Studies over the last twenty years show that understanding the social and economic forces that govern
societies and create vulnerability should have the same emphasis as understanding the physical causes of
vulnerability. This enquiry investigates what social networks and institutions are available (created or
imposed) to people that enable them to cope with large scale crises, such as tropical cyclones, and ‘everyday’
problems such as poverty and illness. The social institutions prevalent in villages in coastal Andhra Pradesh,
south India, have been assessed regarding which variables appear to be the most and least successful in
contributing to the reduction of people’s vulnerability and aiding the coping mechanisms of individuals and
households.Using a mixed methods approach, quantitative and qualitative data were collected from over 300 respondents
in twelve villages, enabling an assessment to be made of each respondent’s access to resources as indicators
of levels of poverty, marginalisation, resilience and social power. Data was obtained using questionnaires,
village cartographic surveys, sociograms and semi-structured interviews. The sociograms used throughout
this research were developed during the fieldwork. They were adapted to suit the requirements of the
research focus and thereby facilitate the assessment of the types and strengths of social networks used by the
respondents in both ‘everyday’ and ‘crisis’ situations. A contextual analysis was conducted to locate concepts
such as ‘community’ and ‘risk’ within the discourse appropriate to the respondents. From this analysis it
became apparent that the village level respondents perceive risk in terms of recurring ‘everyday’ occurrences
such as low crop yields and the lack of basic needs, and not high impact but infrequently occurring events
such as tropical cyclones or floods. The village level respondents, government officials and NGO employees
typically perceive the concept of ‘community’ as defined by caste classifications.
The main variable explaining access to resources is caste. Caste is the dominant social institution that
influences social networks and hence (with poverty) levels of vulnerability. This is because it not only
influences levels of vulnerability directly, through levels of poverty, but also restricts the ability of some
respondents to change their circumstances through enduring caste-defined inequalities with regards to
accessing the resources that might help them to reduce their levels of vulnerability. Given this finding, and
because caste is not predominately an economic phenomenon, vulnerability reduction initiatives that focus on
economic advancement alone are unlikely to usurp patterns of caste discrimination and thereby are unlikely
to reduce endemic levels of vulnerability for the most vulnerable members of society: they are treating
symptoms, not causes. In contrast, social networks are also important: while the ‘lower’ castes are the most
marginalised, powerless and poorest members of the case study areas they attempt to address this by
accessing socio-economic resources that can, and marginally do, increase their resilience to frequent small
scale crises, typically via social networks with informal social institutions such as NGOs, CBOs and kinship
networks.
In view of the dearth in empirical evidence associated to inequalities related to caste in rural India, this thesis
adds to the limited contemporary evidence that suggests that caste defined disparities persist. Consequently,
practitioners involved in vulnerability reduction need to gain a better understanding of the communities and
the ‘political realities’ in which they operate, so that future interventions will be better targeted and
ultimately be more appropriate and sustainable that they have been in the past.