4. Choice and Time Flashcards

1
Q

What is individual dynamic choice?

A

When there is extended time between determination of outcome and receipt of outcome or in the outcome itself

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is sequential choice?

A

When a choice is made at more than one point or uncertainty resolved at more than one point or resolution of uncertainty precedes choice

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is the orthodox economic answer to lifelong trade offs that individuals make between outcomes at different dates

A

Discounted utility model

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What are the components of the discounted utility model?

A

Ut = w(t+j)u(Ct+j)

Where u(.) is instantaneous utility function
Where w(.) is weighting function, giving weights on utility in different periods viewed from period t w(t+j) = d^j where d is a discount parameter

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Features of discounted utility model

A

-Additive separability of Ut
-Instantaneous utility function isn’t time dependent
-Weighting function isn’t time dependent or consumption dependent
-W(t+j) declines exponentially as j rises
-relative weight w(t+j+k)/w(t+j) is d^k
-for given d, it depends only incremental delay on k and is independent of baseline delay j

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Is the discount rate constant in the data?

A

Studies suggest discount rate declines as j rises.
Evidence of sharp discounting of near future relative to present but less sharp discounting to very distant future relative to not quite as distant future. Think of this as diminishing marginal sensitivity to delay

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

How do people choose between smaller-sooner and larger-later options and which paper shows this?

A

Tendency to prefer smaller-sooner option when there is little/no baseline delay, but larger-later when there is a baseline delay
Keren & Roelofsma 1995

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Components of hyperbolic discontinuing Loewenstein-Prelec form

A

W(t+j)= 1/(1+alpha x j)^(y/alpha)
Absolute weighted decline with j
Relative weight is increasing in baseline delay j and decreasing in incremental delay k. These fall sharply but then flatten out compared to exponential discounting

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What are the two key features of quasi hyperbolic discounting?

A

Immediate future can be weighted much less than present
Discount rate declined progressively with baseline delay

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Describe Laibsons quasi hyperbolic discounting model

A

Captures phenomenon of present bias without postulating a steadily declining discount rate. Very simple functional form which nests standard exponential model and is sometimes known as B-d preferences

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What are the different psychological intuitions behind Laibsons quasi hyperbolic discounting?

A

Visceral attraction of present rather than diminishing marginal sensitivity to delay. Only the present is certain

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Describe the quasi hyperbolic present bias model

A

Reduces to exponential model if beta =1.
Beta<1 reduces weight on all future periods by constant factor relative to exponential model with same discount rate
Relative weight is Bd^k when j=0 and d^k when j>0
Beta is an extra discount factor applying to future but only affecting comparisons involving the present

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is a possible welfare implication of the present bias model?

A

Decisions about allocation of expenditure between periods t and t+x should be taken before t arrives in order to avoid influence of present bias. This is an argument in favour of commitment strategies

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Dynamic consistency

A

An agent sticks to and executes the plan she forms at the start

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Forward looking consequential rationality

A

An agent is forward looking consequential if at each point she evaluates the remaining options
-treating bygones as bygones
-treating all feasible plans as possible choices and evaluating them by their outcomes

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

How do non exponential discounting models do when looking at consequential rationality and dynamic consistency

A

They permit conflict between forward looking consequential rationality and dynamic consistency

17
Q

Resolution

A

Form best plan for whole problem at the start and stick to it

18
Q

Myopia/naivety

A

At each choice act in accordance with current view of best plan for remainder of problem

19
Q

Sophistication

A

At each choice, act in accordance with current view of plan that is the best for remainder of problem taking as given that will also do this at every remaining choice

20
Q

Is sophistication dynamically consistent and forward looking consequential rational?

A

Dynamically consistent but violates forward looking consequential rationality by treating future revisions as constraints on current plans

21
Q

What is Rabin &O’Donoghue 1999?

A

Paper comparing agents with and without present bias. And agents with present bias who don’t foresee their own self control problems.
Time consistents, myopic naifs, sophisticates
Have to miss the movies one weekend in 4

22
Q

How does present bias affect costs and pleasure for naifs and sophisticates in Rabin and O’Donoghue 1999?

A

Present bias leads to procrastination of costs and acceleration of pleasure. If true long term welfare is measured by present bias free preferences then present bias has a welfare cost

23
Q

What are the conceptual problems associated with discounting and which papers point then out?

A

There are problems with attempts to infer:
-discount parameter without taking account of shape of u(.). Anderson 2008
-preferences over consumption streams from preferences over income streams. Cubitt & Read 2018
-a single discount parameter for an individual applicable to all their decisions. Cubitt, McDonald, Read 2018

24
Q

What is a consumption stream?

A

It assigns consequences by dates