Loss aversion for time and money: Cross-commodity variation in loss weighting and choice consistency

Loss aversion is often inferred from a prospect-theory parameter (λ) greater than 1 and interpreted as evidence that losses receive greater psychological weight than equivalent gains. Yet this interpretation depends on the assumptions used to translate observed choices into model estimates. We examined whether λ can be interpreted as a measure of psychological loss weighting by comparing observed choices and model estimates across money and time. A total of 124 participants chose between safe and risky options in separate gain-only and loss-only tasks involving both resources and completed a measure of time perspective. Participants chose the risky option more often for losses than for gains in both resources. However, the primary model yielded population-level λ estimates of 0.77 for money and 0.82 for time, with no reliable difference. Individual estimates showed weak correspondence across resources, although recovery analyses indicated it could arise even when the underlying λ is identical. Time perspective did not predict λ. Across alternative modeling specifications, estimated λ shifted from below to above 1 despite unchanged observed choices. Moreover, different combinations of loss valuation and sensitivity to value differences could produce identical predicted choice probabilities. Thus, behavioral differences between gains and losses, and even an estimated λ greater than 1, do not necessarily demonstrate greater psychological weighting of losses.