Last month a police officer in Houston was run over and killed during a traffic stop. The suspect got away. The next day, millions of phones across Texas buzzed with news of the officer’s death after the state’s Department of Public Safety blasted out what’s known as a Blue Alert. This prompted considerable concern and confusion. A man in Odessa, some 500 miles away, spoke for many when he tweeted: “wtf is a blue alert?”

Blue Alerts are mass notifications, now used in 35 states, that are sent to mobile phones and flashed on electronic highway signs when a suspect on the loose is thought to be an “imminent and credible threat to law enforcement.” The hope is that pinging the public will lead to tips for the police, and then a speedier capture. It’s an idea that originated with the better-known Amber Alert program, named after a 9-year-old abductee from Arlington, Texas, who was murdered, which aims to help authorities recover kidnapped children. Along with Blue and Amber alerts, there are Silver Alerts, issued for elderly people who are lost and might be suffering from dementia, and Camo Alerts, dispatched in at least three states when current or former members of the military are missing and thought to be a threat to themselves or others.

The appeal of doing everything we can in the aftermath of a horrific crime is powerful. But there’s little evidence that any of the rainbow of alerts have much impact at all. In fact, these alerts are best described as “crime control theater”—a term criminologists use for programs that merely foster the perception that the government is taking swift and significant action.

Each time a new alert is proposed, the success of Amber Alerts is cited as precedent. In 2018, 161 Amber Alerts were sent out, in cases involving 203 children. Of those, 34 children were recovered based on an Amber Alert tip. That’s about 17 percent. Since it began in 1996, the return of 967 children has been credited to the program. If Amber works, the thinking goes, then Blue, Silver, and Camo should work too. But does the track record of Amber really match that seemingly unassailable reputation?

It depends on how you measure success. Read the detailed annual Amber reports published by the National Center for Missing & Exploited Children and you’ll notice that the majority of child-abduction cases involve those taken by parents or other family members, often in the midst of a custody dispute. That doesn’t mean these situations aren’t threatening, but they are not the stranger-in-a-white-van scenarios we tend to imagine when our phones buzz. If you search through news reporting of the events that led to Amber alerts, you’ll also notice that many appear to fall short of the “imminent danger of serious bodily injury or death” guideline for their use from the US Department of Justice.

Some of these alerts are certainly justified, and rescuing nearly 1,000 children over two decades is far from nothing. But when researchers dig into those numbers, they start to seem less impressive. In a 2016 paper, criminologists examined 448 child-abduction cases in which Amber Alerts were sent out to the public. (In 401 of those cases, the abducted child or children were recovered unharmed; in 88, an alert-inspired tip was credited with the recovery.) The study found that outcomes for the children didn’t vary all that much. That is, children were typically taken by a family member and returned home safely; and this was true regardless of whether the Amber Alert had brought in useful tips. It’s likely that the alerts sometimes led to a speedier recovery of those children—which is clearly a great thing—but the researchers didn’t find support for the assertion that the Amber program saves lives.