Homeschooling, Child Safety, and the Evidence

Claims about abuse need evidence about the children involved. I keep hearing that it is basically a legal loophole for parents to hurt their children, as though saying it confidently supplies the evidence. It doesn't. There are abusive parents who homeschool, and I won't defend them. But an accusation against every family making that educational choice is a much larger claim. You have to establish it.

The study I cited examined adults' retrospective reports of childhood maltreatment. Brian D. Ray and M. Danish Shakeel reported that school sector was not a significant predictor after they accounted for demographic factors in their analysis. That doesn't prove every setting is equally safe, but it is evidence against treating homeschooling itself as an automatic warning sign. Their published study is observational: it cannot establish that choosing a particular educational setting causes a particular safety outcome. We should use the finding for what it actually shows.

School attendance also deserves scrutiny rather than an automatic presumption of safety. In In-Person Schooling and Youth Suicide, Benjamin Hansen, Joseph J. Sabia, and Jessamyn Schaller examined U.S. mortality data from 1990 through 2019 and found that suicides among 12-to-18-year-olds were highest during school months and lowest in June through August. Areas starting school in August saw the increase in August, while September-start areas saw it later. That pattern gives the calendar argument actual evidence. It does not prove that homeschooling prevents suicide, compare homeschool families directly with school families, or predict an individual child's mental health. It does challenge the assumption that putting children in school settles the safety question before we examine what happens to them.

That distinction should change how we listen to an individual child. If they report being frightened of someone at school, attendance is not an answer to the report. If they report harm at home, a defense of educational freedom is not an answer either. The adults need to establish what happened and respond to the person who may be in danger. Favorable aggregate findings cannot do that work for us. I reject the blanket accusation precisely because it bypasses the examination every actual child deserves.

Listen to someone describing abuse without turning them into a representative of every family that used the same curriculum or legal category. Their experience matters in its own right. Equally, another parent's decision to homeschool doesn't become an admission of wrongdoing because someone else used that freedom terribly. We can distinguish those claims without becoming less serious about either child.

A school, a family and a community can each notice things the others miss. None gets to assume a report of harm has been handled simply because another adult was also in a position to hear it. Protecting a child requires someone to take that particular report seriously and follow up. A finding about school sector in an observational study cannot establish that this child is safe, any more than a frightening anecdote can establish that every homeschool family is abusive.

If you believe the educational choice itself makes children unsafe, say what evidence would change your mind. If no finding could, then this isn't an empirical claim anymore. Apply the same test to defenses of homeschooling: we shouldn't dismiss an account of harm because it is politically inconvenient. A serious argument must remain answerable to the children whose welfare it invokes.

A family that wants one-to-one attention, a more suitable environment and an education their child cares about should be able to make that case without first accepting an accusation of concealed abuse. Someone who uses the same freedom to harm a child deserves no defense from me. Those are different people making different uses of an educational choice, and our judgment needs to remain capable of telling them apart. The seriousness of protecting children is a reason to demand evidence, not permission to dispense with it.