3. Group Liability Loans
• Pros (?)
– “Key innova&on responsible for expansion of access to
credit” (Morduch; Armendariz de Aghion, Morduch; Microcredit Summit Campaign 2005)
– Addresses market failures of adverse selec&on and moral hazard
– Peer pressure improves repayment
– Increased access to uncollateralized loans
• Cons (?)
– Tension, harming social networks
– Free‐riding clients
– More costly to good clients; difficult to aract new clients,
dropout
Mo&va&on Defini&ons Study Designs Results
4. Individual Liability Loans
• Pros (?)
– Reduced group tension
– Preferred product for many
– Increased client reten&on and growth as a result
• Cons (?)
– Does not address possibility of adverse selec&on and moral hazard
– Shi^s burden of monitoring and enforcement to credit officer
Mo&va&on Defini&ons Study Designs Results
5. Group vs. Individual Liability Loans
The unanswered ques&ons:
– Is repayment actually higher in group liability than individual
liability lending?
– Is there a tradeoff in terms of client base?
– How do these factors affect overall ins&tu&onal profitability?
IPA and Green Bank designed an experiment to answer these
ques&ons
Mo&va&on Defini&ons Study Designs Results
6. Liability vs. Lending
• Group liability
– Members are borrowers and guarantors
• Group lending
– Group structure
– Individual liability
This project tests removal of group liability, not of other group
borrowing characteris&cs
Mo&va&on Defini&ons Study Designs Results
8. Evalua&on Design: Product Characteris&cs
• Green Bank BULAK product
– Women in rural areas
– Two levels of liability: group, center
– Ini&al loan size: P1000‐P5000
– Loan term: 4‐6 months
– Weekly center mee&ngs
– Weekly deposit into center, group, personal savings accounts
Mo&va&on Defini&ons Study Designs Results
9. Evalua&on Design: Conversion Phase
• Liability structure
– 169 exis&ng centers in Leyte (total)
– Randomly converted centers into individual liability
– Remaining group structure did not change
– Social influences on repayment likely remained
• Savings
– Dissolved group, center savings
– Personal savings remain
– No funds saved toward center ac&vi&es
Because all borrowers were screened under group liability, isolates
impact of group liability peer pressure to encourage payment.
Mo&va&on Defini&ons Study Designs Results
10. Evalua&on Design: Expansion Phase
• Liability structure(s)
– New centers in Cebu, Bohol, Panay, Negros
– Three liability structures
1. Group
2. Individual
3. Phased‐in individual (group liability in the first cycle only)
– Liability structure randomly assigned
Does performance change when clients are screened into the
group under different liability structures?
Mo&va&on Defini&ons Study Designs Results
11. Selected Results
• Repayment
– No measurable difference in repayment rates
– True for both conversion and expansion phases
Mo&va&on Defini&ons Study Designs Results
12. Selected Results: Con&nued
• Dropout
– No significant difference between group and individual liabili&es
– Less likely to dropout in phased‐in individual
• Client appeal
– Individual liability centers aracted more clients, grew larger
– Exis&ng centers converted to individual liability much less likely to
dissolve
• Account Officer appeal
– AOs less likely on average to open centers in individual‐liability or
phased‐in areas than group‐ (not significant on individual)
Mo&va&on Defini&ons Study Designs Results
13. Selected Results: Con&nued
• Time alloca&on of Account Officers
– Time alloca&on did not change between group and individual
centers under conversion phase
– During expansion, AOs spent more &me (90 min per week) on
repayment ac&vi&es in individual liability centers
o Shi^ in &me away from program introduc&on/marke&ng?
Consistent with lower likelihood of establishing individual centers.
– No difference in &me alloca&on between phased‐in individual &
group centers
Mo&va&on Defini&ons Study Designs Results
14. Selected Results: Con&nued
• Center rela&ons
– Reduced monitoring by peers (less likely to be able to predict
defaults)
– Less social cohesion?
o As evidenced by less spent on Christmas par&es
• Center composi&on
– Prior members more likely to know new members well under
individual liability; new members not likely to know each other
o Exis&ng members no longer fear bad rela&onships
– Exis&ng clients were not uniformly pro‐individual; some
supplemented loans through other ins&tu&ons a^er conversion
Mo&va&on Defini&ons Study Designs Results
15. Results: Summary
• Liability structure had no effect on repayment
• Individual liability can lead to larger lending groups
• Credit officers
– Less likely to open individual liability centers
– Spend more &me on individual liability centers
o Might change with greater experience with individual liability?
Expanding individual liability lending products may deepen
outreach and provide desired flexibility to clients
…but, replica&on needed to further understand these effects.
Mo&va&on Defini&ons Study Designs Results