We can’t help but pursue the things we want, even when the chase is fruitless and exhausting. Sometimes, the futility of the chase is inexplicably motivating: After working so hard, giving up always seems so much harder. (Raise your hand if you watched Lost all the way through, even when it went off the rails.) As researchers show in a new Science study, we’re not the only species that go through this irrational process: Other animals do it too, and there may even be a common psychology that unites us all in that cognitive bias.
In the paper, published Thursday, a team of University of Minnesota researchers show that the more time humans, rats, and mice sink into a task without receiving a reward, the harder they’ll stick with it. Anyone who’s taken an introductory economics class will know there’s a name for this phenomenon: the sunk cost fallacy.
The idea of a sunk cost cognitive bias in humans is old news, but the fact that it shows up elsewhere in the animal kingdom is a big deal because it suggests that our irrationality could be the result of a decision-making system that humans share with rats and mice.
“The sunk cost fallacy, by definition, arises from valuing spent resources that cannot be recovered,” write the study’s authors, led by neuroscience doctoral student Brian Sweis. “Our data finds that these sunk costs only accrue under specific situations in mice, rats, and humans.” To simulate these situations, the researchers designed “foraging” tasks for the rat, mouse, and human subjects.
Hungry rodents were placed in an experiment called Restaurant Row, a square maze with a different “restaurant” chamber at each corner. Each chamber offered a different flavor of food pellet — grape, chocolate, banana, or plain. In each chamber, there was an “offer zone,” in which the mouse or rat would hear a fixed-pitch tone that indicated how long they would have to wait for a reward, which could range from one to 30 seconds. If they decided to bail, they moved on to the next restaurant. But if they decided to wait, they entered the “wait zone,” at which point the clock began, and a tone descended in pitch to indicate the wait time passing. They could leave at any point during the waiting period, which would end the trial and give them a chance to check out another chamber. But if they stuck it out, they got a little snack as a reward.
The catch was that they only had a limited amount of time in the experiment, so the longer they waited, the less time they had to explore other options. Despite the ticking clock, the longer the rats and mice spent in the wait zone, the more likely they were to stay there until they got their reward. This result is totally consistent with a sunk cost bias.
Importantly, the researchers note that the countdown didn’t begin until the rodents moved from the offer zone to the wait zone. “This meant that the animal was choosing between distant options and had not yet invested in the offer,” they write. Until they invested by stepping into the wait zone, they did not display the pattern sunk cost bias.
Human participants did a similar task, but instead of a maze and food pellets, they were presented with options for different videos on a computer: dance, landscapes, kittens, or bicycle crashes. They were shown how long the videos would take to load and given the option to “skip” or “stay.” Once they chose to stay, the download bar showed the video’s progress, but just like the mice, the humans could move on to the next video during the download process, cancelling the download and giving them a chance to check out something else.
Sure enough, the humans showed the same pattern as the mice and rats: The longer they waited for the video to load, the more likely they were to wait the full time and watch it.
One potential explanation for these shared behaviors is that, since it’s really hard to accurately calculate future benefits, animals will use prior effort as a substitute for that measure. It’s also possible that the effort spent working for a reward can leave both humans and rodents depleted of energy, “enhancing the perceived value of the yet-to-be-obtained reward.”
Whatever the reasons may be, the authors point out that all the animals behaved predictably in the trials, sinking time and energy into rewards, even if there might be something better out there. For this reason, they say there’s probably a shared neurological basis for how humans, rats, and mice process the costs and benefits of potential rewards.
“We suggest that multiple, parallel decisionmaking valuation algorithms implemented in dissociable neural circuits have persisted across species and over time through evolution,” write the study’s authors. “These tasks and findings may aid future research in education or neuropsychiatry by shedding light on diagnostic or intervention strategies and revealing the roles of neurally distinct decision systems.”