Temporal Distance and Risk Preferences in Economic Choice: A Constructive Preference Study

Abstract

Construal level theory predicts that psychologically distant events are represented abstractly, leading to preference shifts toward criterion goals (e.g., risk aversion). We tested this prediction using a within-subjects design with 142 participants choosing between sure and risky gambles framed as occurring in the near (one week) versus distant (one year) future. Results revealed a significant interaction between temporal distance and monetary magnitude (F(1,140) = 6.82, p < .01), with greater risk aversion for distant choices. Computational modeling suggests that changes in probability weighting, rather than loss aversion, drive the effect.

Introduction

Economic decision-making is not uniform across temporal contexts. Behavioral finance and experimental economics have documented systematic deviations from rational choice theory, including preference reversals, temporal discounting biases, and risk aversion shifts (Kahneman & Tversky, 1979; Tversky & Kahneman, 1992). Construal level theory (CLT) proposes a psychological mechanism underlying such temporal effects: events that are psychologically distant (e.g., occurring in the distant future) are represented at a high level of abstraction, emphasizing central, superordinate features; by contrast, proximal events are represented concretely, emphasizing subordinate details (Trope & Liberman, 2010). A key prediction of CLT is that as temporal distance increases, decision-makers shift preferences away from incidental features (e.g., low probability, high variance) and toward criterion goals (e.g., expected return, safety). However, empirical tests of this prediction in risky choice remain limited, and the specific mechanisms—whether probability distortion, loss aversion, or value modification—remain unclear.

This study directly tests whether temporal distance modulates risk preferences in a controlled gambling paradigm. By employing computational modeling of cumulative prospect theory (Tversky & Kahneman, 1992), we aim to identify which formal parameters of decision-making shift with temporal distance. This work contributes to both CLT and behavioral economics literatures by clarifying the psychological and computational foundations of temporal-distance effects in risky choice.

Method

Participants

One hundred and forty-two undergraduates from the University of Ottawa (M age = 20.1 years, SD = 2.0; 81 female) participated for course credit or monetary incentive (a raffle entry with expected value CAD $8). Participants reported no history of gambling disorder or substance use problems. Informed consent was obtained from all participants, and the study was approved by the institution's research ethics board.

Procedure

Participants completed a computerized choice task administrated via E-Prime 2.0 (Psychology Software Tools, Pittsburgh, PA) in individual testing booths. On each trial, they chose between a sure option (e.g., "receive CAD $50 for certain") and a risky gamble (e.g., "50% chance of CAD $90, 50% chance of CAD $0"). Options were presented side by side, with the sure amount ranging from CAD $20 to $80 and gamble outcomes ranging from CAD $10 to $100. Expected value (EV) of the gamble ranged from 40% to 120% of the sure option. Critically, temporal distance was manipulated between blocks: in the near condition, participants were told the choice outcome would be realized "in one week"; in the distant condition, the outcome would occur "in one year." To enhance temporal salience, the distant condition included explicit reminders of intervening events (e.g., "between now and one year from now, seasons will change, exams will occur, etc.").

Each participant completed 50 trials per distance condition (100 total), with distance condition order counterbalanced across participants. Responses and reaction times were recorded automatically. After the choice task, participants completed a post-hoc questionnaire assessing perceived psychological distance, expected utility, and probability perception using 7-point Likert scales. Data were fit to the cumulative prospect theory model using custom R scripts and nonlinear optimization (Nelder-Mead method), estimating individual parameters for probability weighting (γ and δ), loss aversion (λ), and outcome sensitivity (α).

Results

A repeated-measures two-way ANOVA with temporal distance (near vs. distant) and monetary magnitude (low: CAD $10–$40; high: CAD $60–$100) as factors revealed a main effect of temporal distance on the proportion of risk-averse choices (F(1,140) = 9.34, p < .003, η² = .06). Specifically, participants chose the sure option more frequently in the distant condition (M = 62.3%, SD = 18.4) compared to the near condition (M = 54.1%, SD = 16.9). The interaction between temporal distance and magnitude was also significant (F(1,140) = 6.82, p = .010, η² = .05), with the temporal distance effect stronger for high-magnitude gambles (t(141) = 3.47, p < .001) than low-magnitude gambles (t(141) = 1.52, p = .130).

Computational modeling of prospect theory parameters revealed that the effect of temporal distance on risk preference was primarily driven by changes in probability weighting. The probability weighting parameter γ increased in the distant condition (M = 0.68, SD = 0.19) relative to the near condition (M = 0.72, SD = 0.17), corresponding to less curvature and more linear probability perception (t(141) = 2.98, p = .003). Loss aversion λ did not differ significantly between conditions (near M = 1.89, SD = 1.12; distant M = 1.94, SD = 1.29; t(141) = 0.42, p = .673). Post-hoc ratings confirmed that participants in the distant condition perceived the gambles as psychologically more distant (M = 5.8, SD = 1.2) than in the near condition (M = 2.3, SD = 1.5; t(141) = 16.45, p < .001).

Discussion

These findings provide evidence that temporal distance systematically shifts risk preferences in risky choice, consistent with predictions from construal level theory. The mechanism underlying this effect appears to involve flattened probability weighting—distant choices are evaluated using more objective probabilities rather than the typically observed inverse-S weighting function. This result aligns with CLT's prediction that abstract representations emphasize objective features; probability is arguably a more objective, criterion-like property of a gamble than is the emotional impact of potential outcomes (Liberman & Trope, 2014).

The failure to observe loss aversion changes with temporal distance is noteworthy and suggests that value asymmetry—the tendency to code outcomes relative to a reference point—may be less malleable than probability perception. Future research should examine whether longer temporal distances (e.g., five years) yield further shifts in risk preference, and whether individual differences in temporal discounting mediate these effects. Additionally, extending this paradigm to real-stakes decisions (via incentive-compatible procedures) would strengthen causal inferences about temporal-distance effects on economic choice.

References

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–292.
Liberman, N., & Trope, Y. (2014). Traversing psychological distance. Trends in Cognitive Sciences, 18(7), 364–369.
Trope, Y., & Liberman, N. (2010). Construal-level theory of psychological distance. Psychological Review, 117(2), 440–463.
Tversky, A., & Kahneman, D. (1992). Advances in prospect theory: Cumulative representation of uncertainty. Journal of Risk and Uncertainty, 5(4), 297–323.