Home
This is the second in a series of blog posts responding to an article that calls for a presumptive ban on homeschooling.
A LOOK AT THE DATA
There is a saying that “hard cases make bad law.” This means we should not make broad policy decisions based on a few horrific cases. What kind of evidence should we consider when making dramatic changes in public policy that will affect millions of people? Before looking for answers, it is important to determine if there is, in fact, a problem. It is helpful to look at scientific evidence and data to help us determine if there are trends that need to be addressed.
What does the data show? Analysis of the 2016 Federal Child Abuse statistics show that out of 1750 child abuse deaths, 1479 (84.5%) involved preschool children (ages 0-5). Two hundred sixty six involved children who were not homeschooled (15.2%), and there were 5 homeschool abuse deaths (0.3%). 1 The fact that homeschoolers are safer than public school students was echoed in the Gen2 Survey that concluded that homeschoolers were 257% less likely to be sexually abused. 2
Elizabeth Bartholet, a Harvard law professor, has called for a presumptive ban on homeschooling based in part on the assertion that homeschoolers are at “significant risk for abuse.” On what basis does she make this claim? She cites law student Carmen Green’s quote in a Law Journal article that “a substantial amount of anecdotal evidence show[s] that some abusive parents, who have no intention of educating their children, have taken advantage of lax homeschooling laws to hide their children… .” 3 Carmen Green sits on the board of a pro-regulation group, Coalition for Responsible Home Education (CRHE).
The law professor’s article is filled with references to anecdotal evidence, as well as citations to phone conversations and unpublished articles. Anecdotal evidence is a nonsystematic collection of personal stories unsupported by objective evidence. In a courtroom, anecdotal evidence is hearsay, which is generally inadmissible. Why is it inadmissible? Without an opportunity to cross-examine the source of information, it cannot be considered reliable.
Another unsubstantiated assertion, repeated throughout the article, is that homeschoolers are “isolated.” The author speaks of isolation from the community. What about homeschoolers who attend co-op classes, take independent lessons, and volunteer in their community? Is the isolation from mandatory reporters because they do not sit in the government schools? What about the mandatory reporters they encounter throughout their lives: attorneys, doctors and others who see them on a regular basis? Some states make every person a mandatory reporter so that anyone a homeschooler comes in contact with qualifies as someone who could report suspected abuse. Even those who are not mandated by state law to report abuse may report abuse.
Faced with the lack of evidence, the Ms Bartholet states there is “no way to know” the exact scope of the child maltreatment problem” because not every homeschooler is registered. As another writer pointed out when discussing the illogic of this argument, “No one knows if these parents are beating and starving their children. No one knows if the family is keeping vicious man-eating tigers in their homes. No one knows if the children are at risk of being abducted by aliens. (We could go on ad nauseam, getting more absurd as we go, but you get the idea.)”4
When Bartholet does cite a report, it is to an unscientific paper by the CT Child Advocate. The supplemental report published by the Connecticut Child Advocate 5 fails to make the case for homeschool abuse because:
1. It included only 6 out of 170 school districts in Connecticut. There is no evidence of which towns those were, or why they were selected, aside from the assertion from the Advocate that she was attempting to select a random sample.
2. The reports reviewed spanned a three year period. Without knowing what years were included and why, there is no way to tell if those were typical.
3. The raw data for the report has never been released to the public, so there is no way to analyze the accuracy or sufficiency of the data.
4. The statistics included both substantiated and unsubstantiated reports, which means it included false or inaccurate reports.
Without the information noted above, the report tells us nothing worthwhile to help guide policy. More telling is that DCF was aware of these families. Despite being listed as “homeschoolers”, the state child protection agency knew of the families and was following them. In other words, they weren’t hidden.
Ms Bartholet also implies that Matthew Tirado, upon whose death the Child Advocate reported, was homeschooled. Yet the Child Advocate’s first report made it clear that Matthew was an enrolled (although absent) student at Oak Hill School. He was not homeschooled. He was truant.6 The fact that his mother was “able” to withdraw his sister to homeschool her is not an indication that homeschool laws are lacking, but that DCF and the other agencies tasked with helping the Tirado children failed. 7
In addition to the Connecticut report, Bartholet cites a “similar study in a different state, to date unpublished…” Once again, we are asked to accept authority based on an unnamed study, in an unnamed state, that remains unpublished. But wait! another writing is cited: Invisible Children, a study by Meggan Goodpasture, et al, in 2013. That article looked at four instances of child abuse. One child had been previously reported to CPS, one child was not of school age, and 2 were reported by neighbors to be homeschooled, although there was no homeschooling report on file. A review of 4 cases of abuse are not statistically significant when considering regulation of 2.3 million homeschool students.
A third cited study discusses child torture. This is the Knox study that has been touted by the Coalition for Responsible Home Education for years. The descriptions of abuse are certainly horrific. Is the study pertinent to a discussion of the rates of abuse in the homeschooling community? Let’s take a look at what the authors explained regarding how they approached the study [their methodology]:
“This series and paper is limited in that it is a select and by no means, inclusive series, of abuse cases. They have been chosen to be illustrative of the phenomenon of torture, but cannot be considered a consecutive case series for statistical analysis.”8
Unlike the CT Child Advocate, the researchers acknowledge the limited use of their study to provide a scholarly definition of child torture.
Having exhausted the studies preferred by anti-homeschoolers, we are led back to anecdotal stories that include the Turpin and Hart cases. In these families, the children were seen by mandatory reporters who failed to spot or report abuse. Once detected, they followed the same pattern of moving or fleeing. 9
The final references are to information collected via the internet by Coalition for Responsible Home Education(CRHE) and its sister organizations, HIC (Homeschooling’s Invisible Children) and HARO (Homeschool Alumni Reaching Out). ] CRHE has stated that their “database is not a complete catalogue of all homeschool child abuse fatalities.” Despite that acknowledgement, they then feel free to conclude that there is a “higher rate of death due to abuse or neglect among homeschooled students [than children overall].” 10 Their database is, in fact, a mix of cases that include genuine homeschooled students as well as truants who may or may not masquerade as homeschoolers. HARO’s conclusion of 42% of homeschool graduates who were abused is based on a biased group of self-selected homeschoolers. 11
Bartholet’s conclusion that there is abuse within the homeschooling community that requires a presumptive ban is not supported by reliable statistical evidence. Her frequent reliance on anecdotal stories, phone conversations and unpublished articles to support her position weakens her argument. The published articles cited do not include statistical studies with significant data, but focus on the hard cases that make bad law. Studies that have reviewed reliable data indicate that homeschoolers are at a lower risk for abuse.
1, https://homeschoolingbackgrounder.com/child-abuse-neglect-fatalities-2016/
2. http://www.nheri.org/Gen2SurveyASpiritualandEducationalSurveyonChristianMillennials.pdf
3. Elizabeth Bartholet, Homeschooling: Parent Rights Absolutism vs Child Rights to Education & Protection. (October 7, 2019). Forthcoming, Arizona Law Review, Vol. 62, No. 1, 2020 ; Harvard Public Law Working Paper No. 19-2 Available at SSRN: https://ssrn.com/abstract=3391331 or http://dx.doi.org/10.2139/ssrn.3391331
4. https://pjmedia.com/trending/why-is-fox-news-using-left-wing-talking-points-to-take-cheap-shots-at-homeschoolers/
5. http://www.ct.gov/oca/lib/oca/OCA.Memo.Homeschooling.4.22018.pdf
6. http://www.ct.gov/oca/lib/oca/MT_final_12_12_2017.pdf
7. https://spark.adobe.com/video/53J4ZYVwgcA6w
8. Knox, B.L., Starling, S.P., Feldman, K.W. et al. Journ Child Adol Trauma,(2014) 7:37, available at
https://doi.org/10.1007/s40653-014-0009-9
9. EG, Turpin home in Texas discovered to be filled with excrement and police notified, but no follow up with CPS. https://www.independent.co.uk/news/world/americas/turpin-family-latest-updates-perris-california-couple-david-louise-trial-children-torture-dead-dogs-a8172441.html.
Former friend reported concerns to child welfare officials without response. https://www.foxnews.com/us/surveillance-video-seen-as-possible-clue-to-hart-familys-route-before-california-cliff-crash
10. Bartholet at 15.
11. https://homeschoolingbackgrounder.com/research-indicates-homeschoolers-lower-child-abuse-rate-than-average/