Empirical Evidence On Increasing Savings Flashcards
Saving has been increasing by default:
Somville & Vandewalle main idea
More people have bank accounts, and receive income in banks.
This increases saving since it is the DEFAULT OPTION. Savings will happen by default, unless people make the active decision to withdraw!
Research design in Somville and Vanderwalle
Studied impact on expenditures and savings between a control group receiving income in cash and treatment group receiving money in bank account
Somville and Vandevalle results phase 1:
Increase in savings in treatment group match the equivalent increase in regular consumption in the control group.
Results phase 2: (week 20 onwards)
Account balance of treated (account-based payment) remained higher, but…
No difference in savings and consumption patterns, because both groups deposit and withdraw similar values.
Overall findings: account-based payments are saved by default, whereas cash payments are ready to be spent unless deposited actively.
Demand for deferred payments (Brune, Chyn & Kerwin)
Workers were allowed to defer a fraction of their pay for 3 months at zero interest.
i.e save for lumpy purchases. (so no temptation since wont have the money to spent!)
Findings (2)
Savings significantly increased.
Correspondingly durable investment increased since deferred payment allowed them to save. (deferred payment removed temptation!)
Is this about self control or the safety that the scheme provides?
Self control mainly - a follow up experiment gave the option of the original scheme or a modified version where manual deposits were to be made.
Found manual deposit scheme resulted in substantially lower savings (50% lower)
(Deferred payment guarantees money won’t be spent since we can’t access it, but manual deposits people may not have self control if it is accessible)
Savings for health - Dupas & Robinson
(D&R also looked at open savings account in “savings framework’ - no interest, withdrawal feel, weekdays experiment)
Kenya -
Box with padlock. Randomised whether respondents had a key (safe box) or no key (lockbox).
2 boxes - one for health pot (group savings), other for health saving account (individual savings)
Everyone was encourage to save for health.
Findings: (3)
Health box (group savings) worked best along with safe box.
Lockbox and HSA (individual saving) more muted since people saved more slowly, didn’t want to ‘lock’ their savings. (Kinda constrasts Brune - where deferring payments increased saving!)
Earmarking (lockbox) only worked for individuals frequently taxed by friends and relatives. (Since could lock away and say they can’t repay friends!)
Duflo Kremer and Robinson - saving for fertilisers
Fertiliser can be bought in small quantities, so why can’t farmers buy enough for a small part of their farm?
Income is usually seasonal, rather than a constant monthly.
So they can afford it after the harvest, but the fertiliser isn’t needed then, but for the next planting season. So it requires to be forward looking, realising to increase produce they need to save today.
Which is hard to do if quasi hyperbolic (present-biases) since have to resist until planting season
So if farmers got the chance to put money in towards fertiliser, do they take it up?
Compare situations if
A)not time inconsistent
B)time inconsistent but naive
C)time inconsistent but sophisticated.
D) so when will it work?
If farmers are not time inconsistent, it will not work. Since they are forward looking, so do not see point in buying now, and save money in case of emergencies. (Which is good as forward looking, but not good in this context of improving yields through fertiliser!)
If time inconsistent but naive, would not work. Farmers would think they will do it, but not follow through.
If time inconsistent but fully sophisticated, it won’t work as farmers will solve it on their own.
Can only work if farmers are time inconsistent but partially sophisticated. (Not able to calculate cost and benefits properly)
So the issue seems to be they’ll say they will, but then do not follow through.
So commitment devices were created (2)
Gave farmers at voucher AT THE TIME OF HARVEST, when they have the INCOME!
50% discount and free delivery at planting season
Or business as usual
What were the results
Offering early with discount - in the season increased fertiliser use by 46-60%.
Which is a bigger effect than the free delivery at 50% discount at planting season (since people may have spent the money by then and so still don’t buy!)
What was the main conclusion on why fertiliser use is low
Behavioural barriers explain part of the reason why fertiliser use is low in developing countries.
Main overall conclusion on savings
The reason is not that people are too poor to save, it is behavioural constraints, and also a lack access to financial products (like the discount AT TIME of harvest) that help deal with temptations, which causing them to undersave.
