Science is a social process, subject to heartache and the thousand natural shocks that flesh is heir to

From Why the Most Important Idea in Behavioral Decision-Making Is a Fallacy by David Gal.

Not ready to throw the baby out with the bath water. Gal is not, I think, rejecting the reality of loss aversion. I think what he is doing is rejecting a naive/blanket application of the hypothesized loss aversion. My suspicion is that loss aversion is real but only has narrow application under defined and particular circumstances. It is not the psychological deus ex machina it has been made out to be.

I liked this observation.

If what I am claiming is true, why has belief in loss aversion persisted so strongly? An idealized view of science is that theories are accepted or rejected based solely on empirical evidence. In fact, science is not simply an objective search for truth, but also a social process, in which proponents of a theory must convince other scientists, through logic and argumentation, of how evidence should be interpreted.

However, this process advantages incumbent theories over challengers for a number of reasons, including confirmation bias, social proof, ideological complacency, and the vested interests of scientists whose reputations and even sense of self are tied to existing theories. A consequence is scientific inertia, where weak or ill-founded theories take on a life of their own, sometimes even gaining momentum despite evidence that puts their veracity in doubt.

In the case of loss aversion, contradictory evidence has tended to be dismissed, ignored or explained away, while ambiguous evidence has tended to be interpreted in line with loss aversion. For example, a paper purporting to illustrate that price increases are more impactful than price decreases received 65 citations in Google Scholar in 2016, whereas a follow-up paper challenging this view received only 17 citations.

Moreover, belief in loss aversion has meant that phenomena that have nothing to do with loss aversion have nonetheless been interpreted to reflect loss aversion. For example, the sunk cost effect, the finding that people are more likely to continue an endeavor once an investment in it has been made, has been attributed to loss aversion. While the sunk cost effect might reflect a reluctance to recognize losses, this is not relevant to loss aversion, which requires a comparison be made between losses and gains.

Or as Max Planck put it, acknowledging the social dimension/flaws of science and the capacity of the vested interests to stymie new insight:

Science advances one funeral at a time.