A strange coalition has formed around the police officer-worn body camera.

Their ubiquitous adoption is the sole policy change requested by the family of Michael Brown, the teenager who was fatally shot by an officer in Ferguson, Missouri. Police departments, too, hail body cameras, saying that they will shield officers from false claims of wrongdoing. Endorsed by law enforcement agencies, police reformists, and equipment vendors, body cameras seem to promise accountability at a time when police power seems untouchable.

Faced with cultural and political issues that can seem intractable, the U.S. has suddenly and rapidly adopted the little lenses and, with them, a new surveillance regime. President Obama earlier this week announced $263 million in funding to purchase 50,000 body cameras for local police agencies.

As a Baltimore civil litigator told the Washington Post on Wednesday: “The body camera is here to stay.”

But a debate very similar to the one around body cameras has happened before. Two decades ago, law enforcement agencies—and activists hoping to change them—argued about a different kind of mass video surveillance. That technology was not body cameras but in-car dashboard cameras, tape recorders that filmed the action in front of the car and preserved the audio of an officer’s interactions with citizens.

As in today’s debate, in-car cameras found support from both police chiefs and police reformists. Law enforcement claimed video evidence would protect cops from post hoc citizen grievances, while reformists claimed that surveilling cops would reduce racial profiling.

At the end of the 1990s, a combination of those arguments helped secure federal and private funding to purchase in-car cameras en masse. Dash cams are now everywhere. The most recent federal data, from 2007, states that 67 percent of state or local police departments had at least some squad cars with cameras, and experts say that they’ve only become more popular since.

The body camera debate now, in other words, is where the dash-cam debate was 15 years ago. We can look back at the promises that dash-cam advocates made and see where they fell short—we can, in a limited way, predict the future from the past. And while history doesn’t exactly repeat itself, understanding what was supposed to happen with dash cams—and what actually did—takes us out of the the fanciful future of shining screens and dystopian omniscience and puts us in our own—where cops get tired, camera lenses get scummy, and it’s harder to fix things than it is to buy them.

* * *

1. The adoption of body cameras will not all happen at once.

“I was skeptical of the cameras at first,” Michael Creamer, the chief deputy of the Franklin County Sheriff’s office in Ohio, told a New York Times reporter. “We’re not in the movie business. But they’ve been fantastic for us.”

He was speaking in April 1990 about the power of dashboard-mounted, in-car cameras. During a six-month test run, he boasted, all 17 people who had been arrested for drunk driving pleaded guilty because their arrests were caught on film. (Most people who get arrested for drunk driving, the Times added, plead not guilty because of the crime’s “severe fines and jail terms.”)

“We’ll show the judge, the jury, and the courtroom how they really looked driving on the wrong side, falling down by their car, unable to walk a straight line or recite the alphabet,” Creamer said. “It’s very hard to rebut that kind of testimony.”

With such a strong endorsement, it seems like a no-brainer for the government—at the local, state, or federal level—to immediately pitch in and finish purchasing cameras. Creamer mused to the Times that he wanted cameras in every car in his fleet.

But Franklin County had never purchased those first, test-run cameras. In fact, no government agency had purchased them. Creamer’s dash-cams came from Aetna Life and Casualty, the private insurance company. Aetna was one of a number of insurance firms at the time that hoped to reduce the drunk-driving-related injuries—and their attendant medical bills—by making enforcement of the law much more consistent and severe. A prime way it could accomplish this? Ensuring that drivers arrested for DUIs could be prosecuted for them.

In the United States, in-car cameras have been purchased in two big waves historically. Creamer’s cameras—and those underwritten by Aetna—were part of the first wave, during the 1980s. The cameras in Franklin County weren’t installed to monitor cops or citizens, but to solve a basic enforcement problem: Without hard photographic evidence, it was difficult to prosecute drunk drivers. So insurance companies and the recently formed group Mothers Against Drunk Driving shelled out for police in-car cameras.

The second wave began during the 90s. Additional financial support in that wave came from the DEA, which partnered with local agencies to catch drug trafficking on interstate highways. Cameras, and especially microphones worn by cops, could document suspects consenting to their car being searched—something juries often had a hard time believing if police then found guns or drugs.

But even with DEA assistance, less than 40 percent of police departments had even some cars with dash cams. The second wave could only kick into gear once a certain coalition emerged—a political partnership that will look familiar to contemporary eyes. More on that soon.