Decision effect in multiple dated intertemporal choice and its psychological mechanisms

Hailong Sun, Xinru An, Guanxing XiongAdvances 2021 · DOI

Abstract: Traditional intertemporal choice research often focuses on single-dated outcomes, yet real-world decisions typically involve sequences of rewards or costs over time. This paper reviews the shift toward multiple dated outcomes, identifying key decision effects like the "hidden zero" and "sequence effects" that deviate from standard economic discounting. By comparing attribute-based trade-off models with alternative-based sequence models, the review explores the psychological mechanisms—such as anticipation and the preference for improvement—that drive human behavior in complex temporal contexts.

Introduction

Most research on intertemporal choice (e.g., choosing between $100 today or $115 in a month) focuses on single-dated outcomes. However, real-life decisions are rarely that simple. Whether it is a multi-year salary structure, a mortgage repayment plan, or a long-term corporate investment, outcomes are usually spread across multiple points in time.

Traditional economic models, like the Discounted Utility (DU) model, assume we discount future rewards at a consistent rate. Yet, human behavior often contradicts these models when faced with multiple-dated outcomes. This paper addresses the gap in understanding the unique psychological biases (decision effects) and the underlying cognitive mechanisms that emerge when rewards or losses are distributed over time.

Categorizing Decision Effects

The authors conduct a comprehensive literature synthesis, categorizing decision effects into two main categories: two-point outcomes (where a single choice involves two specific dates) and sequence outcomes (where choices involve a series of events over time).

Decision Effects in Two-Point Outcomes

Several "aha moments" occur where human psychology deviates from "rational" economic expectations in two-point scenarios:

  • The Hidden Zero Effect: Simply making the $0 outcomes explicit (e.g., "$10 today and $0 in a month" vs. "$0 today and $30 in a month") makes people significantly more patient compared to the standard "$10 today vs. $30 in a month" framing.
  • The Mere Token Effect: Adding a small, identical reward (a "token") to both the near and far options at an earlier time point reduces the perceived cost of waiting, leading to more patient choices.
  • Front-end vs. Back-end Effects: The timing of "extra" money matters. Adding money to the beginning of a sequence (Front-end amount effect) can actually increase impatience, whereas adding it to the middle or end (Extra dated-money effect) can have the opposite effect.
Summary of decision effects in two-point intertemporal choices, including the hidden zero and mere token effects.

Sequence Effects

When faced with sequences, people often exhibit a preference for improvement. Standard discounting suggests we should prefer getting more money sooner. However, individuals often prefer an increasing sequence (starting small and ending large) over a decreasing one, even if the total value is the same. This is because the psychological utility of "improvement" and positive anticipation outweighs the mathematical time value of money.

Theoretical Frameworks and Model Comparison

The paper evaluates these effects through the lens of two primary theoretical frameworks:

  1. Trade-off Model (Attribute-based): This model suggests that decision-makers compare the differences in "value" dimensions against the differences in "delay" dimensions between options. It is highly effective at explaining two-point outcome effects like the hidden zero effect.
  2. Sequence Model (Alternative-based): This model proposes that individuals evaluate a sequence as a single holistic unit. The utility of a sequence is composed of discounted utility, improvement utility, and spread utility. This model excels at explaining why we prefer increasing salary trends but is less suited for simple two-point choices.
Comparison of the Trade-off Model and Sequence Model across different outcome domains.

Key Results and Psychological Drivers

The review identifies that attention, anticipation (dread or excitement), and the desire for "improvement" are more powerful drivers in multi-dated choices than simple mathematical discounting.

  • Theoretical Alignment: As shown in the comparison table, the Trade-off Model finds consistent evidence in two-point results but lacks verification in sequence domains. Conversely, the Sequence Model is strongly supported by sequence effects but shows conflicting or unverified evidence in simpler two-point scenarios.
  • Mechanism Synthesis: The paper provides a structured comparison, highlighting which model best explains specific behavioral "anomalies."

Limitations and Future Work

The authors acknowledge several areas where the field needs to grow:

  • Mixed Gains and Losses: Most current research looks at pure gains or pure losses. Real-world sequences are often mixed (e.g., a high-paying job that requires an immediate stressful move), and the weighting of these mixed sequences is not yet fully understood.
  • Process-Tracing Technology: To move beyond just observing final choices, the authors advocate for using eye-tracking and mouse-tracking to reveal the real-time cognitive "weighting" process as people evaluate different points on a timeline.
  • Unified Theory: There is still no single "grand theory" that perfectly explains all intertemporal choice effects across both simple and complex scenarios. Integrating attribute-based and alternative-based models remains a major challenge for future researchers.

Summary updated

Abstract

摘要:
BibTeX
@inproceedings{sun2021decision,
  author = {Hailong Sun and Xinru AN and Guanxing Xiong},
  title = {{Decision effect in multiple dated intertemporal choice and its psychological mechanisms}},
  booktitle = {Advances},
  year = {2021},
  doi = {10.3724/sp.j.1042.2021.02224},
}
Figure