Combining Participatory and Survey-Based Approaches to Poverty Monitoring and Analysis

1
Combining Participatory and Survey-Based
Approaches to Poverty Monitoring and Analysis

• For the purpose of
• Monitoring the Implementation of the PEAP

2
Combining Participatory and Survey-Based
Approaches to Poverty Monitoring and Analysis
Our Starting point

• Consider what Caravalho and White distinguish as
  the approaches to combine the two methods
• 1. Integrating the two approaches into one
  methodology
• 2. Using the two approaches to CONFIRM, REFUTE,
  ENRICH and /or EXPLAIN findings from the other

3
However many more issues were raised and
discussions were much broader
4
Presentation will cover ..

• Objectives (What do we want to achieve by
  combining)
• Strengths and weaknesses of each method
• Issues around combining and refocusing methods
• Conclusions for PPA2
• Conclusions for combining
• Sampling and linking
• UBOS role
• Analytical levels
• And What has been achieved so far

5
Combining Participatory and Survey-Based
Approaches to Poverty Monitoring and
AnalysisObjectives

• To respond to the wish of the Government of
  Uganda to achieve a better articulation of the
  relevant research processes and to obtain a
  better understanding of the results.
• To go beyond joint reporting of findings from
  non-statistical and statistical sources (which is
  well established in Uganda through the Poverty
  Status reports).

6
Combining Participatory and Survey-Based
Approaches to Poverty Monitoring and Analysis
Objectives

• we wish to
• Improve the two way flow of information between
  beneficiaries, service providers and policy
  makers, covering the information on
• Inputs (public spending etc)
• Outputs (The quantity and quality of immediate
  results)
• Immediate outcomes (e.g.access to services)
• Final (Multi dimensional) Poverty Outcomes

7
Combining Participatory and Survey-Based
Approaches to Poverty Monitoring and Analysis
Objectives

• Greater robustness in findings (less likely both
  methods would fail in a given instance)
• Complementarities (seeing a situation from
  different perspectives)
• Increased influence on policy makers

8
Survey Based MethodologySTRENGTHS

• .
• Definite comparative advantage in obtaining
  quantitative data
• Makes aggregation possible (data can be
  generalised)
• Allows systematic disaggregation of data, (we can
  measure trends within sub-groups).

9
Survey Based MethodologySTRENGTHS

• Allows comparison over time (particular strength
  with panel survey data)
• Allows simulation of different policy options
• Provides results whose reliability is measurable

10
Survey Based MethodologyWeaknesses

• Sampling and non-sampling error (particularly in
  under reporting of income and some expenditure).
  Although different sources can help to measure
  these errors
• Miss what is not easily quantifiable
• Fails to capture intra-household allocation (a
  particular problem)

11
Survey Based MethodologyWeaknesses

• Difficult to measure attitudes and behaviour (Has
  typically closed questions)
• Cost and length of time required for analysis
• Some feel this approach is extractive (morally
  questionable)

12
Participatory ApproachesSTRENGTHS

• Provides a richer definition of poverty
• More insight into casual processes
• More accuracy and depth of information on certain
  questions and in certain cases
• There is a possibility of being holistic (looking
  at a set of relationships as a whole)

13
Participatory ApproachesSTRENGTHS

• Ability to go immediately back to data and
  interrogate initial findings/puzzles (further
  interviews and observation)
• A wide range of resources for triangulation
  (systematic cross checking)

14
Participatory ApproachesStrengths Continued.

• It was noted that a PPA is not just a new type
  of study of poverty and its causes, but is a
  process which aims to achieve
• a better understanding of poverty
• New constituencies for anti-poverty action
• Enhanced accountability to poor people
• More effective policies and action
• i.e. Main strength of PPAs lies in identifying
  the range of both final and intermediate
  processes and issues that are important to
  poverty reduction.

15
Participatory ApproachesWeaknesses

• The real Observer bias issues (Lack of guarantee
  of objectivity ?)
• It is unknown how representative the data is of
  the National situation
• It is not suitable for providing definitive tests
  of hypotheses that apply to such wider
  populations.

16
Participatory ApproachesWeaknesses cont

• There are difficulties in verifying information
• No systematic disaggregation
• Perceptions of poverty are relative to changing
  environments

17
Issues around Combining and refocusing methods

• PPAs are not the best tool for monitoring final
  outcomes.
• Opportunities for using survey and participatory
  methods to confirm and refute each other are
  fewer than previously thought.
• Maintaining the essential differences between
  survey-based and participatory approaches is the
  best way to exploit their complementarities.
• Technocrats and politicians like to see a
  diversity in the types of evidence

18
Issues around combining and refocusing methods

• As it is important to generate quick feedback on
  PEAP implementation, there should be a relative
  shift of attention towards intermediate factors
  and policy implementation bottlenecks.
• This would seem to imply
• some change in focus of PPA research, and
• more attention to access to services etc. in the
  analysis of existing data

19
Conclusions for PPA2

• There was a wide agreement on the need to give
  the second national PPA a strong and quite
  focused analytical framework to guide field work
  and reporting. This will take the form of
  addressing unanswered questions from
• The reports from PPA1
• Analytical work on the household survey series,
  particularly its panel component
• A review of key implementation bottlenecks, - and
  intermediate input, output and outcome issues

20
Conclusions for combiningSampling and linkage
for PPA2

• From the discussion, three different reasons for
  exploring a fresh approach to sampling for PPA2
  emerged.
• The growing analytical interest in how people
  become less poor and hence in investigating
  poverty by studying successes
• The wish to be able to make general statements
  that carry weight with policy makers.
• The desirability of maximising design and
  analytical links between survey results and PPA
  themes and findings.

21
Conclusions for combiningSampling and linkage
for PPA2

• So.
• Sites for PPA2 will be chosen to maximise mutual
  linkages with the panel element of the Household
  surveys.
• PPA2 will also have purposively selected sites
• There will be over sampling of cases of special
  analytical interest

22
Conclusions for combiningUBOSs role

• UBOS will become more active in the analysis of
  data. In particular, they will provide summarised
  general and panel data, to provide information to
  help UPPAP in their sample selection. In areas
  selected UBOS will supply fact sheets of
  interesting (possibly conflicting) information
  that can be further researched.

23
Conclusions for combining analytical levels

• UBOS and UPPAP will undertake joint dissemination
  and sensitisation efforts
• They will create a circle of analytical linkage
  (rather than linear , in one direction)
• UPPAP will focus on asking why questions, and
• UPPAP will NOT make participatory work more
  survey-like in order to check other data.

24
Conclusions for combining(and finally)

• UPPAP and UBOS will develop a closer and more
  cooperative relationship by
• More sharing of information and analysis (improve
  communications between partners offices)
• Jointly develop a community based information
  system
• Sharing experience on how to stimulate interest
  in data and how to promote its intelligent use
• Collaborating to provide more reporting back at
  the grass roots level, especially in panel areas

25
What has already been done !

• UPPAP are currently designing PPA2 with
  significantly increased consultation with UBOS.
• Next step when selecting sites is to ensure
  plenty of overlap with the sites of the panel
  element in the household survey. 

26
What has already been done !

• UBOS has created a new research unit which will
  consist of Statisticians, sociologists and an
  economist to improve their analytical
  capabilities.
• This unit will provide information to assist with
  the design and further research for PPA2.