Module

1
Module 1 Q2 History, Concepts Definitions
Training Seminar - Q-Squared (Q2) Combining
Qualitative and Quantitative Approaches in
Poverty Analysis (www.q-squared.ca)
International Development Research Centre (IDRC)
(Canada)

• Paul Shaffer
• Centre for International Studies
• University of Toronto, Canada

Bangladesh Rural Advancement Committee (BRAC)
Dhaka, April 3-4, 2006
2
Objectives and Topics Covered

• Objectives
• To provide an understanding of the history of the
  recent reemergence of Q2
• To provide an understanding of what Q2 means,
  different ways of combining approaches to poverty
  analysis and different stages of poverty
  analysis.
• Overall, to provide historical and conceptual
  context for subsequent modules.
• Topics Covered
• Abbreviated History of Q2 in Poverty Analysis
• Definitional Issues What does Q2 Mean?
• Ways of Combining
• Stages of Poverty Analysis
• Organizing Framework
• A Few Identification Issues
• A Few Causal Issues

3
GROUP WORK 1

• 1. Divide into three groups
• Q What do the terms qualitative and quantitative
  mean to you? Provide a definition and give 3
  examples of what you mean, based on work you have
  done.
• Hint the definition can include a number of
  different elements/dimensions
• 2. Take 5 minutes
• 3. Nominate 1 person to present your answer to
  the group

4
GROUP WORK 1 Rationale

• The Group Work had three objectives
• 1) to illustrate that there are multiple
  potential meanings to be given to the
  qualitative/quantitative terminology and that
  there is no consensus about how to define it
• 2) to set up the discussion of the Kanbur
  typology Topic 2
• 3) to provide support for the decision not to use
  the terminology in the case studies but to refer
  to exactly what is being combined Topic 2

5
1. Abbreviated History of Q2 in Poverty Analysis

• There is nothing new about mixed method (Q2)
  analysis of poverty in the developing world.
• In the 1960s 70s there were a number of
  important publications on this theme and research
  projects oriented in this direction. Classic
  publication is Michael Liptons
  Interdisciplinary studies in less developed
  countries. Journal of Development Studies, 7
  (1), 1970.
• The recent reemergence of interest in
  interdisciplinary studies of poverty in the
  developing world reflects the PPA challenge to
  the way poverty is usually analysed in the
  applied microeconomic tradition (consumption
  poverty). As such, it is very much an applied
  debate, only loosely related to academic
  debates. These two traditions (PPA Consumption
  Poverty) are the backdrop for much of what is
  discussed in the training module 2 on Poverty
  Correlates/ Characteristics (though much of what
  is said generalises beyond this debate) .
• Consumption Poverty has been the gold standard in
  poverty analysis for some time. It has its roots
  in poverty studies undertaken by Roundtree and
  Booth in England in the 19th century, though the
  methods are not the same as present. It draws on
  utility and consumer theory as well as nutrition
  science.
• The PRA challenge to consumption poverty began in
  the 1980s-90s. It was based on some of the
  following considerations 1) peoples definitions
  of poverty should count in poverty analysis 2)
  short-cut methods of data collection should
  supplant/supplement big household surveys 3)
  people should actively participate in poverty
  analysis which may be a catalyst for social
  change. It uses a range of short-cut methods,
  some drawn from social anthropology, such as
  focus group discussions, semi-structured
  interviews, rankings, mapping, etc.

6
1. Abbreviated History of Q2 in Poverty Analysis
(Cont.)

• The reemergence of interest in Q2 is due to the
  recognition that much more attention has been
  placed on conducting PPA and consumption-based
  poverty analyses separately than on integrating
  the two in mutually beneficial ways.
• The present Q2 initiative began with a conference
  organised by Ravi Kanbur at Cornell (books from
  the Cornell conference will be distributed). A
  second conference organised by Kanbur and Shaffer
  was held at the University of Toronto in 2005
  (www.utoronto.ca/mcis/q2). Many of the case
  studies in the modules draw on papers from this
  conference which appear as Q2 Working Papers on
  the Q2 website (www.q-squared.ca).
• The International Development Research Centre
  (IDRC) has supported this initiative via a
  research project housed at the Centre for
  International Studies, University of Toronto. The
  project finances information dissemination
  (website, working papers), training, pilot
  studies as well as a third Q2 Conference which
  will likely be held in Hanoi in 2007.
• NB There are other Q2-related projects on-going
  including a forthcoming special issue of the
  Journal of Development Studies (edited by Toye
  and Hulme) work conducted by the Chronic Poverty
  Research Centre (UK), the Institute of
  Development Studies, UK, etc. etc..

7
2. Definitional Issues What does Q2 mean?

• The qual/quant distinction is used in different
  ways and means different things to different
  people.
• Kanbur has phrased it according to the following
  five dimensions of difference
• Type of Information Numerical/non-Numerical
• Type of Population Coverage Specific/General
• Type of Population Involvement Active to
  Passive
• Type of Inference MethodologyInductive to
  Deductive
• Type of Disciplinary Framework Social Sciences
  to neo-classical economics
• These categories help to unpack the Q/Q
  terminology but they tend to collapse. For
  example
• Much narrative data collected in discussions can
  be transformed numerically by counting, scaling,
  ranking, etc. As shown in Group Exercise 1.
• Small n studies can involve collection of
  standard numerical data (in a household survey)
  or narrative information (e.g. semi-structured
  interviews) as can large n studies (though it is
  more rare for ethnographies/PPAs to be done in
  large n studies)
• All research methodologies combine induction and
  deduction though in different ways.
• Within most of the social science disciplines
  there are established numerical, statistical and
  mathematical traditions in addition to traditions
  focusing on narrative, historical or comparative
  analysis (e.g. sociology, political science,
  economics, history, geography, etc.)

8
2. Definitional Issues What does Q2 mean? (Cont.)

• There are two ways to go
• Go Philosophical (Epistemology). Many of the
  differences in the Kanbur typology are derivative
  of a categorisation in the philosophy of social
  science which distinguishes between
  empiricism/positivism hermeneutics/constructivism
  critical theory/critical hermeneutics. Some
  papers do this (Kanbur and Shaffer, 2005,
  Epistemology, Normative Theory and Poverty
  Analysis Implications for Q-Squared in Practice,
  Q2 Working Paper No. 2, October) but requires
  wading through a great deal of jargon and
  terminology (fact/value subject/object etc.)
• Go Specific. Refer to exactly what is being done
  in different studies. We will adopt this strategy
  in the training. From here on, we will try not to
  refer to qualitative/quantitative terminology
  when distinguishing different aspects of
  approaches to poverty analysis.

9
3. Ways of Combining

• As with the Q2 distinction, ways of combining
  approaches are categorised in many different
  ways. Some examples include
• Sequential vs. Simultaneous Mixing
• Confirming, Refuting/Integrating, Merging
• Primary, Lead, Check/Follow-up
• Parallel, Sequential, Iterative
• Iteration, Linkage, Triangulation, Convergence
• Merging, Sequencing, Mixed Suite
• Etc. Etc.
• The present typology distinguishes between
• Putting Together. Results of different
  approaches conducted separately are put together
  with a view to enrich or confirm/refute each
  other.
• Methodological Integration. The outputs of one
  approach feed into the design or methods of
  another (e.g. using ethnographic results to
  inform household survey questionnaire design) or
  the integration of techniques typically used in
  one approach in another, (e.g. selecting PRA
  sites probabilistically and calculating standard
  errors for the numerical data generated).

10
4. Stages of Poverty Analysis

• Three stages are distinguished according to the
  following issues
• Identification Issues. Analysis of
  characteristics/correlates of poverty. Addresses
  question Who are the poor? Module 2
• Causal Issues Analysis of causes of poverty
  (chronic or transitory). Addresses question Why
  are they poor and/or why do they enter into and
  exit from poverty? Module 3
• Prescriptive Issues Analysis of remedies for
  poverty. Addresses question What to do about it?
  Not addressed in training
• Notes
• Distinctions are not hard and fast (esp. between
  stages 2 and 3) but still useful to bring out
  different types of analyses.

11
5. Organizing Framework

• The Framework Presents the
• Stages of Poverty Analysis (Identification vs.
  Causal)
• Ways of Combining Approaches (Putting Together
  vs. Methodological Integration)
• Q2 Definition (Shows Exactly what is Being
  Combined)

12
6. A Few Identification Issues
There are three separate issues in the
Identification Stage 1) Determining the
dimensions/conception of poverty (Poverty Basket
2) Determining the poverty line (Poverty
Cut-off) and 3) Adding up those below the
Poverty Line (Aggregation). The Consumption and
PPA Poverty Approaches do these differently.

• These issues will be discussed again in Module 2
  on the Analysis of Poverty Correlates/
  Characteristics.

13
6.1 Illustration of the Consumption Poverty
Cut-Off The Food-Share Method
Minimum Exp. to meet caloric require-ments
Food Poverty Line
Total Poverty Line
Approach Determine cost of a basic needs bundle
based on consumer behavior of a reference
population group Stage 1 Calculate Food Poverty
Line 1.1 Select reference population group (e.g.
bottom 20 percent of distribution 1.2 Determine
caloric content of foods consumed 1.3 Scale up
food consumed to a level which meets minimum
caloric intake (holding relative shares
constant) 1.4 Cost out the food basket Stage 2
Calculate the non-Food Component 2.1 Calculate
non-food share in consumption of those whose
total expenditure equals the fool poverty line
(or of any other population group) using averages
or regressions 2.2 Add this portion to the food
poverty line. The key point to note is that this
poverty line is applied uniformly to all
households in the dataset.
14
6.2 Illustration of the PPA Poverty Cut-Off A
Well-Being Ranking Matrix
Re. the Poverty Basket, this would be determined
by participants in dialogue Re. the Poverty
Cut-off, this would represent the distinction
between the worse-off and middle categories in
the first matrix and the poor and just doing OK
categories in the second. Re. Aggregation, if
PPA poverty attempts to aggregate, it represents
the average population percentage falling in the
poor categories of well-being matrices. This
which can be thought of as relative poverty
incidence, i.e relative to the populations in the
individual rankings.
15
6.3 Illustration of Aggregation in Consumption
Poverty The FGT Poverty Measures
The industry standard for the aggregation of
consumption poverty is the Foster, Greer,
Thorbecke (FGT) Class of Measures represented as
follows
where z is the poverty line gi z - yi, the
consumption shortfall from the poverty line of
the ith poor person, q the number of poor persons
and n the total population. The FGT measures
take on different properties as the value of
alpha changes. When ? is set to 0, the index
collapses to q/n or the percentage of the poor
(Poverty Incidence). When ? is set to 1, the
index measures the average consumption gap g,
multiplied by the percentage of poor persons
(Poverty Intensity or Poverty Gap). When ? is set
to 2, the index squares the average consumption
gap g, and then multiplies by the percentage of
poor persons (Poverty Severity or the Squared
Poverty Gap). Aggregation in Consumption
Poverty is facilitated by the Poverty Basket and
Cut-off Consistency (the same poverty dimension
and line is used).
16
7. A Few Causal Issues

• As discussed in Modules 3 (Causal Analysis)
  Module 4 (Impact Assessment), there are
  different empirical ways to try and show
  causality. These differences may be distinguished
  according to whether or not they 1) present a
  counterfactual scenario (the situation that would
  have occurred in the absence of the causal
  intervention 2) an analysis of transmission
  mechanisms, or the actual causal processes and
  pathways.

17
7. A Few Causal Issues (Cont.)

• 1. Experiments Module 4. Used to assess
  poverty impact of development programs. The
  approach is to randomly assign program
  participation and compare the outcomes of program
  participants (the treatment group) and
  non-participants (the control group). There is
  a counterfactual (the control group) but no
  discussion of transmission mechanisms (how impact
  is or isnt generated). Experiments are usually
  viewed as the gold standard in program impact
  assessment.2. Quasi-Experiments Module 4.
  Similar to experiments but assignment is not
  random so a comparison group is constructed
  statistically. The comparison group is supposed
  to be identical to the treatment group in all
  respects except for the effects of program
  participation. There is a counterfactual (the
  control group) but no discussion of transmission
  mechanisms.
• 3. Process Tracing/Causal Mapping Modules 3
  4. Often used in the PRA, social
  anthropological or theory-based evaluation
  traditions where linkages are drawn between
  events or actions which are given a causal
  interpretation, e.g. asking people to explain the
  main events leading to increases or decreases in
  their well-being. There is a detailed account of
  transmission mechanisms but no counterfactual
  analysis.
• 4. Comparative Methods. Similar or dissimilar
  situations are compared over time or space and
  causal inferences are drawn. Often used to make
  establish causality in history, comparative
  sociology, etc. Combines transmission mechanisms
  and counterfactual analysis in an informal way.
• 5. Causal Models Module 3. Transmission
  mechanisms, i.e. linkages between variables are
  formally modelled, which allows for the
  simulation of different scenarios (i.e.
  counterfactual analysis).
• NB. Note these for now. They will be explained
  with examples later.

18
8. Summary Conclusion

• The reemergence of Q2 has been driven by the
  practical objective of improving the integration
  of approaches to poverty (mainly consumption and
  PPA poverty approaches).
• The meaning of Q2 is ambiguous. As a result, we
  ban the Q2 terminology and refer to exactly what
  is being combined in the case studies.
• Ways of combining distinguish putting together
  and methodological integration.
• Stages of poverty analysis distinguish the
  identification stage (the analysis of poverty
  correlates/characteristics) and the causal stage
  (the analysis of why people are poor or
  enter/exit poverty)
• The organizing framework used in Modules 2, 3
  and 4 integrates points 2-4.
• Three key issues in the Identification stage
  concern determining the dimensions of poverty
  (Poverty Basket) the poverty line (Poverty
  Cut-Off) and Adding up those below the line
  (Aggregation). Consumption and PPA poverty
  approaches do these things differently.
• There are different ways to conduct causal
  analysis of poverty which may be distinguished by
  whether on not they provide a counterfactual
  scenario (experiment, quasi-experiments), an
  analysis of transmission mechanisms (process
  tracing/causal mapping) or both (comparative
  methods, causal models).
• The purpose of Module 1 is to provide some
  historical and conceptual context for subsequent
  modules. The conceptual issues will be
  illustrated with Q2 examples and made more
  concrete in the next three modules.