Working PaperPDF Available

Gender and Social Vulnerability to Climate Change: A Study of Disaster Prone Areas in Sindh

Authors:
  • Manzil Pakistan
  • Social Policy and Development Centre
i
Gender and Social Vulnerability to Climate Change: A Study of Disaster Prone Areas in Sindh
Gender and Social Vulnerability
to Climate Change:
A Study of
Disaster Prone Areas in Sindh
SOCIAL POLICY AND DEVELOPMENT CENTRE
2015
Graphics and Designing
Muhammad Rizwanullah Khan
Printed in Karachi by Times Press (Pvt.) Ltd.
The views expressed in this report represent those
of the research team, and do not necessarily
represent the views of the donors.
Climate change, coupled with natural disasters, is linked with the livelihoods of rural communities
and hence with their food security. People experience natural disasters depending on the extent
to which they are vulnerable to the associated risks. Likewise, it is also widely accepted that the
impacts of climate change are not gender neutral. The social structures, power dynamics and the
role that men and women play both at home and within the community influence the perception
and the impact as they are understood and interpreted at societal or individual level. Pakistan is
one of the least studied countries in terms of socioeconomic impacts of climate change. Therefore,
the challenge is to address the knowledge gap and to provide evidence based input to policies and
programmes meant to reduce social vulnerability of the people in comparatively more disaster
prone areas of Pakistan.
This report consolidates the finding of research project entitled ‘Gender and Social Vulnerability
of Climate Change: A Study of Disaster Prone Areas in Sindh’. The project is part of Climate Change
Adaptation, Water and Food Security in Pakistan, which is a research initiative of the International
Development Research Centre (IDRC), Canada. The study has been conducted in four districts of
Sindh –Badin, Dadu, Thatta and Tharparkar. In addition to this consolidated report, working
papers on each study district are also being published, which provide detailed analyses of social
vulnerability in the local and national context.
This report is an attempt to contribute to the knowledge regarding gender and socioeconomic
aspects of climate change in the context of Pakistan. The analysis presented in the report reveals
that communities in the disaster prone areas are already experiencing profound transformations
in their lives – both as a result of climate change and socioeconomic change. It is, therefore,
important to adopt policies that promote adaptation and resilience based upon present day
stresses. It is hoped that the publication will be of interest and value to the various stakeholders
including policymakers, parliamentarians, academics, development practitioners, civil society
activists, donors, and others in Pakistan and abroad.
Dr. Khalida Ghaus
Managing Director
Social Policy and Development Centre (SPDC)
iii
Foreword
Foreword
Khalida Ghaus
Muhammad Asif Iqbal
Nadeem Ahmed
Manzoor Hussain Memon
Meher M. Noshirwani
Naveed Aamir
Tabinda Areeb
Technical Advisors
Daanish Mustafa
Nikhat Sattar
Consultants
Mohsin Iqbal
Danish Rashdi
v
The Project Team
The Project Team
Team for consolidation of report:
Daanish Mustafa
Nadeem Ahmed
Manzoor Hussain Memon
Meher M. Noshirwani
Iffat Idris
Social Policy and Development Centre (SPDC) would like to
acknowledge numerous people and institutions whose collective
efforts and support contributed to successful completion of this report.
SPDC is specially indebted to the members of the Project Advisory
Committee consisting of Mr. Javed Jabbar who very kindly consented
to chair the committee, Mr. Naseer Memon of Strengthening
Participatory Organization and Dr. Uzma Shujaat of Area Study Centre
for Europe, Mr. Iqbal Memon of Government of Sindh, and Dr. Abid
Suleri of Sustainable Development Policy Institute for their support
and guidance. Thanks are due to Dr. Sara Ahmed, Senior Program
Specialist, IDRC, who provided her support during the entire course of
the project.
We would like to thank Dr. Mohsin Iqbal and Mr. Arif Goheer of Global
Change Impact Study Centre for conducting analysis of climate data.
The contribution of ield staff that conducted/ facilitated focus group
discussion cannot be overappreciated. In this regard, we acknowledge
the hard work and dedication of Mr. Aftab Ahmed Mangi, Ms. Tasneem
Bhatti, Ms. Bushra Memon, Mr. Muhammad Iqbal Dars, Mr. Mohram Ali
Channa, Mr. Mukhtiar Ali Panhwar, Mr. Ubaidullah Bhutto, and Ms
Seema Rana. Thanks are also due to Mr. Bilal Brohi of Electric Room
Studio and his entire crew members for preparing the documentaries
based on the research undertaken on the four districts identiied for
the project.
We are thankful to our local partner organizations for providing logistic
as well as technical support for conducting ield research in the
respective districts. These organisations include Society for
Environmental Actions, Reconstruction and Humanitarian Response
(Dadu), Badin Development and Research Organization (Badin), Baanh
Beli (Tharparkar), and Sindh Radiant Organisation (Thatta).
vii
Acknowledgements
Acknowledgements
In particular, we would like to express our gratitude to the women and
men of all those villages that participated in this research and shared
their experiences. This research would not have been possible without
their cooperation and their willingness to talk both with the survey
team members and in front of the camera.
We are grateful to all the government departments that provided the
required data and information. SPDC is indebted to the ofices of
Deputy Commissioners of all the four districts for providing their
support and facilitation as and when needed during the ield work.
Special thanks to IDRC and Royal Norwegian Embassy (Pakistan) for
providing inancial support to undertake research on an important
issue.
viii
Acknowledgements
Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
The Project Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Chapter 1: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Context and Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Rationale and Objectives of Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Outline of Report. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Chapter 2: Methodology and Framework. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Vulnerability and Capacities Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Geographic Scope. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Analytical Framework. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Ethical Considerations, Constraints and Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Chapter 3: Vulnerability and Capacity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Vulnerability Profile of the Study Communities and Households . . . . 20
Authenticating Household Level VCIs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Authenticating Community Level VCIs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Comparing Household (HH) VCIs within the Agro
Ecological zones/Livelihoods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Comparing Community VCIs across the Agro
Ecological zones/Livelihoods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Chapter 4: Vulnerability and Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
MacroDevelopment Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
MicroDevelopment Indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Chapter 5: Climate Change, Perceptions and Adaptation. . . . . . . . . . . . . . . . . . . . . . . . 51
Perceptions of Climate Change/Experience of
Environmental Hazards. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Adaptation / Coping Strategies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
ix
Contents
Contents
Chapter 6: Governance and Policy Environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Institutional and Policy Environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
NonGovernment Entities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Gender Mainstreaming. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Chapter 7: Key Recommendations and Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Annexure A1: VCI for Rural Communities and Households . . . . . . . . . . . . . . . . . . 96
Annexure A2: List of Villages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Annexure A3: List of Participants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Annexure A4: List of Interviewees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Annexure A5: District Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Acronyms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
List of Tables, Charts, Figures and Boxes
Table 2.1: Calculation of score for indicator ‘Income Source’. . . . . . . . . . . . . . . . . . . . . . . . 10
Table 2.2: Number of villages included in the study by agroecological/
livelihood zone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Table 2.3: Number of households included in the study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Table 3.1: Basic statistics for the overall sample population . . . . . . . . . . . . . . . . . . . . . . . . . 20
Table 3.2: Household statistics disaggregated by male and
female headed households. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Table 3.3: Pearson’s correlation results for comparison between community
VCI and aggregated household VCIs for communities. . . . . . . . . . . . . . . . . . 25
Table 3.4: Community VCI and aggregate household VCI scores
for selected communities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Table 4.1: Districtwise access to economic services outside village. . . . . . . . . . 41
Table 4.2: Districtwise availability of social services within village . . . . . . . . . . 41
Table 4.3: Employment rates by sector and by gender. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Figure 1.1: Map showing the study areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Figure 2.1: Interviews being conducted with male and
female respondents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Figure 3.1: The box plot for the overall sample showing the
distribution of the observations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Figure 3.2: Box plots for male and female headed households. . . . . . . . . . . . . . . . . . . . . 20
Figure 3.3: Distribution of categories—resilient, low, moderate,
high, very high and extreme vulnerability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Figure 3.4: A women sitting in traditional housing structure (chora). . . . . . . . 23
Figure 3.5: Box plots for the HH VCI scores for the villages
Varshi Kohli, Besarno, Ghulam Dablo and Ali Patni . . . . . . . . . . . . . . . . . . . . 26
Figure 3.6: Village Ghulam DhabloThatta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Figure 3.7: Guar is a major food crop in Tharparkar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Figure 3.8: Box plots for household VCI scores of villages with
canal irrigation and fresh groundwater. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
x
Contents
Figure 3.9: Box plots for household VCI scores of villages with canal
irrigation and saline groundwater. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Figure 3.10: Box plot for household VCI scores for
agropastoralist villages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Figure 3.11: Box plot for household VCI scores of fishing villages. . . . . . . . . . . . . . . 31
Figure 3.12: A view of village Haji Khair Din Mallah located on
the bank of MNVD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Figure 3.13: A girls school established by an NGO in village Khat Lashkar. . 33
Figure 3.14: Box plots of community VCIs compared across
agroecological/livelihood zones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Figure 4.1: Literacy level by gender by village categories. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Figure 4.2: Education status by gender. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Figure 4.3: Unemployment rate by gender by village categories. . . . . . . . . . . . . . . . . . 45
Figure 4.4: Average monthly monetized income by gender
by village categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Figure 4.5: Average household monthly per capita expenditures by
male and female headed households by village categories. . . . . . 48
Figure 5.1: Fishing community living on the bank of Manchar Lake. . . . . . . . . . . 53
Figure 5.2: Women and girls bear the difficulties and suffering. . . . . . . . . . . . . . . . . . . 55
Figure 5.3: Communities in rural Sindh have limited access to health care. 58
Figure 5.4: A woman managing livestock in a village of Dadu. . . . . . . . . . . . . . . . . . . . . . . 62
Figure 5.5: Necessity and survival: young girls seen fetching water. . . . . . . . . . . 64
Figure 5.6: The water source being shared with animals in Tharparkar. . . 65
Figure 5.7: Women making local handicraft in a village of Badin. . . . . . . . . . . . . . . . . 66
Figure 5.8: A woman making rope with pesh in a village of Dadu . . . . . . . . . . . . . . . 67
Figure 5.9: Women plastering mud walls. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Figure 5.10: A flood affected family living in a tent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Figure 5.11: Houses built on higher ground in Dadu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Box 1.1: Terminology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Box 2.1: Background and advantages of VCI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Box 4.1: Case of high social capital and high vulnerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Box 4.2: Lower vulnerability with a big household, through
higher education and diverse incomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Box 4.3: Unstable income as a source of vulnerability for a
female headed household . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Box 4.4: Education and income diversity tickets to low vulnerability
in Agropastoralist communities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
xi
Contents
Like many other countries in the developing world, Pakistan is
experiencing the impact of climate change at the same time is
undergoing a socio-economic transition - a combination which, poses
significant challenges, particularly for vulnerable populations. This
report looks at the experience of climate and social change among rural
communities in disaster prone areas of Sindh, with particular focus on
gendered impacts. It also assesses people’s capacity at the local level
to cope/adapt to these changes. Based on the findings, it makes a
number of gender specific policy recommendations for inclusion in
disaster management and climate change adaptation strategies at
provincial and local level.
The study, commissioned by IDRC, was conducted in four districts of
Sindh: Badin, Dadu, Tharparkar and Thatta. It covered four agro-
ecological/livelihood zones: canal irrigated with fresh groundwater,
canal irrigated with saline groundwater, agro-pastoral and fishing
communities. A total of 62 rural communities and 1,259 households
were assessed using a mixture of qualitative and quantitative
tools/techniques including community and household level surveys,
focus group discussions and key informant interviews. The main
quantitative tool was the Vulnerability and Capacities Index (VCI),
which looks at twelve drivers of vulnerability, divided into three
categories: material, institutional and attitudinal. A high VCI score
indicates high vulnerability relative to other communities, and
conversely a low VCI score points to relatively low vulnerability. The
SPDC methodology went beyond the VCI to gather additional
information, e.g. about health status. Such data (as well as findings
from the focus group discussions) was used to validate the VCI findings.
Overall the survey households and communities were found to be
highly vulnerable with very limited adaptive capacity. The VCI tool
indicates that stability of income, quantum of assets and social capital
are the key factors that describe the variance in the household level
xiii
Executive Summary
Executive Summary
vulnerability scores. At inter-community level quality of infrastructure
along with overall educational attainment of the community are two
of the largest drivers of variance in VCI scores. Significant difference
was found in VCI for male and female headed households; indeed,
almost all the extremely vulnerable households are female-headed.
Examination of qualitative data about individual households did in
general validate the VCI scores. However, while high VCI scores
represent gendered vulnerability well, low vulnerability scores could
hide constraints to female empowerment within households and
communities, that can be essential for building resilience. This points
to the need for qualitative research to complement VCI findings and
better understand gender differentiated vulnerability profiles.
Pearson’s correlation tests were conducted on the aggregated
household VCI scores for communities (20 households in each village)
and the community VCI scores. The tests found a highly significant but
weak correlation, showing that community level VCI can be used as a
general guide to community vulnerability without having to conduct
household level vulnerability assessment – this does not apply,
however, for communities where there is significant household level
variation in VCI scores. Comparison of community level VCI scores for
the four agro-pastoral zones found little difference in scores for
irrigated freshwater and saline communities. However, the mean
scores for agro-pastoralist were significantly higher, and for fishing
communities higher still. The difference can be accounted for by the
fact that agro-pastoralist and fishing villages are more directly
dependent upon natural ecosystem services than irrigated
communities, and there is a greater presence of the state in the latter.
As noted above, the SPDC study looked at a number of wider (beyond
the VCI) development aspects at household and community level.
Macro-indicators were access to roads and other infrastructure
essential for economic development, and presence of education and
health facilities; micro-indicators were household monthly income and
expenditure, illiteracy and unemployment rates. The study found that
while some development indicators converged with VCI findings, there
was also some divergence. Thus, access to services (health, education)
and roads was positively associated with vulnerability, education
partially so, but quantum of income or poverty level was not a good
indicator of vulnerability – more significant was the stability and
diversity of income. The implications are that income/poverty level
should not be used as a surrogate for vulnerability, and the state –
xiv
Executive Summary
which bears responsibility for service provision – has a pivotal role to
play in determining vulnerability.
The focus group discussions – conducted separately for men and
women – yielded a wealth of information about people’s perceptions
and experience of climate change, the position of women and the
differential impact of environmental and social change on them.
Communities across all four agro-ecological zones reported marked
changes in the climate around them – erratic rainfall, changed onset of
seasons, reduced water, more extreme temperatures – and increase in
the frequency and intensity of natural disasters. As a consequence
traditional sources of livelihood (fishing, livestock, agriculture) are
seriously threatened and people’s quality of life (access to food, water,
health status and so on) has deteriorated. Key coping/adaptation
strategies are migration, seeking alternative sources of livelihood and
borrowing. However, in all agro-ecological zones, options for
alternative sources of livelihood are limited.
What also clearly emerges from the discussions is that women suffer
disproportionately: their already disadvantaged position in society is
made much worse by the challenges posed by climate change. For
example, migration of males in search of alternative sources of income
increases the already heavy workload of women. Similarly, as the
quantity and quality of food available to families is reduced, women’s
share of this falls disproportionately. The focus group discussions also
explained some of the divergence between VCI scores and gendered
vulnerability noted earlier: in Khat Lashker, for example, higher
education levels had been accompanied by greater religiosity, leading
to greater purdah and mobility restrictions for women.
The institutional environment for managing adaptation to climate
change in Pakistan remains fragmented and relatively ineffectual. The
governmental institutions fall into two parallel streams: the older
institutions that have the budgets and legal mandate to [somewhat
indirectly] tackle climate related challenges, e.g., revenue, and
irrigation departments at the provincial levels and the FFC, WAPDA,
PMD and the military at the federal level; and the purpose-built newer
institutions that however, have ambivalent and overlapping legal
authority and virtually no budgets to do their jobs, e.g., NDMA, PDMA,
DDMA, and Climate Change Division. Numerous policies and plans
have been formulated addressing risk reduction as well as emergency
response, but in the absence of strong implementing bodies, they are
far from being translated into practice. The less than optimal response
xv
Executive Summary
xvi
Executive Summary
to the devastating 2010 floods highlighted the shortcomings in disaster
management on the part of the state in Pakistan. These shortcomings
are offset, to some extent, by the efforts of civil society – local, national
and international NGOs – and international donor agencies. However,
their interventions are typically heavily resource constrained, limited
in scope and thus not sustainable in the long-term.
Against this backdrop, it is not surprising that the specific and pressing
needs of women in disaster situations have been neglected. At the
legislative and policy level, the GoP continues to have legal and stated
commitment to mainstreaming gender in its climate and environment
related policies. In reality, though, there is very little understanding or
internalization of gender issues. It is always tacked on almost as an
afterthought and then, too, as something to do with women alone in
every policy document. There is little understanding and perhaps even
sympathy towards how to mainstream it.
In conclusion, the findings of this report clearly show that people in
the study areas are already experiencing profound transformations in
their lives – both as a result of climate change and socio-economic
change (urbanization, modernization, etc). It is therefore vital to treat
climate change as a present and pressing reality - rather than as some
future biophysical threat - and to promote climate adaptation and
resilience based upon present day stresses. In light of the study
findings, a number of general recommendations are made, indicative
of the types of directions that developmental and adaptive
interventions could take. The far from exhaustive list includes:
Incorporating the VCI into the workings of state institutions:
data collection could be carried out by DDMAs, and in turn help
them direct relief aid more effectively in the aftermath of a
disaster and ensuring better provincial level planning;
Focusing developmental and adaptation interventions on agro-
pastoral and fishing communities (including programs for
sustainable fishing livelihoods) because of their significantly
higher vulnerability compared to irrigated communities;
Carrying out more analysis and research to understand
differences in vulnerability and drivers of this for irrigated
freshwater and saline communities;
Focusing on education provision not just on quantitative
measures (number of children in school) but also on quality of
education to avoid the type of perverse gender status outcomes
highlighted in this report;
Carrying out specific interventions to address the issue of
women’s increased workload in the absence of migrant males in
the households;
Promoting and building capacity of traditional health care
workers, e.g. dais, so they can provide basic, affordable health
care to women locally;
It is hoped that the research tools and insights offered in this report
will provide pathways to mitigating gendered vulnerability in the here
and now, so that the future can be met from a position of strength
instead of vulnerability.
xvii
Executive Summary
Context and Literature Review
International scientific opinion has reached consensus (or as close as
it is ever likely to get) about the existence of anthropogenically induced
climate change (IPCC 2013). Simply stated, the carbon emissions from
fossil fuel burning since the Industrial Revolution have measurably
increased the carbon concentrations in the atmosphere, leading to the
‘greenhouse effect’. The increase in carbon in the atmosphere will have
complex and geographically varied impacts on seasonal temperatures,
hydro-meteorology and consequently on dependent ecologies, agro-
ecologies and human production systems. More frequent and severe
extreme weather events such as floods and droughts are one
consequence already being seen across the world.
Most of the knowledge about climate change consequences is surmised
from the computerized model runs of the General Circulation Models
(GCMs), but these vary in their predicted climate scenarios and their
spatial resolution. Nevertheless most model runs for North-Western
South Asia, particularly coastal Pakistan, seem to predict more
frequent higher intensity rainfall events and an increase in average
winter temperatures (IPCC 2013). Regardless of the confidence level
in individual scenarios, it is almost certain that historical hydro-
climatic patterns are not going to hold into the future as a result of
climate change. Pakistan, like most countries in the global South is
also experiencing a socio-economic transition, principally the
transition from purely rural and urban livelihoods towards mixed
urban/rural (desakota) livelihoods (The Desakota Team 2008). The
existing hydro-climatic regimes coupled with the human social systems
offer formidable enough challenges to securing sustainable livelihoods,
ecosystem services and welfare gains. The challenges are likely to
become even more formidable through the double exposure to climatic
1
Chapter 1 | Introduction
Introduction
and socio-economic systemic change (Leichenko and O’Brien 2008).
This is the climate change context within which, the findings of this
study must be viewed.
This study is not going to take the GCM generated climate scenarios as
its starting point of what the vulnerability to climate change is likely
to look like in the future. Instead, following Wescoat (1991) and most
recent scholarship by Hulme (2009) the study will focus on current
water problems and vulnerabilities arising from that, and the present
cultural idioms that mediate the experience of vulnerability. This
approach, anchored in the concern with building resilience to the
vulnerabilities of the present, is more likely to influence change in
behavior to build resilience to future climate change. Furthermore,
such a focus on the present must not assume some sort of
predictability about the future based upon past averages, but rather
be mindful of the existing experiences of hydro-meteorological and
social change and how that change is being negotiated by the
vulnerable populations. This report is an attempt at documenting such
experiences (also for another example of such experience in Nepal see
Manhandar et al. 2011, and NCVST 2009).
The issue of vulnerable populations is a problematic one as well. This
report understands vulnerability, beyond a biophysical condition, as
susceptibility to suffer damage from extreme events and relative
inability to recover from that damage (Mustafa 1998 and Mustafa et
al. 2010). People’s susceptibility to suffer damage is embedded in
everyday power relations and political economy and can be inflected
by class (Mustafa 2005, Pelling 1997), gender (Sultana 2010), and
ethnicity (Bolin 2007) among other factors. Wisner et al. (2004)
formulated a Pressure and Release (PAR) model to illustrate the
progression of vulnerability from root causes, e.g. social power
relations, political economy, developmental discourses and patriarchy,
to dynamic pressures, e.g. weak institutions, corruption, and
environmental degradation, to unsafe conditions, e.g. malnutrition,
occupation of flood plains or other exposed locations, and poor
construction quality. In the PAR model disasters come about when
unsafe conditions converge with extreme physical events, e.g.
earthquakes, floods, climate change. Within this formulation then,
vulnerability originates in the root causes that also are determinants
of the contours of everyday life. Given that gender, ethnicity, class and
other such metrics of identity influence everyday life, they also
inevitably bear upon the vulnerability of individuals and communities.
2
Chapter 1 | Introduction
The Pakistani situation with regard to power fault lines along ethnic,
class and in particular gender lines are well known and highly
problematic, e.g. Pakistan stands 135
th
out of 136 countries in the 2013
Gender Gap Report (Haider 2013).
The rationale and objectives of the study outlined below should be
understood in this context and the conceptual framework.
Rationale and Objectives of Study
Vulnerability to climate change and drivers of it are predicated upon
the present state of development and welfare in the countries of the
global South like Pakistan. Pakistan, like most other countries of the
global South, is experiencing unprecedented environmental and socio-
economic transition. The vulnerability to climate change of this
country therefore, must also be understood in the context of this
unprecedented transition. While there has been an increase in
research on climate change in Pakistan in recent years, the topic is still
significantly under-researched. Moreover, there are major gaps in
understanding of social vulnerabilities differentiated by gender and
their contribution to climate change impacts and coping strategies.
This study aims to address the knowledge gaps in the existing research
and provide the missing information to influence plans for adaptation
and disaster management at local and national level. The project -
3
Chapter 1 | Introduction
Adaptation refers to a suit of policies, interventions, and behavioral changes
that individuals and societies are supposed to implement in order to adapt
to the threats and opportunities to emerge from climate change.
Climate Change refers to the phenomena of the change in regional and
global climatic patterns largely as a consequence of the accumulation of
carbon, from fossil fuel burning since the Industrial Revolution.
Gender beyond the biological sex amongst humans, refers to the socially
adopted roles and mannerisms that define masculinity and femininity. Such
socially determined roles and power relations are often justified in the name
of biological difference, and hence the particularly resilient and insidious
nature of the gender roles.
Hazards by definition can only come about when a vulnerable human
population comes into contact with an environmental extreme.
Vulnerability is the susceptibility to suffer damage from an extreme event
and the relative inability to recover from that damage (Mustafa 1998).
Box 1.1: Terminology
4
Chapter 1 | Introduction
‘Gender and Social Vulnerability to Climate Change: A Study of Disaster
Prone Areas in Sindh’ - was commissioned by IDRC Canada, and
undertaken by the Social Policy Development Centre (SPDC). It is
funded by IDRC and, until February 2014, by the Royal Norwegian
Embassy, Pakistan.
The specific objectives of the study are as follows:
To investigate gender dimensions of socio-economic vulnerability
to climate change among rural communities of disaster prone areas
in Sindh;
To assess the adaptive capacity of men and women at community
level and the social capital available to them;
To formulate a set of gender specific policy recommendations for
inclusion in disaster management and climate change adaptation
strategies and plans at provincial and district levels;
To build awareness and understanding among stakeholders
including communities, civil society, media, academia, government
and international development partners.
Given the environmental and socio-economic transition Pakistan is
undergoing, four districts close to the commercial and business hub
of the country and its largest city, Karachi were chosen to undertake
a climate and hazard vulnerability assessment. The assessment is
directed towards understanding the drivers of vulnerability,
vulnerability profiles and associations of those profiles with other
indicators of poverty, environmental quality, and general well-being
at the household and community level. In addition to the substantive
objectives, the research is also directed towards fine tuning and
validating a methodology for assessing social vulnerability at the
local scale.
The four study districts (Figure 1.1) Dadu, Thatta, Badin and
Tharparkar, in addition to their proximity to Karachi also characterize
the challenges of climate adaptation in the lower Indus Basin. The
lower Indus Basin with its largely arid climate, irrigated agriculture,
saline groundwater—with the exception of tracts along the main stem
Indus, previously rich fresh and salt water fisheries, vibrant pastoralist
and agro-pastoralist communities, and issues of salt water intrusion,
encapsulates the socio-environmental stressors that will influence
experience of climate change in most arid deltaic environments across
the globe, e.g., the deltas of the rivers Nile, Colorado, Tigris/Euphrates
to name a few. The results of the field study presented in this report
are therefore not only relevant to understanding climate vulnerability
and adaptation challenges in Southern Pakistan, but also may be
relevant to many other comparable arid lands and coastal contexts.
Outline of Report
Chapter two details the analytical framework and methodology used
in the research study. The study was conducted in four districts of
Sindh (Badin, Dadu, Tharparkar and Thatta) but the findings have been
categorized on the basis of agro-ecological/livelihood zones: canal
irrigated (fresh groundwater), canal irrigated (saline groundwater),
fishing and agro-pastoral communities. A mixture of qualitative and
quantitative techniques were used, notably the Vulnerabilities and
Capacities Index (VCI) quantitative tool. The limitations of the research
and challenges faced are also detailed here.
5
Chapter 1 | Introduction
Figure 1.1: Map showing the study areas
Dadu
Thatta
Badin Tharparkar
Chapter three ‘Vulnerability and Capacity Analysis’ presents the VCI
findings at community and household level, again categorized by agro-
ecological zones. An initial analysis of the findings is carried out to
account for resilient, low, moderate, high, very high and extreme
vulnerability communities/households, as well as to explain any
anomalies.
Chapter four ‘Vulnerability and Development’ looks at the correlation
between vulnerability to climate change and developmental factors not
necessarily covered in the VCI such as poverty, access to healthcare,
and nutrition. Again the findings are categorized by irrigated (f),
irrigated (s), fishing and agro-pastoral communities. Chapter five
‘Climate Change Perceptions and Adaptation’ details the study findings
for each category with regard to people’s perceptions of climate change
and the coping strategies they have adopted.
Chapter six ‘Governance and Policy Environment’ looks at the
institutional set-up at national, provincial and district level to deal with
climate change, the policies that have been formulated in this regard,
and budgetary allocations. The final chapter ‘Key Recommendations
and Conclusion’ consolidates the research findings to answer perhaps
the key underlying question in this study: what are the drivers of
vulnerability and how can these be addressed? It includes a number
of gender-specific recommendations for DRR/climate change policies
and programs.
Gender has not been considered in a separate chapter because this is
a strong cross-cutting theme across the report: chapter two details the
measures taken to ensure women’s representation in the study;
chapter three gives disaggregated VCI findings by male and female
headed households as well as from other qualitative research
techniques, notably focus group discussions (conducted separately for
men and women); chapter four presents findings on health, nutrition,
etc.; chapter five similarly highlights differences in perceptions and
coping strategies on the part of men and women, and chapter six looks
at how gender is addressed in policies, while chapter seven analyses
the relationship between gender and vulnerability and makes gender-
specific policy recommendations. Finally, a number of individual
stories and personal quotes are given in each chapter to provide a
‘human dimension’ to the findings that have been presented: many of
these are stories of/quotes from women.
6
Chapter 1 | Introduction
7
Chapter 1 | Introduction
NOTES:
1Carbon is one of the main greenhouse gases that traps the long wave solar
radiation reflected back from the planetary surface, to make the Earth’s
surface temperature much warmer than it would be otherwise (15°C at
present, which would be -18°C without the Greenhouse effect).
2
Wescoat (1991) cautioned in the early days of climate change research, the
high science of scenario assessment is unlikely to resonate with the
decision makers and populations of the Indus Basin where the
vulnerabilities are more urgent and the decision horizon more focused on
the present than the future.
3
Hulme (2009) reminded us that high science of climate may seem
convincing near the centers of global power, where climate science
knowledge is generated, but climate delinked from its cultural context is
unlikely to travel across spaces to influence changes in behavior, that it is
intended to.
The research study was conducted in four districts of Sindh – Badin,
Dadu, Tharparkar and Thatta - using a mixture of quantitative and
qualitative techniques, including household and community surveys,
focus group discussions and key informant interviews. Vulnerability
was assessed using the Vulnerability and Capacities Index (VCI), a tool
that looks at material, institutional and attitudinal drivers of
vulnerability. Particular stress was placed on ensuring representation
of women in field research.
Vulnerability and Capacities Index
Social vulnerability, as discussed above, is a very important concept
for understanding who is most likely to suffer from adverse effects of
environmental extremes and climate change, and why. There are
different indices for measuring vulnerability. This research study uses
the Vulnerability and Capacities Index (VCI) (Mustafa et al. 2010) (see
Box 2.1). The VCI identifies twelve drivers of vulnerability, which are
divided into three categories:
material - individual assets, livelihoods, education, and exposure
to hazard;
institutional - e.g. social networks, extra-local kinship ties,
infrastructure, warning systems, employment and minority status;
attitudinal vulnerability - knowledge and empowerment.
The VCI was chosen over other vulnerability indices for a number of
reasons: one, the architecture of the VCI is simple and analytically
encompasses three important dimensions of vulnerability (material,
institutional and attitudinal). Two, it provides a robust comparative
metric of vulnerability that is easy to understand. Three, it is a peer
reviewed and field-tested tool that therefore has academic credibility
(see Box 2.1 for full discussion).
9
Chapter 2 | Methodology and Framework
Methodology and
Framework
Weights assigned to material, institutional and attitudinal
vulnerabilities are 35, 50 and 15 respectively. The maximum VCI score
is 100. While the weightages of the three categories are fixed, there is
flexibility in assigning weightages to the variables within each category.
Annexure A-1 gives summary tables of the rural household level VCI
used to do the vulnerability scoring, and the rural community level VCI.
As an example, Table 2.1 below shows how the score for one variable,
income source (under the material category), is calculated:
The weights assignment to each of the variables for this study were
based upon theoretically informed judgments on their contribution to
vulnerability, and on consultations with field researchers and study
managers. These variables were incorporated in the household and
community survey questionnaires
1
. The quantitative data from the
household and community surveys together with some information
from key informant interviews and focus group discussions (FGDs)
were used to calculate the VCIs at community and household levels.
10 Chapter 2 | Methodology and Framework
Assessing social vulnerability is important to identify who is likely to be most affected by climate change. Most of the social
vulnerability research in the past has been based upon q ualitative research presented a s narratives to capture the nuance s,
complexities and inter-linkages of factors contributing to differential patterns of damage. In the policy world, however, it is
very rare for textual material to be the basis for action. Most decision makers are looking for concise preferably quantitative
information, which can be generalized over larger populations and can help ranking and prioritizing populations and
activities respectively.
Towards that end Mustafa et al. 2010 formulated a theoretically driven but empirically informed quantitative Vulnerabil ities
and Capacities Index (VCI), which could be used to capture household and community level vulnerability profiles. The index
identifies twelve drivers of vulnerability, which are divided into three categories following Anderson and Woodrow (1989)
into material (individual assets, livelihoods, education, and exposure to hazard), institutional (e.g., social networks, extra-
local kinship ties, infrastructure, warning system, employment and minority status) and attitudinal (knowledge and
empowerment) vulnerabilities. The original formulation of Vulnerabilities and Capacities Matrix by Anderson and Woodrow
(1989) is one of the most commonly used instruments for qualitative participatory vulnerability assessments across the
global South (e.g., see Action Aid 2005), and therefore, the VCI builds upon the strengths of this tested instrument.
The twelve drivers of vulnerability, which are a part of the architecture of the VCI, are identified based upon their significance
in the vulnerability literature. Whilst the universe of vulnerability drivers is practically infinite, the assumption of the VCI is
that the twelve drivers that are part of it will explain a preponderance of the variance in household and community level
vulnerability.
There is a proliferation of vulnerability indices making the rounds, but this is one of the few instruments, which has been
peer reviewed and therefore has the affirmation of the academic vulnerability research community that the claims made on
its behalf are in fact, based upon requisite research literature and evidence. Accordingly, SPDC chose to empirically test this
instrument for its vulnerability assessment exercise.
Table 2.1: Calculation of score for indicator ‘Income Source’
Indicator Vul. Cap.
Income Source: start value 10
Start Value represents 100% dependency on a local level productive asset
ȋǡϐǡǡȌǤ
Add 2 to the score if the income sources are unstable (for example, daily labour). +2
Subtract 2 if the local income sources are stable and insensitive to local hazards. -2
Lower score by 1 for every 10% of non-local income reported -1 per
Box 2.1: Background and advantages of VCI
It is important to note that VCI is a tool for comparative analysis rather
than an absolute indicator of vulnerability. A higher VCI score would
mean a higher level of vulnerability and vice versa. However, while
interpreting the results of the VCI survey the following constraints
must be borne in mind:
Vulnerability is a dynamic process but the VCI score can only
capture a snapshot in time of the state of vulnerability;
VCI categories and weightages thereof are based upon the
South Asian experience. Some of the categories and
weightages thereof may have to be modified for different
contexts.
VCI scores are for comparative purposes. The score by itself
does not mean anything. Therefore, it is important that there
is consistency in applying the weightages across field sites.
VCI’s simplicity and ease of use is its strength but could also
be a weakness in the sense that it will inevitably miss some
interlinkages and nuances of the drivers of vulnerability.
VCI scores are meant to be used in conjunction with
narrative/qualitative vulnerability analysis, not instead of
them. Sometimes, there may be a temptation to dispense with
the qualitative analysis altogether, which must be resisted.
The above limitations notwithstanding, VCI scores can provide a
simplified snapshot of differential vulnerability that can be an
invaluable tool for action. Even the limitation of interpreting the VCI
scores by themselves can only partially be overcome with a big enough
sample by running appropriate statistical routines to classify the data
to identify groups of high, medium, low and resilient populations. One
such routine used in the case of this study, Jenks Natural Breaks
Optimization method, is particularly appropriate for classifying the VCI
data. Jenk’s method is a data clustering method that classifies the data
by maximizing the variance between categories and minimizing the
variance within categories. Based upon a numerically high enough
number of observations, one could derive categories through the Jenk’s
method to make confident enough predictions for the boundaries of
extreme, very high, high, moderate, low, and resilient levels of
vulnerability. Such an exercise was undertaken in the case of the data
from this study (see Chapter 3).
11
Chapter 2 | Methodology and Framework
12
Chapter 2 | Methodology and Framework
Methodology
The field methodology adopted consisted of both qualitative and
quantitative techniques. Key tools for primary data collection included
community and household surveys, focus group discussions (FGD),
shared learning dialogues (SLDs) and key informants’ interviews (KII).
a) Community and Household Survey: The household survey looked
at members of the household, formal education, labour force
participation, household income and expenditure, housing conditions,
assets, agricultural and non-agricultural land, social capital and
empowerment, climate change, hazard awareness, and adaptation and
mitigation. In the case of female respondents, the survey looked at
their activity profile (how they spend their time). The community
questionnaire was designed to obtain information about the socio-
economic profile of the surveyed localities, and covered topography,
infrastructure and social services.
The village was the primary sampling unit used for this study. With
the help of local partner organizations, key informants, and available
secondary data, rural Union Councils were divided into two categories:
(a) Union Councils that are more prone to disasters, and (b) Union
Councils that are less prone to disasters. Historical data of the
incidence and frequency of disasters was also used for classification of
localities. Two-thirds of the sample villages was drawn from category
A, and one-third of the sample from category ‘B’. Using the Population
Census Organization (PCO) District Census Reports, villages were
randomly selected from each category. A total of 62 villages were
selected with district-wise and agro-ecological zone wise distribution
as shown in Table 2.2 [see Annexure A-2 for full list].
Within each village around 20 households were randomly selected for
inclusion in the survey. Two adult members from each household, one
male and one female, were interviewed to ensure sex-disaggregated
Tabl e 2. 2 : Number of villages included in the study by agro-
ecological/livelihood zone
Badin Dadu Tharparkar Thatta Total
Canal Irrigated
(fresh groundwater) 2 5 0 4 11
Canal Irrigated
(saline groundwater) 12 7 2 8 29
Agro-pastoral 0 3 13 0 16
Fishing 1 2 0 3 6
Tota l 15 17 15 15 6 2
data (Figure 2.1). Again, in order to capture the gender dimension of
social vulnerability, sample households were divided into male headed
households (MHHs) and female headed households (FHHs). Out of a
total sample of 1,259 households, 1,102 were MHHs and 157 were
FHHs. Table 2.3 gives the breakdown of households surveyed by
village and agro-ecological/livelihood zone:
b) Focus Group Discussions (FGD): These were conducted to elicit
information on men’s and women’s perceptions of climate variability,
sources of livelihoods, exposure to environmental hazards,
coping/adaptation strategies, social networks, warning systems and
empowerment. Checklists were developed to guide the discussions.
Separate male and female FGDs were conducted in each village to
ensure that gender dimensions of vulnerability were captured. A total
of 56 male FGDs and 56 female FGDs were conducted.
13
Chapter 2 | Methodology and Framework
Figure 2.1: Interviews being conducted with male and female respondents.
Tabl e 2. 3 : Number of households included in the study
Badin Dadu Tharparkar Thatta Total
Canal Irrigated
(fresh groundwater) 40 102 0 80 222
Canal Irrigated
(saline groundwater) 241 146 41 162 590
Agro-pastoral 0 61 266 0 327
Fishing 20 41 0 59 120
Total 301 350 307 301 1259
c) Shared Learning Dialogues (SLD): Multi-stakeholder dialogues
were conducted in each district involving representatives from local
government, community organisations, researchers and scientists. On
average, a total of 10-12 participants took place in each dialogue [See
Annexure A-3 for list of participants]. A discussion was conducted
which covered the parameters of social vulnerability in the districts,
and the impacts of climate change, including gender-differentiated
impacts. Views from the different groups were also sought on existing
programmes and strategies to reduce disaster related risks. These
dialogues helped in the development of questionnaires for both the
household and community surveys.
d) Key Informant Interviews (KII): Detailed interviews with key
informants were conducted to inform the institutional analysis.
Officials of district and provincial governments, local community
leaders and notables, and development practitioners were interviewed
[See Annexure A-4 for list of interviewees].
Geographic Scope
Sindh has been particularly vulnerable to climate change: it was the
most affected province during the 2010 floods and large areas remain
prone to high water stress, salinity, desertification, drought, floods and
cyclones. The four study districts—Badin, Dadu, Tharparkar and
Thatta, were selected based on a number of factors, e.g., experience of
changes in environmental conditions; vulnerability to floods and
droughts; diverse agro-ecological zones; diverse sources of livelihood,
including agriculture, livestock and fishing; low human development
indicators; gaps in gender specific research; availability of secondary
information; willingness and interest on the part of both government
and communities; security and access.
A brief profile of each district, focusing on topography and climate
change impact, is given below: [See Annexure A-5 for full profiles.]
a) District Tharparkar: Tharparkar district is located in the Thar
desert to the east of Sindh province.
2
The district consists of barren
tracts of sand dunes covered with thorny shrubs. The climate is of a
tropical desert, with extreme heat in the summer and cold winters. In
arid and semi arid regions, the effects of climate change are particularly
severe because these areas are particularly vulnerable to changes in
temperature, evaporation, and precipitation variability.
3
Water is a
major issue: in most parts of the district people consume brackish
water. Agriculture and livestock are the main sources of livelihood, and
14
Chapter 2 | Methodology and Framework
15
Chapter 2 | Methodology and Framework
both depend on the amount of rainfall, which is erratic and irregular.
The district has been hit periodically by droughts, with the most recent
in 2013-14.
4
Reduced rainfall leads to soil and land erosion, reduction
in vegetation, and thus to less fodder for livestock and food shortages.
b) District Badin: The southern part of Badin district is close to the
delta of the River Indus, and the eastern part is connected with the
sand dunes of Tharparkar. The water table has a depth of 240 cm in
winter, and 150 cm in summer, and the drainage system is inadequate.
As a result, even a nominal increase in rainfall leads to flooding.
Rainfall is erratic and unpredictable, and since the Indus Delta is a low
lying area it bears the full brunt of the south west monsoon.
5
Sea
intrusion and drainage from the Left Bank Outfall Drain
6
(LBOD) have
led to excessive water logging and salinity, affected land and
agriculture, and subsequently the livelihoods of the people. Salinity is
greater in the southern part of Badin, and a mass migration of
fishermen has taken place for the last thirty years, which has ruined
once prosperous communities.
7
c) District Thatta: Thatta district is a low-lying area located about 60
miles from Karachi.
8
Since Thatta is at the tail end of the River Indus
it faces constant water shortages, partially because of excessive
upstream withdrawals, which threaten agriculture and livelihoods.
9
The lack of fresh water to recharge the ground water aquifers and sea
intrusion, has led to rising salt content in the soil, which has risen to
the surface, killing vegetation, making the land unfit for cultivation or
growing natural grasses. This has led to destruction of agricultural
land, a drastic reduction in yield per acre of various crops, and what
were once grazing grounds becoming uncultivable wastelands. Sea
intrusion coupled with rising sea levels has led to flooding and erosion
of coastal areas, affecting fishermen. Thatta is also vulnerable to other
natural disasters such as cyclones and droughts.
10
Finally, it is among
the poorest districts in Pakistan, especially its coastal areas.
d) District Dadu: District Dadu is located in the south-west of Sindh
bordering Balochistan. Climatic conditions in the district are
considered as extreme; intensively hot in summer and moderately cold
in winter. There are three distinct topographic areas: hilly, irrigated,
and low lying riverine land. Average rainfall is 120 mm, and the main
source of water is the River Indus. The district is prone to natural
disasters including floods (due to hill torrents, heavy rains and flooding
in the River Indus) and droughts. The main sources of livelihood are
agriculture and livestock.
Analytical Framework
While the field research for the study was carried out in four districts
on the basis of administrative boundaries – roughly 300 households
per district – the analytical framework is based on agro-
ecological/livelihood zones. There are four main agro-ecological zones
/ livelihoods in the target areas:
canal irrigated with fresh groundwater;
canal irrigated with saline groundwater;
agro-pastoral;
fishing
The canal irrigated villages are divided into two zones, the ones with
fresh and the other with saline groundwater respectively. The villages
with fresh groundwater have the choice of supplementing inherently
scarce canal water with groundwater irrigation, while the villages with
saline groundwater do not. This is important in terms of choice of crops
and sustainability of livelihoods between the two zones, and hence the
distinction. In Sindh the non-canal irrigated regions are invariably
based upon pastoral or agro-pastoral livelihood systems. This
classification captures the issues of villages in those categories. Finally
the category of riverine and estuarine fishing communities captures
the vulnerability profile of these much neglected communities within
this eco-livelihood system.
The findings of the study – VCI findings, views expressed in FGDs and
SLDs, the non-VCI data to emerge from community and household
surveys, etc – have been organized on the basis of the four zones listed
above. This approach has two main advantages over presenting the
findings organized by district. One, the experience and issues faced by
fishing communities will largely be the same in Badin as in Thatta and
Tharparkar; by merging them together a lot of repetition is avoided.
Two, presenting the findings based on agro-ecological zones will
provide far more useful learning for other countries/regions facing
similar challenges. For the intended global audience for this study,
administrative boundaries are largely meaningless. Conversely, of
course, district-wise findings will be of great interest to the domestic
audience, particularly to provincial and district level policy-makers.
However, this need will be addressed through district reports prepared
by SPDC to be published separately.
16
Chapter 2 | Methodology and Framework
Ethical Considerations, Constraints and Challenges
Ethical considerations have been taken into account at every stage of
the research study. Ethical guidelines were prepared at the inception
stage of the project and staff were trained in these. Key ethical
considerations included:
Procedures were adopted to ensure that all participants
(interviewees, community and household members) understood
the process in which they were engaged and they gave voluntary
and informed consent. As well as sending out introductory
letters to respondents explaining the research objectives and
seeking their consent, this was repeated at the time of
conducting interviews/surveys.
It was made clear to participants that they could withdraw from
the study, and that it was not binding on them to provide
information or respond to the questions being asked.
Efforts were made to safeguard confidentiality of information.
The respondents were assured that the information provided by
them (at household level) will only be used for the purpose of
analysis. No specific information (related to individuals) will be
made public. However, it was also explained that if they gave an
opinion their name might be quoted in the report. As far as
possible interviews were recorded (with participants’ consent)
to avoid any misinterpretation.
The constraints of the VCI were explained above. Wider constraints of
the research study are the lack of gender disaggregated secondary data
on disasters. A number of challenges were faced, particularly in
conducting the field research. These included: difficulty in accessing
districts due to recurring floods; diversion of attention of both
communities and officials to relief and reconstruction; lack of authentic
data on climate and agricultural production over the required period
of time (30-40 years); lack of personnel willing to travel to such remote
places, especially in extremely hot weather; and difficulty in sifting
information on changes due to climate change from other factors such
as poorly maintained irrigation infrastructure, poor farming and water
management practices, degradation of natural ecosystems and lack of
overall development. Various strategies were adopted to overcome
these challenges, e.g. having a flexible approach to setting timings for
field research and doing so in consultation with local people; recruiting
17
Chapter 2 | Methodology and Framework
field enumerators and surveyors from within organizations already
working in the target districts and training them; developing
partnerships with research institutions to ensure access to relevant
material; and developing specific criteria to probe the impacts due to
climate change.
NOTES:
1. All survey instruments of this study including household and community
questionnaires and FGDs checklists are available on request from SPDC (email:
spdc@spdc.org.pk).
2. Pakistan Emergency Situation Analysis. District Tharparkar April 2013.
3. Gender and Climate Change: An Introduction, Edited by Irene Dankelman,
Earthscan 2010.
4. Tharparkar Drought 2014, Jaggarta Social Welfare Organisation.
5. Ibid.
6. Down the Drain by Gulmina Bilal Ahmed, Newsline, 5th February 2007.
7. ‘The Sea May Swallow Thatta, Badin, and Sujawal, in 30 years’. The News, 6 June
2014. According to one estimate 3.5 million acres of arable land have been
inundated by the sea in three districts of Sindh, Thatta, Badin and Sujawal. And
800,000 fishermen from these areas have been forced to shift to different parts
of the country.
8. Effects of Climate Change on Thatta and Badin. Sami Khan. Envirocivil.com.
January 25, 2012.
9. Ibid.
10. Disaster Risk Management Plan District Thatta Government of Sindh 2008.
18
Chapter 2 | Methodology and Framework
This research study is probably one of the most extensive field tests of
the VCI undertaken so far. As mentioned in Chapter 2, VCI is not meant
to be a substitute for qualitative vulnerability analysis but rather a
complement to them. Hence, the discussion of VCI results in this
chapter is in tandem with the information gathered through key
informant interviews, focus group discussions and participant
observations. Throughout the chapter, as the VCI results are presented,
it is stressed that the scores are a static generalization of a complex
and changing reality. The static snapshots of vulnerability can,
however, provide the type of generalized actionable information, which
can be the basis for deeper inquiry and directing interventions.
This chapter presents the VCI results from 1,259 households in 62
villages, across four agro-ecological zones/livelihood types. In the first
instance, a few representative household level VCI scores are
presented and deconstructed to illustrate how they were derived.
Furthermore, some additional information is also presented about
those representative households to further validate the scores. The
same process is repeated for community level VCIs. The VCIs across
agro-ecological zones are then compared, supplemented with
qualitative data from other sources. Some general insights about
vulnerability profiles across the agro-ecological zones/livelihood
types are offered based upon the analysis. Lastly, and most
importantly the VCI scores are cross referenced with the qualitative
information to tease out disconnects between VCI scores and
gendered vulnerabilities. Concluding reflections upon disconnects
illuminate the possible points of leverage and interventions to address
the gendered vulnerabilities.
19
Chapter 3 | Vulnerability and Capacity Analysis
Vulnerability and
Capacity Analysis
Vulnerability Profile of the Study Communities
and Households
Before delving into the details of the vulnerability profile of the study
communities based upon the VCI analysis, it would be useful to
stipulate a few notes on interpreting the data statistics that are
presented in this chapter and through the report. The overall
distribution of VCI scores, their descriptive statistics, such as mean,
median and mode are presented. As per the popular prejudice, there
may just be the temptation to privilege the means of the data being
presented. That would be a mistake. The whole notion of climate
change, at its core is about moving away from normal, mean conditions
based thinking and planning paradigms. Therefore, the key statistic
that is highlighted is the modal distribution of vulnerabilities across
the overall sample, agro-ecological zones and study communities.
Table 3.1 presents the basic descriptive
statistics for the overall sample of
households, while Figure 3.1 is a box
plot demonstrating the overall spread of
the VCI scores for the sample.
According to the Shapiro-Wilke test of
normality and the convergence between
mean, median and mode, the overall VCI
data is normally distributed. The
important point here is that the modal
distribution is 63, and the mean
distribution is 63.94. The sample also has a reasonable spread of 63
points with a minimum score of 26 and a maximum score of 89. It
should also be noted that there is a significant difference in the VCI
scores between female and male headed households. Table 3.2 lists the
differences between male and female headed household.
20
Chapter 3 | Vulnerability and Capacity Analysis
Table 3.1: Basic statistics
for the overall sample
population
Statistic Value
N (Valid) 1259
Mean 63.94
Median 63.78
Mode 63
Std. Deviation 9.668
Variance 93.462
Range 63
Minimum 26
Maximum 89
VCI Score
All Households
1
90
80
70
60
50
40
30
20
_
_
_
_
_
_
_
_
o33
o26
Figure 3.1: The box plot for the
overall sample showing the
distribution of the observations.
VCI Score
90
80
70
60
50
40
30
_
_
_
_
_
_
_
Figure 3.2: Box plots for male and
female headed households.
Tabl e 3.2: Household statistics disaggregated by male and female headed
households
Mean Mode N Std. Deviation
All Household Level VCI (HH VCI)
Male Headed Households 62.41 63 1102 8.905
Female Headed Households 74.63 79 157 7.908
Total 63.94 63 1259 9.668
MHH FHH
A Jenk’s maximizing variance routine was conducted on the overall
data to get some sense of the natural breaks in it. The results of the
routine are outlined in Figure 3.3.
Jenk’s routine minimizes the variance within the category and
maximizes it between categories. There are five categories shown in
Figure 3.3, but we divide the data into six categories, with the score of
38 defining the boundary between resilient and low vulnerability
categories. There are two outliers in the low vulnerability category,
therefore the boundary is drawn at 38, above which are the rest of the
observations. Below that number we deem the scores to represent
resilience. The VCI score boundaries for the categories therefore are
as follows:
Resilient 0-37
Low Vulnerability 38-50
Moderate Vulnerability 51-59
High Vulnerability 60-66
Very High Vu lnerabil ity 67-74
Extreme Vulnerability 75 and above
Since VCI is a comparative number, it is important that these
empirically derived numbers be used to determine some generalized
classification of low, medium, high, very high and extreme levels of
vulnerability. Such an empirically derived global classification can
provide a further simplified tool for a policy maker to interpret the
21
Chapter 3 | Vulnerability and Capacity Analysis
0.04
0.03
0.02
0.01
0.00
Figure 3.3: Distribution of categories—resilient, low, moderate, high, very high and
extreme vulnerability.
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Probability Density, p (VCI)
_
_
_
_
_
Vulnerability & Capaciti es Index, (VCI)
Total Vulnerability
results of any VCI exercise. The sample of 1,259 is big enough for
deriving statistically robust boundaries between the categories. The
categories are likely to become more robust with bigger samples, but
given the sample size and the rigor of the sampling, it is not anticipated
that the actual boundaries will vary significantly. The remainder of the
VCI scores presented in this report should be interpreted with these
categories in mind.
Authenticating Household Level VCIs
In terms of the household VCI scores the case of Fehmida
1
, an illiterate
widow from the coastal fishing village of Hashim Mandrio in district
Thatta, is illustrative. Her VCI score is 86, which is in the very high
vulnerability category according to the Jenk’s classification routine.
This is a female headed household in which Fehmida lives by herself
as all her daughters are married. Her house is highly exposed to
coastal flooding, an attribute that she shares with the rest of her village,
though there is a considerable range of vulnerabilities in this village,
with the lowest VCI score of 50. She did not report kinship or
associational ties which could be a source of help and assistance to her
in the event of an emergency. Indeed, she is almost entirely dependent
upon the larger community for her food and shelter. This is a
heartrending case of gendered vulnerability where, despite having
adult daughters, she does not think that they can be a source of solace
or help to her. Because of the particular patriarchal ethos of rural Sindh
it is considered quite shameful to call upon female offspring for any
kind of help or support. Furthermore, it would most likely be
impossible for her daughters to support their mother, because of their
own dependent situation on their husbands’ households.
Whilst Fehmida is alone and dependent upon the community’s largesse
for her upkeep, Bhago is an 80 year old illiterate widow from the agro-
pastoralist village of Mondro in district Tharparkar with a house full
of children and grandchildren, but still has a household VCI of 84. Her
46 year old son is in prison and hence unable to provide for the family.
She lives in a traditional Chora structure
2
(Figure 3.4) with her 45 year
old daughter–in-law and four grandchildren all below the age of 14.
Nobody in her household has ever had formal schooling. Her daughter-
in-law is a daily wage worker and a livestock trader. Her oldest
grandson, aged 14, manages livestock for local landlords and makes
some money that way. Her granddaughter, aged 12, does embroidery
work which is a source of some income. Bhago is also like Fehmida
above, a recipient of income from the Benazir Income Support
22
Chapter 3 | Vulnerability and Capacity Analysis
Programme (BISP) as well as benefitting from her community’s
largesse. The household is not aware of any warning system, has no
extra kinship ties or associational membership. They have no
anticipation of help from any quarter in the event of a disaster. They
have no access to any infrastructure, e.g., mobile phones, electricity,
sanitation. Furthermore, the household is perpetually indebted even
just on account of food expenses.
Whilst almost all the extremely vulnerable households are female
headed, there are instances of male headed extremely vulnerable
households. Mahmud from the coastal island fishing village of Ghulam
Dhablo in Union Council Keti Bandar, district Thatta, heads one such
household. He is a young man of 25 with a 23 year old wife and three
minor children. The wife is a full-time home maker while Mahmud is a
fisherman who reports having to go further and further - up to 90 km
from the coast - to find fish, whereas in the past he just used to fish
locally. He owns his fishing boat and a net, and those are about his only
assets. He and his community have no access to modern or indeed any
infrastructure, but what particularly distinguishes him is his lack of
extra local kinship ties, associational life or access to any leadership
structures at any scale. Being the sole breadwinner living on a highly
exposed island and engaged in a somewhat dangerous profession, his
family is perpetually living on a knife’s edge. Overall his community is
highly vulnerable (minimum score 60), but what makes him
particularly vulnerable is his lack of social capital and lack of access to
23
Chapter 3 | Vulnerability and Capacity Analysis
Figure 3.4: A women sitting in traditional housing structure (chora).
the protection of a larger family. Although this cannot be confirmed,
but given the atypical closeness in the ages of his wife and himself and
his nuclear family, one could safely surmise that he is probably
estranged from the larger community because of marriage without his
family/community’s consent. Here again, possibly the question of
personal choices bears upon the vulnerability of this household.
On the lower end of the spectrum, Abdul Rahim from village Khat
Lashkar in district Dadu has a VCI score of 40. Khat Lashkar is a canal
irrigated village with fresh groundwater. Khat Lashkar is overall a less
vulnerable community and Rahim is one of the least vulnerable
members of that community. Rahim lives with his three brothers, a
wife and mother. His wife has a college education and the mother is
also educated. He has a government job as a stenographer in Karachi,
while one of his brothers has an agricultural income, and the other two
are unemployed. The household is well equipped with electrical
appliances, e.g. a refrigerator, television, and flat iron, and they own a
motorbike for transportation. The household did not report any
associational membership, or extra local kinship ties, but they did
report good linkages to community and local leadership structures.
In the agro-pastoralist village of Dondio Meghwar, in district
Tharparkar, Baghia (with VCI score of 38) is a 67 year old patriarch of
a household comprising his three sons (ages 35, 23 and 20 years), two
daughters (ages 20 and 15 years), two daughters-in-law (ages 28 and
25 years), four grandsons (ages 8, 5, 3 and under one year) and two
granddaughters (ages 3 and 9 years). Both he and his eldest son have
had 13 years of schooling while the other two sons had completed 9
and 12 years of schooling respectively. The two daughters-in-law as
well as the matriarch of the family were illiterate. The two daughters
had two years of schooling, while one of the grandchildren who was
old enough was in the 1st grade. Baghia cultivates his own land, while
his sons earn from government service, livestock business and a
grocery store respectively. The daughters and the daughters-in-law
also bring in income from their embroidery work. The family lives in a
six room house with a cement roof and burned brick walls. The family
uses firewood for fuel, and use a pit latrine. They also have a mobile
phone connection. While the men reported an active associational life,
engaged with the local community based organization, the women
reported no such associations. Furthermore, the men considered
themselves close to local leadership structures and had awareness of
a disaster warning system that they trusted. Women were aware of no
such warning system. Here again, a seemingly less vulnerable
24
Chapter 3 | Vulnerability and Capacity Analysis
25
Chapter 3 | Vulnerability and Capacity Analysis
household has hidden dimensions of gendered vulnerability, where
women are simply not aware of any warning systems and are
dependent upon the men in the family for timely warning, information
and support.
While the above sample household studies can validate the types of
VCI scores derived from the survey, they also caution against over
reliance on these scores for assessing gendered vulnerability. The VCI
does a competent job of capturing gendered vulnerability at the higher
end of the spectrum. But at the lower end, the scores can often
obfuscate higher gendered vulnerability within households. This point
will be thrown into sharper relief as the VCI scores at household level
are validated.
Authenticating Community Level VCIs
The 62 village communities included in the survey were assessed for
their collective VCI scores by aggregating the 20 household level VCIs
calculated in each of the villages and by using the rural community
based VCI table shown in Annexure A-1. The rural community VCI is
an instrument for rapid assessment of VCI when the time and
resources are not available to undertake a household level VCI
assessment exercise. This project provided an important opportunity
for testing the linkages between aggregated household level VCIs and
the directly calculated community based VCI scores. Pearson’s
correlation tests were conducted on the aggregated household VCI
scores for communities and the community level VCI scores. The
results of the Pearson’s correlation are listed in Table 3.3.
Table 3.3 illustrates that community level VCIs (Community VCI) are
significantly correlated with the aggregated household VCIs for
communities (ALL HH VCI) at more than 99% confidence level (Sig. 2
tailed = .000). The correlation is, however, weak, explaining only 36%
Tabl e 3. 3 : Pearson’s correlation results for comparison between
community VCI and aggregated household VCIs for communities
Community VCI ALL HH VCI
Community VCI Pearson Correlation 1 .603
Sig. (2-tailed) .000
N6262
ALL HH VCI Pearson Correlation .603 1
Sig. (2-tailed) .000
N 62 1259
of the variance (test statistic = 0.603). Therefore the community VCI
can be used in lieu of a household level VCI assessment exercise to get
an approximation of the overall vulnerability profile of the community.
However, the community itself at times can have large variations in
vulnerability within it, rendering the community level VCI score a weak
generalization. The community level VCI could be a strong
generalization where the spread of VCI scores within the community
is small. Table 3.4 lists the communities that are profiled in detail in
this section to validate the community VCI scores. Figure 3.5 gives the
box plots of the household level VCI scores for the same communities.
The village Ghulam Dhablo is an island in
the Keti Bandar creek (Figure 3.6), district
Thatta. Being a coastal island it is highly
exposed to tropical cyclones and marine
flooding. The village has virtually no
modern infrastructure, e.g., roads,
electricity, schools, hospitals or water
supply system. Drinking water is brought
into the village by boats from Keti Bandar
Port, so if the connection with the main
town is temporarily severed in the event of
a coastal storm, the approximately 400
people in the village would not even have
drinking water. The nearest government
funded boys’ and girls’ schools are about 90
km from the village on the mainland, as are
any basic health units with a qualified physician. A hospital for any
major illnesses is about 150 km away. Almost all the residents of this
village are engaged in fishing and are progressively having to travel
longer and longer distances chasing after dwindling fish stock. The
reasons for the dwindling fish stock are many, the chief amongst them
being off shore commercial fishing by foreign and domestic fishing
trawlers. But unsustainable fishing techniques used by the local
fishermen are not exactly helpful either. Furthermore, subsistence
26
Chapter 3 | Vulnerability and Capacity Analysis
Table 3.4: Community VCI and aggregate household VCI scores for selected communities
Village Zone Comm. HH. Mean HH. Mode HH. Min HH. Max
Ghulam Dhablo Fishing 70 74 73 60 85
Besarno Agro-Pastoralist 57 62 63 48 70
Ali Patni Irrigated-Fresh GW 65 67 63 58 84
Varshi Kohli Irrigated-Saline GW 73 74 75 66 89
VCI Score
Villages
Varshi Kohli Besarno Ghulam Dhab lo Ali Patni
90
80
70
60
50
40
_
_
_
_
_
_
_
_
_
_
o89
o87
o48
o84
o61
o
Figure 3.5: Box plots for the HH VCI scores for the villages
Varshi Kohli, Besarno, Ghulam Dhablo and Ali Patni.
fishing is increasingly being replaced by commercial fishing even
amongst local fishermen resulting in greater reliance on unsustainable
fishing techniques. This fragility of the livelihood base, particularly in
terms of its susceptibility to commercial exploitation and
environmental degradation, is one of the main reasons that the fishing
communities in the sample in particular, and in Pakistan in general, are
highly vulnerable.
The VCI range of 25 with a minimum of 60 illustrates the type of high
vulnerability characteristic of this community. The only thing
distinguishing households with lower vulnerability from high
vulnerability is the strength of the social capital that the people are able
to call upon, or cases of more than one bread earner in the household.
Otherwise the level of infrastructure, educational attainment and assets
is generally the same across the village. In this instance the closeness
between the household VCI mean and mode, and the community VCI
illustrates that the community VCI is a reasonably accurate reflection
of the relative vulnerability of the village.
Surrounded by sand dunes village Besarno is located in district
Tharparkar, within walking distance of an all weather metalled road. The
nearest town of Mubarak Tarr, with a secondary school and a basic
health unit (BHU) is about 10 km away. The approximately 350 residents
of the village largely live in adobe mud houses. Some of the residents of
the village have employment with the military and government service;
this constitutes a stable source of income for about 36% of the
households. Although there is no electricity in the village, there is a boys’
27
Chapter 3 | Vulnerability and Capacity Analysis
Figure 3.6: Village Ghulam Dhablo-Thatta.
primary school and three deep wells that are the source of fresh drinking
water for the residents. The low modal HH VCI score of the village, with
the highest score of 70 and a small VCI score range of 22, is indicative of
the lower vulnerability of the village, primarily on account of the diverse
income sources of the residents, and extra kinship ties of more than 50%
of the residents. Many of the residents, because of a household member’s
employment in the government, also report having access to local
leadership structures. Agricultural production in the village has been
good for the past few years, though in 2013 the main crop of guar
(cluster bean: Cyamopsis tetragonoloba) was spoilt because of untimely
rains (Figure 3.7). That was reported to be the first instance of spoilage
of a crop from untimely rains in living memory.
The canal irrigated village of Ali Patni is located in Taluka Keti Bander
in district Thatta. The village is canal irrigated, though it is only 1.5 km
from the sea, and consequently coastal flooding is one of its main
hazards. The main sources of drinking water for the 600 residents of
the village are a hand pump, and a local stream with slightly brackish,
low quality drinkable water. The village does not have any electricity,
schools or medical facilities. There are schools and medical facilities 5
km from the village, but most people cannot afford the transport fare
for the children to travel to those schools. In addition, the motorable
road is not passable during the rainy season. There is not much in
terms of agricultural productivity in the village. It is at the tail end of
the irrigation canal and irrigation water access is erratic at best. With
land degradation from salt water intrusion and salinity, agricultural
productivity is quite marginal. The extreme poverty of the village is
manifest in the relatively high community VCI score of 65 and the
household VCI scores ranging from a low of 58 to a high of 84.
The highly vulnerable village of Varshi Kohli (community VCI of 73)
is in district Badin and all of its approximately 450 residents are
Hindu migrants from neighboring Nagarparker. The reason behind
their migration was water scarcity in their traditional area. The
village is characterized by extremes of poverty and vulnerability
the maximum VCI score being 89. Apart from a boys’ primary school
and a seasonal motorable road, the village has no infrastructure or
facilities, e.g., electricity. The village does have some hand pumps,
which are functional, but they mostly deliver brackish water. The
residents of the village migrated from Nagarparker in the 1970s
because of drought. The residents of the village are largely associated
with day labour in the area, though some people still maintain
landholdings in Nagarparker from which they derive an income. The
28
Chapter 3 | Vulnerability and Capacity Analysis
Figure 3.7: Guar is a major food crop in
Tharparkar.
village is quite close to the Left Bank Outfall Drain (LBOD), which is
a source of flooding for the village. Since the initial migration the
entire population of the village has had to temporarily migrate back
to its homeland in Nagarparker because of flooding in 1984, 1994
and 2011. Thus, ironically, while they migrated because of lack of
water, they periodically have to retrace their steps because of too
much water.
The above case studies illustrate the stories behind the community
VCI scores and their correlations with the HH VCI scores. The
criterion for the selection of these cases studies was the closeness of
the HH VCI scores. Whilst on the one hand these case studies convey
how the community VCI score can be indicative of the relative
vulnerabilities of the communities, on the other hand these same case
studies underline the need to go beyond the numbers to understand
the specific drivers and configurations of community vulnerability.
The numbers draw attention to the fact that something of interest is
going on, but it is the qualitative information, which provides specific
guides to action.
Comparing Household (HH) VCIs within the
Agro-Ecological Zones/Livelihoods
In this section the box plots are presented for household (HH) VCI
scores for the entire sample of 62 communities by agro-
ecological/livelihood