Quasi-Hyberbolic Discounting And ROSCAs

Question 1

Exponential discounting concept

Answer 1

Things in future are less value than present

Question 2

How is this idea captured in the model

Answer 2

Constant discount factor to weigh utilities in different periods

δ

Question 3

Expression for present discounted value of a stream of income over the next t periods (skip)

Answer 3

u(x₀) + δu(x₁)+δ²u(x₂)+…δ to the t u(xt)

Where
xt is the investment decision made at time t
u(xt) is the amount of money earnt in time t

Question 4

£1000 today or £1500 in 5 years.

What would you choose? How is this modelled with the discount factor

Answer 4

Choosing £1000 today implies:

u(£1000) > δ⁵u(£1500)

Question 5

£1000 in 30 years or £1500 in 35 years. What would you choose?

What is the issue we have found

Answer 5

Choosing £1500 in the 2nd choice implies:

δ³⁰u(£1000) < δ³⁵u(£1500) =⇒ u(£1000) < δ⁵u(£1500)

Exponential discounting is violated; people are time inconsistent. We will patient in the future (as in this choice), but present bias (impatient in present)

Question 6

What model captures present bias

Answer 6

Quasi hyperbolic discounting model

Question 7

Quasi-hyperbolic discounting model captures present bias: How

Answer 7

Adds parameter β to account for the bias

u(x₀) + β[δu(x₁) + δ²u(x₂) + … + δT u(xT )]

So basically the same but bracket all the future periods with β. (Since x₀ is the current period so not included)

Question 8

Consider the 2 choices we made, how have their expressions been adjusted?

Answer 8

u(£1000) > βδ⁵u(£1500)

And for the second choice , no change; no need to add β since both options are in future

δ³⁰u(£1000) < δ³⁵u(£1500) =⇒ u(£1000) < δ⁵u(£1500)

So exponential discounting not violated!

Question 9

Numerical example of QHD with/without β

Question 10

How can we overcome the time-inconsistency problem and save? (3)

Answer 10

Microfinance
Excess purchase of durables
ROSCA participation

Question 11

ROSCA - Rotating Savings and Credit Association

B) how are payouts decided (3)

Answer 11

Everyone contributes a fixed sum of money to the pot.

B) Payouts of the pot can be made in several ways:
Random
Bidding
Pre-determined

Question 12

Applications :

Korean ROSCA is called…

How many people participate in ROSCAs in Kenya

Answer 12

Kye in Korea

45-50% participation in Kenya

Question 13

Why ROSCA’s exist: 4 theories

Answer 13

Lumpsum purchases
Protect female income
You can’t save alone
Insurance

Question 14

1st theory for why ROSCA’s exist: Lumpsum purchase

Answer 14

People save ‘x’ every period , but need ‘Nx’ to buy a cow. (the lumsum purchase)

So forming a ROSCA with n other people, and each meeting someone gets ‘Nx’.

Question 15

Why is this good, and evaluation

Answer 15

Good as can they can receive ‘Nx’ amount quicker than saving alone (ex-ante, gains are bigger if random allocation)… EXCEPT

The last person is no better off, since they are last to receive the ‘Nx’

Question 16

So gains are bigger when random, as could get money very fast!

But in pre-determined ROSCA’s, why does the last person to get Nx agree?

Answer 16

May be better off part of the scheme, free benefits.
Out of pressure, or good of heart.

Question 17

2nd theory why ROSCA’s exist:

Protect Female Income

Answer 17

Women save in ROSCAs to protect their income from their husbands.

Question 18

Why would they protect income from husbands?

Answer 18

Asymmetric preferences:
Women spend more on family needs, men spend for personal use.

Thus, women are better of saving in a ROSCA than at home. Can save at a higher rate than otherwise

Question 19

Anderson and Baland findings on women in ROSCAs in Kenya

(Baland same person finding borrowing and asaving at same time to make it look like under pressure)

B) are ROSCAs overall welfare improving

Answer 19

Looked at slums in Kenya. Found 84% of Rosca members are women.

B) women save at a higher rate, but husband worse off. (so not necessarily welfare-improving)

Question 20

Issue with the ROSCA pot for women

Answer 20

When the woman comes home with the ROSCA pot, husband may want to buy an indivisible good since it is immediate consumption.

Question 21

3rd theory of why ROSCAs exist:

You can’t save up alone (Gugerty) :

Firstly, how does it contranst Anderson and Baland’s findings

Answer 21

Another case in Kenya.

Found a high ROSCA participation rate even among widows. (contrasts with Anderson & Baland, since women arent under household pressure yet still use)

Question 22

2 key Gugerty findings

Answer 22

41% of women got money from their husbands to join ROSCA cycle. (Constrants A&B of asymettric preferences…)

48% gave some money to husbands after receiving the pot was allocated to them.

Question 23

60% of ROSCAs have a committment feature. what is this?

Answer 23

A spending arrangement - agree beforehand on what they money should be used for, and monitor!

Question 24

Main conclusion of 3rd theory

Answer 24

New reasoning for ROSCA participation: a commitment device to solve the problem of quasi-hyperbolic discounting (present bias issues)

Hence the name ‘its hard to save alone’

Question 25

What was the evidence behind this reasoning:

Answer 25

when asked why participate in ROSCA, most say ‘it is hard to save alone’

Question 26

4th theory on why ROSCAs exist: ROSCA participation as insurance

And criticisms of this theory (2)

Answer 26

Bidding ROSCAs help deal with raise funds for big expenditures like funerals.

Criticisms
However bidding ROSCAs are rare. (So cannot be primary motive of why ROSCAs exist!)

Women often reinvest profits back into the ROSCAs later on. (So not just for large expenditures)

Question 27

Conclusion of this theory

Answer 27

insurance is one of the motives but not the primary motive

Question 28

What does recent neuroscience show us?

Outline cases (2)

Answer 28

Stressful environments can increase present bias (and poor decision making more generally)

B) McClure et al: Decisions made in prefrontal cortex: highly susceptible to stress

Lichand & Mani: Farmers reduced cognitive performance when considering weather. (Lichand and Mani)