Behavioural Economics 5: Intertemporal Choice Flashcards
(35 cards)
Define intertemporal choice and list factors that influence it.
Intertemporal choice = decisions involving trade-offs between costs and benefits at different times.
Influencing factors:
Anticipated changes in needs/preferences
Affordability now vs. later
Trust in future availability
Opportunity cost of waiting
Patience
Examples: saving vs. spending, studying now vs. later
What are the standard theories of intertemporal choice?
Historical foundations:
Arthur Pigou (1921): Future pleasures are seen “on a diminished scale.”
Adam Smith (1759): We are more interested in pleasures today than in 10 years.
William Stanley Jevons (1871): Concern diminishes as delay increases.
These early views emphasized psychological realism.
Later models focused on normative mathematical frameworks, often ignoring behavioural nuance.
look at picture
In normative theory, people should optimise where the marginal rate of substitution (MRS) equals the slope of the budget constraint.
Optimal condition:
−(1+r)=−(1+δ)⇒r=δ
Interpretation:
The interest rate (r) compensates for the discounting (δ) of future consumption.
In experiments:
People should prefer larger-later (LL) over smaller-sooner (SS) only if the experimental interest rate exceeds their personal discount rate.
Explain the exponential discounting model and its implications.
- The first equation is ‘the present value of the outcome’
How do we calculate personal time preferences?
“t”’s units are in years
List and briefly describe experimental techniques used to measure time preferences.
Matching tasks: fill in the blank (e.g., £100 now = £X in 1 year)
Single binary choice: one SS vs. LL decision
Series of binary choices: multiple SS vs. LL decisions
Iterating tasks: staircase method to find indifference
Convex time budget: allocate tokens across time
Self-reports: rate willingness to delay gratification
Your note:
Works less well with goods that have diminishing marginal utility (e.g., snacks)
Describe ‘matching tasks’ as an experimental technique to elicit time preference
Describe ‘single binary choice’ as an experimental technique to elicit time preference
Describe ‘series of binary choices’ as an experimental technique to elicit time preference
Describe ‘iterating time preference task’ as an experimental technique to elicit time preference
Each question depends on the previous answer.
Describe ‘convex time budget task’ as an experimental technique to elicit time preference
Participants allocate 100 tokens between two time periods.
Tokens for later periods are worth more (reflecting interest rate).
Describe ‘Self-Reported Patience’ as an experimental technique to elicit time preference
Describe ‘Delayed Compensation Method (DCM)’ as an experimental technique to elicit time preference
Describe ‘Marshmallow task’ as an experimental technique to elicit time preference
Describe the key difficulties and limitations in measuring time preferences through experiments.
⚠️ Challenges in Eliciting Time Preferences
Despite the variety of experimental methods, several practical and conceptual issues can affect the accuracy and reliability of elicited time preferences.
- Trust in Future Rewards
Participants may not believe they will receive delayed payments.
This lack of trust can artificially inflate discount rates.
Especially problematic in lab settings with unfamiliar researchers. - Transaction Costs
Delayed rewards often require returning to the lab or waiting for a bank transfer.
These inconveniences can bias participants toward smaller-sooner (SS) options.
Easier to manage in online experiments. - Difficulty with Non-Monetary Outcomes
Most experiments use money, but real-world decisions often involve non-monetary goods (e.g., health, experiences).
Problems:
Can’t offer half a Ferrari now vs. a full one later.
Many goods are not divisible or not linearly valued.
Hard to calculate a discount rate for such goods.
💸 Hypothetical vs. Real Rewards
Does it matter if rewards are hypothetical?
Kirby & Marakovic (1995):
Compared real vs. hypothetical matching tasks.
Found lower discount rates for hypothetical rewards.
Coller & Williams (1999):
Compared real vs. hypothetical choice tasks.
Found lower raw discount rates for hypothetical rewards.
But after regression controls, the difference disappeared.
Conclusion: Evidence is inconclusive. Some studies show differences, others do not.
❓ Are Elicited Discount Rates Meaningful?
Fuchs (1982):
Found no correlation between elicited time preferences and behaviours like smoking, credit card debt, or exercise.
Chapman & Coups (1999):
Found that employees who chose to get a flu shot had lower discount rates.
Munasinghe & Sicherman (2000):
Found that smokers had flatter wage profiles, suggesting less investment in human capital.
Addiction Studies:
Heroin addicts consistently show higher discount rates:
Alvos et al. (1993)
Kirby, Petry & Bickel (1999)
Madden et al. (1997)
Murphy & De Wolfe (1986)
Petry, Bickel & Arnett (1998)
Caveats:
Consider budget constraints.
Be aware of publication bias.
Overall: Mixed evidence on whether elicited discount rates predict real-world behaviour.
What is domain specificity in intertemporal choice?
Domain specificity means that time preferences may vary across contexts.
Challenges the assumption that discounting is domain-general (i.e., stable across money, health, experiences, etc.).
Evidence suggests people may discount differently depending on the type of good or decision context.
Reference: Ubfal (2016)
What are the challenges in measuring time preferences for non-monetary goods?
To calculate a discount rate, the good must be:
Divisible – so you can vary the amount.
Linearly valued – utility should increase steadily with quantity.
Not quickly satiating – e.g., not like food or pleasure.
Without these, we can still study preference consistency and reversals, but not compute a precise discount rate.
What methods have been used to study discounting beyond money?
🏥 Health:
Trade-offs between short illness now vs. longer illness later.
Social discounting: fatalities prevented now vs. later.
References: Van der Pol & Cairns; Cropper et al. (1992); Bleichrodt & Johannesson (2001)
🍫 Snacks:
Small snack now vs. bigger snack later.
Healthy now vs. unhealthy with future consequences.
Reference: Mischel et al. (1988); Read & van Leeuwen (1998)
🎬 Experiences:
Movies (high-brow vs. low-brow), electric shocks, kisses, noise, dinners, weekends, drugs.
Focus on preference reversals and sequence planning.
References: Loewenstein (1987); Read et al. (1999); Bickel et al. (2015)
What is the significance of studying discounting beyond money?
Unlocks understanding of real-world behaviours (e.g., health, addiction, procrastination).
Helps design better policies and interventions.
Encourages development of new methods to measure time preferences in diverse domains.
Explain the common difference effect and how it challenges standard discounting models.
People often prefer:
£110 in 31 days over £100 in 30 days (LL)
But prefer £100 today over £110 tomorrow (SS)
Standard theory predicts no change in preference since the time interval and reward difference are the same.
This preference reversal is called the common difference effect.
Reference: Green et al. (1994)
Describe hyperbolic discounting and how it differs from exponential discounting.
In hyperbolic discounting, the discount rate declines over time.
People discount near-future rewards more steeply than far-future ones.
Example (Thaler, 1981):
£150 now vs. £X in 1 month → high discount rate
£150 now vs. £X in 1 year → lower discount rate
£150 now vs. £X in 10 years → even lower
Explains preference reversals and dynamic inconsistency.
What is the sign effect in intertemporal choice?
People discount gains more than losses.
Example (Thaler, 1981):
£15 now = £X in 3 months (gain)
-£15 now = -£X in 3 months (loss)
People are more willing to delay losses than gains.
Suggests asymmetry in how we treat positive vs. negative outcomes.
What is the magnitude effect in intertemporal choice?
Larger amounts are discounted less steeply than smaller ones.
Example (Thaler, 1981):
£15 now = £X in 1 year
£250 now = £X in 1 year
£3000 now = £X in 1 year
People are more patient with larger sums.
What is the magnitude effect in intertemporal choice?
Larger amounts are discounted less steeply than smaller ones.
Example (Thaler, 1981):
£15 now = £X in 1 year
£250 now = £X in 1 year
£3000 now = £X in 1 year
People are more patient with larger sums.
What is the delay/speedup asymmetry in intertemporal choice?
People react differently to delaying vs. speeding up outcomes.
Example (Loewenstein, 1988):
Delay: How much to accept to delay a £20 voucher by 8 weeks?
Speedup: How much to pay to get it 8 weeks earlier?
People demand more to delay than they’re willing to pay to speed up.
Shows framing matters.
