Improving the Impact of Microfinance on Poverty

1
Improving the Impact of Microfinance on
Poverty Action Research Programme Presentation
to SEEP Poverty Assessment Working Group 22nd
October, 2003 Anton Simanowitz, Programme
Manager
2
Action Research Programme 20012004

• Imp-Act is a three-year action research programme
• 30 partners in 21 countries very diverse range
  of MFIs
• Initiated and funded by Ford Foundation
• Implemented by a team from three British
  universities
• the Institute of Development Studies (IDS), the
  University of Bath, and the University of
  Sheffield

3
Why collect poverty outreach data?
MFI need understand who is reached to be able to
deliver services targeted to the needs of
specific clients Donors need for transparent
information on social performance to make
resource allocation decisions Monitoring changes
in poverty status allows for action to be taken
to respond Measuring changes over time to be
able to determine impact and value added by MFI
4
Different needs in measuring poverty outreach

• What is poverty? Income or broader affects how
  poverty is measured and work of MFI
• Comparability and simplicity. Measuring income
  is relatively simple, but this is only one aspect
  of poverty
• Contextualisation. Poverty varies in nature
  according to location, therefore all methods must
  be locally adapted.
• Usefulness. How to balance MFI need for useful
  data with requirement for comparability?

5
Work of Imp-Act

• Developing systems based on needs of partners and
  their stakeholders
• Strengthen internal poverty assessment systems
• Verify and test these systems using CGAP poverty
  assessment
• Linkages to national/international poverty lines
• Promote examples of good practice based on the
  credible and useful systems developed by partners
  eg. BRAC,
• LAPO, PRADAN, PRIZMA, SAT, SEF, SHARE

6
Examples of poverty assessment tools
1) Single proxy indicator linked to
national/absolute poverty level

• Indicator links to poverty, but does not
  necessarily correlate with poverty statistics.
• Grameen Bank use of land holding
• Cashpor Housing Index

7
Examples of poverty assessment tools

• Indicator can be selected with strong linkage to
  locally contextualised poverty
• ? Second stage needed to make the linkages with
  national/international poverty lines.
• ?/? Quite easily replicated, but need to
  contextualise eg. Failure of CHI in South Africa
  where many people have homes in other places.

8
Examples of poverty assessment tools
2) Score-cards closely correlated to national
poverty indicator(s)
Collecting an indicator(s) correlating directly
with those in national poverty survey eg. PRIZMA,
LAPO
9
Examples of poverty assessment tools
2) Score-cards closely correlated to national
poverty indicator(s)

• PRIZMA Poverty Score card risk of being poor
• Track client poverty status as part of the credit
  scoring process
• 2. Locally relevant indicators are linked to
  national poverty data through LSMS poverty survey
• 3. Data is integrated into on-going monitoring
  system to allow for tracking of client status and
  assessment of impact.

10
(No Transcript)
11
Examples of poverty assessment tools
PRIZMA poverty score card key points

• Indicator development
• - multiple data sources LSMS, UNDP, Early
  warning systems research, PAT, Focus groups
• - simple picture of income poverty
• 2. Simple to implement
• - half of data already being collected
• - testing of questions to only include those
  that are easy to collect
• 3. Credibility
• - indicator selection process
• - weighting of indicators determined by LSMS
• - multiple indicators compound accuracy

12
Examples of poverty assessment tools
PRIZMA poverty score card key points
3. Cost-effective - four additional, simple
questions - once in place virtually without
cost 4. Data quality - use of LOT quality
assurance - part of internal audit function 5.
Management use for social performance -
integrated into MIS - cross-tabs with 30 MIS
variables - mixture of indicators with different
sensitivity to change
13
Examples of poverty assessment tools
2) Score-cards closely correlated to national
poverty indicator(s)

• LAPO Participation form (work in progress)
• NB similar to many Grameen means test forms
• Participation form includes a weighted list of
  poverty related variables, which provide an
  overall poverty score
• 2. Collected by field staff for all clients
• 3. Indicators correlated with national LSMS
  poverty data
• 4. On-going monitoring systems utilising the
  poverty score
• 5. Base-line information is also provided for
  future impact assessment work.

14
(No Transcript)
15
Examples of poverty assessment tools

• LAPO Participation form (work in progress)
• Verification and strengthening of the
  participation form using CGAP PAT
• Key stage is to test the relevance of indicators
  selected for the participation form and the
  weighting given to them PAT important step in
  this process

16
Examples of poverty assessment tools
2) Score-cards closely correlated to national
poverty indicator(s)

• Can provide accurate picture of poverty, which
  can be easily correlated to national poverty
  statistics
• Problematic if good national data is not
  available
• Fairly complex system, as it is normally
  necessary to include a number of indicators and
  to then decide if they should have different
  weightings
• ? Process of testing rigour of indicators in each
  context is quite difficult
• ?/? Can be contextualised, but correlation is
  basically with income poverty

17
Examples of poverty assessment tools
3) Geographic targeting Use national and local
data to identify the poorest areas in which to
work, and use national poverty data to make
comparisons

• PRADAN
• PRADAN works holistically to develop and support
  self-help groups.
• Very successful in using a detailed system of
  geographical targeting reaches a wide range of
  population all but the bottom 5 and top 20.
• Using national data on the relative poverty of
  the operational areas can relate this to the
  national poverty line and /day comparators

18
Examples of poverty assessment tools
PRADAN

• National data. Select the poorest regions in the
  country
• Local data. Select most marginalised communities
  with these areas
• Degraded natural resources
• Adherence to traditional technologies and modes
  of production
• Low productivity of labour and capital
• Lack of access to or control over productive
  assets
• Lack of access to financial services, business
  support services, knowledge resources and markets
• Lack of social and economic infrastructure

19
Examples of poverty assessment tools
PRADAN

• 3. Focus on reaching the poorest within the areas
  identified
• Simple wealth ranking to understand
  characteristics and needs of very poor, and to
  identify poverty pockets
• Use of visual methods to exclude the visibly
  better-off
• 4. Attempt to cover the whole of the poor and
  very poor populations there

20
Examples of poverty assessment tools
3) Geographic targeting
Results National Poverty line is much lower
than US/day 33 of rural population below
national poverty line (0.23/ day) 44 below
international poverty line (1/day) PRADAN
operational area (Jharkhand) 57 below
national poverty line 80 below 1/day
21
Examples of poverty assessment tools
3) Geographic targeting
?/? In very poor countries where a substantial
number of the population live below US1/day this
can be effective. In other contexts, this
would not be sufficient, but might make a very
effective first stage of poverty assessment.
22
Examples of poverty assessment tools
(4) Locally defined poverty measures

• SEFs PWR
• Mapping the village and listing all households on
  cards
• Wealth ranking by card sorting with reference
  group of 3-6 people
• Repeat of reference group with new groups two
  more times
• Rigorous checking of results
• Use of information give by participants to set a
  cut-line for programme inclusion

23
Examples of poverty assessment tools
PWR

• 1. Locally defined ie. not just money, therefore
  is likely not to completely correlate with /day
• 2. Very accurate through triangulation
• 3. Cost-effective and is widely applied. SEF has
  now ranked in excess of 300,000 people
• 4. Main cost is skilled facilitators
• facilitation skills also used widely in the work
  of loan officers
• staff time involved in PWR is partly or fully
  off-set by the marketing impact of holding a PWR
  exercise in a community.

24
Examples of poverty assessment tools
Linking PWR to international measures

• Reference indicators. eg. the pension line in SA
  rough estimation, but reasonably accurate
• 2. Geographic. Using national data calculate
  poverty of clients based on based on percentages
  in the area below certain poverty level and set
  the PWR cut-line accordingly

25
Examples of poverty assessment tools
Linking PWR to international measures

• 3. Calibrating PWR results.
• Sample of clients at different levels on PWR
  build into PWR process
• Sample survey to calibrate a region
• Eg. van de Ruit and May
• Comparison with CGAP PAT shows strong
  correlation
• PAT can be correlated to national/international
  poverty lines
• By inference PWR is similarly correlated.

26
Examples of poverty assessment tools
Linking PWR to international measures
4. PIR (participatory income ranking). Would
probably be possible ie. telling people to rank
by income not other poverty definition, but
defeats the whole purpose!
27
Lessons from Imp-Act Experience

• Support the needs of MFI
• 1. Monitoring not just assessment
• Information with which to manage day-to-day
  social performance
• Potential to see changes in client status and to
  be able to respond to this

28

Source Imp-Act partner
29
Lessons from Imp-Act Experience
Maximum flexibility to fit with MFI objectives
and context very different contexts of
organisations very different social
objectives very different definitions of poverty
impact need for context specific flexible
systems
30
Lessons from Imp-Act Experience
Optimal ignorance credibility for whom? Focus
on what is practical and useful, not on 99.99
accuracy
31
Improving the Impact of Microfinance on
Poverty Action Research Programme www.Imp-Act.o
rg www.microfinancegateway.org/impact