Weathering Changes: Livelihood Adaptation to Weather Shocks in Rural India by Disadvantaged Social Groups

Abstract

In this article, we examine how households from disadvantaged social groups in India adapt through migration to climate-related shocks. We examine the relative importance of factors like social networks and public intervention in enabling adaptation to slow-onset climate change. We use household- and village-level data from two consecutive waves of the Indian Human Development Survey and gridded weather data from CRU at the University of East Anglia for our analysis. Our results indicate that, in India, major changes in dryness significantly increase migration, but disadvantaged social groups facing climate change are less likely to migrate. Social networks do not play any significant role in the migration behaviour of disadvantaged groups facing these changes. Efficient implementation of poverty alleviation programmes does improve the probability of migration among these groups.

JEL Codes: O15, Q54

Keywords

Climate Change Social Groups Migration India

1. Introduction

Climate change is no longer a distant fact. Its effect can be seen all over the world as extreme weather events have become more frequent and severe (Christensen et al., 2007; Masson-Delmotte et al., 2018). Such changes in climate have important consequences in developing countries where the population is often dependent on agriculture and other climate-sensitive natural resources for their livelihood (Dell et al., 2014; Harrington et al., 2016; Millock, 2015; Skoufias et al., 2011). Given the immediacy and severity of the effects of climate change, there has been considerable interest in understanding the ability of these societies to adapt. Such adaptation is often difficult among disadvantaged social groups in these countries, as they are more likely to be exposed to such conditions and have limited resources to mitigate those conditions. In addition, they are more likely to be further impoverished as a result of climate hazards, which reduces their ability to cope and recover from its consequences (Field & Barros, 2014; Islam & Winkel, 2017).

India experienced an exponential increase in extreme climate events during the period 1970–2019, with the last two decades showing significant acceleration (Mohanty, 2020); in 2018, India was ranked the fifth most climate-vulnerable country in the world (Eckstein et al., 2019). One of the most important consequences of climate change in India is its effect on the agriculture sector. Although agriculture and allied sectors contribute to only 15 per cent of GDP, nearly 75 per cent of Indian families rely on rural income. The agriculture sector remains a major source of livelihood for more than 70 per cent of the rural workforce, and approximately 70 per cent of the poor in India are concentrated in rural areas (Himanshu et al., 2013).

In this article, we focus on adaptation to climate change among disadvantaged social groups in India. India has several social groups that were either historically excluded from its development process or for whom the development process remained distant because of their remote location. Two of the largest groups are the scheduled castes (SCs) and scheduled tribes (STs). SCs have been historically segregated primarily by occupational opportunities, and STs mainly live off natural resources, but their livelihood is often threatened by encroachment from the rest of society. According to the 2011 census, SCs and STs comprised 16.2 and 8.2 per cent of the population, respectively, yet accounted for 40.6 per cent of the poor in the 2004/2005 household expenditure survey. These groups have the highest incidence of poverty in India along with poor outcomes in the areas of physical and human capital. Being a member of these groups is considered one of the primary risk factors for being poor in India.

Several studies in India have highlighted the effect of climate change on health and mortality. Burgess et al. (2017), using district-level daily weather and annual mortality data from 1957 to 2000, find that hot days lead to substantial increases in mortality in rural but not in urban India. Others note how climate change also results in the misallocation of resources such as lower investment in human capital. For example, Garg et al. (2020) observe that hot days during the growing season reduce agricultural yields as well as test score performance, with comparatively modest effects of hot days in the non-growing season. Carleton (2017) shows that fluctuations in temperature significantly influence suicide rates during India’s agricultural growing season, when heat also lowers crop yields.

The effect of climate change in India corroborates studies in other countries that show how climate-related shocks can force vulnerable households to pursue unsustainable coping strategies as they are often based on imperfect information (Schipper, 2020). Warner and Van der Geest (2013) provide several evidences of such maladaptation. For example, Rabbani et al. (2013) show how preventive measures for flooding introduced in Bangladesh were often not adequate for repeated or more severe events. Further, the introduction of saline-resistant crops was found to be inadequate because salinity increased over time. In Bhutan, farmers were exposed to price fluctuations in non-standard agricultural products after they switched to drought-resistant crops following a decline in rainfall (Kusters & Wangdi, 2013). In Gambia, a reduction in income and rising food prices as a result of drought led to reduced food consumption with future consequences on productivity (Yaffa, 2013). In Kenya, following flooding and the consequent destruction of crops, farmers sold off their farm animals, thereby losing production capabilities in the next crop cycle (Opondo, 2013). Generally, such imperfect adaptation strategies add to the original effect of climate change on sustainable development.

Under such circumstances, migration to diversify livelihood spatially or in sectors that are not sensitive to climate change plays a very important role in mitigating its effects (Adger & Adams, 2013; Barnett & Webber, 2009; Black et al., 2011; Renaud et al., 2007; Sabates-Wheeler & Waite, 2003; Warner, 2010). Historically, migration has played the role of a natural adaptive strategy for adverse environmental conditions (Cattaneo et al., 2019; Hugo, 1996; McLeman & Smit, 2006). There are numerous studies on the increase in migration probability due to extreme heat or a lack of precipitation (Feng et al., 2010; Gray & Bilsborrow, 2013; Gray & Mueller, 2012a, 2012b; Jessoe et al., 2018). Yet others note that climate shocks can also decrease migration due to the adverse effect they have on the resources required to finance migration journeys (Cattaneo et al., 2019). Several studies also highlight social group-based differences in the ability to migrate (Kartiki, 2011; McLeman & Smit, 2006; Nielsen & Reenberg, 2012). It is often the case that those who are most vulnerable to climate change are most constrained to move and smooth consumption over time (Black et al., 2013).

Migration is generally low in India compared to countries at similar stages of development (Keshri & Bhagat, 2010; Munshi & Rosenzweig, 2016; Topalova, 2010). Though there has been some increase between decennial censuses, the pattern of migration has generally remained the same in the last few decades. According to the 2001 census, approximately 3.2 per cent of the population were migrants for economic reasons, which increased to approximately 4.2 per cent in the 2011 census (Nayyar & Kim, 2018). The bulk of the movement is within the same district followed by those within the state (Bell et al., 2015). In India, the cross-district migration rate was around 2.8 per cent (2001 Census) compared to 9.9 per cent (between prefectures) in China (2000 census) that had restrictions on such movement. For the same period, interstate migration was slightly above 1 per cent in India compared to 4.7 per cent (between provinces) in China (Kone et al., 2018).

However, temporary or circular migration for work has increased over the years (Deshingkar & Farrington, 2009). The majority of these migrations involve a few members leaving the household for work while the household remains in the state of origin. The percentage of households with at least one circular migrant in rural areas increased from 3.2 per cent in 1999–2000 (55th round) to 5.8 per cent in 2007–2006 (64th round) and further to 6.5 per cent in 2013 (70th round) of the National Sample Survey (Bhagat & Keshri, 2020).

But the poor are less likely to be among the group of migrant households. Poverty makes households risk-averse to such enterprise. Also, their low human capital makes it less likely for them to be employed in more productive sectors of the economy. For poorer groups, migration may also increase vulnerability and reinforce poverty, as when such migration is debt-financed or when their only employment possibilities are in the precarious urban unskilled labour market. As a result, they often decide against migration (Banerjee & Newman, 1991; Kanbur, 1979).

For disadvantaged social groups in India, migration takes special importance when facing the effects of slow-onset climate change because other avenues such as switching crops or investment in irrigation are not pertinent for they are often landless or with very little landholdings (Government of India, 2015; Mohanty, 2001). However, migrants are less likely to be members of scheduled groups (Bhattacharya, 2002; Deshingkar & Start, 2003; Hnatkovska & Lahiri, 2015; Mosse, 2010). This is not only because of the higher incidence of poverty among them but also the discrimination faced by members of those groups. Iversen et al. (2014) show that SCs do better in villages where they are in the majority. In addition to the barriers to migration highlighted earlier, the scheduled groups face barriers due to affirmative action programmes which reserve jobs and educational opportunities for disadvantaged groups based on their state of original residence (Kone et al., 2018). These unexpected consequences of affirmative action policies may not directly affect circular migration associated with slow-onset climate change but can affect migration indirectly through the size of migrant networks of members of these social groups.

Several studies have examined the relationship between the effects of climate change and migration in the Indian context. Viswanathan and Kumar (2015) examine the three-way linkage between weather, agricultural performance and internal migration. They also note that weather-induced drop in agricultural productivity increases interstate migration. Dallmann and Millock (2017) show that drought frequency in the origin state increases interstate migration in India. Such a migration pattern is stronger in agricultural states. Sedova and Kalkuhl (2020), using household survey data, show that adverse weather shocks decrease rural–rural and international migration and increase rural–urban migration. However, none of the studies on India specifically look at the impact of climate change on migration patterns of disadvantaged groups, even though their lives are more likely threatened by slow-onset climate change.

2. Research Questions

Several studies in India have examined the impact of climate change on agricultural output and jointly its effect on irrigation (Guiteras, 2008; Taraz, 2018). While such an approach acknowledges the effect of changes in weather conditions on plant stress or groundwater access, the variables used to identify the climate events, like meteorological drought (usually measured by deviation in temperature or precipitation from a long-term average), are limited in capturing the effect of climate–crop relations. Alternatives like the standardised precipitation index, though widely used, do not account for the impact of rising temperature and several other related environmental factors. Such relations are better captured by measures of agricultural drought identified by plant water stress, reduced biomass and reduced yield due to soil moisture deficits (Wilhite & Glantz, 1985). In this article, we identify these characteristics using the standardised precipitation–evapotranspiration index (SPEI) (Vicente-Serrano et al., 2010). Instead of using only temperature, precipitation and so on, SPEI uses time series data on ‘water balance’, which is the difference between precipitation and potential evapotranspiration. Potential evapotranspiration is the amount of evaporation and transpiration that would occur as a result of temperature, vapour pressure, cloud cover and wind-field value if a sufficient water source were available. SPEI is also used by others: for example, Defrance et al. (2023) in their study of drought-induced rural–urban migration in Mali.

Slow-onset climate change evolves gradually through incremental changes occurring over a long period of time as against rapid-onset events like a major storm or flood. The migration that is associated with these events is therefore quite different. While a rapid-onset climate event may displace the entire household, slow-onset events are not likely to have similar migration outcomes. As a result of the gradual nature of the depletion of their livelihood, labour migration following slow-onset climate change is mostly a livelihood diversification strategy (Fussell et al., 2014). Our first research question examines whether climate change measured using changes in water balance at shorter durations is associated with changes in migration among households. As we are interested in the heterogeneity of this migration pattern by social groups, we also examine whether households from disadvantaged social groups migrate less even when facing slow-onset climate change.

Next, we look at two different sources of intervention with possible consequences on migration probability among those affected by climate change. First, we look at the efforts that households can take themselves by supplementing resources from social networks. Then we look at the effect of a public intervention.

In developing countries with poorly functioning markets and inadequate welfare programmes, households often turn to social networks in times of income shocks (Gough, 2004). In India, several studies demonstrate evidence of how social networks provide informal insurance in times of income shocks (Munshi & Rosenzweig, 2016; Rosenzweig & Stark, 1989; Townsend, 1994). Such social networks can be built proactively by being members of different groups or through kinship and reciprocal arrangements and it may reduce the need to migrate in the face of climate shocks (Adger, 2010; Wolf, 2011). On the other hand, for those unable to participate in migration due to resource constraints, social networks may supplement resources needed to migrate and increase migration probabilities. So, our next research question examines whether access to social networks decreases or increases the chances of migration in households facing slow-onset climate change. We are also interested in examining whether the effect is heterogeneous by social group.

Several studies have highlighted the role of social protection programmes in climate change adaptation (Béné, 2011; Coirolo et al., 2013; Kuriakose et al., 2013; Weldegebriel & Prowse, 2013). Here we specifically look at the implication of the National Rural Employment Guarantee Scheme (NREGS) introduced in 2005. It is a workfare programme that guarantees manual employment for 100 days for a member of a household who is willing to work (Khera, 2011). By incorporating work requirements as a screening device, it ensured proper targeting (Drèze & Khera, 2011). The programme did achieve some success like increasing consumption expenditure (Deininger & Liu, 2013; Imbert & Papp, 2015; Ravi & Engler, 2015). However, its impact varied by state, and poorer states lagged in their capacity to implement the scheme (Dutta et al., 2014; Liu & Barrett, 2013; Stahlberg, 2012). Hagen-Zanker and Himmelstine (2013) highlight a complex and context-sensitive relationship between NREGS and internal labour migration. Some studies note a decline in distress migration during agricultural lean seasons following the introduction of the scheme (Deshingar et al., 2010; Imbert & Papp, 2015; Liu & Barrett, 2013), while others observe mixed results (Das, 2015; Datar, 2007; Khan & Saluja, 2007; Novotný et al., 2013; Solinski, 2012). Some of it is because the studies were conducted at different stages of the implementation, while some are possibly due to state-level differences in the efficiency of programme implementation. While several studies have looked into the effect of the programme on migration, their focus was specifically on its implication on lean season migration. More importantly, none of these studies specifically explored its effect on migration among scheduled groups in areas facing adverse slow-onset climate change. Our final research question examines whether poverty alleviation programmes affect migration among those facing climate shocks and whether the effect varies by social group.

3. Data

The ‘New Economics of Labor Migration’ framework, pioneered by Stark and Bloom (1985),provides the relevant framework to conceptualise migration behaviour associated with slow-onset climate change. Here the interest is not in whether the entire household migrates, but whether households diversify livelihood by sending a few members as a household-level income diversification strategy. While the decennial census captures migration between waves, they are not useful for capturing short-term or temporary migration where only a few members of the household leave for migration. Other sources like the National Sample Survey do capture household-level data (e.g., 55th round and 64th round), but they do not track the same household over the waves. This is the main reason for using India Human Development Survey (IHDS) data. The IHDS survey is a nationally representative multi-topic panel survey in India and is conducted jointly by the University of Maryland and the National Council of Applied Economic Research (Desai et al., 2018, 2019).

Migrants within a household can be identified in different ways in the IHDS surveys. The first is the ‘tracking sheet’ for the survey in the second wave which reports if anyone is not present in the household as they have migrated for economic or other reasons. While this is the only place where a migrant is directly identified, it does not clearly state the economic attachment through remittances or transfers. This, however, can be found in the non-resident files in both waves where it is possible to identify the spouse of any household member or parent of any children who reside outside and regularly transfers money to this household. Such individuals meet our conceptualisation of livelihood diversification at the household level and are identified as migrants for this analysis.

Our primary goal is to understand how disadvantaged groups facing structural discrimination react to slow-onset climate change. So, the analytical strategy is to essentially compare them with those who are not from those social groups. However, the intersectionality of different types of discrimination and exclusion in India poses a challenge in simply comparing the scheduled groups to those who are not. So, we also consider the ‘other backward classes (OBC)’ who face similar economic disadvantages to the scheduled group, but, unlike the latter, they do not face social barriers like untouchability and they have been more successful in altering their conditions through political and social organisations in recent years (Thorat & Joshi, 2020). There are still other sources of discrimination and exclusion which makes the group who are neither Scheduled or OBC quite heterogenous in terms of disadvantages. An important source of such disadvantage is based on faith (Thorat, 2010). So, we further break down the ‘Other’ group into those belonging to the majority Hindu religion and those who are not. We use a variable available in the survey which combines the intersectionality of these discriminations to create our social group variable classified as Hindus, non-Hindus, OBCs and scheduled groups. As a result of the nature of the constructed variable, the scheduled groups and the OBCs also include Hindus, but the Hindu group identified here includes those who do not face disadvantages as the scheduled groups or the OBCs.

We restrict our analytical sample to households residing in rural areas in Wave I of the survey. The attrition rate for the rural subsample is low (around 9 per cent), giving us an analytical sample of 25,570 households for which relevant information is available.

Table 1 shows the distribution of migrant households among different social groups. The overall percentage of migrant households has increased between Wave I of the survey conducted in 2005 and Wave II of the survey conducted in 2011–2012. The tabulation shows that non-Hindus and OBCs are more mobile in Wave I, but Hindus are more mobile in Wave II of the survey. The proportion of migrant households from the scheduled groups increased between the waves, but their proportion is still the lowest among all social groups.

Table 2 shows the distribution of different characteristics of the households where the number of migrants increased between the waves and where they did not. Households which saw an increase in the number of migrants are more likely to have a larger number of members and lower dependency ratio. They are more likely to be literate and less likely to be poor. The occupation of the household head for households which saw an increase in the number of migrants is less likely to be agricultural labour and more likely to be non-agricultural occupations.

Table 1.

Per Cent of Migrant Households in the IHDS Sample

	Social Groups
	Other Hindu	Other Non-Hindu	OBC	SC/ST	Total
Wave I	4.7	5.5	5.1	3.9	4.7
Wave II	12.3	11.5	10.2	8.2	10.1

Source: Computed by the author.

Table 2.

Household Characteristics in the IHDS Sample

	Change in Number of Migrants in HH
	Same/Lower	Increase	Total
HH size	5.98	6.53	6.03
Dependency ratio	0.59	0.58	0.59
Literate	74.2%	76.5%	74.4%
Poor	22.4%	21.7%	22.4%
Main occupation
Cultivation	39.7%	37.5%	39.5%
Ag labour	19.8%	14.6%	19.3%
Non-ag labour	16.4%	16.5%	16.4%
Others	24.2%	31.4%	24.8%

Source: Computed by the author.

We use TS4.04 data for monthly gridded precipitation and potential evapotranspiration for 1970–2010 from the Climate Research Unit of the University of East Anglia (Harris et al., 2020) to construct the SPEI. Calculation of the SPEI involves fitting the water balance data with a distribution with heavier tails, transforming the probability distribution into a standardised normal distribution, and then computing the inverse probability to obtain the index. We use R library SPEI created by Beguería and Vicente-Serrano (2017) for the calculation. SPEI can be calculated for different time scales like three, six or more months. For the purpose of this article, we use six months to reflect short-term agricultural drought. SPEI values are classified as wet if they are greater or equal to 1 and dry if they are less than or equal to negative 1. Values between –0.99 and +0.99 are considered normal water balance.

Figure 1 shows the distribution of SPEI6 values across districts used in this study for each month between 2000 and 2010. Each line represents a box plot with the thicker part showing the interquartile range. For several months between 2002 and 2003, the interquartile ranges dipped below –1 indicating the incidence of agricultural drought. A similar distribution can be seen between 2009 and the first half of 2010. Also, between 2006 and 2007 and during the second half of 2010, the interquartile range crossed +1, indicating periods of wetness.

Figure 1.

Source: Computed by the author.

Figure 2 shows the change in the distribution of the number of dry months measured using SPEI6 among districts between the two waves. It shows a decline in the number of dry months in the western and southern parts of the country and an increase in the central and northern parts of the country. Figure 3 shows the same for wet months. Here we see a noticeable increase in the number of wet months in most of the southern part of the country.

Figure 2.

Source: Computed by the author.

Figure 3.

Source: Computed by the author.

4. Analytical Strategy

Our approach to identify the effect of slow-onset climate change on migration focuses on the effect of the change in district-level water balance for the preceding five years on the change in the number of household members identified as economic migrants between the two surveys. We use the following regression specification to identify the effect of slow-onset climate change on household-level migration.

\begin{array}{l} Δ M i g r a t i o n_{i} = β_{0 j} + β_{1} Δ W e t n e s s_{j} + β_{2} Δ D r y n e s s_{j} + \\ \begin{matrix} \begin{matrix}  \end{matrix} \end{matrix} β_{3} S o c G r o u p_{i} + β_{4} S o c N e t_{i} + β_{5} N R E G S a d v_{i} + β_{6} Z_{i} + ϵ_{i_{j}} \\ \begin{matrix}  \end{matrix} β_{0 j} = γ_{00} + u_{0 j}, \end{array}

(1)

where i indicates a household-level variable and j indicates a district-level variable. For estimation purposes, we use multilevel modelling using a random intercept at the district level (Aguinis et al., 2013; Gelman & Hill, 2006; Mathieu et al., 2012).

Our dependent variable, ΔMigration_i, is calculated as a binary variable indicating whether the household shows an increase in the number of migrants over the surveys. This is calculated by comparing the number of economic migrants in the household across the two waves. It is a household-level measure. Since the outcome variable is a binary variable indicating an increase in migration, the coefficient of the independent variables can be interpreted as in a linear probability model (Horrace & Oaxaca, 2006).

Since wetness and dryness can potentially have different effects, they are accounted for separately. ΔWetness_j and ΔDryness_j refer to a change in the count of months with SPEI6 values greater than positive one or less than negative one, respectively, in the 60 months before the surveys in the respective districts. It is the nature of slow-onset climate change that the magnitudes of effects of a single month’s change on migration probabilities are low, even if they are significant and robust. So, we consider an alternative way to characterise the change in climate between the waves. We identify districts with lower or above median (for the specific wave) count of wet and dry months in the five years preceding the survey. Then we create a binary variable indicating whether or not the district shows a movement between the waves from lower to higher than the median count.

SocGroup_i refers to the social group of the household head and is classified as mentioned earlier. SocNet_i here indicates whether the household has access to social networks. Here we use two alternative definitions based on IHDS instruments. In one a binary variable is created based on whether the household has anyone among their acquaintances and relatives any professionals associated with medical care, education or government services who also belong to the same social group. This will capture the quality of the household’s social network quality (SocNetQ). We also create an alternative variable to indicate whether anyone in the household belongs to a religious or social group, a festival society or a caste association. This is expected to capture the household’s social network membership (SocNetM). We use the binary variable NREGSadv_i which indicates whether, for the village in which the household resides, the NREGS wages are at least as high as prevailing female unskilled worker wages.

The regressions control for the effect of several other variables like the number of household members and a quadratic term (to account for non-linearity), the dependency ratio (based on the number of persons less than age 15 versus those above (because in developing countries it is often the case that people work until they are fully incapacitated). Other controls include occupation, which is the main source of income for the household, categorised as cultivation, agricultural labour, non-agricultural labour and others; whether anyone in the household is literate, and whether the household is poor based on per capita consumption expenditure. The regression also controls several characteristics of the village where the household is a resident. It controls for the distance from the nearest town (as a measure of remoteness), whether villagers leave this village for a seasonal job (as a measure of the presence of a migrant network), and the proportion of villagers belonging to the same social group as the household head. The specification also controls for the quarter of the survey in either wave to account for seasonal effects.

Subsequently, to examine the heterogeneous pattern of the effect of climate variables on social groups we consider the following specification.

\begin{array}{l} Δ M i g r a t i o n_{i} = β_{0 j} + β_{1} Δ W e t n e s s_{j} + β_{2} Δ D r y n e s s_{j} + β_{3} S o c G r o u p_{i} + \\ \begin{matrix} \begin{matrix}  \end{matrix} \end{matrix} β_{4} Δ W e t n e s s_{j} ´ S o c G r o u p_{i} + β_{5} Δ D r y n e s s_{j} ´ S o c G r o u p_{i} + \\ β_{6} Δ S o c N e t_{i} + β_{7} N R E G S a d v_{i} + β_{8} Z_{i} + ϵ_{ij} \\ \begin{matrix}  \end{matrix} β_{0 j} = γ_{00} + u_{0 j}, \end{array}

(2)

Next, we examine the effect of two types of intervention variables, one that can be proactively carried out by the household—their social network (SocNet_i) measured by the quality of the household’s social network (SocNetQ_i) and household’s social network membership (SocNetM_i), and one which is a consequence of a public intervention—household’s residence in a village where public poverty alleviation is effective (NREGSadv_i). Here we introduce a three-way interaction of the intervention variable with the district-level climate variable and the social group variable along with relevant interactions at lower levels.

5. Regression Results

The regression on the effect of changes in climate variables on change in migrants among households (Appendix Table A1) shows that an increase of one more dry month, measured here using SPEI6, is associated with a 0.1 percentage point increased likelihood of migration. Increase in one more wet month does not have any significant effect on migration. Using the alternative climate variable, measured by change from being less than median months of dryness (months with less than –1 of SPEI6 values) in Wave I to more than median months in Wave II, is associated with a 3.2 percentage point increase in the likelihood of migration among households. An increase in the number of wet months measured in a similar manner does not show any significant effect on migration probability. The results also capture another important aspect of social group-wise variation in migration. Those belonging to scheduled groups are 2.6 percentage points less likely and those belonging to OBCs are 1.3 percentage points less likely to migrate compared to Hindus. The non-Hindus do not show any significant difference from the Hindus in their migration pattern between the two waves of the survey. For the rest of this article, we will refer to this increase in the number of months from less than median to more than median for both wet and dry months as excess wetness and dryness, respectively.

These migration patterns by different social groups might vary by their exposure to slow-onset climate change variables. So, next, we consider the interaction between social groups and the climate variables to capture the heterogeneity of the effect of climate variables on migration patterns (Appendix Table A2). Figure 4 shows the marginal effect of excess wetness and dryness over different social groups. The results show no significant effect of excess wetness on migration probabilities for any of the social groups. However, excess dryness is associated with a significant increase in migration probabilities of 4.9 percentage points among Hindus and OBCs. For the scheduled groups, the marginal effect of excess dryness between waves is not significantly different from zero. Given the lack of significant effect of excess wetness overall as well as by social groups, we will focus only on interactions with excess dryness in the rest of the article. Regression specifications will still control for excess wetness.

Figure 4.

Source: Computed by the author.

Next, we conduct regressions to examine the effect of social networks on migration probabilities (Appendix Table A3a and A3b).

Figure 5 shows the marginal effect of social network quality and social network membership on migration outcomes for different social groups contrasted between households from districts which experienced excess dryness to those that did not experience excess dryness. In either case, there is no significant effect of social networks—either when measured by quality or by membership, on migration probabilities for households belonging to different social groups.

Figure 5.

Source: Computed by the author.

Finally, we examine the effect of the NREGS programme on changes in migration (Appendix Table A4). Figure 6 shows the marginal effect of NREGS wage advantage on migration outcomes over different social groups contrasted between those from districts with excess dryness versus those that were not. Here we can see that the NREGS wage advantage shows a significant positive effect on migration among households belonging to scheduled groups in districts which experienced excess dryness compared to those who did not experience excess dryness. Households belonging to other social groups show no significant difference in migration probabilities in those circumstances.

Figure 6.

Source: Computed by the author.

6. Conclusion

The main purpose of this research is to examine how households belonging to disadvantaged social groups adapt their livelihood strategy in the face of slow-onset climate change in India. Using data from two waves of IHDS and corresponding environmental data from CRU, we find that slow-onset climate change, measured by changes in SPEI6 values, is significantly associated with an increase in migration probabilities among households in rural India. While the results highlight the importance of migration as an adaptation strategy for households, it also shows that it is not common among households from disadvantaged social groups facing slow-onset climate change. Even though households from socially disadvantaged groups show a lower probability of migration when exposed to slow-onset climate change, our results show a significantly positive contribution of a targeted poverty alleviation programme on their ability to migrate in those circumstances.

An important result in this study is the absence of any significant effect of social networks either to reduce or to increase migration among any social groups when exposed to excess dryness. This is contrary to the possible negative effect expected from climate-related income shocks as has been observed in earlier studies. On the other hand, social networks also do not show a positive effect on livelihood diversification through migration, assuming migration was limited by resources. Our results indicate that even the more privileged groups, who are more likely to have access to better quality social networks and groups, do not benefit from these networks when facing slow-onset climate change. One possible explanation may be that slow-onset climate change in itself might deplete the abilities of the household to invest in such a network by eroding social and cultural assets (De la Fuente, 2007). The inability to contribute to such a network can reproduce preexisting disadvantages. In addition, slow-onset climate change can potentially affect all those in the network (Carter & Maluccio, 2003; Dercon, 2005). The utility of social network-based risk sharing is highest when the income sources of these households are not similar. One way it can be achieved in the context of slow-onset climate change is through spatial diversification. However, distance often raises of cost of maintaining such links. Fafchamps and Gubert (2007) argue that geographic proximity is a major determinant of interpersonal relationships, as it facilitates the development of such relationships and also ensures monitoring and enforcement. Consequently, most networks are local and face similar risks. In addition to spatial diversification strategies, social networks can also be useful when they are created across groups whose income sources are different or uncorrelated from a climate perspective. However, it is rarely the case that kinship is developed across communal or occupational divide, and thus social networks play a limited role in insuring consumption (Carter & Maluccio, 2003).

A more important result is how public policy can improve livelihood diversification among those facing climate change. Our results show that targeted poverty alleviation programmes improve the probability of migration among disadvantaged social groups exposed to excess dryness. In a randomised controlled trial in Bangladesh, Bryan et al. (2014) assigned a $8.50 incentive to households in rural Bangladesh to temporarily out-migrate during the lean season. The incentive induced 22 per cent of households to send a seasonal migrant, their consumption at the origin increased significantly, and treated households were 8–10 percentage points more likely to re-migrate one and three years after the incentive was removed. To keep it in perspective, the average annual payment under NREGS in neighbouring India was approximately $100. Similar evidence can be seen in the case of migration of prime-aged adults following South Africa’s social pension programme (Ardington et al., 2009) or a means-tested pension scheme in China (Eggleston et al., 2018), where such transfers relaxed household credit constraints. There is similar evidence in Mexico where public transfers (Oportunidades) reduced financial constraints for migration (Angelucci, 2015). If disadvantaged social groups cannot take advantage of migration opportunities for risk aversion, such a poverty alleviation programme can help them choose migration as a diversification option. Evidence of NREGS reducing risk aversion can be seen in the case of farming too as there is evidence that farmers shifted to riskier and more profitable crops following the implementation of NREGS (Gehrke & Hartwig, 2018).

The main limitation of this study is how we define the migrant household. The non-resident instrument was created to account for the transfer of income to and from households and not to account for migration. Additionally, we use only two waves of the IHDS survey which may limit our ability to truly account for slow-onset climate change. Even with these limitations, our results provide valuable insights into how disadvantaged social groups negotiate slow-onset climate change.

When one thinks of policies towards climate change and its effects on the rural agricultural sector, the first thing that is often considered is improvement in irrigation or crop switching. However, Taraz (2017) notes that adaptation by choosing different crops or investment in irrigation recovers only 14 per cent of the profits lost due to harmful changes in India. Similarly, a recent paper by Fishman (2018) argues that sustainable use of irrigation water can mitigate less than a tenth of climate change impact in India. An alternative that is also often considered to address the effects of climate change on the rural population is reducing the dependence on the agricultural sector by encouraging the growth of non-farm sectors as productivity in this sector is higher and less sensitive to climate. However, recent studies indicate that because of their preexisting limitations, like low educational attainments as well as discrimination, such changes are bypassing disadvantaged social groups (Bera & Dubey, 2020; Himanshu et al., 2013). This once again highlights the importance of social protection in conceiving a policy on adaptation to climate change.

In conclusion, the effects of slow-onset climate change are beyond the point where steps to reverse climate change through conservation measures will be of any relief to large populations for whom the vagaries of slow-onset climate change have become a way of life. At this point, the urgency in terms of public policy is to enable those who are most affected by slow-onset climate change to adapt and overcome the adverse effects. Socio-economic disadvantages not only make them the most vulnerable, but they also make them least capable of adapting to such events.

This article highlights how targeted poverty alleviation programmes can still be an effective tool to enable the vulnerable population to weather the challenges of slow-onset climate change.

Footnotes

Declaration of Conflicting of Interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.

Appendix

Note: All regressions control for the number of household members and a quadratic term, dependency ratio, main occupation for the household, binary variable to indicate whether the household head is literate or poor, a control for both measures of social network and one for NREGS wage advantage. Other variables include distance from nearest town, whether villagers leave this village for seasonal job, proportion of villagers belonging to same social group as the household head. The specification also controls for the quarter of the survey in either wave.

Unless stated otherwise, all following regressions also control for the same set of variables.

References

Adger

W. N.

(2010). Climate change, human well-being and insecurity. New Political Economy, 15(2), 275–292.

Adger

W. N.

, & Adams

(2013). Migration as an adaptation strategy to environmental change. World Social Science Report, 261–264.

Aguinis

, Gottfredson

R. K.

, & Culpepper

S. A.

(2013). Best-practice recommendations for estimating cross-level interaction effects using multilevel modeling. Journal of Management, 39(6), 1490–1528.

Angelucci

(2015). Migration and financial constraints: Evidence from Mexico. Review of Economics and Statistics, 97(1), 224–228.

Ardington

, Case

, & Hosegood

(2009). Labor supply responses to large social transfers: Longitudinal evidence from South Africa. American Economic Journal: Applied Economics, 1(1), 22–48.

Banerjee

A. V.

, & Newman

A. F.

(1991). Risk-bearing and the theory of income distribution. The Review of Economic Studies, 58(2), 211–235.

Barnett

, & Webber

(2009). Accommodating migration to promote adaptation to climate change. Commission on Climate Change and Development. Policy Research working paper no. WPS 5270. World Bank Group. http://documents.worldbank.org/curated/en/765111468326385012/Accommodating-migration-to-promote-adaptation-to-climate-change

Beguería

, & Vicente-Serrano

S. M.

(2017). SPEI: Calculation of the standardised precipitation-evapotranspiration index. R Package Version, 1(7).

Bell

, Charles-Edwards

, Ueffing

, Stillwell

, Kupiszewski

, & Kupiszewska

(2015). Internal migration and development: Comparing migration intensities around the world. Population and Development Review, 41(1), 33–58.

10.

Béné

(2011). Social protection and climate change. IDS Bulletin, 42(6), 67–70.

11.

Bera

, & Dubey

(2020). Structure and changes in household income and employment across social groups in rural India. The Indian Journal of Labour Economics, 63, 407–435.

12.

Bhagat

R. B.

, & Keshri

(2020). Internal migration in India. Internal Migration in the Countries of Asia: A Cross-National Comparison, 207–228.

13.

Bhattacharya

P. C.

(2002). Rural- to- urban migration in LDCs: A test of two rival models. Journal of International Development, 14(7), 951–972.

14.

Black

, Adger

W. N.

, Arnell

N. W.

, Dercon

, Geddes

, & Thomas

(2011). The effect of environmental change on human migration. Global Environmental Change, 21, S3–S11.

15.

Black

, Arnell

N. W.

, Adger

W. N.

, Thomas

, & Geddes

(2013). Migration, immobility and displacement outcomes following extreme events. Environmental Science & Policy, 27, S32–S43.

16.

Bryan

, Chowdhury

, & Mobarak

A. M.

(2014). Underinvestment in a profitable technology: The case of seasonal migration in Bangladesh. Econometrica, 82(5), 1671–1748.

17.

Burgess

, Deschenes

, Donaldson

, & Greenstone

(2017). Weather, climate change and death in India. University of Chicago.

18.

Carleton

T. A.

(2017). Crop-damaging temperatures increase suicide rates in India. Proceedings of the National Academy of Sciences, 114(33), 8746–8751.

19.

Carter

M. R.

, & Maluccio

J. A.

(2003). Social capital and coping with economic shocks: An analysis of stunting of South African children. World Development, 31(7), 1147–1163.

20.

Cattaneo

, Beine

, Fröhlich

C. J.

, Kniveton

, Martinez-Zarzoso

, Mastrorillo

, Millock

, Piguet

, & Schraven

(2019). Human migration in the era of climate change. Review of Environmental Economics and Policy, 13(2), 189–206.

21.

Christensen

J. H.

, Hewitson

, Busuioc

, Chen

, Gao

, Held

, Jones

, Kolli

R. K.

, Kwon

W. T.

, & Laprise

(2007). Regional climate projections. In Fu

& Giorgi

(Eds.), Climate change, 2007: The physical science basis. Contribution of working group I to the fourth assessment report of the intergovernmental panel on climate change (pp. 847–940). Cambridge University Press.

22.

Coirolo

, Commins

, Haque

, & Pierce

(2013). Climate change and social protection in Bangladesh: Are existing programmes able to address the impacts of climate change? Development Policy Review, 31, o74–o90.

23.

Dallmann

, & Millock

(2017). Climate variability and inter-state migration in India. CESifo Economic Studies, 63(4), 560–594.

24.

Das

(2015). Can the rural employment guarantee scheme reduce rural out-migration: Evidence from West Bengal, India. The Journal of Development Studies, 51(6), 621–641.

25.

Datar

(2007). Failure of national rural employment guarantee scheme in Maharashtra. Economic and Political Weekly, 3454–3457.

26.

De la Fuente

(2007). Climate shocks and their impact on assets. Occasional Paper for UNDP.

27.

Defrance

, Delesalle

, & Gubert

(2023). Migration response to drought in Mali. An analysis using panel data on Malian localities over the 1987–2009 period. Environment and Development Economics, 28(2), 171–190.

28.

Deininger

, & Liu

(2013). Welfare and poverty impacts of India’s national rural employment guarantee scheme: Evidence from Andhra Pradesh (Vol. 1289). International Food Policy Research Institute.

29.

Dell

, Jones

B. F.

, & Olken

B. A.

(2014). What do we learn from the weather? The new climate-economy literature. Journal of Economic Literature, 52(3), 740–798.

30.

Dercon

(2005). Insurance against poverty. Oxford University Press.

31.

Desai

, Vanneman

, & National Council of Applied Economic Research, N. D. (2018). India Human Development Survey (IHDS), 2005.

32.

Desai

, Vanneman

, & National Council of Applied Economic Research, N. D. (2019). India Human Development Survey Panel (IHDS, IHDS-II), 2005, 2011–2012.

33.

Deshingar

, Akter

, Sharma

, & Farrington

(2010). The impacts of social protection on labour markets and migration with particular reference to the NREGA. Overseas Development Institute London.

34.

Deshingkar

, & Farrington

(2009). Circular migration and multilocational livelihood strategies in rural India. Oxford University Press.

35.

Deshingkar

, & Start

(2003). Seasonal migration for livelihoods in India: Coping, accumulation and exclusion (Vol. 111). Overseas Development Institute London.

36.

Drèze

, & Khera

(2011). Employment guarantee and the right to work. In Jayal

N. G.

& Mehta

(Eds.), The battle for employment guarantee (pp. 3–20). Oxford.

37.

Dutta

, Murgai

, Ravallion

, & Van de Walle

(2014). Does India’s employment guarantee scheme guarantee employment? Economic and Political Weekly, 47(16), 55–64.

38.

Eckstein

, Künzel

, Schäfer

, & Winges

(2019). Global climate risk index 2020. Who suffers most from extreme weather events? Weather-related loss events in 2018 and 1999 to 2018. Germanwatch.

39.

Eggleston

, Sun

, & Zhan

(2018). The impact of rural pensions in China on labor migration. The World Bank Economic Review, 32(1), 64–84.

40.

Fafchamps

, & Gubert

(2007). The formation of risk-sharing networks. Journal of Development Economics, 83(2), 326–350.

41.

Feng

, Krueger

A. B.

, & Oppenheimer

(2010). Linkages among climate change, crop yields and Mexico–US cross-border migration. Proceedings of the National Academy of Sciences, 107(32), 14257–14262.

42.

Field

C. B.

, & Barros

V. R.

(2014). Climate change 2014–Impacts, adaptation and vulnerability: Regional aspects. Cambridge University Press.

43.

Fishman

(2018). Groundwater depletion limits the scope for adaptation to increased rainfall variability in India. Climatic change, 147(1), 195–209.

44.

Fussell

, Hunter

L. M.

, & Gray

C. L.

(2014). Measuring the environmental dimensions of human migration: The demographer’s toolkit. Global Environmental Change, 28, 182–191.

45.

Garg

, Jagnani

, & Taraz

(2020). Temperature and human capital in India. Journal of Environmental and Resource Economists, 7(6), 1113–1150.

46.

Gehrke

, & Hartwig

(2018). Productive effects of public works programs: What do we know? What should we know? World Development, 107, 111–124.

47.

Gelman

, & Hill

(2006). Data analysis using regression and multilevel/hierarchical models. Cambridge University Press.

48.

Gough

(2004). Welfare regimes in development context: A global and regional analysis. In Gough

, Wood

, Barrientos

, Bevan

, Room

, & Davis

(Eds.), Insecurity and welfare regimes in Asia, Africa and Latin America: Social policy in development contexts. Cambridge University Press.

49.

Government of India. (2015). All India report on agriculture census 2010–11. New Delhi: Ministry of Agriculture and Farmer’s Welfare.

50.

Gray

, & Bilsborrow

(2013). Environmental influences on human migration in rural Ecuador. Demography, 50(4), 1217–1241.

51.

Gray

, & Mueller

(2012a). Drought and population mobility in rural Ethiopia. World Development, 40(1), 134–145.

52.

Gray

, & Mueller

(2012b). Natural disasters and population mobility in Bangladesh. Proceedings of the National Academy of Sciences, 109(16), 6000–6005.

53.

Guiteras

(2008). The impact of climate change on Indian agriculture. University of Maryland.

54.

Hagen-Zanker

, & Himmelstine

C. L.

(2013). What do we know about the impact of social protection programmes on the decision to migrate? Migration and Development, 2(1), 117–131.

55.

Harrington

L. J.

, Frame

D. J.

, Fischer

E. M.

, Hawkins

, Joshi

, & Jones

C. D.

(2016). Poorest countries experience earlier anthropogenic emergence of daily temperature extremes. Environmental Research Letters, 11(5), 055007.

56.

Harris

, Osborn

T. J.

, Jones

, & Lister

(2020). Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Scientific Data, 7(1), 1–18.

57.

Himanshu, Lanjouw

, Murgai

, & Stern

(2013). Nonfarm diversification, poverty, economic mobility, and income inequality: A case study in village India. Agricultural Economics, 44(4–5), 461–473.

58.

Hnatkovska

, & Lahiri

(2015). Rural and urban migrants in India: 1983–2008. The World Bank Economic Review, 29(suppl_1), S257–S270.

59.

Horrace

W. C.

, & Oaxaca

R. L.

(2006). Results on the bias and inconsistency of ordinary least squares for the linear probability model. Economics Letters, 90(3), 321–327.

60.

Hugo

(1996). Environmental concerns and international migration. International Migration Review, 30(1), 105–131.

61.

Imbert

, & Papp

(2015). Labor market effects of social programs: Evidence from India’s employment guarantee. American Economic Journal: Applied Economics, 7(2), 233–263.

62.

Islam

S. N.

, & Winkel

(2017). Climate change and social inequality. United Nations, Department of Economics and Social Affairs.

63.

Iversen

, Kalwij

, Verschoor

, & Dubey

(2014). Caste dominance and economic performance in rural India. Economic Development and Cultural Change, 62(3), 423–457.

64.

Jessoe

, Manning

D. T.

, & Taylor

J. E.

(2018). Climate change and labour allocation in rural Mexico: Evidence from annual fluctuations in weather. The Economic Journal, 128(608), 230–261.

65.

Kanbur

S. M.

(1979). Of risk taking and the personal distribution of income. Journal of Political Economy, 87(4), 769–797.

66.

Kartiki

(2011). Climate change and migration: A case study from rural Bangladesh. Gender & Development, 19(1), 23–38.

67.

Keshri

, & Bhagat

R. B.

(2010). Temporary and seasonal migration in India. Genus, 66(3), 25–45.

68.

Khan

A. U.

, & Saluja

(2007). Impact of the NREGP on rural livelihoods. New Delhi: India Development Foundation.

69.

Khera

(2011). The battle for employment guarantee. Oxford University Press India.

70.

Kone

Z. L.

, Liu

M. Y.

, Mattoo

, Ozden

, & Sharma

(2018). Internal borders and migration in India. Journal of Economic Geography, 18(4), 729–759.

71.

Kuriakose

A. T.

, Heltberg

, Wiseman

, Costella

, Cipryk

, & Cornelius

(2013). Climate‐responsive social protection. Development Policy Review, 31, o19–o34.

72.

Kusters

, & Wangdi

(2013). The costs of adaptation: Changes in water availability and farmers’ responses in Punakha district, Bhutan. International Journal of Global Warming, 5(4), 387–399.

73.

Liu

, & Barrett

C. B.

(2013). Heterogeneous pro-poor targeting in the national rural employment guarantee scheme. Economic and Political Weekly, 46–53.

74.

Masson-Delmotte

, Zhai

, Pörtner

H.-O.

, Roberts

, Skea

, Shukla

P. R.

, Pirani

, Moufouma-Okia

, Péan

, Pidcock

, Connors

, Matthews

J. B. R.

, Chen

, Zhou

, Gomis

M. I.

, Lonnoy

, Maycock

, Tignor

, & Waterfield

(2018). Global warming of 1.5°C. An IPCC Special Report on the Impacts of Global Warming of, 1, 1–9.

75.

Mathieu

J. E.

, Aguinis

, Culpepper

S. A.

, & Chen

(2012). Understanding and estimating the power to detect cross-level interaction effects in multilevel modeling. Journal of Applied Psychology, 97(5), 951–966.

76.

McLeman

, & Smit

(2006). Migration as an adaptation to climate change. Climatic Change, 76(1–2), 31–53.

77.

Millock

(2015). Migration and environment. Annual Review of Resource Economics, 7(1), 35–60.

78.

Mohanty

(2020). Preparing India for extreme climate events: Mapping hotspots and response mechanisms. Council on Energy, Environment and Water.

79.

Mohanty

B. B.

(2001). Land distribution among scheduled castes and tribes. Economic and Political Weekly, 36(40), 3857–3868.

80.

Mosse

(2010). A relational approach to durable poverty, inequality and power. The Journal of Development Studies, 46(7), 1156–1178.

81.

Munshi

, & Rosenzweig

(2016). Networks and misallocation: Insurance, migration, and the rural-urban wage gap. American Economic Review, 106(1), 46–98.

82.

Nayyar

, & Kim

K. Y.

(2018). India’s internal labor migration paradox: The statistical and the real. World Bank Policy Research Working Paper (8356).

83.

Nielsen

J. Ø.

, & Reenberg

(2012). Exploring causal relations: The societal effects of climate change. Geografisk Tidsskrift-Danish Journal of Geography, 112(2), 89–92.

84.

Novotný

, Kubelková

, & Joseph

(2013). A multi-dimensional analysis of the impacts of the Mahatma Gandhi national rural employment guarantee scheme: A tale from Tamil Nadu. Singapore Journal of Tropical Geography, 34(3), 322–341.

85.

Opondo

D. O.

(2013). Erosive coping after the 2011 floods in Kenya. International Journal of Global Warming, 5(4), 452–466.

86.

Rabbani

, Rahman

, & Mainuddin

(2013). Salinity-induced loss and damage to farming households in coastal Bangladesh. International Journal of Global Warming, 5(4), 400–415.

87.

Ravi

, & Engler

(2015). Workfare as an effective way to fight poverty: The case of India’s NREGS. World Development, 67, 57–71.

88.

Renaud

F. G.

, Bogardi

J. J.

, Dun

, & Warner

(2007). Control, adapt or flee: How to face environmental migration? UNU-EHS.

89.

Rosenzweig

M. R.

, & Stark

(1989). Consumption smoothing, migration, and marriage: Evidence from rural India. Journal of Political Economy, 97(4), 905–926.

90.

Sabates-Wheeler

, & Waite

(2003). Migration and social protection: A concept paper. Institute of Development Studies, Sussex, December, 1645, 1980–2000.

91.

Schipper

E. L. F.

(2020). Maladaptation: When adaptation to climate change goes very wrong. One Earth, 3(4), 409–414.

92.

Sedova

, & Kalkuhl

(2020). Who are the climate migrants and where do they go? Evidence from rural India. World Development, 129, 104848.

93.

Skoufias

, Rabassa

, & Olivieri

(2011). The poverty impacts of climate change: A review of the evidence. World Bank Policy Research Working Paper (5622).

94.

Solinski

(2012). NREGA and labour migration in India: Is village life what the ‘rural’ poor want? The South Asianist Journal, 1(1), 17–30.

95.

Stahlberg

(2012). India’s latest and largest workfare program: Evaluation and recommendations. Center on Democracy, Development, and the Rule of Law, Stanford.

96.

Stark

, & Bloom

D. E.

(1985). The new economics of labor migration. The American Economic Review, 75(2), 173–178.

97.

Taraz

(2017). Adaptation to climate change: Historical evidence from the Indian monsoon. Environment and Development Economics, 22(5), 517–545.

98.

Taraz

(2018). Can farmers adapt to higher temperatures? Evidence from India. World Development, 112, 205–219.

99.

Thorat

(2010). Ethnicity, caste and religion: Implications for poverty outcomes. Economic and Political Weekly, 47–53.

100.

Thorat

, & Joshi

(2020). The continuing practice of untouchability in India. Economic & Political Weekly, 55(2), 36–45.

101.

Topalova

(2010). Factor immobility and regional impacts of trade liberalization: Evidence on poverty from India. American Economic Journal: Applied Economics, 2(4), 1–41.

102.

Townsend

R. M.

(1994). Risk and insurance in village India. Econometrica: Journal of the Econometric Society, 62(3), 539–591.

103.

Vicente-Serrano

S. M.

, Beguería

, & López-Moreno

J. I.

(2010). A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index. Journal of Climate, 23(7), 1696–1718.

104.

Viswanathan

, & Kumar

K. K.

(2015). Weather, agriculture and rural migration: Evidence from state and district level migration in India. Environment and Development Economics, 20(4), 469–492.

105.

Warner

(2010). Global environmental change and migration: Governance challenges. Global Environmental Change, 20(3), 402–413.

106.

Warner

, & Van der Geest

(2013). Loss and damage from climate change: Local-level evidence from nine vulnerable countries. International Journal of Global Warming, 5(4), 367–386.

107.

Weldegebriel

Z. B.

, & Prowse

(2013). Climate-change adaptation in Ethiopia: To what extent does social protection influence livelihood diversification? Development Policy Review, 31, o35–o56.

108.

Wilhite

D. A.

, & Glantz

M. H.

(1985). Understanding: the drought phenomenon: The role of definitions. Water International, 10(3), 111–120.

109.

Wolf

(2011). Climate change adaptation as a social process. In Ford

J. D.

& Berrang-Ford

(Eds.), Climate change adaptation in developed nations: From theory to practice (pp. 21–32). Springer Netherlands.

110.

Yaffa

(2013). Coping measures not enough to avoid loss and damage from drought in the North Bank region of the Gambia. International Journal of Global Warming, 5(4), 467–482.