AMBER Alerts were designed to recover children in the most serious abduction cases, but they might be ineffective at saving lives, and could carry hidden costs.

With the new and poorly implemented Amber alert system frequently making news in Canada and setting off acrimonious firestorms on social media I was curious about my (and many others') assumption that Amber alerts as a whole save lives. Common sense would say they obviously do but the research, which is all that really counts, paints a different picture.

The abstract of the article discussed in the linked interview:

A sample of 448 child abduction cases in which America's Missing: Broadcast Emergency Response (AMBER) Alert was issued was examined to determine the extent to which AMBER Alert is successful in rescuing abducted children, and whether the successes suggest ‘lifesaving’ rescues. We reached conclusions consistent with the scant available prior research on AMBER Alert: although over 25% of the Alerts facilitated the recovery of abducted child(ren) and are thus arguably ‘successful’ by that standard alone, there was little evidence AMBER Alerts ‘save lives.’ In fact, AMBER Alert success cases are in almost every measurable way identical to AMBER Alert cases in which the child(ren) were returned unharmed but the Alert had no direct role in that outcome: they typically involve abduction by family members and other (apparently) non-life-threatening abductors, and the vast majority of recovery times are over 3 h. The implications for the public discourse regarding AMBER Alert and directions for future research are discussed.

source