Individual differences in time preferences for losses

This study was an investigation of choice behavior regarding individuals’ time preferences for immediate or deferred aversive outcomes. Subjects were 32 Rice University students randomly assigned to the experimental (loss) or control (gain) conditions. In each condition subjects made choices among alternatives differing only in variable outcome receipt time (constant value-uncertainty pairs) and among other pairs of options varying also in value and probability (non-constant value-uncertainty pairs). The former was the most important for purposes of the present study, however, consistency of previously stated time preference was examined by the latter. A definition of the generic form of time preference behavior was offered, and theories relating to temporal considerations were discussed. Methodological difficulties involved with previous research were discussed and hopefully obviated in order to effect a more precise measurement of time preference behavior. In choosing between amounts of money that could be lost at various points in the future, subjects preferring outcomes differing in variable outcome receipt time were almost evenly divided in their preferences for losses. On the other hand, choices between non-constant value-uncertainty pairs options evidenced the relative unimportance of temporal factors with subjects maximizing expectations or avoiding sure losses. The first observed relationship between trait anxiety and the tendency to advance losses was noted. Some situational and organismic factors influencing advancing or deferring aversive outcomes were outlined and practical implications discussed