A new study involving rat restaurants and cat videos has shown that the longer you wait for something, the less likely you are to give up on it.
University of Minnesota neuroscientist A. David Redish and his team published a study on the sunk-cost fallacy in rats, mice, and humans last week in Science.
The sunk-cost fallacy is a phenomenon where a creature sticks with a decision based on what they’ve already invested in it rather than its probable future outcome, Redish said. It’s what happens when you stick with a terrible film because you’ve already watched half of it, or stay in an endless line since you’ve already spent 20 minutes there in the rain.
“Whether you go home or whether you stay doesn’t change anything. You’ve still spent those 20 minutes either way,” Redish said.
The fallacy is often a problem as it leads to poor decisions, but researchers have yet to figure out its cause or whether it happens in non-humans, Redish said.
Redish and his team set up two tests based on the natural behaviour of foraging to see if they could observe the sunk-cost effect.
They took brown Norway rats and mice and put them in a four-room maze. Each room was a restaurant, and rodents could receive a grape, banana, chocolate or plain pellet if they waited in the appropriate room for enough time. A tone played outside each room to tell the rodents what the wait time was for that pellet, with the time varying between trials. The team judged how much the rodents liked each pellet based on the amount of time they lingered in a restaurant after eating it.
The team set up a parallel experiment with students. Instead of pellets, they “foraged” for videos of kittens, bike accidents, landscapes or ballroom dancing by choosing between four web galleries, with each gallery telling them how long they’d have to wait to get that video. They also graded each video afterwards to show how much they liked it.
The team found that subjects in both experiments were more likely to stick with a wait for a reward the more time they had already spent waiting for it – even if they could give up to switch pursuit to a reward they liked more. This shows that the sunk-cost fallacy was in effect.
Surprisingly, the team found that time spent prior to committing to a wait did not influence the likelihood that a subject would stick with it. Redish said this suggests that two different decision-making processes were at work. Pre-commitment, subjects used a forward-looking system (“What will I get?”) where sunk costs did not apply. Post, they switched to a backward-looking one (“What have I spent?”) where they did.
“In the real world, accurately predicting the future is hard,” Redish said, and time spent to get an outcome is sometimes a decent proxy for its value. Mammals may have evolved different decision-making systems for different circumstances. Studying how these systems work could teach us how to treat addictions and other mental illnesses where people demonstrate poor judgment.