Child Abuse in Public Schooling, Private Schooling, and Homeschooling: A New Study and Past Research

The Context

One new study and a review of previous work offer new insights on abuse and neglect of children in public schools, private schools, and homeschooling. Various scholars and others have claimed that either homeschool children are at more risk than other children of experiencing child abuse or neglect or that whatever the risk is to homeschool children, the government should control (regulate) parent-led home-based private education – homeschooling – more (e.g., Bartholet, 2020).[i] These claims have been made, however, without any empirical evidence that home-educated are at any more risk than public school and private (independent) school children.

At the same time, there is limited evidence that home-educated students are less likely to be abused than are public school and private school children (e.g., Ray, 2015; Williams, 2018).[ii] Also, one study revealed that whatever the apparently low rate of abuse and neglect of homeschool there is, more government control over homeschooling is not associated with lower rates of abuse or neglect (Ray, 2018).[iii]

On this topic, the recent series of weekly presentations entitled, Post-Pandemic Future of Homeschooling,” was sponsored by the Harvard Kennedy School, Taubman Center for State and Local Government, and included a session in June 2021 entitled, “Is child abuse greater at school or homeschool?” The session description also included the questions, “What do we know about the incidence of child and sexual abuse that occurs in schools and in homeschool households?”[iv]

The panelists were Dr. Charol Shakeshaft, professor, Virginia Commonwealth University, and Dr. Angela Dills, professor, Western Carolina University, while the commentator was Dr. Martin West, professor, Harvard Graduate School of Education. Dr. Shakeshaft’s presentation was entitled, “How widespread is school employee sex abuse?” on public and independent schools, and researcher Dills presented “Homeschooling and child safety: Are kids safer at home?”

The Research

Shakeshaft summarized that 9.6 percent of all school students in grades 8 to 11 report contact and/or non-contact educator sexual misconduct that was unwanted. Some 8.7 percent report only non-contact sexual misconduct (e.g., lewd behavior or transmittal of sexual photos via the Internet by school employees, taking pictures of the students) while 6.7 percent report contact (i.e., physical contact) sexual misconduct.

This means, she said, “… that [of today’s school students] about 5.6 million public and independent students are subject to sexual misconduct by employees of a school sometime between kindergarten and 12th grade.” Dr. Shakeshaft also pointed out that these statistics are based on older studies (e.g., 2002 and older). There have been no national or regional studies done to update these statistics. There is a data lack in the public and independent schools, she said.

“There’s no reason to believe from the number of reports that have happened, in terms of reports to police and others, that sexual abuse by school employees has decreased. On the other hand, we just don’t know because we don’t have that data.”

Dr. Shakeshaft then made some conjectures.

    And what I would say is that based on what I know about schools and from what I’ve read about homeschooling and what happens it is highly unlikely that there is a higher incidence of sexual abuse of kids in the homeschooling world than in the public school world. But again, I can’t substantiate that through great datasets because we don’t have those data. But my research on schools and predators and cases over the years have looked very closely at the patterns and how it happens. What I will say, for instance, is that predators in schools, many of them have children and it’s very rare that they sexually abuse their own children. So that is something that turns up in the data that we find even when that’s investigated.”

Dr. Dills, an economist, cited some of the research noted by Shakeshaft and then went on to report on her new original study. She was trying to estimate a causal effect between homeschooling laws and child safety (or maltreatment). “The approach I’m going to take is I’m going to rely on states’ adoption of homeschool rights legislation so laws that were primarily adopted in the 1980s and 1990s.” One study found that these laws, ones that made it easier for families to homeschool, increase the prevalence of homeschooling of children ages 5 to 8 years old. She decided to estimate the effects of these laws (“homeschool rights legislation”) on measures of child safety.

Dills’ dependent variables were three. The first was violent deaths of 5-9- and 10-14-year-olds during 1979-1998, based on data from the Centers for Disease Control and Prevention (CDC). The other two variables were on child maltreatment fatalities and cases during 1990-2008, based on data from the National Data Archive on Child Abuse and Neglect (NDACAN). She noted that “… most cases of child maltreatment are of neglect, and the bar for what constitutes neglect may change over time and across states …” Dr. Dills explained the complexities of these measure and limitations to using them as indicators of child safety related to homeschooling.

Dr. Dills presented her findings this way: “To summarize my results, what I find is fairly consistent across a variety of specifications. Most of the estimated effects are statistically insignificant.”

She went on to say the following:

    [I find] “… very slightly higher rates of violent deaths following the adoption of homeschool rights, although again, those differences are not statistically significant. Estimates using the child maltreatment data are somewhat contradictory. So I show a significant decline in fatalities due to child maltreatment and a significant increase in reported child maltreatment cases. However, neither of these results, neither that decline in fatalities nor the increase in child maltreatment, are robust to some standard empirical checks” [e.g., event study, Goodman-Bacon Decomposition]. ..… [And] effectively, what I find with those are no significant effects in the event study and an average of roughly zero effects when I decompose it in the different kinds of toolmakers. I also looked specifically at whether there’s a change in reporter type following homeschool rights. And one of the things that NDACAN provides is who reported the suspected abuse. And I find no statistically significant or large changes in reporter type.

Dr. Dills’ summary statement was as follows: “Overall, the picture is one of little to no effect of homeschool rights on child safety.”

Concluding Thoughts

Dr. Dills offered some concluding opinions, as follows:

    Whether proposals to restrict homeschooling are likely to improve child safety depends on at least two things, whether homeschooling reduces child safety and whether the stated requirements actually improve child safety. My estimates suggest that there’s little causal evidence for the former. Whether additional regulation of homeschooling is likely to improve child safety is also unclear. Existing laws prohibit child maltreatment as well as assault, rape, and homicide by non-caregivers. And relying on school employees as mandatory reporters when in many cases, they simply aren’t reporting their own colleagues for their sexual misconduct seems fairly ineffective. And evidence on, for example, home visits for infants shows somewhat mixed effects on outcomes of child abuse and neglect. We may have other concerns with oversight of homeschooling, such as concerns for parental rights, disparate enforcement of laws against families of color or non-mainstream religious groups, effects of homeschooling on other outcomes, and the like are all going to affect any cost-benefit analysis of additional regulation.

Moderator Dr. West asked of the presenters, in a direct way, whether the states should control (regulate) homeschooling more to prevent homeschool students from being abused? Dr. Shakeshaft replied as follows:

    I can’t see where regulation [of homeschooling] is going to help much [in the area of sexual abuse of students and kids]. It hasn’t helped much in public and independent schools, as far as we can tell. …..

    I don’t see that we have a great track record in public schools of regulation that indicates that that would necessarily be an appropriate way to go about reducing harm to kids. What I would say is that in schools everybody is fingerprinted, they have background checks, they do those things. None of those background checks ever turn up, among professionals in schools, really hardly ever turn up somebody who has been accused or been found to sexually abused a kid. So while they’re [regulations] there, they aren’t a regulation that is really helping us to weed out people who might abuse or harm students.

Dr. Dills reflected back on her study that considered whether rules (laws) that made it easier to homeschool one’s children was associated with or caused more child maltreatment. She explained the following:

    So if really what you were doing was abusing your children and sort of outside the letter of the law and going to do that [abuse] anyways, then that probably was already happening prior to these homeschool rights legislation being passed. And those are not the children and families being affected.

One of the series’ organizers, Dr. Paul Peterson asked the moderator and panelists why the government is not collecting data on public and independent school employee maltreatment of students. Why don’t we have the needed information? Dr. Shakeshaft replied as follows:

    I think, A, we don’t value children. B, I think we value adults more than children. ….. What I would say is that we have a lot of mechanisms and in 2004, I did a report to the U.S. Senate ….. and we could add questions [to many already-existing government data collections ….. [and] there are lots of ways to do this. ….. But what I’m saying is that my folks [of similar political perspectives to myself] are beholden to teachers’ unions as well as other things. And so this isn’t maybe their favorite topic.

In sum, Dr. Shakeshaft reported that more and updated data are needed on the extent to which public school and independent school employees are maltreating (e.g., sexually, other ways) students. She also voiced her view that more government control over homeschooling would not likely reduce whatever amount of abuse and neglect is being done by homeschool parents. Dr. Dills’ pointed to her new research findings that most of the estimated effects between increasing homeschool rights laws and child safety were statistically insignificant.

Keywords, Categories, Tags:

Public school, independent school, private school, homeschooling, home education, child abuse, child neglect, law, legal, research, statistics


[i] Bartholet, Elizabeth. Homeschooling: Parent rights absolutism vs. child rights to education and protection, Arizona Law Review, 62(1) (2020), retrieved April 30, 2020 from

[ii] Ray, Brian D. (2015, January 30). Gen2 Survey: A spiritual and educational survey on Christian millennials. Salem, Oregon: National Home Education Research Institute. Retrieved October 28, 2019 from; Williams, Rodger (2018). Research evidence indicates homeschoolers have a lower child abuse rate than average. Retrieved July 7, 2021 from

[iii] Ray, Brian D. (2018). The relationship between the degree of state regulation of homeschooling and the abuse of homeschool children (Students). Retrieved June 3, 2021 from

[iv] Retrieved description June 25, 2021 from

and video-recording from