APPENDIX B: MESO ANALYSIS mapping across layers of marginalization By: Mary Picard
Table of Contents B.1 Layering Analyses ................................................................................................ 2 B.2 Concluding the Meso Level Analysis .................................................................... 5
List of Figures Figure 1: Illustration on Layering Macro Analyses, Identifying Marginalized Areas .... 6
Introduction
The macro-level analysis furnishes a lot of information about broad trends in the country, such as issues of poor governance, gender inequity, a political economy analysis writ large, effects of climate change, etc. Many, but not all, of the phenomena that explain poverty and social injustice will apply to the country as a whole. Spatial differentiation – differences from region to region – are likely to exist for many social, economic, demographic phenomena, and certainly for poverty itself. Thus, before deciding which locality to choose for the micro-level analysis and to guard against the urge to go where the Country Office is already working, the design team should be aware of regional inequalities and how these might explain why vulnerability exists where it does and who the vulnerable populations are. Thus, some investigation is warranted to see whether regional disparities exist and who they affect. This analysis may not need data gathering additional to what has been done for the macrolevel. How sufficient the data is for this analysis will need to be assessed, but any additional efforts to gather data should be kept to a minimum. The meso level seeks to organize the information and apply what is called spatial analysis to the data. For this type of analysis, the leading questions for the meso-level include: •
Where are the most marginalized population groups located in space?
•
What defines the boundaries around these groups and what are the intersections between different forms of marginalization?
In the end you are likely to find where multiple forms of marginalization co-exist.1 The possibilities for completing this entire exercise will vary considerably, owing to constraints such as the availability and format of the data as well as the availability of GIS capacity (not necessary but extremely valuable). Teams are encouraged to make the effort to apply spatial analysis, however, you are likely to find that some data are more readily available in map format than others – population size and density, climate change (i.e., natural resources or satellite imagery maps), livelihood systems, and conflict mapping. Even geographic marginalization can be deduced from topography/roads maps. You may want to begin with these and proceed to the other data layers in accord with data availability.
1
Readers might also be interested to know about a growing body of research on the geography of economic development. It looks at three geographic dimensions that describe the transformation of economies as they develop – economic density, distance, and division. It reveals that economic unevenness over space is not bad for growth; it is the connection (or integration) of lagging areas of the country with the leading areas of the country that will enable ‘unbalanced’ economic growth with inclusive development. This can be done through improved transportation and other means of linking laggard areas. See 2009 World Development Report.
1
B.1 Layering Analyses A series of analysis layers or maps is proposed for this purpose. The use of geographic information systems is ideal for this, where the digital data exists but even without the technology, plain maps of the country can be used as a base to sketch and identify marginalized areas. In doing this exercise, be aware that where disparities are low or insignificant, a map may not be necessary.
Data Layer 1: Social and Economic Development Using data available from your macro-level analysis, consider the spatial distribution of these phenomena. Use the data considered most reliable. •
What is the distribution of poverty across the country?
•
In choosing 4-5 of the top social indicators (e.g., maternal mortality, primary school completion, the number one cause of morbidity), where are they the most acute? Use your own scale (1-3, 1-5) to show most acute to the least acute. The social indicators may not align with each other but it is simply to give some sense of where regional disparities might exist.
Data Layer 1
Areas differentiated by poverty line and social indicators
Data Layer 2: Demographics Populations are not static nor is poverty. One common pitfall in using poverty data is that the analysis stops at ‘where’ poverty exists (using an indicator such as“below $1 a day"). It must also ask: •
What is the size of the population considered poor and what is their spatial distribution? In other words, where are they concentrated?
•
This also needs to be viewed in tandem with a map of population size and density. Often the numbers of poor are higher in urban areas because of their larger population size than rural areas, even though a larger proportion of rural areas is known to be poor. It is important to count the number of poor in addition to looking at the proportion of poor in rural vs. urban areas.
•
What are the migration trends? Where is the population growing? The hypothesis is that, even as cities, for example, may have a better quality of life and higher per capita income, the numbers of people living in urban squalor are also rising. This becomes important in projecting where vulnerability is likely to grow in the future; vulnerability is not a static picture.
Data Layer 2
Areas differentiated by size, growth and density of the population; and of the poor population
2
Data Layer 3: Political Economy Analysis Given the history and political economy of the country: •
How do disparities in political power manifest? Prompts: are some regions more powerful or influential, while others are more marginalized? Are some regions more able to negotiate with the center and command more resources (through nepotism, patronage networks, political favors, etc.)? Have alliances with a political figure or a political party privileged some regions over others, over time?
•
Are there certain lobby groups or civil society groups (e.g., the church) that have wielded power and continue to? And then are they concentrated in space?
Data Layer 3
Areas differentiated by political power (marginalized vs. privileged)
Data Layer 4: Socio-cultural Analysis This is to do with the phenomenon of social exclusion. •
Are there population groups, often minority groups, discriminated by society at large? Are they geographically concentrated and, if so, where?
•
Are there areas of the country where cultural groups or systems practice social exclusion of specific population groups (on the basis of caste, kinship, ethnicity, a social stigma, or other distinction they make)? How are they spatially distributed?
Data Layer 4
Areas where socially excluded groups exist; or enduring forms of social exclusion are practiced
Data Layer 5: Geography and Accessibility Geography (no fault of humanity) is often one of the causes of marginalization, i.e., certain parts of the country are very difficult to reach. This is due to physical terrain – jungle, rivers, mountains, etc. – that may literally cut off access seasonally or, given the country’s low level of development, are the areas where infrastructure development is the most difficult, expensive, and thus neglected. These are typically remote areas with low integration in the economy. This is of paramount importance in understanding the geography of economic development (see the 2009 World Development Report). •
What are the areas of the country that are very difficult to reach (geographically marginalized) but do have resident populations?
Data Layer 5
Geographically marginalized areas
3
Data Layer 6: Livelihoods and Climate Change Other data layers will identify areas where vulnerability is high due to low social and economic development indicators. Here, two critical factors are the focus: (a) livelihoods that depend on the natural resource base, and (b) areas that are vulnerable to disaster. They are interlinked because the potential to develop and sustain the livelihood systems depend on the condition of the natural resources, the variability of weather factors, and the threats to livelihoods posed by proneness to natural disasters. Access to and control over those resources should derive from the power analysis in other data layers above and be integrated here. •
In which areas of the country are livelihoods most threatened by vulnerability to disasters? And which major livelihood systems are these?
•
In which areas of the country is the natural resource base dwindling or in a state of decline such that the sustainability of the livelihoods is threatened? Which livelihoods?
•
In which areas of the country are major livelihood systems in the hands of a few and depriving the resident population of a livelihood option? In a way, this is the same as asking: in which livelihood systems (and their location) are forces of impoverishment and/or exclusion most prevalent?
Example: Livelihood Systems and Exclusion In many countries, particular livelihood systems are monopolized by a few elites, to the exclusion and exploitation of the poor. The regional capacity-building exercise took place in Tanzania’s sugarcane belt. Through analysis, teams found that due to high capital investments required, the poor did not engage in sugarcane farming. In fact, the study area hosted many landless poor, while absentee landowners held large swaths of land for sugarcane farming. Further, many of the labor opportunities in sugarcane production were contracted out to seasonal migrant laborers, leaving few livelihood options for people there as well as affecting local wage rates given the lower earnings of outside contract labor. Within the area, a large sugarcane factory also housed an industrial plantation where laborers worked under difficult conditions, with few benefits, and pay determined by demanding quotas of land worked per day (Bode and Wu, 2010). Across the world – not only in Tanzania, but also Latin America/the Caribbean – sugarcane cultivation has been characterized by industrialized plantation systems worked by poor laborers under sub-standard conditions. In addition to the case of sugarcane, other livelihood systems often linked with exclusion or exploitation of the poor include mining areas, as seen in DR Congo; as well as the tea plantation systems of Sri Lanka.
Figure 1
As with all the data layers, the objective is to try and identify major disparities where they exist and identify the most marginalized populations as a result of a specific driver. This can also mean identifying a scaling – most, moderate, least – as appropriate.
Data Layer 6
Areas where livelihood systems (name them) are most threatened by disaster or a declining natural resource base; areas where livelihood systems perpetuate impoverishment or exclusion
4
Data Layer 7: Conflict The conflict analysis will reveal where tensions (hidden or overt) exist or have the potential to manifest; where violence or conflict exist and is likely to continue or worsen. It also identifies areas where conflict does not exist or is not likely to occur. The conflict mapping exercise may be guided by instructions in the macro analysis or, for local level, in the micro analysis appendices. In this instance, marginalization relates to not only violence, but living in a state of insecurity and/or instability. Open conflict is destructive (and development therefore retrogressive), while a state of insecurity may simply arrest development.
Data Layer 7
Areas of conflict will produce insecurity, instability and/or difficult accessibility (showing degrees)
B.2 Concluding the Meso Level Analysis Even with rough sketches for each data layer, the next step is to arrive at an analysis of the most marginalized areas (which can be big or small) or areas with large proportions of marginalized people, by overlaying the data layers. Again, a GIS suits this purpose perfectly but this is rarely available both in terms of the data and the technology. It answers the question – where do the disparities in development exist across the national landscape? It should also help identify where the more obvious pockets of poverty exist. Again, this depends on the disaggregation of data from the macro level analysis and the quality of the secondary literature available. This is not intended as a sophisticated model for doing spatial correlations. Data issues will probably limit the analysis for every one of the data layers. So the expectation is not to produce causality in linking one data layer with another. At best it will roughly identify where multiple forms of marginalization (or multiple drivers thereof) apply and then explanations can be sought. Finally, the analysis cannot be concluded before examining the most marginalized areas and putting them through this filter: •
In the next five years (roughly), which are likely to be targets for investment (government, donors, private sector, foreign)? Which are likely to develop more quickly because of certain advantages or incentives they offer or priorities set by donors or national governments? These are very contextspecific.
5
The pitfall of merely identifying marginalized areas is that designers often forget to take the trends into consideration, trends that may deepen marginalization or alleviate it. Consider the effects of any of these – conflict, the construction of a road, a major foreign investment scheme (that can spur growth but exploit at the same time), migration to urban areas, rejuvenation of an industry (oil, minerals), a major resettlement scheme. This discussion should be revisited after undertaking the micro level analysis, at which point the team will be deciding on the final impact groups. Consider the development potential of diverse regions in making any final assessment of where CARE will become operational. Layering Macro Analyses, Identifying Marginalized Areas
Figure 1
As indicated earlier, the set of overlain maps above is the ideal. Teams should work with the material they have and optimize use of those materials to render the best possible analysis within the timeframe allotted.
6
