Is loss aversion a bias?


Jason Collins


November 5, 2011

From the Journal of Personality and Social Psychology:

Much research shows that people are loss averse, meaning that they weigh losses more heavily than gains. Drawing on an evolutionary perspective, we propose that although loss aversion might have been adaptive for solving challenges in the domain of self-protection, this may not be true for men in the domain of mating. Three experiments examine how loss aversion is influenced by mating and self-protection motives. Findings reveal that mating motives selectively erased loss aversion in men. In contrast, self-protective motives led both men and women to become more loss averse. Overall, loss aversion appears to be sensitive to evolutionarily important motives, suggesting that it may be a domain-specific bias operating according to an adaptive logic of recurring threats and opportunities in different evolutionary domains.

Unfortunately, I cannot access the article beyond the abstract, but one of the article’s authors, Douglas Kenrick, has blogged about the paper at Psychology Today (HT: Åse). Kenrick writes:

[R]esearch by several members of our team shows that loss aversion waxes and wanes in flexible ways, depending of whether or not the person is experiencing different fundamental motivational states (such as self-protection or mating motivation). Research participants were asked how happy or unhappy it would make them to gain or lose $100, for example, or to experience a 30-percentile boost in their financial assets. As in the previous research, losses typically loomed slightly larger than gains. But all that changed for participants who answered the questions in a mating frame of mind (after imagining themselves having a romantic encounter with someone they found highly attractive). …

According to Jessica Li, who was the first author of the study: “For men in a mating frame of mind, loss aversion completely disappeared, so that they became more focused on wins than losses. For women, on the other hand, mating motivation led them to be even more loss averse.”

In another evolutionary scenario, where losses were clearly more costly than gains, participants become even more loss averse.

When we put participants in a self-protective frame of mind (by having them imagine being alone in the house on a dark night and hearing an intruder breaking in), both men and women became more loss averse in their judgments.

The way I prefer to think about loss aversion is to consider what the objective of the person is - and that is normally an evolutionary objective. Once the gains and losses are framed in evolutionary terms, what is loss aversion in one dimension is not loss aversion in the dimension that matters. For example, if a small cash win will not increase the number of mates but a small cash loss might cost them the mate that they have, a loss averse response to the potential cash pay-offs can be contrasted to the zero-gain potential in the mating dimension. It is not loss aversion to reject a bet with no upside.