Impact Of Microcredit

Question 1

The microcredit revolution - how did formal institutions find a way to lend to the poor at reasonable rates?

(Remember why they dont lend to poor, e.g legal interest ceiling, cannot threaten, expensive to screen/monitor, far away)

Answer 1

Use social capital rather than physical capital as collateral!

Used neighbours/peers knowledge to screen and monitor without incurring cost themselves!

Question 2

What was the model they used to do this called?

Answer 2

Joint liability model

Question 3

Joint liability model: social capital as collateral

Answer 3

5+ borrowers jointly liable for each others loans.

If someone repayments missed, group has to pay for them.

If someone defaults, whole group is cut from future loans

Question 4

What is the ideology behind this? How does social capital act as collateral (2)

Answer 4

Shame and gossip are powerful deterrents - people can risk their relationships if fail to pay cos affects others.

Also cheap for peers to monitor compared to high costs for formal banks

Question 5

Why is rigid repayment structure so effective (3)

Answer 5

Easy accounting to check total group repayment correct.

Good for borrower discipline - constant/simple

Makes monitoring easier for peers

Question 6

As a result of the cost savings from rigidity and social capital,

Did this allow micro finance institutions to lend to the poor at more reasonable rates?

Answer 6

Yes - charged 2% a month following these cost savings

Question 7

So does microcredit work? Reasons for yes (4)

Answer 7

Higher business investment and profits
More consumption (later debates this)
Better health & education
Gender empowerment (since usually targets women e.g Field)

Question 8

MFIs themselves on their impact, and caveat to their belief

Answer 8

They believe as people keep coming, their services must be useful.

Caveat - not all MFI’s are profitable, the industry is heavily subsidised, and people may be making mistakes

Question 9

Morduch test on effectiveness of microfinance

Answer 9

Compared villages with/out microfinance.

Question 10

Why is this naive?

Answer 10

Selection effect - results can be influenced by other factors - remember why we use random controlled trials instead so we know results are purely from treatment!

Villages would not have identical outcomes if hadn’t been implemented - so not a valid counteractual.

Question 11

Banerjee - miracle of microfinance

Answer 11

Spandana microfinance team - identified 104 neighborhoods willing to work right away but willing to wait.

50% operations right away
50% start after 2 years.

Women given starting loan of 10,000 rs, repaid weekly.

Question 12

Findings on Spandena takeup on overall microfinance takeup within areas

Answer 12

26% takeup in treated areas vs 18% control areas after 15-18months.

Microfinance not that popular as seems!

Question 13

We are not interested in the eect of Spandana branches (there were other MFIs). We want to know the effect of microcredit!

We can use randomised encouragement design to find this.

What did the impact of Spandana have on use of MFI overall, and informal

Answer 13

Overall MFI credit used increased 8.4% (18 to 26!)

Reduce informal lenders credit by -5.2% (good as informal expensive!)

Question 14

What were the reasons to take up a loan

Answer 14

Start new business
Repay old loan
Expand

Question 15

Spandana on business’ and consumption (3)

(Between Spandana areas and non treated)

Answer 15

Business creation increased by 1.6% - not a lot (COMPLIES WITH OUR LOAN TO MAINTAIN NOT EXPAND)

No change in business profits

Overall consumption unchanged. Durable consumption increased and temptation goods fell

Question 16

When using encourage design Wald estimator we can then see the effect on microfinance overall from our instrumental variables design (not just Spandana’s effect).

We saw the difference between Spandana vs non only led to a 1.6% increase in business creation.

How much overall having microfinance increase probability to start a new business?

Answer 16

12% (for those who choose to get a loan)

So upon people who get the loan, 12% more now start a business

Question 17

When is this estimate wrong?

Answer 17

Underestimated if competition exists: e.g learn from others so get loan

Overestimated if competition exists so decide not to pursue.

Question 18

Consumption by type: they used propensity score: based on what?

Answer 18

How likely someone is to start a new business.

Question 19

Consumption for those likely to start a business vs unlikely

Answer 19

Likely - They reduce temptation goods

Unlikely increase their spending on temptation goods.

Question 20

So overall benefits of microcredit (2)

Answer 20

Helps people get out of expensive informal loans (MFI cheaper 2% a month following joint liability model benefits)

Allows a committment to save - since obligation to put money aside.

Question 21

However… (2)

Answer 21

No broader change on household behaviour (overall consumption remained)

Business impacts are small - no change in business profits and only 12% more likely to open business

So not a miracle!

Question 22

Other RCTs support the ideas of Banerjee (3)

Answer 22

Low take-up of microcredit - not as popular as perceived

Business investment increases somewhat

No change in OVERALL consumption or household spending