In every American state, a teacher who suspects a child is being abused or neglected is required by law to report it. Failure to report is itself a crime. This is mandatory reporting. It is one of the most ambitious legal architectures in American family policy: a system designed to convert millions of adults—teachers, doctors, social workers, coaches, daycare providers—into sensors for child harm. It was built in the 1970s on a wave of horror at battered child syndrome and the previous era's reluctance to intervene in family privacy. The architects expected it to save lives. It does save some. It also produces a parallel harm that the architects did not foresee, and that the system has been slow to acknowledge.
The harm is this: mandatory reporting, as currently structured, generates an enormous volume of reports—roughly seven and a half million calls a year in the United States, concerning about four million children. The vast majority do not result in substantiated abuse. They result in investigations, home visits, family disruption, and a permanent record in a state database. The children who are reported most often are poor, Black, Native, and disabled. The reporters who generate the most reports are educators. The mechanism designed to protect children has become, in many communities, a mechanism that surveils them.
This is not an argument against the underlying impulse. Children do get hurt. Some of them get hurt at home. Many would not be discovered without an outside adult noticing and acting. The teacher who calls because she sees a bruise pattern that does not match the explanation is doing exactly what the law was designed to enable, and sometimes saving a life. The question is not whether to have a reporting system. The question is whether the current calibration is producing more good than harm, and whether the harm falls on the families it was supposed to protect.
Law 4 is planning. A plan that worked in 1974 may not work in 2026, not because the values changed but because the implementation environment did. In 1974, the typical reporter was a doctor or a teacher who knew the family, considered the call carefully, and reached a hotline staffed by a social worker who could exercise judgment. In 2026, the typical reporter is often a school employee following a mandated training that emphasizes "when in doubt, report" and threatens criminal liability for under-reporting. The hotline is overwhelmed. The judgment has been replaced by a triage algorithm. The downstream investigation is conducted by a child welfare agency that is itself overwhelmed and under-resourced. Each layer optimizes for not missing the next high-profile fatality. None optimize for not harming the families who are not the next fatality.
The disparity is the strongest evidence of system failure. Black children in the United States are more than twice as likely to be reported to child welfare than white children, even after controlling for income. Native children are reported at rates that approach the historical removal rates of the boarding school era. These are not the rates at which Black and Native parents abuse their children. They are the rates at which Black and Native parents are watched. The act of watching, in a system where watching triggers investigation, becomes the harm.
Schools are the largest single source of reports. Teachers see children daily, are trained extensively, and operate under criminal liability if they fail to report. The school as institution has become a frontline of child welfare surveillance in a way it was never originally designed to be. This produces specific dynamics: a teacher who builds trust with a struggling parent must weigh that trust against the reporting obligation. A parent who tells a teacher about a difficult home situation may discover the conversation triggers a CPS visit. The result is that many parents, particularly those who already distrust institutions, stop talking to teachers about anything substantive.
The reform conversation is split. Some advocates push for narrower reporting requirements, restricting the threshold to actual evidence of harm rather than suspicion. Some push for differential response—a system where lower-risk reports trigger family support services rather than investigation. Some push to abolish mandated reporting entirely, on the grounds that it cannot be calibrated and the harm to overpoliced communities exceeds the benefit. The middle ground argues for keeping the obligation but radically retraining reporters, raising the threshold, and rebuilding the downstream system so that an investigation is not the only response available to a hotline.
Whatever direction the reform takes, the current system cannot be defended as designed. It is producing a million children a year subject to investigation who are not being abused, concentrated in communities that have historically suffered the worst forms of state intervention in family life. A plan that does not look at this honestly is not a plan; it is inertia. The children whose lives the system was built to save deserve a system that actually saves them, not one that confuses surveillance with care.