of short-term homeschoolers agreed, as did 45 percent of non-homeschoolers.

Family Formation

Perceptions of homeschooled children—whether they are from conservative or religious, or from progressive or secular backgrounds—are often driven by popular media. 24 24 D.C. Hauseman, “‘Nerdy Know-It-Alls’ and ‘Paranoid Parents’: Images of Alternative Learning in Films and Television Programs,” Journal of Unschooling and Alternative Learning 5, no. 9 (2011): 1–17, https://jual.nipissingu.ca/wp-content/uploads/sites/25/2014/06/v5191.pdf. Regarding the conservative or religious group of homeschoolers, some scholars have theorized that parents from these families inculcate traditional views of marriage, gender, childbearing, and family. Indeed, data from the 2011 and 2014 CES administrations indicated that religious respondents who were homeschooled were more likely to endorse those traditional views. 25 25 D. Sikkink and S. Skiles, “Homeschooling and Young Adult Outcomes: Evidence from the 2011 and 2014 Cardus Education Survey,” Cardus, 2015, https://cardus.ca/research/education/research-report/homeschooling-and-young-adult-outcomes-evidence-from-the-2011-and-2014-cardus-education-survey/. Nonetheless, other studies suggest a more complicated picture. For instance, a separate analysis of the 2014 CES data found that the ages at which one gets married and has children were not different for respondents primarily homeschooled in high school and respondents primarily educated in traditional public high schools. 26 26 J.E. Uecker and J.P. Hill, “Religious Schools, Home Schools, and the Timing of First Marriage and First Birth,” Review of Religious Research 56, no. 2 (2014): 189–218, https://doi.org/10.1007/s13644-014-0150-9. Such a finding undermines the popular perceptions and common theories about conservative or religious homeschoolers. Nonetheless, the study authors acknowledge that they were unable to distinguish conservative or religious homeschoolers from their progressive counterparts in the data.

Data to distinguish between these two groups of homeschoolers remain unavailable in the 2023 CES. However, data from this administration do allow for describing marriage and childbearing outcomes for respondents who were homeschooled for different lengths of time.

Marriage

When marriage rates by duration of homeschooling are considered, respondents who were homeschooled for between three and seven years were most similar to respondents who were never homeschooled. Fifty-nine percent of medium-term homeschoolers had never married, compared to 56 percent of non-homeschoolers. In contrast, 41 percent of respondents who were homeschooled for one or two years and 35 percent who were homeschooled for eight to thirteen years had never married.

To show these results another way, the data can be presented in terms of the proportion of respondents who were currently married, by years of homeschooling. About two out of every three long-term homeschoolers and about half of short-term homeschoolers were currently married. One out of every three medium-term homeschoolers were currently married, a rate that is slightly lower than the 40 percent of non-homeschoolers who were currently married.

The above results are consistent with respondents’ self-reports of the importance of marriage. When asked how important marriage was personally, more than half of long-term homeschoolers said it was very important. In contrast, just over one third of the other respondents provided the same answer.

 

A key question remains, however, whether respondents who have married stay married. Almost one tenth of non-homeschoolers who had been married were currently divorced or separated. The divorced-or-separated rate for long-term homeschoolers was about half that rate. On the other hand, divorced-or-separated rates for short-term and medium-term homeschoolers were higher, at 17 and 13 percent, respectively.

Notably, the 2011 CES found higher divorce rates among religious homeschoolers relative to other respondents. The findings regarding divorce in the 2023 data suggest that the higher rates of divorce among homeschoolers occurred among short- and medium-term homeschoolers. Stable marriages appeared to be quite common among long-term homeschoolers.

Children

The CES provides insight into another aspect of family formation, namely, having children. When asked how important having children was as a personal value, 36 percent of long-term homeschoolers indicated that it was very important. The proportion of medium-term homeschoolers, short-term homeschoolers, and non-homeschoolers who shared the same view was seven to ten percentage points lower than that of long-term homeschoolers.

Furthermore, long-term homeschoolers were more likely than other respondents to have had children, though the likelihood of having children for short-term homeschoolers and non-homeschoolers was not much lower. Forty-six percent of long-term homeschoolers have had children, compared to 40 percent of short-term homeschoolers and 42 percent of non-homeschoolers. Medium-term homeschoolers, in contrast, were much less likely to have had children. Only 29 percent of those respondents had children.

The CES also examined the number of children that respondents with children currently had. Respondents with children and who were never homeschooled had on average 0.84 children, while short-term and medium-term homeschoolers with children had on average 0.75 and 0.61 children, respectively. Long-term homeschoolers, however, appeared to have more children than other respondents. Conditional on having children, they had about 1.23 children on average. Thus long-term homeschoolers were not only more likely to have children but also had a larger number of children.

Conclusion

Studies of homeschoolers have identified many dimensions of diversity among homeschooling families, such as their pedagogical approach, reasons for homeschooling, and duration of time spent homeschooling. Any two homeschoolers may have vastly different later-life outcomes because of these differences. However, researchers have not always been able to separately compare educational outcomes for different types of homeschoolers, in part due to the challenges of collecting representative samples. Using data from the 2023 CES, this report compares respondents who have been homeschooled for various durations of time for their primary and secondary education on a range of outcomes, including educational attainment, preparation for employment, mental health, civic engagement, faith, and family formation. This report differs from prior CES reports, which focus on comparing all homeschooled respondents to respondents from traditional public, Catholic, Protestant, and nonreligious private schools.

As demonstrated in this report, respondents who were homeschooled for a short period of time often looked most like respondents who were never homeschooled, while respondents who were homeschooled for longer periods of time often had markedly different outcomes from those who were homeschooled only briefly. Furthermore, outcome differences were not always linear with respect to the amount of time a respondent was homeschooled. Although faith outcomes, for instance, tended to have a linear relationship with the amount of time a respondent was homeschooled, the likelihood of being married, and various mental health outcomes revealed a U-shaped pattern, in which long-term and short-term homeschoolers were more alike, compared to medium-term homeschoolers.

It is important to reiterate that the associations between outcomes and amount of time someone was homeschooled should not be interpreted as causal. The analytic methods used in this report cannot account for every potential factor that leads to the differences observed among homeschoolers. Likewise, the findings in this report cannot speak to why homeschooling for different lengths of time might lead to different outcomes. For instance, what is it, if anything, about long-term homeschooling that leads to greater levels of religious engagement? Is it an effect of long-term homeschooling, or do families that tend to engage in long-term homeschooling engage in other practices that promote religious engagement? Similarly, are some outcome differences explained directly by homeschooling, or by homeschooling’s indirect effects on an intermediate outcome? For example, is the greater likelihood of marriage among long-term homeschoolers due to homeschooling itself, or to greater levels of religiosity that arise from homeschooling? Such caveats aside, describing these differences among the homeschooled population is necessary for better understanding that population and identifying important lingering questions.

The spirit of the CES has always been to examine the ways in which different educational sectors and educational approaches contribute to the common good. Although each sector and approach has its particular weaknesses, each also has particular strengths. Homeschooling is no exception, where encouraging findings were observed such as higher rates of volunteering, charitable giving, religiosity, life satisfaction, and marital stability among respondents who were homeschooled for a large majority of their primary and secondary education. On the other hand, the CES also found lower rates of educational attainment among medium- and long-term homeschoolers, a finding that will hopefully prompt not only additional study but also discussion of implications for practice within the homeschooling community. Meanwhile, policy proposals aimed at addressing access to homeschooling and improving the practice need to take into account the many distinctions within the larger homeschooling population that were uncovered in this report. Policy considerations for short-term homeschoolers may be different from those for long-term homeschoolers.

It is important to understand that these findings provide the outcomes for respondents who were homeschooled one to three decades ago. If the amount of time children were homeschooled varied then, it should certainly vary more now that homeschooling participation has increased in number and diversity. The expansion of Education Savings Accounts (ESA) programs across the United States, together with the proliferation of educational entrepreneurs who are establishing new forms or models of education, are providing families with access to more educational options than ever before in recent history. How homeschooled children growing up in this new educational environment fare later in life, and how they contribute to the common good within a pluralistic society, will be important questions to consider in the years to come.

References

Bennett, A.P.W. “Still Christian(?): What Canadian Christians Actually Believe.” Cardus, 2024. https://www.cardus.ca/research/faith-communities/reports/still-christian/.

Caplan, B. The Case Against Education. Princeton University Press, 2019.

“Cardus Education Survey 2018: From the Classroom to the Workplace.” Cardus, 2019. https://cardus.ca/research/education/research-report/cardus-education-survey-2018-from-the-classroom-to-the-workplace/.

“Cardus Education Survey 2018: Spiritual Strength, Faithful Formation.” Cardus, 2019. https://cardus.ca/research/education/research-report/2018-us-cardus-education-survey-spiritual-strength-faithful-formation/.

Dee, T. “Where the Kids Went: Nonpublic Schooling and Demographic Change During the Pandemic Exodus from Public Schools.” Urban Institute, 2023. https://www.urban.org/research/publication/where-kids-went-nonpublic-schooling-and-demographic-change-during-pandemic/.

Diener, E., R.A. Emmons, R.J. Larsen, and S. Griffin. “The Satisfaction with Life Scale.” Journal of Personality Assessment 49 (2010): 71–75. https://doi.org/10.1207/s15327752jpa4901_13.

Goldin, C., and L.F. Katz. The Race Between Education and Technology. Harvard University Press, 2008.

Green-Hennessy, S., and E.C. Mariotti. “The Decision to Homeschool: Potential Factors Influencing Reactive Homeschooling Practice.” Educational Review 75, no. 4 (2023): 617-36. https://doi.org/10.1080/00131911.2021.1947196.

Guzman, G., and M. Kollar. “Income in the United States: 2023.” United States Census Bureau (Sept. 2024). https://www2.census.gov/library/publications/2024/demo/p60-282.pdf.

Hamlin, D., and A. Cheng. “Homeschooling, Perceived Social Isolation, and Life Trajectories: An Analysis of Formerly Homeschooled Adults.” Journal of School Choice 16, no. 2 (2022): 332–59. https://doi.org/10.1080/15582159.2022.2028338.

Hanushek, E., J.F. Kain, and S.G. Rivkin. “Disruption Versus Tiebout Improvement: The Costs and Benefits of Switching Schools.” Journal of Public Economics 88, no. 9–10 (2004): 1721–46. https://www.sciencedirect.com/science/article/abs/pii/S004727270300063X?via%3Dihub.

Hauseman, D.C. “‘Nerdy Know-It-Alls’ and ‘Paranoid Parents’: Images of Alternative Learning in Films and Television Programs.” Journal of Unschooling and Alternative Learning 5, no. 9 (2011): 1–17. https://jual.nipissingu.ca/wp-content/uploads/sites/25/2014/06/v5191.pdf.

Hudson, L., T. Kaatz, D. Battle, L.J. Hall, S. Bahr, and S. Eyster. “2019 Homeschooling and Full-Time Virtual Education Rates.” National Center for Education Statistics, US Department of Education. 2023. https://nces.ed.gov/pubs2023//2023101.pdf.

Kroenke, K., R.L. Spitzer, and J.B.W. Williams. “The PHQ-9: Validity of a Brief Depression Severity Measure.” Journal of General Internal Medicine 16, no. 9 (2001): 606–13. https://doi.org/10.1046/j.1525-1497.2001.016009606.x.

Kunzman, R. “Homeschooling and Religious Fundamentalism.” International Electronic Journal of Elementary Education 3, no. 1 (2010): 17–28. https://www.iejee.com/index.php/IEJEE/article/view/236.

Kunzman, R., and M. Gaither. “Homeschooling: An Updated Comprehensive Survey of the Research.” Other Education: The Journal of Educational Alternatives 9, no. 1 (2020): 253–336. https://www.researchgate.net/publication/374195331_Homeschooling_An_Updated_Comprehensive_Survey_of_the_Research.

Murphy, J. Homeschooling in America: Capturing and Assessing the Movement. Corwin, 2012.

National Institute of Mental Health. Youth Suicide Rates Increased During the COVID-19 Pandemic. Research Highlight. 2023. https://www.nimh.nih.gov/news/science-news/2023/youth-suicide-rates-increased-during-the-covid-19-pandemic.

Pennings, R., and K. Wiens. “Cardus Education Survey: Do the Motivations for Private Religious Catholic and Protestant Schooling in North America Align with Graduate Outcomes?” Cardus, 2011. https://www.cardus.ca/wp-content/uploads/2023/05/Cardus-Education-Survey-Phase-I-report.pdf.

Reeves, R.V. Of Boys and Men: Why the Modern Male Is Struggling, Why It Matters, and What to Do About It. Brookings Institution Press, 2022.

Reich, R. “The Civic Perils of Homeschooling.” Educational Leadership 59, no. 7 (2002): 56–59. https://ascd.org/el/articles/the-civic-perils-of-homeschooling.

Roeloffs, M. “Teen Suicide Plummeted During Covid-19 School Closure, New Study Finds.” Forbes, July 19, 2023. https://forbes.com/sites/maryroeloffs/2023/07/19/teen-suicide-plummeted-during-covid-19-school-closures-new-study-finds/.

Sikkink, D., and S. Skiles. “Homeschooling and Young Adult Outcomes: Evidence from the 2011 and 2014 Cardus Education Survey.” Cardus, 2015. https://www.cardus.ca/research/education/research-report/homeschooling-and-young-adult-outcomes-evidence-from-the-2011-and-2014-cardus-education-survey/.

Swaner, L.E., A. Cheng, and J. Eckert. “School-Sector Influence on Graduate Outcomes and Flourishing: Findings from the 2023 Cardus Education Survey.” Cardus, 2024. https://www.cardus.ca/research/education/reports/school-sector-influence-on-graduate-outcomes-and-flourishing/.

Uecker, J.E., and J.P. Hill. “Religious Schools, Home Schools, and the Timing of First Marriage and First Birth.” Review of Religious Research 56, no. 2 (2014): 189–218. https://doi.org/10.1007/s13644-014-0150-9.

US Centers for Disease Control and Prevention. Youth Risk Behavior Survey Data Summary and Trends Report: 2011–2021. 2022. https://www.cdc.gov/yrbs/dstr/pdf/YRBS_Data-Summary-Trends_Report2023_508.pdf.

Valiente, C., T.L. Spinrad, B. Ray, N. Eisenberg, and A. Ruof. “Homeschooling: What Do We Know and What Do We Need to Learn?” Child Development Perspectives 16, no. 1 (2022): 48–53. http://dx.doi.org/10.1111/cdep.12441.

VanderWeele, T.J. “On the Promotion of Human Flourishing.” Proceedings of the National Academy of Sciences 114, no. 31 (2017): 8148–56. https://doi.org/10.1073/pnas.1702996114.