As many of you know, much of my work focuses on user motivation and self-delusion, and how we can design user experiences that account for both of these. A recent study in the Proceedings of the National Academy of Sciences led by Adam Waytz at Kellogg School of Management brings all of these together in an interesting situation.
This research demonstrates a fundamental cognitive bias driving such conflict intractability: Adversaries attribute their ingroup’s actions to ingroup love more than outgroup hate and attribute their outgroup’s actions to outgroup hate more than ingroup love.
A finding that we have long known about it is that people often debase the motives of our opponents when their self-identity is at stake. When we have a visceral need to be right, it helps to attribute the worst of intentions to our opponents. We feel better about ourselves by opposing the “evil other.”
Another finding that we have long known about is that extrinsic rewards like cash give people a burst of good feelings. If you have ever won a bet, you know what this feels like. It is even better if the bet is based on displaying your superior knowledge or ability. There are clear downsides of extrinsic rewards, but those are not relevant to the PNAS research.
What happens if we place these two impulses in direct opposition? That is what the new study investigated. They asked liberals and conservatives to attribute motives to the other side when making political decisions. The assessments were much more extreme than a control group of moderates. Both sides considered their own motives to be based on love and caring, but the other side’s motives to be based on hate and fear. This made it easier to viscerally oppose the other side and feel better about themselves for opposing them.
The experimenters wanted to know what would happen if they offered financial incentives for accuracy. The financial incentives would make participants feel good about themselves for attributing more moderate motives to their opponents, which deep down the participants must know is more accurate. But this would be in conflict with the first effect. Which one would win out? Would the assessments be extreme to feel good about opposing evil? Or would the assessments be moderate to feel good about winning the cash?
As you would imagine, they split the difference. The assessments were still somewhat more extreme than the control group’s assessments. But not nearly as much as they were without the financial incentives.
Image credit: fotometin