Most people are unaware how much public policy is either lacking in any empirical-research support or driven by bad research. Political actors motivated by ideology or donor/constituent demands propose new government practices—in areas ranging from policing to funding of treatments for gender dysphoria in youth to welfare-qualification rules—that will greatly impact people’s lives, and if anyone asks what basis they have for thinking the impact will be good, they can readily find some study to support their case. Especially when powerless populations are put at risk, neither the legislative process nor peer review in the publication process provides a real check on reckless experimentation and incompetent or corrupt research.

This Article argues that, at least with respect to social services for vulnerable populations, innovation and scientific study should be subject to constraints analogous to those for introduction of new drugs and vaccinations. These include pre-implementation assessment of evidentiary basis by panels of independent experts, piloting, and assurance of scientific rigor as well as protections for human “subjects”—a concept that, even in medicine, should be expanded, to include anyone substantially impacted by an experimental intervention and not just those whom researchers choose to study. In addition, agencies and research institutions must become more circumspect about who provides proxy consent for non-autonomous subjects.

As illustration of the problem and how the solutions might be implemented, the Article focuses on the repeated innovations over the past forty years in state response to child maltreatment, a pattern sure to continue indefinitely unless discipline is imposed. A voiceless population with no reliable surrogates, too often treated as distributable goods rather than persons, children in the child protection system present the perfect storm of conditions conducing to unethical behavior among policy makers and social scientists. In this realm, “fake news” destroys lives.

Document Type


Publication Date

Winter 2022

Publication Information

87 Missouri Law Review 1-94 (2022)