Reference dependence is the foundational assumption of Prospect Theory: people evaluate outcomes not in terms of final states (as expected utility theory assumes) but as deviations from a reference point. What counts as a "gain" or "loss" depends entirely on where the reference point is set, making the reference point a crucial determinant of choice behavior.
Reference Point Determination
The reference point is typically the status quo — the current state of affairs. However, it can also be set by expectations, aspirations, social comparisons, or the framing of the decision problem. Kőszegi and Rabin (2006) proposed that reference points are determined by rational expectations about outcomes, creating a formal model where the reference point is endogenous rather than assumed.
The same objective outcome can be perceived as a gain or a loss depending on how it is framed. Tversky and Kahneman's Asian disease problem showed that describing outcomes as "lives saved" (gain frame) versus "lives lost" (loss frame) dramatically shifts preferences — from risk averse to risk seeking — even though the objective outcomes are identical. This demonstrates that the reference point is influenced by the description of the problem, not just its substance.
Adaptation and the Hedonic Treadmill
Reference points adapt over time: after a raise, the new salary becomes the reference point, and only further increases feel like gains. This adaptation — the "hedonic treadmill" — explains why absolute levels of wealth are poor predictors of happiness, while changes in wealth are powerful predictors. Reference point adaptation has important implications for well-being, policy evaluation, and the design of incentive systems.