The evidence is real and the argument about it is still running: how strong it is, how far it travels, or whether it repeats. Trust the direction, and hold the numbers loosely.
Imagine a coin flip. Heads, you win 150 dollars. Tails, you lose 100. The math says take it, the average outcome is positive. Most people refuse. In the 1970s, two researchers built a whole theory around that refusal. They found that the pain of losing is roughly twice as strong as the pleasure of gaining the same amount, so a loss of 100 hurts about as much as a gain of 200 feels good. This one idea reshaped economics and won a Nobel prize, because it explained a thousand things at once: why people hold losing stocks too long, why they overpay for insurance, why the safe salary beats the better gamble. But here is the honest part. The 1979 paper established the asymmetry without ever fixing a ratio to it. The famous 2.25 arrived in a follow up thirteen years later, fitted to the choices of twenty five graduate students at Berkeley and Stanford, and it went into the textbooks as though it were a constant of human nature. When researchers later estimated the same parameter across the wider literature, the mean came out nearer 1.3, and most individual studies found no significant effect at all. Gal and Rucker went further in 2018, arguing that many of the classic demonstrations confused losing something with having to take an action, and that separating the two makes the effect often disappear. Others replied that it has moderators rather than being dead, and they are probably right. The direction is not in doubt. Losing does hurt more. The size is contingent, weakest when the stakes are small, and still argued over.
If you notice the safe choice keeps winning even when the numbers favor the risk, this is the reason, and it is worth knowing the feeling is real, not weakness. But it is also worth knowing the effect may be smaller than the famous version says, which means some of your caution is protecting you and some is just fear talking louder than the math. The only way to tell them apart is to make a few small bets you can afford and watch how you actually feel, rather than trusting the theory or the fear.
There is one setting where the asymmetry is not a bias at all. If a loss can remove you from the game entirely, avoiding it is arithmetic rather than fear. A bet that pays well on average is still a bad bet when losing it means there is no next round, because the average only arrives for people who survive long enough to collect it. That distinction is worth holding carefully, because it is also the setting the research deliberately excluded. The studies use amounts small enough that no outcome threatens anybody, which is how they isolate the psychology, and it is why their findings say nothing about the times caution is simply correct. So the question to ask of your own hesitation is whether losing would end something. If it would, trust the flinch. If it would not, the flinch is running on a number that may not hold.
The endowment effect says owning a mug triples what you think it is worth, and the premium is invisible from the inside. Loss aversion says losing hurts more than an equal gain feels good. The endowment effect holds; loss aversion is still argued over. That is an odd shape, because loss aversion is the usual explanation given for why ownership inflates value: selling feels like a loss and buying feels like a gain. So the effect is solid and its explanation is contested. What that leaves you with is the effect on its own terms. You will defend what you authored partly because it is good and partly because it is yours, whatever the mechanism turns out to be, and the mechanism being uncertain does not make the premium any smaller.
Source: Kahneman and Tversky, Prospect Theory, Econometrica, 1979. The 2.25 coefficient comes from the 1992 follow up, Advances in Prospect Theory. The challenge: Gal and Rucker, The Loss of Loss Aversion, Journal of Consumer Psychology, 2018.
Richard Thaler, 2015
Thaler's own account of finding loss aversion and the endowment effect, and of the years economists spent insisting people were rational anyway.
Draw your own card. It does not take long, and it rewards taking your time.