January 25, 2018
On January 14, 2018, news broke that the thirteen children of David and Louise Turpin, ages 2 to 29, had been imprisoned for years in a California home-turned-prison under the guise of homeschooling. Nine days later, Brian Ray published an article on his website purporting to find that homeschooled children experience abuse at a lower rate than do other students. Ray argues that no laws should be changed to protect homeschooled children like the Turpins’ from the prolonged torture they experienced. However, to reach this conclusion, Ray both misrepresents evidence and leaves out studies that contradict his conclusions.
Before we begin, a word about Ray. Brian Ray is a former homeschool father who runs the National Home Education Research Institute, which effectively acts as a research arm for the Home School Legal Defense Association, an organization that opposes oversight of homeschooling. Ray has been widely criticized by researchers for relying too heavily on his own research and at the same time wildly misrepresenting his own findings.
In this piece, Ray sets out to examine whether child abuse rates among homeschooled students compare to child abuse rates among other students. “A recent thorough search of available literature has resulted in identifying only three published reports relevant to whether the abuse of public school and private school children happens at a lower, similar, or higher rate than for homeschool children,” he writes. What are these three studies?
The first study Ray refers to is one conducted by Rodger Williams that Ray calls “original research.” The second study is Ray’s own survey of homeschool graduates, a study that did not use a random sample (meaning that it cannot be generalized to all homeschoolers) and asked only one question related to abuse. Ray refers to the third study, CRHE’s analysis of cases from our Homeschooling’s Invisible Children (HIC) database, as “a report using publicly available data.” But far from conducting original research, Williams actually relied on the incomplete numbers from CRHE’s analysis. Bafflingly, Ray keeps this information from his readers, presenting the two as entirely independent studies.
Ray writes that Williams “used information from one organization that promotes more government regulation of homeschooling.” After devoting ten paragraphs to discussing Williams’ analysis, Ray spends a single paragraph discussing CRHE’s treatment. He does not identify CRHE as the organization Williams drew his data from. This is important because by leaving out the fact that Williams drew on CRHE’s data, Ray leaves the impression that Williams’ analysis represents a third independent study with its own data source.
Either Ray was careless when putting together his article and did not realize that the two data sources were the same, or he left out the source of Williams’ analysis intentionally to create the impression of multiple independent studies with their own separate data sources.
In 2013, we began collecting cases of severe and fatal child abuse in homeschool settings and posting them in a database, Homeschooling’s Invisible Children. In 2015, we tallied the homeschool fatalities in our database for each year. We then used data from the Child Welfare Information Gateway and the National Center for Education Statistics to determine the rate of child fatalities among children ages 6-17 and create an estimate of how many children were homeschooled each year. We used this data to compare child fatality rates.
We found that there were slightly more fatalities in our database than would be expected given the overall child fatality rate for their age group (84 versus 73). We then ran statistical significance testing to find out whether we could eliminate random variation as a possible cause for the higher number of fatalities we were finding. We found that the difference between the two numbers was not large enough to eliminate the possibility that our higher homeschool fatality number could be the result of random variation. Our 2015 analysis relied on figures from 2000 to 2012; we have added a number of cases since then.
As we outlined in our article, our database is not a complete catalogue of all homeschool child abuse fatalities. This is because there is no official data that provides these numbers. Statistics on the rate of child abuse are typically compiled by a state’s social services division and draw on its case files. To our knowledge, no state currently includes information on a child’s educational status when tabulating child abuse statistics. Instead of relying on a nonexistent official tally, our database is created using internet searches and is entirely dependent on what cases hit the news, and which news articles remain on news websites (many newspapers archive old pages, meaning that news stories disappear from the internet).
Ray does not mention that we found a higher number of child fatalities than would be expected given homeschoolers’ share of the population. Instead, he merely states that we “summarized various disparate sources of information” and “found no statistically significant difference between the general public and homeschool fatality rates.” Ray also does not inform his readers that our database is emphatically not comprehensive, as we stated directly in our study.
Rodger Williams is a former homeschool father and a lobbyist for the Oregon Christian Home Education Association Network. He posts his own statistical analysis on a website called “the homeschool effect,” which has no landing page or “about” section. In a posting dated July 28, 2017, Williams used our estimates (he did not try to run his own) and our database (which he did not identify as incomplete) to run his own analysis. Williams’ innovation was to remove cases from our list which he determined should not be considered to involve homeschooled children.
Williams looked at 61 homeschool fatalities from 2003 to 2012 and determined that 21 were legally homeschooled, 19 were truant, and 21 were low regulation cases. He does not state how he determined this. Our HIC team relies on published news sources, legal records, and in some cases correspondence with investigating agencies to determine that a case involved homeschooling. In order to investigate the legal status of each child’s homeschool as Williams claims to have done, he would need to contact the investigating agency (child services, law enforcement, or the courts), possibly using a FOIA request, to gain access to information which is rarely made public in the media. It is unlikely that Williams did this given that he did not report his methodology.
News reports of the horrific child abuse cases we include on HIC often do not state whether state law was followed. No state offers a comprehensive list of homeschooling families, which would allow for easy verification. This is one reason we made the decision, when we created our database, to include all cases where the parents claimed to be homeschooling. We have been transparent about this; you can read our explanation in our FAQs.
Another reason we include all cases where the parents claim to be homeschooling in our database is that we created our database in order to look for themes that can help guide policies to prevent such things from happening in the future. With this in mind, it is important to understand limits in enforcement in states that do require some form of oversight, with an eye toward closing loopholes and fixing enforcement problems.
We are less concerned with determining which fatalities should “count” as homeschooled than we are in investigating the ways in which parents in our database have used homeschooling to cover up their abuse and prolong their maltreatment, and finding ways to stop future parents from doing the same. That makes cases where parents told neighbors they were homeschooling relevant regardless of whether the proper paperwork was completed.
Toward the end of his study, Ray states that the United States Commission to Eliminate Child Abuse and Neglect Fatalities did not mention the type of schooling children received or recommend regulating any type of educational environment. This is disingenuous. If no state currently collects data on a child’s educational status when tabulating child abuse statistics, how could the Commission identify any one educational environment as an area of concern?
In its report, the Commission did identify social isolation as a risk factor for abuse; as we have discussed here, there is no more effective way for an abusive parent to socially isolate a child than to homeschool. (Consider the thirteen Turpin children, raised in near complete isolation.) Furthermore, the Commission did not purport to be the official standard for child abuse prevention. Instead, the Commission called on individual states to conduct their own child fatality reviews.
In some states, child fatality review boards have already raised concerns about abusive parents’ use of homeschooling. In 2011, a review panel in Florida called on the Florida Department of Children and Families to “work with the school system and the Department of Education to devise an efficient alert system, with appropriate follow-up inspections, for at risk children removed from the school system and placed in ‘home schooling.’” Their recommendations followed a case where a child with previous contact with child welfare officials was removed from school to be homeschooled, and then tortured and killed. In 2012, Pennsylvania’s Task Force on Child Protection reported that “[t]he Task Force received testimony from prosecutors and physicians regarding specific instances of child abuse and neglect occurring in homeschool settings where families have had prior involvement in the child welfare system, which suggests that there may be instances where it would be prudent for additional safeguards to be put in place to ensure the health and safety of these children.”
The Missing Study
Ray claimed that a “thorough search of available literature has resulted in identifying only three published reports” relevant to child abuse rates among children in various school settings. The three studies were his own, CRHE’s, and Williams’ (which used our data). Ray completely leaves out Barbara Knox’s 2014 study of child torture. Either he knew about the study and decided not to include it, or his search was not as thorough as he suggests.
In her study, Knox looked at child abuse so severe that it could be termed torture. Knox found that 47% of the school-aged child torture victims in the cases she examined were removed from school to be homeschooled. Knox reported that in these cases homeschooling “appears to have been designed to further isolate the child” and that “their isolation was accompanied by an escalation of physically abusive events.”
It is possible that Ray failed to include Knox’s study because, as a non-random sample of cases, its findings cannot be applied statistically to all homeschooled children. However, this critique applies equally to the three studies Ray cites, including his own. Furthermore, Knox’s study is useful because it illuminates the fact that when abuse does occur in homeschool settings, there are fewer safeguards to catch it or prevent it from escalating.
In his rehashing of our data, Williams wrote that “there is no reason to impose ‘protective’ regulations on families who already are prone to a lower fatality rate than the rest of the nation.” This claim is absurd. Do we let hunger in our communities go unaddressed because the U.S. has a lower rate of food insecurity than the rest of the world? No, we do not.
Ray began his survey of research on homeschooling and child abuse by looking at abuse perpetrated by school personnel. When such abuse occurs in school settings, we as a society take steps to prevent future cases. Some states are considering improved background checks which screen for past allegations of misconduct whether or not there was a criminal prosecution, for example. All we are asking is that abuse in homeschool environments be addressed the same way—with an eye not toward accepting abuse, but preventing it.
For us, the question has never been whether abuse overall is more common among children who are homeschooled or children who attend school, but rather what happens when abuse does occur. CRHE Executive Director Rachel Coleman spoke recently with a crime reporter at a large regional newspaper who said that of all of the cases of abuse and child fatality she had investigated, the homeschool cases were by far the more horrific. This is because, as Knox’s study underlined, homeschooling allows abusive parents to isolate their children and conceal their abuse to a greater extent than they could if their children attended school.
Every child, no matter how they are educated, has the right to be protected against abuse and neglect. This is not about singling anyone out—we already take steps to identify abuse of children who attend public school. It is about ensuring that the same laws that give dedicated homeschooling parents the ability to create positive, innovative learning environments do not also allow parents like the Turpins to create houses of horror.