D. Mallick and M. H. Nabin. Cost effectiveness or serving the poor? Factors determining program placement of NGOs in Bangladesh. Economic modelling, 69:281–290, 2018.
Notes from Mallick Nabin 2018
consensus on its efficacy as a poverty alleviation tool is still elusive (for a recent survey of the literature, see Banerjee et al. (2015))
microfinance programs in Bangladesh target non-poor households without adhering to screening devices (Amin et al., 2003; Matin, 2005; Morduch, 2011)
evidence that microfinance institutions target program locations (villages and communities) based on criteria other than poverty incidences (Fruttero and Gauri, 2005; Salim, 2013).
mission drift: trade-off between serving the poor and cost effectiveness of the microfinance institutions. (Copestake, 2007; Gutiérrez-Nieto et al., 2007; Mersland and Strøm, 2010)
useful information about local characteristics is lost due to data aggregation at the higher level. Only data at the micro level can uncover the specific nature of, and reasons for, the “mission drift” at the local level.
Emprical results:
NGO coverage:
- decreases with village distance from NGO branch
- decreased with lack of physical infrastructure (ie, longer distance from paved road)
- is higher in villages with localised marketing opportunities (greater density of shops, closer proximity to local bazaar)
- increases with adoption of modern irrigation methods
Location choice impacts both profitability (loan repayment, loan disbursement transaction cost), and poverty alleviation. (Poorer people live in less profitable locations.)
Results suggest that location choice determined by cost effectiveness and loan recovery concerns, rather than the poverty alleviation objective.
Empirical Results
%%{init: {'securityLevel': 'loose', 'theme':'base', 'themeVariables':{'darkmode': 'false', 'background': 'Gray' }}}%%
flowchart RL
subgraph Geography
D[Distance from Marketplace]
I[Poor physical infrastructure]
MO[Localised marketing opportunities]
end
subgraph Production Function
PO[Better productive opportunities]
end
D[Distance from Marketplace] --> |decreases|N[NGO coverage]
I[Poor physical infrastructure] --> |decreases|N[NGO coverage]
MO --> |increases|N
PO --> |increases|N
Empirical Hypothesis
NGO coverage determinants
%%{init: {'securityLevel': 'loose', 'theme':'base', 'themeVariables':{'darkmode': 'false', 'background': 'Gray' }}}%%
flowchart RL
T[Transaction costs] ==> |increases|N[NGO coverage]
P[Profitability] ==> |decreases|N[NGO coverage]
T1[distance from NGO] --> T
T2[Physical infrastructure] --> T
P1[Marketing cost] --> P
P11[Distance from market] --> P1
P12[Market Thickness] --> P1
C[Borrower's cost] --> P
C1[Adoption of modern irrigation methods] --> C
C2[Soil quality of agricultural land] --> C
D[Distance] --> T1
D[Distance] --> P11
D --> P12
D --> T2
PR[Productivity] --> C1
PR --> C2
graph RL
style P fill:PowderBlue,stroke-width:0px
style P1 fill:PowderBlue,stroke-width:0px
style P11 fill:PowderBlue,stroke-width:0px
style P12 fill:PowderBlue,stroke-width:0px
style C fill:PowderBlue,stroke-width:0px
style C1 fill:PowderBlue,stroke-width:0px
style C2 fill:PowderBlue,stroke-width:0px
P11[number of shops in the village] --> P1[Marketing cost]
P12[distance from the local bazaar] --> P1
P1 --> P[Profitability]
C[Borrower's cost] --> P
C1[adoption of modern irrigation methods] --> C
C2[soil quality of agricultural land] --> C
graph RL
style T fill:DarkKhaki,stroke-width:0px
style D fill:DarkKhaki,stroke-width:0px
style I fill:DarkKhaki,stroke-width:0px
D[Distance from NGO] --> T[Transaction cost]
I[Physical Infrastructure] --> T[Transaction cost]
graph RL
style N fill:Tan,stroke-width:0px
style N1 fill:Tan,stroke-width:0px
style N2 fill:Tan,stroke-width:0px
style N3 fill:Tan,stroke-width:0px
N[NGO coverage] --> N1[Percentage of NGO member households]
N --> N2[number of NGOs operating]
N --> N3[NGO density in a village]
