RRA had already taken in insights and methods from these other sources, and provided a quick way of gathering information on local realities building from local people's insights. RRA questioned the urban biases implicit in outsiders' role as development consultants (the so-called "development tourism") by giving a more significant role to local knowledge, nonetheless still elicited for analysis by outsiders. Further, RRA challenged the way knowledge was generated, and responded to the challenges of "hard science" (McGee, 1997) by stressing the two key principles of "optimum ignorance" (find out as much as you need to know now) and "appropriate imprecision" (there is no need to know everything exactly) (Cornwall, 2000). RRA also opened the door to methodological experimentation, appearing in different forms, among which Participatory RRA emerged later developing into PRA.

The core difference between RRA and PRA is not only in the extent to which local people are included in the research, but in their ultimate purpose. "A PRA is intended to enable local people to conduct their own analysis, and often to plan and take action" (Chambers, 1994a). By becoming a way in which participation was enacted, the qualitative and often visual tools used in PRA acquired a new and distinctive characteristic. In practice, however, the extent to which these tools effectively brought about participation in all the phases of the project cycle has been questioned leading to many criticisms from those who sought a radical change in the way development efforts were conducted.

Without dwelling on the pros and cons of PRA and the way it was implemented, it is important to stress that the flexibility of the methods meant that it was possible to use them within alternative methodologies. Often in practice, therefore, it was their cost effectiveness and the timeliness with which they produced results, rather than their empowering effects, which underpinned the support they were given.

If the widespread adoption of participatory techniques challenged the extent to which their distinguishing features were maintained in practice, a further challenge was posed by the "scaling up" of PRA from project planning to input into policy making. The most evident form in which this scaling up has taken place has been the Participatory Poverty Assessments (PPA) performed by the World Bank, introduced as complement to Poverty Assessments in the face of criticisms of their exclusive money metric focus. These PPAs have spread rapidly. By 1998, half of the completed poverty assessments performed by the World Bank included a participatory component (Robb, 1999).

By delinking participatory techniques from the direct involvement with community projects and planning, the road was open for more extractive uses of the PPAs. And indeed the emphasis was initially in providing information which could provide information for better policies. A new generation of PPAs, however, seems now to have taken up the challenge of "influencing the policy process". A wider spectrum of actors have been drawn into these processes, which place as much emphasis on the impact of their learning on their agency within the policy processes as on the information that is produced (Cornwall, 2000).

The Methodology of Participatory Poverty Assessments

From the methodological point of view, PPAs can be classified as contextual methods of analysis (Booth et al., 1998) i.e., data collection methods which (taking a poverty-related definition) "attempt to understand poverty dimensions within the social, cultural, economic and political environment of a locality" or of a group of people, by prioritizing local people's perceptions. Though different research methods can be contextual to different degrees, this categorization juxtaposes participatory methods with methods which aim to standardize data collection and analysis, as for example in large household surveys. This way of classifying approaches offers the advantage of breaking away from the quantitative-qualitative dichotomy which is generally seen as characterizing the comparison of survey and participatory data, but which does not consider the potential of obtaining quantitative information from PPAs (through rankings for example; other ways of quantifying information are more debatable).

Participatory approaches, however, are not only contextual, they also emphasize poor people's creativity and ability to investigate and analyze their own reality (Chambers, 1994a). So, they try not only to understand reality at the local level, but they do so through local people's own analysis. For a researcher, this involves not only adopting a set of different tools, but also completely different behaviors and attitudes. By recognizing their role as outsiders, researchers need to redefine themselves as facilitators who have to share in local knowledge and be willing to review their own values and perceptions critically. These behavioral elements are central to the success and truthfulness of the exercise, though they are also among the most difficult to standardize and to verify ex post, when looking at existing research.

One important challenge to participatory poverty assessments is that the nonextractive nature of the exercise and the efforts not to raise expectations that cannot be met is not easy to reconcile with the policy focus of the poverty assessment, when those policies are remote from the local level.

The Tools

Various tools are used in PRA. A classification into visualized analysis, interviewing and sampling, and group and team dynamic methods has been suggested by Cornwall et al. (1993) quoted by Estrella and Gaventa (1998). Examples include:

The variety of these methods and their flexibility distinguishes them from other methods which elicit self-perception data through structured questionnaires (as for example in identifying the minimum level of income necessary for the poverty line, e.g., Pradhan and Ravallion, 1998). As these tools are often adopted in a sequence, the assessment can be tailored to fit the context and the issues to be analyzed appropriately. Further, different tools are used to triangulate (i.e., validating through cross-checking) the results which might allow different insights to emerge.

In performing a PPA, care is needed to choose tools and sequences which are well suited to capture the core elements of deprivation in the specific context and the specific aspects of interest in the assessment. This might imply, for example, adopting different sequences for urban and rural contexts. As an example of the variety of issues which might be investigated in a PPA, we present a description of the issues and methods considered in the World Bank's Zambian PPA in rural areas (Table 1).

Looking through the table, two main features stand out: one is the variety of issues discussed, the other is the number of methods suggested for treating every issue. The detailed breakdown shows that different issues can be dealt with jointly or sequentially, which reinforces the importance of careful planning of the sequences to be adopted, not least to avoid repetition which would be time wasting as well as boring for the participants. It is also clear that, though a PPA is meant to inform policies, not all types of poverty-related research would be equally concerned with discussing policy-options, especially if the research is unlikely to have a direct bearing on the options available. It could therefore raise expectations which could not be fulfilled.

The Analysis

The final phase of a participatory poverty assessment is the analysis of the outputs, and feeding back the results to the community as well as to other final users and in the case of PPAs, those involved in poverty reduction strategies. This last stage entails making sense of all the outputs produced by different groups on different media to arrive at an assessment. Ideally, it should be local people themselves who synthesize the results, but this is not necessarily the case. It is more common that field reports are given back to the community and scrutinized by them. New generations of PPAs are trying to make this "final phase" the beginning of a process of change in the community and in the way they participate into the policy process, going well beyond the production of one snapshot of poverty at a given point in time.

Understanding how the results are arrived at should be helped by the fact that participatory approaches include the documentation of all the stages of the process. In fact, the documentation of the process might help in understanding features like the interpretation given to quantitative outputs (ordinal or cardinal), or the role played by different groups (whether all have been given the same exercise or whether different people have been asked to synthesize outputs).

Documenting the process is rather ineffective as a mechanism to ensure quality – scrupulous and committed researchers will perform and document a process characterized by attentive questioning of their own assumptions, while others will perform all these tasks as steps in a recipe, without the critical awareness which characterizes good processes. Further, not unlike the case of long methodological annexes in monetary poverty assessments, it is unlikely that the readers will have much time or expertise to delve into the procedural details, so that the key findings as highlighted by the researchers are the ones which are going to have an impact.

Important challenges to how truly participatory the process is arise in this final phase – the conversion of local reality as analyzed by poor people into final reports. And the problems which arise seem to be hard to face even when following best practices. Inevitably, every assessment is done for some purpose, and that purpose is likely to be reflected in the elements which are highlighted. And even the more committed and soul-searching of the researchers adopts some criteria to understand the reality she is faced with in order to make sense of its complexity.

A recent analysis of the findings of the first wave of PPA in Africa found that "certain themes were noted (in the PPAs) that have not been highlighted in the main overviews of PPA results. However, what is perhaps more striking is what is missing even from this expanded list. There are a range of other issues for poverty analysis that seem important a priori but that are notably absent in the majority of cases. There are several possible ways of explaining the particular pattern of emphases and absences in the first round PPAs." (Booth et al., 1998). They cite "obvious" selectivity at various levels due for example to "pressures on writers of country synthesis reports to highlight findings that have immediacy for policymakers" or the indirect influence of the strategic policy framework adopted by the World Bank on the way themes are organized. It is hard to see how PPAs can get away from those kind of constraints, which represent both a natural need of the researchers to refer to some known context and a logical consequence of doing analysis for a particular purpose. This raises an important foundational point: can a truly participatory approach deal with a priori held beliefs? Should it, and if so how much? To give a practical example, should a researcher prompt local people to discuss an issue which they have not mentioned on the basis of some prior-held belief that the issue is of importance?

The notion of capital is a powerful entry point for causal explanations of poverty. Capital is understood in a broad sense as any "stock" which is capable of being stored, accumulated, exchanged or depleted, and which be put to work to generate a "flow" of income or other benefits" (Booth et al., 1998).

New Frontiers: Combined Methods

From the debate on PRA, new insights have been gained that led to questioning the original characterization of participatory approaches as antithetical to the collection of household survey data and to attempts to use participatory and nonparticipatory techniques interactively, exploiting their respective strengths. Carvalho and White (1997) synthesizes these possibilities in terms of:

An interesting example is provided by Carter and May (1999) whose identification of households in a large survey was based on a livelihood classification scheme derived from a participatory assessment (similarly, Scoones, 1995 suggests using wealth ranking as cost-effective research tool for examining issues of wealth and poverty in rural context, perhaps setting the agenda for subsequent, more detailed and focused studies into particular aspects).

However, this development could be criticized as bringing participatory techniques into the mainstream poverty analysis tool kit, therefore, changing their non-extractive nature, making it respond to outsiders' priorities as well as bringing about the danger of a routinization of the process (Chambers, 1994b). This does not need to be the case, however. As some of the participatory methods lead to quantitative evidence, usually of an ordinal nature, there is a potential for linking qualitative and quantitative methods (Booth et al., 1998) without reducing the insights from participatory methods only to those which can be quantified.

The central challenge faced in using participatory methods for poverty analysis is implicit in moving participatory techniques from the project level to policy processes. While in fact there have been examples of PPAs which have greatly contributed to the policy debate at the national level, many examples of cosmetic participatory research, performed for extractive purposes and without a commitment to empowering local people to have a greater say in policy processes, can also be found. It is important to consider instances of comparative research adopting different methods of analysis, both participatory and non-participatory, highlighting their relative strengths and weaknesses. In the light of these debates, a move to new and integrated frameworks for the analysis of poverty seem almost inevitable.

References

Booth, D., J. Holland, J. Hentschel, P. Lanjouw and A. Herbert. 1998. Participation and Combined Methods in African Poverty Assessments: Renewing the Agenda. DFID Social Development Division, Africa.

Carter, M. R. and J. May. 1999 Poverty, Livelihood and Class in Rural South Africa. World Development, No. 27, pp 1-20.

Carvalho, S. and H. White. 1997. Combining the Quantitative and Qualitative Approaches to Poverty Measurement and Analysis: The Practice and the Potential. World Bank Technical Paper No. 366. World Bank, Washington, D.C.

Chambers, R. 1994a. The Origins and Practice of PRA. World Development, Vol. 22 No. 7.

Chambers, R. 1994b. PRA: Challenges, Potentials and Paradigm. World Development, Vol. 22 No. 10.

Cornwall, A. 2000. Beneficiary, Consumer, Citizen: Perspectives on Participation for Poverty Reduction, IDS.

Cornwall, A., I. Gujit and A. Welbourn. 1993. Acknowledging Processes: Challenges for Agricultural Research and Extension Methodology. IDS Discussion Paper No. 333.

deGraft Agyarko, R. 1997. Influencing Policy through Poverty Assessments: Theoretical and Practical Overview of a Changing Process. IDS Working Paper.

Estrella, M. and J. Gaventa. 1998. Who Counts Reality? Participatory Monitoring and Evaluation: A Literature Review. IDS Working Paper No. 70.

IDS. 1998. PPA Topic Pack. International Development Studies.

McGee, R. 1997. Looking at Poverty from Different Points of View: A Colombian Case Study. Doctoral Thesis, Faculty of Economic Studies, University of Manchester. Unpublished.

McGee, R. 1999. Technical, Objective, Equitable and Uniform?: A Critique of the Colombian System for the Selection of Beneficiaries of Social Programmes. SISBEN. Unpublished.

McGee, R. 2000. Analysis of Participatory Poverty Assessment (PPA) and Households Survey Findings on Poverty Trends in Uganda. Mission Report. 10-18 February 2000.

Pradhan, M. and M. Ravallion. 1998. Measuring Poverty Using Qualitative Perceptions of Welfare. Policy Research Working Paper No. 2011. Washington D.C.: The World Bank.

Robb, C.M. 1999. Can the Poor Influence Policy? Participatory Poverty Assessments in the Developing World. Washington D.C.: The World Bank.

Scoones, I. 1995. Investigating Difference: Applications of Wealth Ranking and Household Survey Approaches Among Farming Households in Southern Zimbabwe. Development and Change Vol. No. 36. p 67-88.

Contributed by:
Caterina Ruggeri Laderchi
Email: cruggeriladerchi@worldbank.org