Microfinance: Learning What Works, What Does Not, And Why

1
Microfinance Learning What Works, What Does
Not, And Why
2
How do we Reach 1 Billion?
Massive untapped market Why are so many yet to
be reached?

• Is it flexibility?
• Is it price?
• Is it institutional?

3
Whats the Secret?How do we Design Effective
Microfinance?

• How do we

• The answer is

• Create the most impact?
• increase incomes, improve health, education
• Reach more clients?
• Reach poorer clients?
• Increase savings balances?
• Improve client retention?
• Improve profitability?

• We Dont
• Really Know
• Yet

4
We Have a Lot to Learn

• Example Interest rates are very high
• In the US, we call this usury and regulate it. 
• Abroad, we call this charity and donate to it. 
• Impact
• How do interest rates affect impact?
• Outreach
• Can we target only those clients who are likely
  to benefit (likely to earn high returns)?
• Business
• Can we lower interest rates and make more profit?

5
Example Credit with Education

• Why do many microenterprises plateau? How can we
  help them grow?

Is credit with education a solution?
6
Who do You Trust?

• Well-known microfinance leader A
• The poor do not need us to teach them how to
  survive they already know. So rather than waste
  our time teaching them new skills, we try to make
  maximum use of their existing skills.
• Well-known microfinance leader B
• Group-based lending/saving services provide an
  unusual opportunity to provide transformative
  educational experiences along with financial
  services.

7
There is No Magic Solution

• Learning what works in microfinance requires
  methodical experimentation across settings
• Not guesswork based on anecdotes
• We need high-quality research methods to produce
  reliable results
• Randomized controlled trials are the most
  reliable research methodology.

8
Evaluation is for Managers!

• Evaluation is for microfinance programs, not
  annual reports
• Evaluation is much more than does the program
  help people?
• We use rigorous evaluations to learn which
  products and services work best by trying out
  different versions and seeing which are most
  effective
• most impact, most profitable, etc.

9
RD

• Key question for donors/investors

• How do we leverage our investments to maximize
  social welfare, while maintaining private
  returns?
• RD can help figure out how to tap that market
  profitably.
• Make sure the approach, if successful, gets
  replicated and scaled.

10
Why Randomize Example from Get out the Vote
Method Estimated Impact
Reached vs. Not Reached 10.8 pp
11
Why Randomize Example from Get out the Vote
Method Estimated Impact
Reached vs. Not Reached 10.8 pp
Regression Control for differences between the groups 6.1 pp
12
Why Randomize Example from Get out the Vote
Method Estimated Impact
Reached vs. Not Reached 10.8 pp
Regression Control for differences between the groups 6.1 pp
Also control for voting history 4.5 pp
13
Why Randomize Example from Get out the Vote
Method Estimated Impact
Reached vs. Not Reached 10.8 pp
Regression Control for differences between the groups 6.1 pp
Also control for voting history 4.5 pp
Matching similar voters 2.8 pp
14
Why Randomize Example from Get out the Vote
Method Estimated Impact
Reached vs. Not Reached 10.8 pp
Regression Control for differences between the groups 6.1 pp
Also control for voting history 4.5 pp
Matching similar voters 2.8 pp
Randomized Experiment 0.4 pp, statistically insignificant
15
How to RandomizeExperimental Credit Scoring
?
?
applicants scored 0-100
100
100
T
C
x
0
0
16
Credit Scoring Findings

• What is the impact of consumer credit?
• Positive effect on employment, wages and hunger
• 7 percentage-point reduction in poverty level
• Marginal loans are profitable for the lender

17
How to Randomize Group vs. Individual Liability
18
Group vs. Individual Liability Findings

• 150 joint-liability centers
• 75 randomly assigned to convert to individual
  liability
• 75 randomly assigned to remain as-is
• Outcomes
• No change in repayment
• No change in savings
• Higher client retention
• More new members joined

19
Innovations Product Development
Randomize different versions of a product/service
to find the best version Peru Teaching
Entrepreneurship
Credit with Education Credit with Education
Yes No
Voluntary Yes 34 101
Voluntary No 104 101
20
Teaching Entrepreneurship Findings

• Strong benefits for both the client and the
  institution.
• Repayment higher
• Client retention up
• Clients sales higher
• Profitable to offer free
• education

21
Current Work

• Replication
• Price
• Behavioral Savings using psychology to help
  clients save more
• Insurance
• Health insurance
• Crop price insurance
• Targeting the Ultra Poor