Climate-induced Migration in South Asia: Migration Decisions and the Gender Dimensions of Adverse Climatic Events

Full text

Journal of Rural and Community Development
ISSN: 1712-8277 © Journal of Rural and Community Development
www.jrcd.ca
Journal of Rural and
Community
Development
Climate-induced Migration in South
Asia: Migration Decisions and the
Gender Dimensions of Adverse
Climatic Events
Authors: Gopal Datt Bhatta, Pramod Kumar Aggarwal, Santosh
Poudel, & Debbie Anne Belgrave
Citation:
Bhatta, G. D., Aggarwal, P. K., Poudel, S., & Belgrave, D. A. (2015).
Climate-induced migration in South Asia: Migration decisions and the
gender dimensions of adverse climatic events. The Journal of Rural and
Community Development, 10(4), 1-23.
Publisher: Rural Development Institute, Brandon University.
Editor: Dr. Doug Ramsey
Open Access Policy:
This journal provides open access to all of its content on the principle that
making research freely available to the public supports a greater global
exchange of knowledge. Such access is associated with increased readership
and increased citation of an author's work.

Journal of Rural and Community Development
ISSN: 1712-8277 © Journal of Rural and Community Development
www.jrcd.ca
Climate-induced Migration in South Asia:
Migration Decisions and the Gender Dimensions of
Adverse Climatic Events
Gopal Datt Bhatta
CGIAR Program on Climate Change, Agriculture and Food Security
(CCAFS), South Asia
and
Research & Strategy, Community and Neighborhood Services
The City of Calgary, Alberta, Canada
bhattagopal@gmail.com
Pramod KumarAggarwal
CGIAR Program on Climate Change, Agriculture and Food Security,
International Water Management Institute
New Delhi, India
p.k.aggarwal@cgiar.org
Santosh Poudel
Department of Bioresource Policy, Business and Economics
University of Saskatchewan, Saskatchewan, Canada
sap91@mail.usask.ca
Debbie Anne Belgrave
Research & Strategy, Community and Neighborhood Services
The City of Calgary, Alberta, Canada
Debbie.Belgrave@calgary.ca
Abstract
There is significant interest in determining the role of climate-induced shocks as a
prominent driver on migration decisions of different groups of farmers in South
Asia. Using data from a survey of 2,660 farm-families and focused group
discussions in Bihar (India), Terai (plains) (Nepal) and coastal Bangladesh, we
employed logistic regression to investigate household response towards migration
and gender dimensions of adverse climatic events. The results suggest that migration
decisions depend on farmers’ unique resource profiles: (a) households that use
migration to improve their resilience, mostly resource rich households; (b)
households that have no alternative but to migrate, mostly poor farmers; and (c)
households who cannot migrate due to different socio-economic obligations, mostly
farmers with intermediate level of income that also includes women, children and
elderly of different income profiles. These profiles represent a spectrum with
households within a profile being closer to one or the other of the profiles on either
side. They are not mutually exclusive and serve as a point of departure for further
research to refine key explanatory variables. Given that some members of the
household pursue migration as a result of adverse climatic events, government
strategies are required to mitigate risks at destinations and create opportunities for
the trapped populations.

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 2
Keywords: distress migration, climatic risks, extreme events, rainfall variability,
gender dimensions, South Asia
1.0 Introduction
South Asia, one of the highly vulnerable regions to climate change, is the hardest hit
by climate-induced events and climatic change (Bhattacharyya & Werz, 2012;
Battisti & Naylor, 2009; Intergovernmental Panel on Climate Change, 2007;
Mendelsohn, Dinar & Williams, 2006; Aggarwal, Joshi, Ingram & Gupta, 2004).
This region experiences a high degree of exposure to environmental risks (World
Bank, 2009). There is also heavy reliance on climate sensitive sectors such as
agriculture and fisheries (Ahmed, Hassan, Etzold & Neelormi, 2012; Kumar &
Viswanathan, 2012). Increasing temperatures, more adverse climatic events, rising
sea levels, increasing cyclonic activity (Bhattacharyya & Werz, 2012; World Bank,
2009), higher inter-annual variability of precipitation (Giorgi & Bi, 2005; May,
2004; Lal, Meehl & Arblaster, 2000), as well as water-related hazards induced by
climate variability are characteristics of the region (Ahmed, 2009). For instance,
every 3-5 years, two-thirds of Bangladesh is inundated by floods, and approximately
every 3 years a cyclone hits the country (World Bank, 2009). It has been widely
recognized that global climate change will instigate greater climate variability which
in turn will increase people’s exposure to extreme events such as intense rainfall
(Trapp et al., 2007; Diffenbaugh, Pal, Trapp & Giorgi, 2005), greater heat stress
(Battisti & Naylor, 2009; Diffenbaugh et al., 2005), and more storm events
(Emanuel, 2005).
Both sudden-onset events—such as flooding, cyclones, and storms—and slow-onset
events—such as changes in precipitation, temperature, and sea level rise—will have
detrimental impacts on key economic sectors. Over time these events will influence
decisions to move to access greater economic opportunity (Bhatta & Aggarwal,
2015; Bhattacharyya & Werz, 2012). Over the next decade, a considerable number
of people could be affected by climatic hazards in South Asia and the bulk of
climate-induced migration will involve poor people— the most vulnerable segment
to climate change (Asian Development Bank, 2009).
Literature highlights that poor people are the ones to bear the brunt of climate change
and they are the ones to migrate under adverse events due to lower adaptive capacity
(Kumar & Viswanathan, 2012). However, studies investigating the climate-induced
migration pattern——short-term distress migration—across different income and
resource profiles of farm households are very limited. The gender dimension of
adverse weather events and male out-migration as an aftermath of these events is
under-explored. With increasing climatic risks, it is a matter of significant interest
to investigate which group of farmers—based on on-farm livelihood diversification,
income, education, rainfall variability, asset profiles etc.—decide to migrate under
adverse events and what the gender dimensions of adverse climatic events are. We,
therefore, hypothesize that, (a) the middle class farmers—those with a moderate
level of income—including women, children and elderly of different income profile
are vulnerable to climatic shocks in the long-run as they are unable to migrate for a
variety of socio-economic and cultural reasons; (b) farmers with more access to
assets, income and/or education are more likely to migrate under adversity, and (c)
poor farmers who live in more variable rainfall areas are more likely to migrate in
the event of adversity. The current study also presents the gender dimensions of

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 3
adverse weather events, particularly the impact—of adverse climatic events and
male out-migration—on women, children and elderly. We test our hypotheses
through household surveys of 2,660 farm-families (19 sites covering 133 villages)
and focused group discussions in different locations across climatic risk areas of
India (Bihar state), Nepal (Terai), and Bangladesh (south-western coastal areas).
2.0 Literature Review
2.1 Climatic Risks and Migration
Climate change causes serious problems, including extreme weather events, the rise
of sea levels and environmental degradation. Each of these events is expected to
trigger short- or long-term migration. South Asia is burdened with a high level of
exposure to risks from climatic events including the risk of migration. Every year,
large numbers of people migrate (temporarily or permanently) to cope with extreme
weather events such as floods, droughts, salinity and cyclones. Such events are
predicted to become more common with climate change (Asian Development Bank,
2012). While many of those displaced return to their homes as conditions become
conducive, others may struggle to survive in new environment after incurring
substantial losses. The climate-induced migrants are those who leave their habitual
homes, or choose to do so temporarily because their livelihoods have been
negatively affected by adverse weather events (ADB, 2012; McLeman & Smit,
2006). Migration flows as a result of climatic events are often categorized as
‘distress migrations’. The term ‘distress migration’ in this study hence refers to the
household members who pursued migration as a short-term coping strategy in
response to adverse weather events. This movement could be either a labour
migration or migration to make a family living under climate-induced adverse
events. The climate-induced distress migration from agriculture could largely
manifest in the short-term (Kumar & Viswanathan, 2012). While climate will always
interact with numerous push-pull factors to compel migration, a more nuanced
understanding of the role of adverse events in migration decisions of different groups
of farmers will help shape adaptation investments.
2.2 Migration as a Livelihood Strategy
Despite various climate- and non-climate-related stresses and shocks, smallholder
farmers in South Asia have been coping with and adapting to the local circumstances
over many years (Ojha et al., 2014). However, adaptive capacity of smallholder
farmers in the face of increasing climatic risks is usually low due to their dependence
on natural resources and poor human and physical capital (Gukurume, 2013; Salau,
Onuk & Ibrahim, 2012; Ahmed et al., 2012; Shewmake, 2008; Kelkar & Bhadwal,
2007). The situation becomes worse if suddenly occurring hazards destroy the
potential harvest and jeopardize livelihood. While relatively affluent households—
or members of a household—generally choose permanent migration (Kumar &
Viswanathan, 2012; Deshingkar & Akter, 2009), persons with lower income
undertake short-term migration as a livelihood strategy (Kumar & Viswanathan,
2012). Short-term migration to cope with climatic shocks has the potential to
ameliorate the food insecurity to some extent (Ahmed et al., 2012). Cyclic migration
for a short duration in response to climatic adversity may perpetuate or even grow
in the future as long as climatic variability increases (Deshingkar & Start, 2003).

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 4
Migration is often seen as a coping strategy against rapid onset events (Etzold,
Ahmed, Hassan & Neelormi, 2014) and a mode of adaptation to climate change
(Brockhaus, Djoudi & Locatelli, 2013; Kumar & Viswanathan, 2012; Ahmed et al.,
2012; Warner, 2010; Brooks, Grist & Brown, 2009; McLeman & Smit, 2006). The
decisions on migration are driven to maximize income as well as to minimize risks
(Meza, 2015; Massey et al., 1993). Rademacher-Schulz et al. (2012) identified
migration as a major coping strategy to address unfavourable economic and
unexpected environmental conditions, including the local implications of rainfall
variability. For instance, labour migration in response to climatic risks, food and
livelihood insecurity is one of the most important strategies of farmers in Bangladesh
(Etzold et al., 2014). Individuals with some resources, but not an adequate amount,
undertake short-term migration to supplement their livelihoods (Kumar &
Viswanathan, 2012). Economic inequality as a driver of migration also interacts with
environmental change to impel movement from the origin to other areas (Geddes &
Somerville, 2013). The principal motivation of the richer farmers to migrate is for
educational reasons or to obtain better-paid jobs in the urban centres (Ahmed et al.,
2012). Migration for this group contributes to a further diversification of income and
attainment of a higher social status (Etzold et al., 2014).
2.3 Gender Dimension of Climatic Events
The effects of climate change are not gender neutral and impact the poor,
marginalized and vulnerable population groups such as women (Hunter & David,
2009; Canadian International Development Agency, 2002). Unequal gender
relations and access to resources may make women more vulnerable to climate
change than men. Climatic events may not only directly impact women and
vulnerable populations, but also make them more vulnerable because of their
interaction with socio-cultural factors. Furthermore, adaptation, that is, the ability to
adapt to and cope with changes due to climate change, is also gendered. Adaptive
capacities of individuals greatly depend on income, education, health and access to
natural resources. Given that women tend to be poorer, less educated, have a lower
health status and have limited direct access to or ownership of natural resources,
they are disproportionately affected by climatic risks (Demetriades & Esplen, 2010).
Women in Bangladesh were found to be more vulnerable during cyclones because
cultural norms prevented them from leaving their homes in time and learning to
swim (Nelson, Kate, Terry & John, 2002). Water stress due to climate change is
expected to cause further difficulties for women in West Africa (Denton, 2002). In
many cases displaced women are forced into labour-intensive and low-paying jobs
due to low levels of education. Women from the fishing community in the
Philippines facing climate disasters were forced to work as domestic helps due to
lack of skills (United Nations Population Fund, 2009). Although migration, short or
long term and different factors promoting it, bridges the income gaps to some extent,
it also has social costs such as inequality within the community and a substantial
increase in workload of women (Meza, 2010).
Climate-induced women migrants are at a greater risk of sexual and gender-based
violence (Brown, 2008). Many women state lack of safe shelters upon being forced
to migrate as one of their primary concerns after climatic events (Mitchell, Tanner
& Lussier, 2007). There are also other issues of safety and security arising from
women's health status and disintegration of social networks. Mitchell et al. (2007)
observe that women suffer from psychosocial impacts of natural disasters to a
greater degree as compared to men. The extra burden of looking after their family

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 5
members under distress situation results in many women suffering from anxiety and
post-traumatic stress.
Further, the breaking of social ties and separation of families also has a severe impact
on women. In addition, often women are not allowed direct access to relief aid
because they are not the 'head of the household' (Spring, 2008). Such exclusion is
likely to make them more vulnerable. Similarly, in Nepal, as more and more males
migrate from vulnerable areas to the cities, women are becoming heads of
households. These women are highly vulnerable to adverse climatic events as they
have to survive in already fragile landscapes (UNFPA, 2009). If the migrant is the
father or household head, the family is seasonally separated and the woman becomes
the de facto head of household. Although this increases women’s autonomy and
decision making power (Brown, 2008), it places additional burden to care for the
household, the children, elderly, and themselves. In particularly hard times, girls
drop out of school to help their mothers and the probability for domestic violence
increases (Bernabe & Penunia, 2009).
3.0 Research Methods
3.1 Study Areas and Sampling Process
A household survey was done in three contrasting sites of South Asia: India (Bihar
State), Nepal (Terai) and Bangladesh (south western coastal area) in 2011-2012. The
selected sites (Figure 1) represent climatic ‘hotspots’ in South Asia (Bhatta & Aggarwal,
2015; Bhatta, Aggarwal, Shrivastava & Sproule, 2015) and a high degree of socio-
economic vulnerability to climate change (Wood, Jina, Jain, Kristjanson & DeFries,
2014). The key characteristics of the regions surveyed are depicted in Table 1.
Table 1. Description of the Regions
Regions Description
Bihar, India
Hot and humid climate, subsistence farming, very small land size
per household, dominance of rice-wheat pattern, subsistence
farming, a frequent occurrence of drought and floods and erratic
rainfall pattern,
average temperature increasing, rainfall
decreasing, depth of water table increasing and soi
l fertility
decreasing.
Terai, Nepal
Tropical climate and highly fertile flood plain, the major supplier
of food in the country, rice-wheat cropping pattern dominant, a
distinct rainfall gradient from east to west, frequent floods and
droughts, feminization in agriculture, westerly wind during wheat
grain filling period, cold spells increasing, overall rainfall almost
the same but winter droughts common in many parts of the Terai.
Coastal
Bangladesh
High amount of annual rainfall, predominance of aquaculture and
rice-rice based livelihoods, mostly low-
income agricultural
workers, highly vulnerable to climate change and highly
susceptible to climate-induced hazards, high level of salinity in
many areas restricting normal pr
oduction of agricultural crops,
salinity intrusion increasing due to sea-
level rise, frequent
occurrence of floods, cyclones and droughts.

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 6
The sampling process was done in three stages: (a) areas with contrasting climatic
risks were selected (Bihar, Terai and the coastal area); (b) key districts were chosen
based largely on rainfall and salinity gradients. A sampling frame, consisting of a 10
km x 10 km area, was then overlaid in the chosen district purposively (Bhatta et al.,
2015). All villages within the chosen frames were enumerated and seven villages
were selected randomly. A complete list of households within each selected village
was generated with the help of village authorities; (c) using simple random sampling,
20 households within each village were chosen. From 19 sites in the region, we
selected 133 villages and 2,660 farm households. The selected households have
engaged in farming either in their own farms (land owned farmers) and/or
contributed their labour in others’ farms (landless households). Whether or not the
households own land, they all qualify as being considered as farmers because
farming is their dominant profession.
Figure 1: Surveyed sites in South Asia (values in parentheses indicate average
annual rainfall in mm).
Source: Authors.
3.2 Survey Instruments
A structured questionnaire was designed, tested and implemented across all the sites.
Trained enumerators were deployed to ensure a consistent level of precision on
sampling and data collection. The survey questionnaire was first translated into the
local languages of the respective areas and then back translated into English to
ensure consistency in interpretation of each question. The interviews of one or more
individuals primarily the head of the household and/or spouse within the household
was carried out. Questions related to impact of adverse events on women, children

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 7
and elderly were administered in a focused group discussion (FGD) with women
and local government officials. There were two FGDs in each site, one for women
and one for local government officials. A standard checklist was used to facilitate
FGDs. The key components of checklist were: who migrates in the households and
why; aftermath of adverse climatic events on women, children and elderly, food,
nutritional and general security issues. The same checklist was used in a FGD with
local officials to capture their different perspectives. The participants were selected
purposively based on their willingness to participate in the discussions. Local
government officials were approached in their office (mainly in the district
agriculture development office) in each site and FGD was conducted.
The structured questionnaire for household surveys has several components:
socio-demographic information; sources of livelihoods; changes made to farm
practices including livestock over the last 10 years; food security; household asset
profile; community groups; coping strategies under adverse weather events;
distress short-term migration in the adverse year during the last 10 years; and,
factors preventing households from migrating even under adverse events. The
questionnaire also has questions like household income before and after distress
migration, length of migration, number of family members who migrated and the
nature of work post migration.
3.3 Data Analysis
A simple descriptive analysis was done across the three broader areas (Bihar,
coastal Bangladesh and Terai) to see cross-site variations in terms of socio-
demographic variables. The Kruskal-Wallis test was performed to test the
significance of the variables. This test is performed for the dataset that violates
assumption of normality (Field, 2005). Rainfall data from 1961 to 2010 was
extracted from the Climate Research Unit (CRU) gridded database (0.5 degree
resolution) and annual rainfall variability (%) was calculated for 50 years. The
annual rainfall variability in the surveyed sites ranges from 13 to 26%. Variability
less than 19% was considered moderate (n= 1,120) and more than 19% was
considered a highly variable rainfall area (n= 840). Since rainfall variability is
considered one of the most important factors to agriculture livelihood in South
Asia, we included this variable in logistic regression to observe how it categorizes
the migration decisions of the farmers under adversity.
Migration is considered to be a very important coping strategy during adverse events
in some regions in South Asia. How do different groups of farmers behave in terms
of making decisions to migrate under adversity? Are families from a relatively more
deprived household (marginal farmers, farmers with low income, fewer assets, etc.)
more likely to migrate compared to those from a relatively less deprived household?
In order to test the hypothesis that individuals under adverse weather events migrate
due to their resource constraints, we used a logistic model. A set of qualitative and
quantitative variables were used to explain migration which is dichotomous (1=
migration as a coping strategy under adverse event and 0, otherwise). Logit and
probit are the two most commonly used techniques for estimation of models with
dichotomous dependent variables. Estimation of ordinary least square may give
the correct sign; however, it doesn’t fit considering the nature of the data and
dependent variables we had. We found that the distribution of residuals is logistic
and hence logistic regression was employed. The probit and logit models produce
qualitatively similar results. The choice between these two is largely one of the

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 8
statistical conventions since the substantive results are generally indistinguishable
(Long, 1997).
A logistic model was run using data from Bihar and coastal Bangladesh only as none
of the respondents in Nepal considered migration as a coping strategy. Bhatta &
Aggarwal (2015) also reported formal and informal credit markets as the most
important coping strategy of farmers under adversity in Terai while migration in
Bihar and coastal Bangladesh. The reason could be that most of the family farms in
Terai are run by women due to out-migration of men. The Maoist conflict that
occurred from 1996 to 2006, a key driver of out-migration of male members of the
household and subsequent exodus of men from villages, put an additional burden on
women to take the sole responsibility of agricultural production (Gartaula, Niehof
& Visser, 2010; Women Organizing for Change in Agriculture and NRM, 2010).
Since most of the family farms have already been managed by women and their
counterparts are temporarily away from home, out-migration and further women’s
involvement in agriculture during adversity cannot be considered a feasible coping
strategy in Terai (Bhatta & Aggarwal, 2015).
A set of categorical variables was created to group farmers based on their socio-
demographic and resource characteristics. Following Bhatta, Singh & Kristjanson
(2013) and Kristjanson, Garlick, Ochieng, Förch & Thornton (2011), farm
households with less than four on-farm sources of livelihood (on-farm
diversification) were considered poorly diversified (n= 692), 5-8 as intermediate
diversification (n= 880), more than 8 as highly diversified farming (n= 388).
Households with diverse livelihood sources are considered to be better adapted to
climatic risks than those who rely on one of few sources (Cunguara, Langyintuo &
Darnhofer, 2011). The number of farm practices changed, including livestock and
fisheries, over the last 10 years is considered as a proxy of adaptation (or
innovativeness) under changing circumstances including climate change.
Accordingly, farm households with less than 5 changes in farming practices were
categorized as poorly adapted (n= 504), those with 5-10 as intermediate adaptation
(n= 417) and with more than 10 as high adaptation (highly innovative) (n= 1,039).
Participants were also asked the highest level of education of the household head.
Based on the responses, respondents were grouped in to four classes: household with
no formal education (n= 154), primary level (up to 5th grade) (n= 600), secondary level
(up to 10th grade) (n= 691) and post-secondary level (more than 10th grade) (n= 515).
Different types of household assets such as information, transportation, energy,
production and luxury assets were summed up and households with less than 3 assets
were considered to have basic asset profiles (n= 568), those with 3-5 assets as
intermediate asset profiles (n= 764) and more than 5 as higher asset profiles (n=
628). The household’s annual income was used to form three categorical variables:
poor, lower middle class and upper middle class. The World Bank’s criterion of
poverty analysis, which states 1.25 US Dollar ($) a day as poor, was used to define
the poor class in our analysis. Our calculation shows that a farm household with less
than $1,500 in a year falls in the poor class (n= 1,379) followed by lower middle
class with $1,500-3,000 income (n= 410) and upper middle with more than $3,000
annual income (n= 171). However, it should be noted that the current classification
simply serves as a basis of comparing migration decisions of households with
different income profiles.
A multivariate-correlation analysis was done to determine the co-linearity effect of
the independent variables. Any independent variables with a high degree of

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 9
correlation (r >0.5) with each other and a low degree of correlation with the
dependent variable were excluded from the model. The logistic regression
coefficients and marginal effects (dy/dx) were calculated to see which variable
affects the most on migration under stressful events. In logistic regression, a positive
coefficient indicates that each unit increase in the independent variable is associated
with an increase in the dependent variable by a given coefficient.
4.0 Results and Discussion
4.1 Farm-family Resources
This section presents the description of the farm household’s socio-economic
characteristics including resource endowment. On average, all farm households
surveyed in the region are smallholders; farmers in Terai possess relatively bigger
land size (1.37 ha) compared to Bihar (0.86 ha) and coastal Bangladesh (0.63 ha)
(Table 2). This means that a large proportion of the sampled households in coastal
Bangladesh are landless and marginal. Despite higher on-farm diversification
compared to Bihar, farmers in the coastal Bangladesh face hunger months for a
quarter of the year while most of the farmers in Bihar and Terai face around a month
of food deficit in a year.
Farmers in Bihar seem to be somewhat more innovative as reflected by a greater
number of changes made in farming during the last 10 years. Our data reflects a
number of changes made during specified period of time (2001-2010) which
conferred some benefit to the farmers who made such changes. Therefore, number
of changes is a proxy measure of potential adaptation (Wood et al., 2014). Farmers
in Terai have more household assets compared to those in Bihar and coastal
Bangladesh. Average income of the farm household in coastal Bangladesh is
significantly higher (US $1542) compared to Bihar (US $1462). Although rainfall
variability is higher in Bihar and Terai compared to coastal Bangladesh, because of
its physical location, coastal Bangladesh is more vulnerable to climatic hazards.
Despite frequent adverse events such as floods and cyclones, the numbers of coping
strategies farmers follow in coastal Bangladesh are less than that in Terai.
4.2 Distress Migration in the Past
Households were asked whether they had engaged in distress migration (short term)
in the past because of extreme climatic events. The results reveal that distress
migration occurred more often in coastal Bangladesh (35%) as compared to Bihar
(23%). Because of the frequent occurrence of climatic events in Bangladesh,
climate-induced migration is quite common (Gray & Mueller, 2012; IPCC, 2007).
The average number of household members who migrated during adverse events is
also higher in coastal Bangladesh compared to Bihar (see Table 3). In Bihar, almost
all households reporting distress migration involved male migration while in coastal
Bangladesh both male (97%) and female (3%) migration have occurred. In general,
male migration is predominant during agricultural distress in South Asia. Prevalent
social norms, lack of education and skills, fear of harassment and social exclusion
are the prime reasons for no or low rates of female migration (Ahmed et al., 2012).

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 10
Table 2. Average of Different Quantitative Variables in the Study Areas
Key variables Bihar,
India Terai,
Nepal Coastal
Bangladesh
Household size 7.80a
(0.153) 7.40b
(0.145) 5.00c (0.066)
Land size (ha) 0.86b
(0.053) 1.37a
(0.070) 0.63c (0.028)
Food available months in a year 11.09a
(0.074) 10.82b
(0.089) 9.47c (0.102)
On-farm sources of livelihoods 4.69c
(0.087) 7.86a
(0.082) 6.61b (0.091)
Off-farm sources of cash 1.56b
(0.035) 1.44c
(0.040) 1.64a (0.032)
Number of changes made in farming 10.81a
(0.200) 8.90c
(0.143) 9.88b (0.206)
Household assets 5.00b
(0.110) 8.18a
(0.119) 4.01c (0.087)
Annual income per household (US $) 1462b
(61.26) - 1542a (51.47)
Annual rainfall variability (%) 19.69 19.83 15.91
Number of Coping strategies under
adversity 3.58c
(0.043) 4.18a
(0.071) 3.65b (0.040)
Note: Values in parentheses indicate standard error of mean. Letters in the superscript show significant
difference between the areas at 5% level of significance according to Kruskal-Wallis test.
Climate-induced migrants both in Bihar and coastal Bangladesh after migrating
engaged in unskilled work as reported by 66% of the respondents. More than 50%
of the distress migrants reported that they had higher incomes after migration. This
infers that while short-term migration is considered a survival strategy under
adversity, it also enhances the income profile of the migrants and possibly enhances
their livelihood. The migrants, whatever the duration of migration may be, generally
put their efforts into a gradual improvement of their families’ livelihoods (Etzold et
al., 2014). It has also demonstrated that the very poor move to “low-return”
destinations, or to other high-risk marginal places nearby. In Bangladesh, for
instance, people living on chars often need to move regularly to another char, though
these are places of extreme environmental vulnerability (ADB, 2012). However,
detailed investigation on the changes in post migration economy and livelihood
resilience of the migrant families is beyond the scope of this study.

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 11
Table 3. Descriptive Information Related to Distress Migration
Variables Bihar, India Coastal Bangladesh
Number of households with distress migration 221 (23) 345 (35)
Average number of household members
migrated 1.48 2.19
Who migrated
Male
Female
221 (100)
0
336 (97)
9 (3)
Nature of work after distress migration
Skilled
Semi-skilled
Unskilled
30 (14)
42 (19)
149 (67)
41 (12)
76 (22)
228 (66)
Income after distress migration
Same as before
Less than before
More than before
84 (38)
11 (5)
126 (57)
79 (23)
72 (21)
194 (56)
Note: Values in the parentheses indicate percentage.
4.3 Determinants of Migration Under Adverse Weather Events
Table 4 depicts the estimated coefficients and marginal effects of different variables
based on logistic regression analysis for migration as a coping strategy under adverse
weather events. The result shows that the probability of migration increases with an
increase in the number of dependents and number of adults in the household.
Availability of surplus labor within the household enables members of the family to
move out for a short period to supplement the family’s income (Kumar &
Viswanathan, 2012). Agriculture becomes a risky endeavor with rainfall variability,
and households with less land but with surplus labor have a tendency to seek off-
farm income (Gebru & Beyene, 2012). The likelihood of migration decreases if the
household has taken credit in the recent past. This provides interesting insight that
an easy to avail credit mechanism could provide the short-term means of survival in
the event of weather adversity. Bhatta & Aggarwal (2015) also noted that farmers in
Bihar, coastal Bangladesh and Terai avail credit facility (through formal or informal
sources) under adverse weather events.
There is less likelihood of households with irrigation facilities to migrate under
adverse events. Households with no irrigation facility tend to be reluctant to adopt
new agricultural technologies even when expected net returns are high (Yesuf &
Bluffstone, 2008). Families dependent on rain-fed smallholdings are additionally
vulnerable during seasons of adverse weather. Availability of irrigation increases the
options for diversification (Singh, Kumar & Woodhead, 2002). Subsistence farmers
often lack irrigation facilities. Weather events such as droughts hit subsistence
farmers the hardest. Such farm households pursue short or mid-term labor mobility
to secure households’ access to food (Findlay & Geddes, 2011). Compared to the
households with no formal education, households with primary and secondary levels

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 12
of education are less likely to migrate while those with post-secondary education are
more likely to migrate. Kumar & Viswanathan (2012) reported that the probability
of short-term migration is higher among the non-literate category and people with
more education are more likely to be long-term migrants.
The likelihood of migration under adverse weather events is significantly higher
with poor and upper middle class families when compared to the reference category
(lower middle class). Therefore, in the short run, poor farm families seem to be much
more affected by the adverse event which increases their (or at least some members
of the household) probability of migration from the areas of difficult situation to the
new environment. Similarly, the higher income farm families can also afford
migration. However, because of several socio-economic, cultural and emotional
obligations to the community, a large proportion of farmers with a middle level of
income have a tendency to stay in the difficult situation. Cultural constraints and
social networks are also very important determinants of migration decisions
(Sherbinin et al., 2008). Social stigma and security of assets such as livestock,
household items, and land, as well as emotional attachment with the land resources
(Bhatta & Aggarwal, 2015) preclude middle income farmers from migrating even
under adverse weather events. Even if middle income farmers lose a significant
portion of their farm produce due to the environmental crisis, they have hope to
recover (Gray & Mueller, 2012). While people at the middle end of the socio-
economic spectrum may be strongly tied up with their household capital or are
emotionally attached to their land resources which precludes them from resorting to
migration under distress situations (Ahmed et al., 2012), the people at the lower end
of the spectrum (poor, landless, and low asset owners) may easily be displaced by
climate hardships (Ahmed et al., 2012; Kumar & Viswanathan, 2012). The well-off
farmers, on the other hand, temporarily migrate from areas with adversity to safer
places because they can easily afford it. If this is the case, novel adaptations and risk
management strategies are called for, should we expect such middle income farmers
to adapt to the growing risks.
Individuals with a lower adaptability index have a lower probability of migration
while the opposite seems true for the higher adaptability category. Similarly,
households with poor asset profiles are less likely to migrate while the opposite is
true with the high asset class as compared to the intermediate asset profile. The
households with higher number of assets may be better able to reduce economic
vulnerability with off-farm changes (migration, for instance). Importantly, rainfall
variability—captured through the average over a fifty year period—has significant
influence on distress migration. The marginal effect of rainfall variability is very
high (around 12%). In South Asia, inter- and intra-annual rainfall variability is a key
determinant of agricultural productivity. The increase in climate variability means
that the range of weather conditions experienced annually will be higher, exposing
the region to a less predictable and more variable rainfall season. Such changes will
bring about impacts on weather-related events and adverse environmental conditions
(Arendse & Crane, 2011). Greater weather variability in the area of residence will
increase the probability of an individual to undertake migration for alternative
livelihoods (Kumar & Viswanathan, 2012). An associated issue is the potential link
between the weather variability and agricultural production, which in turn would
influence the migration decision of the farmers who predominantly depend on rain-
fed farming (Warner & Afifi, 2014; Rademacher-Schulz et al., 2012). The current
analysis does not aim to capture this three-way linkage.

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 13
Table 4. Determinants of migration under adverse weather events: A logit estimates
Variables Migration as a coping strategy
β
SE
dy/dx
Number of dependents in the household 0.072* 0.039 0.016*
Total land (ha)
-0.066
0.055
-0.015
Group membership (n=354) (yes=1)
-0.054
0.143
-0.013
Credit taken (n=381) (yes=1)
-0.507**
0.149
-0.111**
Farm types (n=470) (Irrigated=1)
-0.260*
0.141
-0.061*
Food availability (months)
-0.029
0.020
-0.006
Household type (n=1898) (male headed =1)
0.484*
0.307
0.111*
Number of adults in the household
0.058**
0.019
0.013**
Number of off-farm sources of income
0.059
0.040
0.014
Number of farm items sold
-0.185**
0.040
-0.042**
Number of coping strategies
0.633**
0.045
0.145**
Education- no formal education (n= 154) (reference)
Primary (n=600)
-0.350*
0.205
-0.079*
Secondary (n=691)
-0.196
0.213
-0.045
Post-secondary (n=515)
0.463*
0.235
0.102*
Household annual income (lower middle class- reference)
Poor
0.419**
0.144
0.094**
Upper middle class
0.207*
0.227
0.048
On-farm diversification (5-8 sources of on-farm sources of livelihoods- intermediate-
reference)
Low (<5)
0.175
0.144
0.040
High (>8)
0.112
0.174
0.025
Adaptability index (5-10 number of farm practices changed- intermediate-reference)
Low (<5)
-0.422**
0.168
-0.094**
High (>10)
0.027
0.139
0.006
Household asset profile (3-5 household assets- intermediate-reference)
Poor (<3)
-0.294*
0.136
-0.066
Rich (>5)
0.092
0.140
0.021
Annual rainfall variability (high=1)
0.549**
0.130
0.122**
Intercept
-3.39**
0.519
Log likelihood
-1098
LR chi2(23)
410**
Pseudo R2
0.157

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 14
The results provide for three distinct household profiles using migration as a short-
term coping strategy under adversity. These profiles represent a spectrum with
households within a profile being closer to one or the other of the profiles on either
side. They are thus not mutually exclusive and serve as a point of departure for
further research to refine key explanatory variables.
• Households that use migration to improve their resilience: The households
use migration in ways which improve their resilience, such as earning more
to invest in climate-resilient livelihood opportunities. These households use
migration as one of a variety of adaptation strategies under adversity,
moving seasonally or temporarily, often to non-agricultural jobs. This
strategy is not only common among relatively richer households but also
households with more adults, those with diverse assets and those which have
access to a variety of adaptation, or risk management options (Warner &
Afifi, 2014), and, hence they use migration in ways that enhance resilience
to changing circumstances including climate-induced adverse events.
• Households that use migration as a survival strategy: This group is risk
averse and has meagre adaptation options or resources. This group may
often move during the adversity to other areas in their region in search of
alternative income sources. Households in this class are mainly poor, have
subsistence-oriented farming and no or very marginal land with more food
deficit months in a year. These households are highly vulnerable to climatic
events and they use migration as a risk management strategy in response to
climatic variability and livelihood insecurity (Warner & Afifi, 2014).
• Households that cannot migrate: The third group includes trapped
populations that struggle to survive under adversity and cannot easily use
migration to adapt to the negative impacts of climatic events. They may have
strong social ties and are emotionally attached to their resources which
restrain them from moving even under adverse events. These households
have an intermediate level of income and land resources, and also include
women headed households, women, children and elderly.
4.4 The Gender Dimensions of Extreme Events
We present results based mainly on focussed group discussions with women and
local officials. Women are affected in two ways due to adverse events: (a) by the
direct impact of adverse events which jeopardize the livelihood security of women,
children and the elderly and (b) by forced migration of the male counterpart which
increases the workload for the women and put them at additional risk. Women are
generally less likely to move, due both to cultural and economic reasons, as are
children and the elderly (Ahmed et al., 2012; Tacoli, 2009). Around 20, 65 and 85%
of the household in Terai, Bihar and coastal Bangladesh respectively reported that
the past adverse events have significant impact on women. Terai of Nepal exhibits
a slightly different picture as compared to Bihar and coastal Bangladesh since only
20% of the households reported impact on women. Since agriculture in many parts
of Nepal is mostly feminized, women are accustomed to farming activities (Bhatta
& Aggarwal, 2015). Another reason could be that even though the male counterpart
is at home, women in Nepal contribute significantly in agriculture irrespective of
climate-induced events. Close to 20, 50 and 70% of the households respectively in
Terai, Bihar and coastal Bangladesh reported that past extreme events have an
impact on children, 10, 20 and 40% reported an impact on pregnant and lactating

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 15
women (see Figure 2). Women’s insecurity associated with male out-migration due
to extreme events is a big issue in coastal Bangladesh as reported by around 60% of
the households.
Since women bear the biggest burden from climate change impacts (Mnimbo,
Mbwambo, Kahimba & Tumbo, 2015), we asked women members of the household
about the consequences of male migration on women, children and elderly. The
socio-economic, cultural and religious values in the region have resulted in its
women and children being more vulnerable to adverse situations in comparison to
men. These dynamics form the vicious cycles where women and their children get
trapped. In general, workload of women both in the household and on the farm,
health risks and security issues increase due to adverse events and climate-induced
migration of male members while nutritional status, educational attainment and food
consumption decrease (see Table 5).
Figure 2: Effect of Adverse Weather Events on Women and Children.
Table 5. Impact of Extreme Weather Events and Male Out-migration on Vulnerable
Population
Group
Workload
Nutritional
status
Education
Health
risks
Security
issues
Food
consumption
Women
Children
Elderly
Overall effect
Severe
increase/decrease
Moderate
increase/decrease
Increase/decrease
slightly
Minor/no effect
0
10
20
30
40
50
60
70
80
90
Women members Children Pregnant and
lactating women
Security of the
women
% of the household
Effect of adverse weather events on
Terai, Nepal Bihar, India Coastal Bangladesh

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 16
Once male members out-migrate, women have to do all the farm work which many
women are not accustomed to. In this circumstance, women become the de facto
heads of the household, and this increases their workload, as they have to manage
the farm in addition to conducting the usual household chores (Shah & Shah, 2005;
Rogaly et al., 2002). Moreover, when women have to spend more time on productive
and reproductive tasks, it limits their time that could be spent on alternative
livelihood development activities. It reinforces already skewed customary gendered
divisions of labour. In Nepal for instance, the out-migration of men from the rural
areas has feminized the agriculture system in which women are extensively involved
in agricultural activities compared to their male counterpart (WOCAN, 2010). It is
estimated that women’s average contribution to the agricultural sector in Nepal is at
55-66% of the total labor (Gurung, Karki, Bista & Oh, 2012). It is found that women
not only have to work harder to meet their family’s food requirements; they also
have to work to pay off the debts that the household incurred. Some women reported
that remittance sent by their husbands relieve some of the debts, however, women
have to listen to the moneylenders which creates mental tension. Increasing
workloads may also result in families withdrawing daughters from schools to help
at home reducing their future opportunities.
Women suffer much from food and nutritional insecurity when there are extreme
events such as floods and/or droughts leading to a shortfall in production. They cope
with the decreased food availability at home by consumption smoothing in order to
feed their children. Prolonged malnutrition and increasing workloads during adverse
events affect women’s physical and mental health, especially of pregnant and
lactating mothers. Infants and young children also suffer adversely as a result
(Bhandari et al., 2007). In addition, women often face the most negative economic
implications of crop failure as they usually have fewer economic resources to fall
back on in times of crisis. This also has implications for the health of many women
and girls, as malnourishment increases so too does the risk of contracting infections.
Furthermore, women and girls’ lower socioeconomic status make it more difficult
for them to access and pay for treatment.
The most visible impact on children is the increased dropout rate from school.
Children, who drop out because of the workload, and who have no opportunity to
attend schools, support their parents by taking charge of household and farm
activities. Extreme events increase the time it takes to collect water in rural areas,
a task mainly done by women and girls in the region. This provides women with
less time to take part in other developmental activities. Furthermore, women and
children are exposed to a health risk from inadequate water supplies during
drought and also predicted changes in vector-borne diseases. Children are also at
the risk of malnutrition which has a long-term implication for overall development.
Children may be at a risk of early entry into work and exploitation in order to cover
lost income from agriculture. Ali, Begum, Shahabuddin & Khan (2006) noted that
women and children in Bangladesh are especially vulnerable to the health effects
of flooding.
Some women reported harassment and even sexual abuse in the absence of their
male counterpart. Ahmed et al. (2012) also reported sexual harassment of women
and girls. A young unmarried girl who has been sexually harassed is subject to
social stigma, for which the parents have to pay a large penalty, often in the form
of dowry at marriage (Ahmed et al., 2012). Thus, the head of household’s
migration under adverse events causes further costs and threatens the security of

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 17
the female members. Furthermore women face social and mental health stress due
to disasters and displacement.
The above-mentioned gender dimension of adverse climatic events demonstrate the
vicious cycle in which the structures in society keep women low in terms of
education, employment opportunities and mobility. The major gap here is that the
socially constructed dependency of women on male family members turns against
women in the absence of male. These gender nuances in climate-related
vulnerabilities vary notably based on caste, class, ethnicity, age and the level of
development of the society. The social barriers occur not as a result of their
femininity, but rather through the institutional and cultural environment that governs
acceptable behaviour and entitlements towards women (Jones & Boyd, 2011). There
remain significant caste inequalities in access and entitlement to key social safety
nets and resources in many countries of South Asia. This inequality may have
considerable implications for lower caste people to adapt amid climate change
(Jones & Boyd, 2011). Addressing the root causes and social conditions of
vulnerability needs to be dealt with at two levels; at the fundamental level, and at a
practical level. At the fundamental level the issues of entitlements, access to basic
needs, and the requirements of the resource poor to develop their full potential are
key issues. To address these fundamental issues, the will and commitment at all
levels is required. At a practical level, the initiatives which create space for men and
women to enhance their capacities and to enhance their survival strategies need to
be implemented. There is also a need to identify and create risk and vulnerability
reduction possibilities within the existing livelihoods. This study makes the case that
any policy intervention aimed at addressing climate-induced migration should care
not only for those who leave but also for those who stay, as they are often the most
vulnerable.
5.0 Conclusions
Study of general migration abounds in the literature, and there is growing interest
on climate-induced migration of farm households (Rademacher-Schulz et al., 2012;
McLeman & Smit, 2006) and the impact of adverse climatic events on vulnerable
populations. In the areas surveyed, around 28% of the households used migration as
a coping strategy in response to climate-induced events in the past. The probability
of migration is higher for individuals residing in areas experiencing greater rainfall
volatility. More deprived households (economically poor, landless and marginal
farmers, those with less number of months of food availability in a year) have a
higher probability to migrate under adversity. At the same time, the likelihood of
migration is higher with higher income.
The results provide for three distinct household profiles using migration as a short-
term coping strategy under adversity: (a) households that use migration to improve
their resilience; (b) households that have migration as a survival strategy; and (c)
households that cannot migrate (trapped population). These profiles represent a
spectrum with households within a profile being closer to one or the other of the
profiles on either side. They are thus not mutually exclusive and serve as a point of
departure for further research to refine key explanatory variables.
The results signal that relatively wealthier households are better adapted to a
changing climate because they can employ multiple measures including migration
to cope with extreme events. The poor, on the other hand, can find migration the
only alternative to evade suffering in the event of adversity. The trapped group has

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 18
intermediate income including women who are generally less likely to move, due to
socio-cultural and economic reasons, as are children and the elderly, and therefore
local adaptation efforts should target these groups specifically. Given that men’s
migration could contribute to the further deprivation and risk that women face, the
trapped population needs to be provided with greater social protection and
opportunities for livelihood security. Accordingly, investments that improve options
for the trapped group, such as improved agricultural technology, financial
instruments and off-farm income opportunities, will likely be critical for adapting to
a changing climate.
The current study is expected to improve the understanding of climate-induced
migration, and stimulates policy debate on how to tackle the anticipated movement
of people due to changing weather patterns in the coming years. It also opens an
avenue for further research deeply covering several socio-demographic and
biophysical factors as potent drivers of migration under distress situations.
Acknowledgements
The authors would like to thank all members of site survey team and all interviewees
in the local communities (Bihar state of India, Coastal Bangladesh and Terai of
Nepal). We appreciate the support from CCAFS’s numerous investors and CGIAR
centre colleagues and partners. CCAFS data collection work was carried out with
funding by the European Union (EU) and with technical support from the
International Fund for Agricultural Development (IFAD). We also extend our thanks
to Amit Shrivastava for preparing a nice map of surveyed sites. The anonymous
reviewers deserve special appreciation for providing their valuable comments on our
manuscript. Last but not the least Robert Rivers gets special thanks for providing his
generous support in going through the manuscript and editing the language.
References
Asian Development Development Bank (ADB). (2009). Policy options to support
climate-induced migration. Technical Assistance Report. Manila, Philippines:
Asian Development Bank.
Asian Development Development Bank ABD. (2012). Addressing climate change
and migration in Asia and the Pacific. Final Report. Manila, Philippines: Asian
Development Bank.
Aggarwal, P. K., Joshi, P. K., Ingram, J. S. I., & Gupta, R. K. (2004). Adapting food
systems of the Indo-Gangetic plains to global environmental change: key
information needs to improve policy formulation. Environmental Science and
Policy, 7 (6), 487-498.
Ahmed, A. U. (2009). Implications of climate change in relation to water related
disasters in South Asia. Earth and Environmental Sciences, 6, 292020.
Ahmed, A. U., Hassan, S. R., Etzold, B., & Neelormi, S. (2012). Where the rain
falls, case study: Bangladesh. Bonn, Germany: United Nations University and
Institute for Environment and Human Security (UNU-EHS).
Ali, Z., Begum, S., Shahabuddin, Q., & Khan, M. (2006). Rural poverty dynamics
2005/2006: Evidence from 64-village census plus Dhaka. Dhaka, Bangladesh:
Bangladesh Institute of Development Studies (BIDS) and Chronic Poverty
Research Centre (CPRC).

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 19
Arendse, A., & Crane, T. A. (2011). Impacts of climate change on smallholder
farmers in Africa and their adaptation strategies: What are the roles for research?
In CIAT Working Document No. 221: International Center for Tropical
Agriculture.
Battisti, D., & Naylor, R. (2009). Historical warnings of future food insecurity with
unprecedented seasonal heat. Science, 323(2), 240-244.
Bernabe, D., & Penunia, E. (2009). Gender links in agriculture and climate change.
Women in Action, 11, 19–22.
Bhandari, H., Pandey, S., Sharan, R., Naik, D., Hirway, I., Taunk, S.K., & Sastri, A.
S. R. A. S. (2007). Economic costs of drought and rice farmers’ drought-coping
mechanisms in eastern India. In S. Pandey, H. Bhandari, & B. Hardy (Eds.),
Economic costs of drought and rice farmers’ coping mechanisms: A cross-
country comparative analysis (pp. 43-111). Los Baños, Philippines.
International Rice Research Institute.
Bhatta, G. D., & Aggarwal, P. K. (2015). Coping with weather adversity and
adaptation to climatic variability: A cross-country study of smallholder farmers
in South Asia. Climate and Development. DOI:
10.1080/17565529.2015.1016883
Bhatta, G. D., Aggarwal, P. K., Shrivastava, A. K. & Sproule, L. (2015). Is rainfall
gradient a factor of livelihood diversification? Empirical evidence from around
climatic hotspots in Indo-Gangetic Plains. Environment, Development and
Sustainability. DOI: 10.1007/s10668-015-9710-6
Bhatta, G. D., Singh, R. K. P., & Kristjanson, P. (2013). Summary of baseline
household survey results: Vaishali site, Bihar state (Northeast India). Available
online at: www.ccafs.cgiar.org Copenhagen, Denmark: CGIAR Research
Program on Climate Change, Agriculture and Food Security (CCAFS).
Bhattacharyya, A., & Werz, M. (2012). Climate change, migration and conflict in
South Asia. Washington, DC: Centre for American Progress.
Brockhaus, M., Djoudi, H., & Locatelli, B. (2013). Envisioning the future and
learning from the past: Adapting to a changing environment in northern Mali.
Environmental Science and Policy, 25, 94-106.
Brooks, N., Grist, N., & Brown, K. (2009). Development futures in the context of
climate change: Challenging the present and learning from the past.
Development Policy Review, 27(6), 741-765.
Brown, O. (2008). Climate change and forced migration: observations, projections
and implications. Background Paper for the 2007 Human Development Report.
Geneva, Switzerland: United Nations Development Programme.
Canadian International Development Agency [CIDA]. (2002). Gender equality and
climate change: Why consider gender equality when taking action on climate
change? Quebec, Canada: Canadian International Development Agency.
Cunguara, B., Langyintuo, A., & Darnhofer, I. (2011). The role of nonfarm income
in coping with the effects of drought in southern Mozambique. Agricultural
Economics, 42(6), 701-713. DOI: 10.1111/j.1574-0862.2011.00542.x

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 20
Demetriades, J., & Esplen, E. (2010). The gender dimensions of poverty and climate
change adaptation. In R. Mearns, & A. Norton (Eds.), Social dimensions of
climate change: Equity and vulnerability in a warming world (pp. 133-143).
Washington, DC: The World Bank.
Denton, F. (2002). Climate change vulnerability, impacts, and adaptation: why does
gender matter? Gender and Development, 10, 10–20.
Deshingkar, P., & Akter, S. (2009). Migration and human development in India.
Human Development Research Paper: United Nations Development
Programme.
Deshingkar, P., & Start, D. (2003). Seasonal migration for livelihoods in India:
Coping, accumulation and exclusion. London: Overseas Development Institute.
Diffenbaugh, N., Pal, J., Trapp, R., & Giorgi, F. (2005). Fine-scale processes
regulate the response of extreme events to global climate change. Proceedings
of the National Academy of Sciences, 102(44), 15774-15778.
Emanuel, K. (2005). Increasing destructiveness of tropical cyclones over the past 30
years. Nature, 436(7051), 686-688.
Etzold, B., Ahmed, A. U., Hassan, S. R., & Neelormi, S. (2014). Clouds gather in
the sky, but no rain falls. Vulnerability to rainfall variability and food insecurity
in Northern Bangladesh and its effects on migration. Climate and Development,
6(1), 18-27, DOI: 10.1080/17565529.2013.833078
Field, A. (2005). Discovering statistics using SPSS. New Delhi: Sage Publications.
Findlay, A., & Geddes, A. (2011). Critical views on the relationship between climate
change and migration: some insights from the experience of Bangladesh. In: E.
Piguet, A. Pecoud, & P. de Guchteneire (Eds.), Migration and climate change
(pp. 138-159). UNESCO/Cambridge University Press.
Gartaula, H., Niehof, A., & Visser, L. (2010). Feminisation of agriculture as an effect
of male out-migration: Unexpected outcomes from Jhapa District, Eastern Nepal.
The International Journal of Interdisciplinary Social Sciences, 5(2), 565-578.
Gebru, G. W., & Beyene, F. (2012). Rural household livelihood strategies in
drought-prone areas: A case of Gulomekeda District, eastern zone of Tigray
National Regional State, Ethiopia. Journal of Development and Agricultural
Economics, 4(6), 158-168.
Geddes, A., & Somerville, W. (2013). Migration and environmental change:
Assessing the developing European approach. Policy Brief Series. Brussels:
Migration Policy Institute Europe.
Giorgi, F., & Bi, X. (2005). Regional changes in surface climate interannual
variability for the 21st century from ensembles of global model simulations.
Geophysical Research Letter, 32(13), L13701.
Gray, C. L., & Mueller, V. (2012). Natural disasters and population mobility in
Bangladesh. Proceedings of the National Academy of Sciences, 109(16),
6000-6005.
Gukurume, S. (2013). Climate change, variability and sustainable agriculture in
Zimbabwe’s rural communities. Russian Journal of Agricultural and Socio-
Economic Sciences, 2(14), 89-100.

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 21
Gurung, A., Karki, A., Karki, R., Bista, R., & Oh, S. E. (2012). Consequences of
water induced disasters to livelihood activities in Nepal. Korean Journal of
Environmental Agriculture, 31(2), 129-136.
Hunter, L. M., & David, E. (2009). Climate change and migration: Considering the
gender dimensions. Working Paper, Population Program. Boulder, CO: Institute
of Behavioral Science, University of Colorado.
IPCC. (2007). Climate change 2007: Working Group II: Impacts, adaptation and
vulnerability. Appendix I: Glossary. Fourth Assessment Report. In Climate
Change 2007. Intergovernmental Panel on Climate Change.
Jones, L., & Boyd, E. (2011). Exploring social barriers to adaptation: Insight into the
role of caste in Western Nepal. Global Environmental Change, 21, 1262-1274.
Kelkar, U., & Bhadwal, S. (2007). South Asian regional study on climate change
impacts and adaptation: Implications for human development. Occasional
Paper, UNDP: Human Development Report Office.
Kristjanson, P., Garlick, C., Ochieng, S., Förch, W., & Thornton, P. K. (2011).
Global summary of baseline household survey results. Copenhagen, Denmark.
Retrieved at: www.ccafs.cgiar.org CGIAR Research Program on Climate
Change, Agriculture and Food Security (CCAFS).
Kumar, K. S. K., & Viswanathan, B. (2012). Weather variability and agriculture:
Implications for long and short-term migration in India. Working Paper No. 220.
Delhi, India: Delhi School of Economics.
Lal, M., Meehl, G. A., & Arblaster, J. M. (2000). Simulation of Indian summer
monsoon rainfall and its intraseasonal variability. Regional Environmental
Change, 1(3-4), 163-179.
Long, J. S. (1997). Regression models for categorical and limited dependent
variables: Advanced quantitative techniques in the Social Sciences. Thousand
Oaks, CA: Sage Publications.
Massey, D. S., Arango, J., Hugo, G., Kouaouci, A., Pellegrino, A., & Taylor, J. E.
(1993). Theories of international migration: A review and appraisal. Population
and Development Review, 19(3), 431-466.
May, W. (2004). Simulation of the variability and extremes of daily rainfall during
the Indian summer monsoon for present and future times in a global time slice
experiment. Climate Dynamics, 22(2-3), 183-204.
McLeman, R., & Smit, B. (2006). Migration as an adaptation to climate change.
Climatic Change, 76(1-2), 31-53.
Mendelsohn, R., Dinar, A., & Williams, L. (2006). The distributional impact of
climate change on rich and poor countries. Environment and Development
Economics, 11(02), 159-178.
Meza, L. E. R. (2015). Adaptive capacity of small-scale coffee farmers to climate
change impacts in the Soconusco region of Chiapas, Mexico. Climate and
Development, 7(2), 100-109. DOI: 10.1080/17565529.2014.900472
Meza, L.E.R. (2010). Climate change, poverty and migration processes in Chiapas,
Mexico. International Journal of Labour Research, 2(2), 187-210.

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 22
Mitchell, T., Tanner, T., & Lussier, K. (2007). We know what we need: South
Asian women speak out on climate change adaptation. Brighton, United
Kingdom: ActionAid and Institute of Development Studies (IDS) at the
University of Sussex.
Mnimbo, T. S., Mbwambo, J., Kahimba, F. C, & Tumbo, S. D. (2015). A gendered
analysis of perception and vulnerability to climate change among smallholder
farmers: the case of Same District, Tanzania. Climate and Development. DOI:
10.1080/17565529.2015.1005038.
Nelson, V., Kate, M., Terry, C., & John, M. (2002). Uncertain predictions, invisible
impacts, and the need to mainstream gender in climate change adaptations.
Gender and Development, 10, 51–59.
Ojha, H., Sulaiman, R., Sultana, P., Dahal, K. R., Thapa, D., Mittal, N., et al. (2013).
Is South Asian agriculture adapting to climate change? Evidence from the Indo-
Gangetic Plains. Agroecology and Sustainable Food Systems, 38(5), 505-531.
Rademacher-Schulz, C., Afifi, T., Warner, K., Rosenfeld, T., Milan, A., Etzold, B.,
et al. (2012). Rainfall variability, food security and human mobility. An
approach for generating empirical evidence. Intersections No. 10, Bonn,
Germany: UNU-EHS.
Rogaly, B., Daniel, C., Rafique, A., Rana, K., Sengupta, A., & Biswas, J. (2002).
Seasonal migration and welfare/illfare in eastern India: a social analysis. Journal
of Development Studies, 38(5), 89-114.
Salau, E. S., Onuk, E. G., & Ibrahim, A. (2012). Knowledge, perception and adaptation
strategies to climate change among farmers in Southern Agricultural Zone of
Nasarawa State, Nigeria. Journal of Agricultural Extension, 16(2), 199-211.
Shah, D. C., & Shah, A. (2005). Migration in remote tribal areas: evidence form
southwestern Madhya Pradesh. Indian Journal of Agricultural Economics,
60(2), 184-204.
Sherbinin, A. D, VanWey, L., McSweeney, K., Aggarwal, R., Barbieri, A., Henry,
S., et al. (2008). Rural household demographics, livelihoods and the
environment. Global Environment Change, 18(1), 38-53.
Shewmake, S. (2008). Vulnerability and the impact of climate change in South
Africa’s Limpopo River Basin. IFPRI Discussion Paper 00804. Washington,
DC: International Food Policy Research Institute.
Singh, R. B., Kumar, P., & Woodhead, T. (2002). Smallholder farmers in India:
Food security and agricultural policy. RAP Publication: 2002/03. Bangkok,
Thailand: Food and Agriculture Organization of the United Nations, Regional
Office for Asia and the Pacific.
Spring, U. O. (2008). Gender and disasters: Human, gender and environmental
security. A huge challenge studies of the University: Research, Counsel,
Education. Bonn,Germany: UNU Institute for Environment and Human
Security (UNU-EHS.
Tacoli, C. (2009). Crisis or adaptation? Migration and climate change in a context
of high mobility. In J. M. Guzman, G. Martine, G. McGranahan, D. Schensul,
& C. Tacoli (Eds.), Population dynamics and climate change (pp. 104-118).
New York: UNFPA; London: IIED.

Bhatta, Aggarwal, Poudel, & Belgrave
Journal of Rural and Community Development, 10, 4(2015) 1-23 23
Trapp, R., Diffenbaugh, N., Brooks, H., Baldwin, M., Robinson, E., & Pal, J. (2007).
Changes in severe thunderstorm environment frequency during the 21st century
caused by anthropogenically enhanced global radiative forcing. Proceedings of
the National Academy of Sciences, 104(50), 19719-19723.
United Nations Population Fund [UNFPA]. (2009). State of world population
2009—Facing a changing world: Women, population, and climate. New York:
United Nations Population Fund.
Warner, K. (2010). Global environmental change and migration: Governance
challenges. Global Environment Change, 20, 402-413.
Warner, K., & Afifi, T. (2014). Where the rain falls: Evidence from 8 countries on
how vulnerable households use migration to manage the risk of rainfall
variability and food insecurity. Climate and Development, 6(1), 1-17, DOI:
10.1080/17565529.2013.835707.
Women Organizing for Change in Agriculture and NRM [WOCAN]. (2010).
Scoping study on women’s leadership in the agriculture sector in Nepal.
Kathmandu, Nepal: Women Organizing for Change in Agriculture and NRM.
Wood, S. A., Jina, A. S., Jain, M., Kristjanson, P., & DeFries, R. S. (2014).
Smallholder farmer cropping decisions related to climate variability across
multiple regions. Global Environmental Change, 25, 163-172. DOI:
10.1016/j.gloenvcha.2013.12.011.
World Bank. (2009). South Asia: Climate change strategy. Washington, DC: The
World Bank.
Yesuf, M., & Bluffstone, R. (2008). Risk aversion in low-income countries:
experimental evidence from Ethiopia. IFPRI Research Brief 15-16, Washington,
DC: International Food Policy Research Institute.