You know the routine. A tough cop berates and browbeats a suspect until his partner sends him out. The partner is reassuring, and the suspect spills his guts. But does it work? Scientists did a few rather odd experiments to find out.
The Good Cop, Bad Cop routine is supposed to work, but why? It might be an emotional response to a sympathetic authority figure. It might involve seeing the logic in complying with the police. But that's not what a group of Polish scientists thought. They thought logic, or emotional appeal, had nothing to do with it. Instead, they believed the move from one emotion to the next was so extreme that people became disoriented, and their ability to think critical simply switched off. It was called the "fear-then-relief response."
It's a fine theory, but how does one test it? Hauling people into a police station and scaring-then-relieving them out of their minds wasn't going to get by an ethics board. (Unless the ethics board was the first group of people the scientists were planning to scare.) Scientists couldn't viciously interrogate people so they did the next best thing — pranking the populace. Instead of committing to intimidation, the scientists decided to commit to whimsy, and devised a weird, wonderful set of experiments with fake blind people asking math questions and police whistles distracting people on the road.
The Good Cop Bad Cop Experiments
The police whistle made an early appearance in the series of experiments. A researcher simply sat by a road and waited until a person jaywalked across it. One loud whistle had that person spinning around, sure they were going to be fined or ticketed. When they saw the whistle being wielded by a civilian, they continued on their way.
But they found their way blocked by those most-feared of all street predators, the clipboard people. These clipboard people had a collection box and one of three spiels. The first spiel was a simple "Excuse me, would you please give me money?" The second was,"We are collecting money. Would you please give us some because we need to collect as much as possible." The last spiel included an actual reason; they were collecting for a holiday camp for mentally disabled children. People who had been allowed to peacefully cross the road only got their wallets out for the last pitch. People who had experienced whistle-induced fear, then relief, were more likely to hand their money to whoever asked for it, regardless of justification.
Then the scientists decided to get theatrical. One of the researchers was outfitted with a white cane and a pair of dark glasses, so he looked like a blind man. They put him near the exit to a covered marketplace, watching people come out. He was to let four people walk by him, but when every fifth person emerged he was to dramatically grab their shoulder, because some experiments are conducted like jump-scares in horror movies. Once the innocent member of the public had jumped out of their skin and turned around, they saw a blind man. The experimenters thought this would fill them with relief. Half the time, that would be it — the man would apologize, and the unwitting subject would keep moving, only to be approached by more clipboard people asking them to fill out a survey about life in Poland.
The other half of the time, things got crazy. Once the person had seen the "blind man" and felt relieved, the blind man would ask them a math question. He would ask how many minutes he had until an upcoming appointment. Adding up the minutes in their head would cause them to re-engage their critical thinking, at which point they'd walk on and also be accosted by the clipboard people with a survey about life in Poland. (Sometimes, the blind man would only ask, "Is that you?" The explanation, scientists thought, would help people jumpstart their critical thinking before they were asked to fill out a survey, at which point they wouldn't be so compliant.)
This series of experiments provided more complicated responses. The people who had their critical thinking skills re-engaged through math or identity questions were much more likely to refuse to fill out the survey. However, the group that had been scared-then-relieved was no more likely to fill out the survey than a third group of people who had simply been approached by survey people, without a fake blind man scaring them first.
For the last experiment, the researchers mounted a full production. They had groups come in and do simple tasks, like spotting a target or doing basic addition. One third of the time, that would be it. Another third of the time the scientists announced that people who fell behind would receive electric shocks. And for the final third of the time, they announced that people would receive electric shocks, then had a researcher come in, announce that this was the wrong group, and they would not be getting shocks.
This time, fear-and-relief seemed to be in full swing. The control group, who did their problems without any threat of shocks, got through the task fastest. The soon-to-be-shocked group was a little slower. Lagging well behind both, however, was the discombobulated fear-then-relief group, unable to mentally get it together.
Good Cops and Bad Cops
So what happened? Some of the experiments indicate that people who experience fear then relief get the sense knocked out of them. They become compliant, and go along with suggestions without thinking.
In the "blind man" experiments, though, fear-then-relief did nothing. The only thing those experiments indicated was that people who have just helped a blind man tell time aren't interested in filling out a survey. And what about re-engaging the critical thinking with a math problem? If a blind man asking a simple question allowed subjects to reassert their critical thinking skills, shouldn't a test do the same thing? Why, then, did the fear-then-relief group taking the test lag behind the other groups? Surely a complicated task like that would allow people to re-engage their brains so that they wouldn't lag behind everyone else.
All we can say in the end is that the "good cop, bad cop" routine can work — but it's not particularly reliable. Sometimes it just makes you really slow at math.