Transcript Temporal Discounting of Real vs Hypothetical

Temporal Discounting of Real vs
Hypothetical Gains and Losses
David J. Hardisty, Elke U. Weber, & Galen Treuer
Stanford University & Columbia University
Abstract
While the majority of research on intertemporal
choice has examined hypothetical financial gains,
many real-world outcomes of interest (such as
credit card debt or drug use) involve real future
losses. Using both within and between subject
designs, we compared discount rates for real and
hypothetical gains and losses, in internet and
student populations. Other than a main effect of the
web sample showing higher discount rates, results
indicate that real and hypothetical outcomes show
similar discount rates, and very similar effects of
framing (including sign, magnitude, direction, and
their interactions). Furthermore, both real and
hypothetical scenarios are correlated with real
world outcomes of interest, including savings and
tobacco use.
Sample Stimuli
Medium magnitude:
• Small Gain Delay:
• Medium Loss Accelerate:
Methods
• 2 samples: 30 MTurk participants and 30
Columbia University undergrads
Spearman correlations:
Results
• 2 task versions: hypothetical and consequential
• In the consequential task, two choices (one gain
and one loss) were randomly selected and paid out
• Amounts ranged from $4.90 to $100.10
• The undergrad sample completed both the
hypothetical and consequential tasks, and the
MTurk sample completed only the hypothetical task
• The following were manipulated within subjects:
- Sign (gain vs loss)
- Direction (delay vs accelerate)
- Magnitude (small vs medium)
• 28 trials per condition, for a total of 224 in the task
• For ease of comparison and interpretation,
choices were converted into discount rates using
the continuously compounded exponential formula
V=Ae-kD, such that a k of .10 indicates a 10%
discount rate
Small magnitude:
Gain
Disco
Loss
Disco
Savings Smoke
Gain
Disco
1
Loss
Disco
0.24*
1
-0.28*
0.29**
1
0.36**
0.07
-0.36**
Savings
Smoke
1
Discussion
• Real and hypothetical, online and student
samples all showed classic descriptive
anomalies, including the sign effect, direction X
sign effect, and magnitude X sign effect
• The online hypothetical sample showed a main
effect of higher discount rates
• Future research can safely continue to employ
hypothetical outcomes
Contact: [email protected], http://davidhardisty.info
Support: NSF grants SES-0820496 and SES-1024599