Population and Environment

, Volume 32, Issue 2, pp 238–262

Migration and climate change: examining thresholds of change to guide effective adaptation decision-making


    • School of Social SciencesThe University of Adelaide
  • Graeme J. Hugo
    • School of Social SciencesThe University of Adelaide
Original Paper

DOI: 10.1007/s11111-010-0126-9

Cite this article as:
Bardsley, D.K. & Hugo, G.J. Popul Environ (2010) 32: 238. doi:10.1007/s11111-010-0126-9


The implications of environmental change for migration are little understood. Migration as a response to climate change could be seen as a failure of in situ adaptation methods, or migration could be alternatively perceived as a rational component of creative adaptation to environmental risk. This paper frames migration as part of an adaptation response to climate change impacts to natural resource condition and environmental hazards. Thresholds will be reached by communities after which migration will become a vital component of an effective adaptation response. Such changes to migration patterns have the potential to undermine migration policy unless appropriate preparations are undertaken. This paper describes an approach to assist researchers to frame how climate change will influence migration by critically analysing how thresholds of fundamental change to migration patterns could be identified, primarily in relation to two case studies in Nepal and Thailand. Future policy for internal and international migration could be guided by the analysis of such thresholds of non-linear migration and resourced effectively to ensure that socio-economic and humanitarian outcomes are maximised.


MigrationClimate changeAdaptationThresholdsEnvironmental riskNepalThailand


There are significant challenges emerging for the globe in relation to both future human mobility and the broader impacts of climate change. The interaction between these two processes is little understood and significantly under-researched with regard to its global importance, as the potential for fundamental changes in the migration of people both within countries and across international borders could be enormous (Reuveny 2007; Renaud et al. 2007; de Sherbinin et al. 2008; Hugo 2008a; Piguet 2008). In particular, the debate on climate change migration has been dominated by a perception that if individuals move as a result of climate change, somehow they would have failed to adapt effectively (Adamo 2008; Heine and Petersen 2008). We argue that this debate needs to be re-framed because migration as a result of perceived or experienced climate change impacts will, in many cases, be a very effective adaptation response in the light of the experienced or perceived future impacts of climate change (McLeman and Smit 2006; Brown 2008; Tacoli 2009). As Boano et al. (2007) and Kniveton et al. (2008) note, migration is often less a function of immediate stress resulting from the onset of a natural disaster than a proactive diversification strategy taken in anticipation of such events in the future, or to cope with long-term declines in livelihood. In the important example of Bangladesh, where environmental hazards and increasing landlessness are seen to be major drivers of migration decision-making, an estimated 20 million people have migrated internationally to seek better opportunities in India, the Middle East, South-east Asia and elsewhere (Siddiqui 2005; Nasreen et al. 2006; Reuveny 2007). If climate change migration could be governed and managed effectively, humanitarian crises will be minimised, conflicts avoided, and opportunities provided for countries to benefit from the processes of environmental and social change—the social limits to adaptation would be considerably expanded (Barnett 2003; Adger et al. 2009). Here, effective migration policy would be seen to facilitate the mobility of people when required to enhance their well-being and, where possible, maximise social and economic development in the places of both origin and destination. However, the development of such an effective response would require a re-conceptualisation of the role of mobility in relation to environmental risk.

The paper initially reviews the manner in which the debate on climate change migration and the policy response to environmental risk has been framed. We critically examine conceptualisations of experienced migration in relation to environmental change. We move on to examine an approach that could assist researchers to frame how climate change will influence migration by critically analysing how thresholds of fundamental change to migration patterns could be identified. Future migration policy risks are presented in relation to climate change impacts in two Asian nations, namely Nepal, a relatively poor country of net emigration, and Thailand, a relatively wealthy, net-immigration country, to outline how future climate change–driven human mobility can be conceptualised by applying the concepts of linearity and non-linearity in human migration to scenarios of potential environmental risk. It is not the aim of this work to present new information on climate change risk and migration for Nepal and Thailand. Rather, we examine possible linear and non-linear migratory responses of large human populations to key climate change vulnerabilities according to future environmental scenarios for the two countries and go on to describe how such an approach can provide vital insights into important potential risks to migration policy. We argue that by identifying and articulating such climate change vulnerabilities to current national and internal migration policy, changes could be made in a timely manner to minimise suffering and maximise potential benefits.

In different contexts, the debate on climate change migration is being lead by environmental scientists, the humanitarian community or demographers with extensive experience in migration studies. Not surprisingly, these different groups are reviewing and articulating the issue through the lens of their areas of expertise and concern, and as a result, holistic reviews of the implications of projected climate change on human mobility are rare. It could be argued that the dominant perceptions of the role of environment on migration are skewed by a broader discounting of the role that the environment has on societal development and the fact that attribution of decision migration to issues of environmental change is not immediate or easy (Brown 1995; O’Hear 1997; Bardsley 2003). For that reason, governments and other institutions involved in responses to environmental hazards or change readily discount the role of the environment on society. Partly as a result, the dominant experience of decision-making in relation to future environmental risk has largely been associated with reaction to management neglect or failure. Moreover, the decision makers do not wish to be seen to be speculating about the future—they do not want to be seen to not respond enough and fail to adapt effectively; to be overly cautious in regard to an evident risk and expend too many resources and over-adapt; or make incorrect decisions and mal-adapt. The result is a focus on the socio-economic drivers of migration decision-making and a tendency to discount the importance of the environment on socio-economic well-being and consequently human mobility, even though the environment is the ultimate cause of much socio-economic opportunity or deprivation, especially for the rural poor (de Sherbinin et al. 2008).

Migration governance is already a highly contested area of policy in Asia-Pacific, the focus region of this article, and the management systems in regard to international migration are quite poor (Hugo 2008a; Warner et al. 2009). Yet decisions need to be made, especially as both research on “dangerous climate change” and experienced impacts of change provide increasing evidence of the risks of governance failure (Dessai et al. 2004; Adger et al. 2005). Already, such failures are manifest in policies in developed countries, which struggle to respond to the socio-ecological implications of increasing risks of hazards or environmental change, such as in the examples of Hurricane Katrina in New Orleans (Elliott and Pais 2006), or in relation to the prolonged drought in south-east Australia (Fritze et al. 2008). To support better decision-making so that policy is able to effectively manage both short- and long-term adaptation, researchers need to provide the best information possible and make it available in appropriate forms to as many people as possible. So, what is the best information to inform the governance of climate change migration?

Framing climate change migration

Mobility in relation to dominant environmental drivers is complex. Changes to existing migration patterns in the Asia-Pacific as a result of climate change will take many forms, but would principally involve an increase in the “push” forces that lead to out-migration and emigration (Hugo et al. 2009). People move temporarily or permanently to avoid and recover from extreme climatic events; they move internally within countries on a temporary or a more permanent basis to establish new livelihood opportunities, and they move temporarily or permanently to other countries to meet similar goals. It is possible to frame the discussion by arguing that there are three main processes through which climate change impacts are likely to manifest themselves on migration patterns:
  • increasing experiences of risk of environmental hazards and associated socio-ecological events;

  • changing resource condition trends through time that alter access and effective utilisation of natural resources; and

  • perceptions of risk of impacts of climate change, irrespective of real experiences.

Much migration occurs along social networks established by earlier generations of migrants (Massey et al. 1993). The majority of movement of people due to climate change impacts, even in extreme cases, will likely follow or reinforce such established channels of movement or at least depend upon established networks and relationships (Hugo 2010). That suggests that just as most past environmentally induced migration has involved internal movement (Hugo 1996), the majority of climate change–related mobility will also be within countries. Although the anticipation of such changes is difficult, it is becoming increasingly clear that environmental pressures resulting from climate change will be so fundamental to societal structures that they will also drive the development of new corridors and new scales of migration. Local adaptation policies could in themselves lead to displaced populations and lead to new types of migration. Furthermore, it is apparent that climate change will impinge upon some important destination areas and hence act to reduce the ‘pull’ of potential destinations.

Environmental change has been, and will increasingly be, an important driver of migration. Undoubtedly, one of the barriers to increasing our understanding of the relationship between environmental change and migration has been what Castles (2002) has referred to as the “conceptual fuzziness” regarding environmentally induced migration. Much of this fuzziness arises from the fact that the environment has received relatively little interest from demographic studies or monitoring programmes (Huguet 2008; Piguet 2008), and when it has been examined, it is clear that the environment has rarely been the only cause of migration. In situ adaptations, defined here as climate change adaptation that is undertaken within or for a place rather than involving the movement of people away from that place, are likely to dominate at least in the early stages of environmental change. However, human migration has also been a vital component of adaptation to changes in natural resource condition or risk of environmental hazard historically, and this will not change in the future (McLeman and Smit 2006). As Fig. 1 suggests, the extent to which environmental change is a cause of migration varies from not being significant at all to being the overwhelming cause of migration, such as when an area is inundated and people are forced to move. Environment has been identified as a mostly indirect or ultimate cause of migration (Hugo 1996). This means that distinguishing “environmental migrants” from others is actually difficult, except at the extreme end of the continuum in Fig. 1. It is at this forced end of the continuum that a designation of migrants as “environmental migrants” has been most applicable, but that largely associates migration only with reactive adaptation to climate change.
Fig. 1

Environment as a cause of migration 2

One of the key distinctions implicit in the relationship between environmental change and migration is between mobility as a strategy for adapting in an anticipatory manner to the impacts and as a displacement when environmental deterioration becomes so extreme that people are forced to leave an area as a reactive adaptation (Fankhauser et al. 1999; Renaud et al. 2007). While the “forced displacement” component of a migration response to impacts such as rising sea level is likely to be significant, it is the use of migration as a precautionary adaptation to climate change based on increased perceptions of risk, which may be increasingly important in the foreseeable future and needs to be recognised and supported (Dessai et al. 2004). In any case, it is highly likely that the anticipated influence of climate change, or the perception of its likely future impact, in most cases will not be the sole driver of migration but mixed with a range of other motivations (Castles 2002; Black et al. 2008). It is essential then that environmentally induced population mobility can be seen as:
  • a wide array of mobility types and not just displacement;

  • only one of the responses available among an array of potential mitigation and adaptation strategies; and,

  • only one influential component of complex decisions whether to migrate.

A relatively simple spectrum is being presented of estimates of future climate change migrants from the alarmist (Myers 2002) to the contended (Black et al. 2008; Castles 2002, 2006), which vary by an order of magnitude from hundreds of millions of people by the middle of the twenty-first century to tens of millions (Renaud et al. 2007; Adamo 2008; Perch-Nielsen et al. 2008). However, most agree that projected climate change will influence the scale and scope of human migration, and whether the change can be managed effectively both internally within countries and internationally will depend on the timing, scale and scope of the movement and our preparations (Barnett and Webber 2009). All projections of future numbers of people that will migrate due to future climate change are likely to be highly inaccurate (Boano et al. 2007). The key point is that climate change will influence migration even if the scale, location and severity of the influence cannot be accurately established. Hence, it is imperative to develop approaches that provide the best information possible to inform the governance of migration in the future.

We argue that this raises a fundamental question for researchers: Will migration be influenced by climate change in a linear manner or will there be thresholds or tipping points where fundamental changes to migration levels and patterns result (Lonergan and Swain 1999; Jones 2001; Warner et al. 2008)? Granovetter (1978, p. 1422) argues that different people will make different decisions in response to stimuli and that a “threshold is simply that point where the perceived benefits to an individual of doing the thing in question exceed the perceived costs.” The concept has also been widely applied in relation to natural systems (Semenov and Porter 1995; Arnell 2000) and in fact climate change itself, with a climatic tipping point reached when “the climate system is forced to cross some threshold and it subsequently changes dramatically beyond that which would be caused by the initial forcing” (Kniveton et al. 2008, p. 24). Here, we use threshold in a slightly different manner to either of these definitions to identify a situation where a significant change in collective social behaviour results. In this case, a threshold of non-linear migration would be defined generally as a point at which the impacts of climate change are so severe or so frequent that the resilience of socio-ecological systems is breached, or that existing in situ adaptation options either fail or are perceived as inadequate, so that people make use of migration as an adaptation option in a manner that will fundamentally alter the form migration is taking. Resilience here would be defined as “The ability of a social or ecological system to absorb disturbances while retaining the same basic structure and ways of functioning, the capacity for self-organisation, and the capacity to adapt to stress and change” (Parry et al. 2007, p. 880). Adaptation would also follow the Intergovernmental Panel on Climate Change definition, which is “Adjustment in natural or human systems in response to actual or expected climatic stimuli or their effects, which moderates harm or exploits beneficial opportunities” (Parry et al. 2007, p. 869). While in situ adaptation refers to adaptation “in place”, it contrasts with ex situ adaptation, of which human mobility is a component. Thus, ex situ adaptation means the adaptation to climate change impacts that involves the movement of people, systems and/or assets from a place of vulnerability.

Environmental change often works both indirectly on migration patterns via an economic process of natural resource exploitation and directly through perceptions and experiences of human security. Meze-Hausken (2008, p. 3000) states, in relation to human adaptation to environmental change, that “the driver of the response is the evolution or variation in socio-economic conditions that determine whether a specified climate change is experienced as acceptable or problematic.” As a society’s collective experience or collective perception of natural resource condition and environmental hazards is altered by a changing climate, fundamental changes in migration levels or patterns will also result in many places. Much of that change will occur as linear changes to migration, which we define here as changes that follow established patterns or a relatively minor new path of human mobility. Of course, such a definition is highly contextual, and specific local definitions would need to incorporate the complexity of migration itself, including, but not exclusively, numbers of people moving; drivers of out-migration; broader socio-economic conditions; distance or permanency of travel; destination; ethnicity and gender; and money flows between sending and receiving areas. The important point is that linear migration will not “surprise” policy or rapidly undermine institutional capacities to deal with change. In contrast, a non-linear change in migration would involve new migration patterns generally involving large numbers of people and/or changes to current numbers following established patterns that are of an order of magnitude greater in relation to current population movements.

A vital research question emerges from this discussion of uncertainty of the change in socio-ecological condition: Will it be possible to identify points at which the impacts of climate change are so severe or so regular that the resilience of socio-ecological systems is breached, or in situ adaptation options fail and people make use of migration as either a reactive or anticipatory adaptation option (see Fankhauser et al. 1999) in a manner that will fundamentally alter the form migration is taking?

Too often, it is assumed that all environmental impacts will result in displacement migration. For most voluntary “environmentally motivated” migrants, the change to their socio-ecological systems is unlikely to be critical before it influences their decision to move. Thus, the threshold level of environmental change that will influence migration patterns may be relatively low and less attributable to particular events. Rather, voluntary out-migration results in part from a perceived reduction in the value of remaining within a place. Again, the types of issues here are highly contextual and will involve a reduction in perceived lifestyle amenity, a reduction in earning capacity or an increased perception of a risk of hazard (Kniveton et al. 2008). Such changes are projected to eventuate from climate change across the Asia-Pacific, but are not likely to fundamentally alter existing patterns of migration in the short term, except to increase the scale of movement (Hugo et al. 2009). That said, the component of the decision to move, the permanency of that move and the destination of a possible movement are all likely to be more greatly influenced in the future by environmental deterioration resulting from climate change.

What might a threshold of non-linear migration involve? The thresholds at which involuntary or forced migration, or in fact displacement, occurs are likely to lead to significant re-ordering or re-scaling of migration patterns and will differ across migration types and for different socio-cultural and geographical contexts. Some gradual environmental change is likely to lead to experiences where a lack of long-term planning and ineffective management measures will weaken the in situ adaptive capacity of affected groups. This could, in turn, result in a shift over time from linear changes to non-linear changes in migration patterns. A significant increase in the recurrence or impact of hazards, or the perceptions of those hazards, could also lead to a major increase in the number of people migrating from affected or potentially affected areas. Such non-linear responses will primarily be associated with forced migration due to major events or tipping points being reached in resource condition, each of which have the possibility to result in major, negative humanitarian impacts. Much of this forced migration will contain at least an element of economic, socio-cultural or political forcing, and thus, it will still remain difficult to classify a mover definitively as a “climate change–forced migrant”. While the thresholds or tipping points are always likely to be highly contextual, an attempt needs to be made to identify the level of tolerance of a society to environmental change, because key risks will emerge to societies in general and migration policy in particular, when the non-linear thresholds of migration are met.

One example of a group that might be perceived as relatively easy to categorise as climate change migrants would be defined as “climate change–displaced people”—as a subset at the extreme end of ‘environmentally migrant’ (see IOM 2007). “Climate change–displaced people” could thus be defined as those people who are forced to leave their place because of the complete and irreversible loss of their habitable terrestrial space or source of livelihoods resulting from climate change. People who apply ex situ adaptation approaches would primarily be those who lose their island or other coastal or riverine terrestrial spaces due to changing environmental conditions, namely permanent immersion, such as through sea level rise, or erosion, through major events including floods, storm surges or landslides. There will also be some groups that effectively lose their terrestrial space through severe desertification, lack of access to fresh water, or displacement due to pollution, development or even climate change adaptation policies, such as the permanent exclusion resulting from the establishment of dams or managed coastal retreat policies. The population affected by these groups, even given the extreme definition of “climate change displacement”, could be quite significant in coastal and riverine environments of Asia and the Pacific, most importantly in the low-lying densely populated delta regions (Hugo et al. 2009).

Yet, given the above definition, the component of the ‘environmentally displaced’ resulting from climate change, and not natural climate variability or other forms of environmental change, will be problematical to differentiate, even if environmental change plays a significant role in their displacement (Castles 2006; Perch-Nielsen et al. 2008). Coastal areas and riverine islands in the Asia-Pacific have changed historically and will continue to change their size and form irrespective of environmental change brought on by climate change (Mimura 2008). The component of such erosive or inundating events that could be attributed specifically to climate change would be difficult to determine and, in fact, may be so contestable that it is likely to remain an academic interest for some time, without policy or legal implications. There has been a failure, moreover, to create and maintain systems that lead to environmentally sustainable management of coastal settlements on many islands and on continental landmasses (Harvey 2006). Thus, inappropriate development of coastal atolls or urban centres in a lower elevation coastal zone must be seen to have a major influence on the future displacement of people resident in these areas. Urban development on marshes, islands, delta areas or sand dune systems that need protection and constant soil or sand replenishment to remain at the same elevation are widespread in the Asia-Pacific. Similarly, the over-withdrawal from freshwater aquifers; the disturbance of sand drift patterns; the mining of sediments; the destabilisation of natural protection such as reefs, dunes or vegetation along coasts or in oceans; the changing sedimentary loads from rivers; the changing erosive capacities of rivers; and the development of engineered structures can and do increase the vulnerability of coastal systems to permanent inundation. In contexts of high incidence of poverty and pressure on land resources in both rural and urban areas, people settle in marginal and vulnerable coastal locations because they have no other choice (Hugo 1996). It becomes very difficult to differentiate the influence of climate change on the condition of the resource and consequently the influence on ‘climate change displacement’ of populations, even in the relatively extreme and simplistic example of total loss of terrestrial space due to sea level rise. Another level of complexity emerges when it is considered that many people can, and do, live permanently over water bodies on the borders of coasts, rivers or lakes, and so the definition of “habitable terrestrial space” will also vary considerably.

In contrast to the longer term changes in resource condition discussed above, sudden disasters can be very destructive and cause major displacement of population, but that displacement is usually temporary. For example, people displaced by disasters like the Asian tsunami of 2004 (Laczko and Collett 2005) mostly return and rebuild their communities. Indeed, the influx of resources and the magnitude of the task of rebuilding can lead to an in-migration of workers. Where the deterioration in the environment is more gradual, the migration response is more complex. A common coping mechanism to a decline in local productive resources in rural Asia and the Pacific is for families to deploy some of their labour to other labour markets, especially in cities, but also increasingly internationally. The remittances which they send back can help support the remaining family members and assist with in situ adaptation (Adger et al. 2002). There also may be some permanent out-migration, perhaps preceded by a preliminary temporary migration, especially if the period of disruption in the place of origin is prolonged. This out-migration can have the effect of reducing the pressure on resources and services locally, but also, through remittances sent back, can reduce the vulnerability of those left behind. In both the temporary and permanent migration strategies, it is usually the young adult population that is selectively involved.

From this discussion, we have identified four important distinctions that need to be recognised in considering environmental (including climate change) impacts on migration:
  • dramatic sudden impacts vs slow onset environmental change impacts;

  • in situ versus ex situ adaptation responses to environmental risk;

  • moving as a result of perceived threats vs actual deterioration of the environment; and

  • ex situ adaptation as forced displacement vs a voluntary movement.

Too often, the relationship between environmental change and migration is depicted as a simple cause and effect, but the analysis here contends that the relationship is actually far more complex, and researchers must respond to this complexity. We have gone into some detail of the importance of definition and context, because it is necessary to tease out the complexities of migration issues before the uncertainties of the inter-related spectrums of climate change influencing environmental conditions that lead to impacts on human populations and their migration patterns can be understood. Moreover, a conceptual framework for analysing the potential impact of climate change on migration emerges from the discussion. Of key interest will be those impacts of climate change, or perceptions of likely future impacts of climate change, which either:
  • Increase the numbers of people migrating using established patterns both internally and externally in a linear manner, primarily via voluntary mechanisms or

  • Lead to non-linear changes to migration that result as thresholds of resilience or tipping points being met, primarily via involuntary mechanisms.

In the latter case, tipping points may lead to new migration patterns and/or changes to current numbers following established patterns that are greater than current population movements. These people will primarily be forced migrants, with perhaps a significant number being environmentally displaced by the effective loss of terrestrial space or livelihoods. For the remainder of this paper, we will use the concepts of linear and non-linear responses to climate change impacts to frame a discussion of the risks to the governance of migration in the Asia-Pacific region.

Examples of potential linear and non-linear change to Asia-Pacific migration

Over the last five decades, there have been massive changes in the form and scale of human mobility in the Asia-Pacific (Hugo 2006; Skeldon 2006; Jones and Douglass 2008). While economic drivers of mobility have been dominant, there is a growing appreciation that environmental change plays a significant role. Projections of future climate change across the region suggest that environmental drivers of migration will become an increasing proximate or contributory factor to human mobility throughout the twenty-first century. A number of factors suggest that climate change impacts on migration in the Asia-Pacific could be particularly important, and these will be outlined at some length below, but briefly:
  • the demographic pressures and environmental management issues in the region are already enormous;

  • climate change impacts are projected to be large in the region, particularly as monsoonal rainfall patterns could become less reliable and cyclonic activity intensify; and

  • a large percentage of the population are poor, and their well-being is highly vulnerable to environmental hazards and degradation of natural resources.

The population of the Asia-Pacific was over 4.1 billion in 2008, representing considerably more than half of the global population (UNESCAP 2008). Thus, what affects societies and the associated mobility of populations in the region will have significant implications for the Earth.

Climate change is projected to have a significant impact on the environments and societies of the Asia-Pacific region. It is not the intention here to repeat the extensive work reviewing the likely impacts of climate change on the region by the IPCC and others (Dilley et al. 2005; Preston et al. 2006; Cruz et al. 2007; Solomon et al. 2007; Levy et al. 2008). However, one way of conceptualising the impacts of climate change is to divide the likely impacts into the form the change will take:
  1. 1.

    Primary impacts, which are resulting immediately from changes to the climatic patterns themselves.

  2. 2.

    Secondary impacts, which are changes to environmental systems resulting from the primary impacts of climate change.

  3. 3.

    Tertiary impacts include the broader impacts on societal systems, including the implications for migration patterns.


The vulnerability of people to the impacts will be associated with the site and situation of where they live, as well as their capacity to manage the threat of or the experience of climate change impacts. For example, Asia contains some of the most densely settled areas in the world such as Java in Indonesia, the Mekong delta, or the Chao Phraya valley in Thailand, the deltas of Southern Asia, coastal China and river valleys of Pakistan, India and Bangladesh—some of the areas most exposed to future climate change impacts (Jones and Douglass 2008). The Asian region has experienced rapid urbanisation, with the percentage of the population living in urban areas increasing from 24% in 1970 to 42% in 2008, and within the next decade, it is likely half of the population in the region will be living in urban areas, many of which will be strongly concentrated in coastal areas (McGee 2008; UNESCAP 2008). Within these areas, the sensitivity of human populations to climate change impacts and the capacity to adapt effectively to climate change will vary significantly in association with their socio-economic situation, as well as specific socio-cultural capacities to manage environmental risk. While there is enormous diversity across the region, many of those living in the region already live in poverty, with gross national income PPP per capita in 2007 at or below US$1,500 in countries such as Nepal, Bangladesh and Papua New Guinea. Thus, the livelihoods of many millions of people across the region are already highly vulnerable due to their socio-economic situation, and their capacity to respond to climate change impacts in an effective manner either in situ or ex situ is limited (Brooks et al. 2005; Parks and Roberts 2006).

Given the potential importance of climate change impacts on migration, there has been relatively little empirical research on the topic for the Asia-Pacific, and it remains very difficult to provide guidance on how many people may move or provide any details on their sources or destinations (Hugo et al. 2009). In fact, there are so many gaps in current knowledge on climate change, migration and the relationships between them that it could be seen to be irresponsible to speculate on the specific numbers of people likely to migrate due to future climate change. Yet, there is a counter-argument that suggests that, while acknowledging the complexity, such informed speculation is vital now to better inform future planning and policy, even if the projected numbers and patterns are likely to be inaccurate.

Here, we wish to apply the concept of linear/non-linear migratory responses in relation to a critical examination of risks of Nepal and Thailand—Asian nations with populations highly vulnerable to projected climate change—to outline how it is possible to follow lines of reasoning to argue whether a climate change issue is or likely to be problematical to the governance of migration, and thereby inform risks to current migration policy. To achieve that aim, it is important now for research to examine and more fully articulate the particular migration issues that will arise in different contexts, and such a review is, in part, beyond the scope of this paper. Rather, the remainder of the paper will focus on what may become a particular problem for policy makers and practitioners wishing to effectively manage processes of migration, both internally within states and internationally—fundamental change in human mobility that leads to non-linear increases in migration as a result of climate change impacts—and reveal how a critical review of important thresholds could effectively guide applied research on risks to migratory policy from climate change impacts. The investigation and identification of such “tipping points” or thresholds of non-linear migration could be a key target for futuring migration research and, although not attempted here, could be potentially explained mathematically (Granovetter 1978).


The current population of Nepal is something above 23 million people (Government of Nepal 2009). Nepal has had much historical rural–urban migration, almost solely to the Kathmandu valley, and rural–rural migration, primarily due to lack of opportunities for development in some marginal agricultural regions and particularly from the hill and mountain regions (Shrestha et al. 1993; Bhandari 2004). Although Nepal’s population remains primarily rural, Kathmandu has grown considerably as an urban agglomeration to over 1 million people (Pradhan 2004; Haack and Rafter 2006). The out-migration from rural areas is both on permanent and on temporary bases, with a considerable number of people working seasonally as farm labourers (Fafchamps and Shilpi 2003). That said, ongoing rural poverty for large numbers of the Nepali population remains a significant potential future force for migration both internally and internationally.

Nepal has had a large traditional international migration to India, facilitated by a long-standing open-border policy. Skeldon (2006, p. 26) notes that “the international migration from Nepal to India is, in reality, little more than a spatial extension of internal circuits of migration.” Thieme and Müller-Böker (2004, p. 343) similarly state that “Nepal–India migration is facilitated by the free border agreement between the two countries. Official statistics from 2001 estimate that about 7,60,000 people were absent from Nepal. Out of this number, 77% were in India. But official figures grossly underestimate the number of migrants in India and the real value of remittances coming back to Nepal since money is sent back home mainly by hand, carried by the wage earners themselves, or sent via their friends. According to some studies, the number of Nepalis working in India ranges from 0.5 to 1.3 million. On the other hand, the Nepalese immigrant associations interviewed in Delhi estimated that around 2,00,000 Nepalis work in Delhi alone.” Links to more developed states are also well established, particularly to the United Kingdom, the Middle East, Japan and throughout South-east Asia (Yamanaka 2000; Seddon et al. 2002). Seddon et al. (2002, p. 20) estimate “that well over 1 million Nepalis are working abroad and that the value of their remittances to the Nepali economy as a whole may be as great as between 15 and 20 percent of gross domestic product.” Such labour migration corridors are likely to provide an ongoing international migration corridor for Nepalis attempting to find greater opportunities than are available within their country, because networks are established and the employment opportunities across the open border to India remain (Thieme and Wyss 2005).

There is some evidence from Nepal that environmental security is an important influence on internal and international migration, particularly via impacts on agroecosystems and production, but the issue has rarely been explicitly examined. Shrestha and Bhandari (2007) found that increased environmental insecurity measured in terms of access to forest resources increases the likelihood of labour migration. Moreover, they provide evidence that labour migration diversifies the family’s portfolio of earning opportunities and is an important mechanism for coping with environmental insecurity. While other studies suggest that the important proximate migratory push factors from the hills and mountain valleys of Nepal are poverty, food insecurity and lack of local development opportunities, there is increasing evidence that the environment is a significant ultimate cause of local under-development and, therefore, indirectly of migratory push factors. Menon (2009), for example, notes that where rainfall is not reliable across agricultural areas of Nepal, households employ a sustainable livelihood strategy of buffering their resources by having some family members involved in off-farm employment. As Menon (2009, pp. 867–868) states, “push factors include the scarcity of land and the need to self-insure by engaging ex-ante risk mitigating and income smoothing mechanisms”, and continues (p. 886), “in a country where poverty is already widespread, uncertainty in income arising from weather shocks can lead to increased vulnerability and deprivation.” One of the authors undertook 26 open-ended, face-to-face interviews with rural development practitioners and farmers in Nepal in 1999–2000 (Bardsley 2001). This work also linked migration to the environmental situation via the dominant form of socio-ecological interaction of local agroecosystems. For example, one rural development practitioner stated, “The people are migrating from the hills. Why? It takes 9 months to grow a maize crop, in the Terai it takes only three or four months maximum, so they can get two or three crops in a year in the Terai compared to the hills. People say in the Terai that they only have one plot of land but they get three times as much as the hills, because they get three crops: I can grow rice, I can grow maize, I can grow wheat. In the hills it is rainfed conditions, in the Terai it is mostly well irrigated, accessible to fertiliser, whatever agricultural products are easily dispersed in the market, you can just send it. In the hills people are migrating, they are looking for new jobs compared to agriculture because the soils remain very poor.”

Movement along existing corridors of migration is likely to continue to intensify as impacts of climate change limit development or undermine livelihoods through western Nepal, in particular. South Asian monsoonal patterns are expected to change significantly, including more extreme rainfall events, higher average rainfall in the summer monsoon and significant drying during the winter months becoming more common (Vörösmarty et al. 2000; Solomon et al. 2007, p. 884). Climate change impacts such as more extreme monsoonal rainfall and associated landslides and flood events would exacerbate the impoverishment of many rural Nepalis in the hill and mountain valley regions. However, for so many people living in the margins of Nepal, and especially western Nepal, the experience of poverty and the threat of malnourishment in the winter months are so significant (Adhikari 2000; Bardsley and Thomas 2005) that climate change is unlikely to make a significant difference. Therefore, it could be argued that established networks are likely to frame a linear increase in migration response to climatic barriers to development in the mid-hills and mountain valleys for some time.

A major threshold for a non-linear increase in migration may originate from a new movement of large numbers of people having to respond to greater intra-annual river flow variability in general and increasing flood risk in particular, in the relatively low-lying frontier Terai region (Agrawala et al. 2003). The Terai borders with India in the south, and in comparison to the hills and mountains, is densely populated (Shrestha et al. 1993). The Terai has been a region of net in-migration from other parts of Nepal since the 1950s, because as malaria risk diminished, agricultural development opportunities and access to land in particular have been comparatively abundant and supported by government (Shrestha 1990; Panday 1999; Bhandari 2004).

Due in large part to in-migration to the Terai population, growth rates have exceeded the national average and landlessness or near landlessness have increased to about 50% of the Terai population of about 7 million people (Shrestha et al. 1993). The Terai plain already experiences regular flooding. However, when considering climate change, the future increase in scale or regularity of extreme flood events in the Terai could lead to a new and significant increase in out-migration from relatively recent settlers who wish to escape regular inundation, crop and stock losses and associated impoverishment and malnourishment. Therefore, with climate change in association with limited flood risk management capacity (Dixit 2003), the Terai could both lose effective agricultural space and also provide less of a buffer within the country to absorb emigrants from mid-hill and mountain valley regions. While there is significant complexity in relation to local decision-making (see Massey et al. 2007), there is the potential for a threshold to be reached for many people to maintain sustainable livelihoods in the Terai in the face of more regular and intensive flood events. Given such a scenario, a non-linear migratory response may result, where there is a significant increase in the number of people that may have to move to other parts of the country or internationally.


Thailand, with a population of 63 million in 2008 (UNESCAP 2008), has experienced considerable internal migration, particular people moving from rural or secondary urban areas to the capital of Bangkok. Since the 1970s, Thailand has passed through a rapid transition to industrialisation, which has been highly centralised in and around the capital and relied upon the seasonal and permanent migration of people from all over the country (Garip 2008; Hirsch 2009). Garip (2008, p. 595) notes that “Much of this labor was provided by rural migrants from the north-eastern part of the country, where 40% of the population lived in poverty. Most of these migrants were in their teens or early twenties, and half of them were women.” There has been a significant environmental component to the tendency of people to move out of the north-east province, because it has comparatively few natural resource advantages in comparison to other regions, and the Central Plains in particular.

In much the same manner as the Nepali example, the impact of the environment on migration is manifest via agricultural development in the north-east (Isaan), which has been limited by the lack of irrigation opportunities and the consequent importance of dryland agriculture in an area of marginal rainfall for rice production (Parnwell 1988; Vanwey 2003). In work by one of the authors in 1995, a number of responses collated during 84 open-ended, face-to-face interviews with rural development practitioners and farmers in the north-east support the view that the long local dry season is fundamental to circular and permanent migration from the region (Bardsley 1996). Examples of the statements made are presented here from three rural development workers in different parts of the region, include: “It is very difficult to stop the young people from going into the town because in the dry season we cannot do anything—sometimes even the water for drinking is not enough”; “Most of the farmers move to the cities, especially Bangkok during the winter to work before returning to the fields when the rains arrive, because they have nothing to do in the dry season. The problem is that the crops are not cared for or weeded, but there is still a very strong commitment to family during the growing season”; and “Most farmers are working in the city during the dry. The new generation moves to the town, but they will come back to be farmers to help their parents. When their parents die they will not come back, the younger generation will sell because they don’t want to grow rice. At the factory they receive more money.” Village migration networks can be vital in supporting the movement of young people to urban centres from the rural north-east (Garip 2008). Thus, the process of urbanisation in Thailand has been driven by a complex mix of push forces associated with a lack of opportunities in many rural areas, but particularly the north-east plateau, and the lack of urban development away from a few major centres, and the pull forces of other rapidly growing areas and particularly the megacity, Bangkok.

Thai people have moved internationally but not in a similar pattern to most other countries across South-east Asia, which have experienced an extended period of colonisation (Huguet and Punpuing 2005; Numnak 2005). Over 1,50,000 Thai migrants have been leaving via legal channels from Thailand each year for the last decade. Huguet and Punpuing (2005, p. 25) note that “The great majority (82%) are males and most have low levels of education, although the average level of education has been increasing in the recent past.” Hugo (2006, p. 160) suggests that Thailand’s skilled workers are less likely to emigrate to OECD countries than those from other developing countries. However, there is still a considerable illegal movement of Thai workers internationally (Huguet and Punpuing 2005; Sciortino and Punpuing 2009). Hugo (2006, p. 165) further states that “Thai women are among the largest illegal migrant groups in Japan, numbering more than 1,00,000. Thai illegal migrants, mostly male, work in 3D (difficult, dirty and dangerous) jobs in Singapore, the Middle East, Malaysia, Taiwan, Japan and South Korea.”

The fact that South-east Asia’s megacities have been targets for rural–urban migrants is reflected in their substantial growth (Jones 2004; Jones and Douglass 2008; McGee 2008). Thailand, and the capital of Bangkok in particular, is just as likely to act as a stepping stone for these migrants on their way to other countries, as to export Thai workers (Skeldon 2006). Thailand is in fact a net-immigration country with around two million foreigners currently living and working in the country (Sciortino and Punpuing 2009). Since Thailand’s fertility level has been below replacement level for more than a decade, there are labour shortages in many low-skilled occupations, especially in sectors such as agriculture and fishing. The numbers of people moving to Thailand are considerable via numerous corridors, and both legally and illegally: including economic migrants from neighbouring states such as China, Burma, Laos and Cambodia (over 12,00,000 registered migrants); political refugees from Burma (over 1,00,000 people); and professionals from all over the world including many OECD countries (Huguet and Punpuing 2005; Balbo and Marconi 2006). Hugo (2006, p. 165) notes that “Thailand is a hub of illegal migration in the region. It is the destination of substantial illegal migration from Burma, but there are also significant numbers from China, Laos, Cambodia and South Asia” and continues “Thailand has become a major transit country for Indians, Pakistanis, Bangladeshis, Sri Lankans and Nepalese headed for the US, Canada, Europe, Japan and Korea while many stay in Thailand or go to Malaysia. Police estimate that at any one time 50,000 such illegals are in Bangkok.”

The impacts of climate change on Thailand could be considerable. A number of specific impacts are likely to reduce local opportunities for development, which in turn will enhance migration linearly along established corridors of migration. The variability in the timing and amount of rainfall associated with the monsoon is a significant factor in the viability of dryland rice farming, particularly in the north-east and north Thailand (Kundzewicz et al. 2009). If, as projected, the rainfall patterns become less predictable and/or reliable, cropping systems in some of the poorest rural areas of Thailand could become less productive. Already, the north-east provides many workers for urban areas, often on a seasonal basis, and internationally, and that migration pattern could be enhanced. Flood risk is also significant in the north and central areas of Thailand, and if extreme rainfall events occur more regularly, more people way leave hazardous areas, potentially following established corridors of migration (Kundzewicz et al. 2009).

One significant non-linear change to migration patterns may occur in Thailand as a result of increasing flood/sea level rise risk in Bangkok itself. Environmental hazards resulting from sea level rise, increasing cyclonic intensity and associated storm surges will be a great concern for many low-lying regions in South-east Asia, and Bangkok has been identified as one of the cities most vulnerable to coastal inundation (Webster et al. 2005; McGranahan et al. 2007; Vecchi and Soden 2007; von Storch and Woth 2008; Yusuf and Francisco 2009). The core of the city is built around the Chao Phraya River, and associated canals (klongs) are effectively at sea level, and therefore, Bangkok is highly vulnerable to sea level rise (Ericson et al.2006; Thaitakoo and McGrath 2008). Syvitski (2008, p. 27) notes that already “Excess groundwater pumping in the Bangkok area has caused rapid subsidence, to more than 100 mm/year, and the delta’s surface has subsided by more than 2 m, not only in the Bangkok metropolitan area, but also in the coastal region south of Bangkok. Between 1970 and 1990, sea level rose 0.5 m in the coastal region and the shoreline retreated 0.7 km.”

There is the potential that the interaction of sea level rise, storm surges and riverine flood risk associated with climate change could lead to a significant displacement of people from the low-lying sectors of Bangkok. Without in situ adaptation, thresholds may be reached where parts of Bangkok may no longer fulfil key roles in supporting dense local populations or sustaining important migration corridors. Structural measures have already been undertaken to reduce the rates of coastal erosion including building storage dams, constructing barrages, diverting channels and dykes, as well as planning future measures such as the development of pumping stations (Vitoolpanyakij 2009). There is a further concern, however, that some of these structural adaptation responses could in themselves cause significant human displacement, as they have in the past (World Commission on Dams 2000). Wahid (2009) proposes a range of non-structural measures that are now also required to reduce the risk of flooding to Bangkok, including smarter operations of upstream reservoirs; reducing land subsistence through control of groundwater extraction; installing a flood warning system; disaster management; mainstreaming climate change into city and land use regulations; flood fighting activity organisations; and a number of institutional mechanisms.

As described, the dominant existing corridor of out-migration from much of South-east Asia is through Thailand, and the strong networks maintained between the countries would mean that any climate change could induce more movement along that corridor. However, climate change has the potential to undermine the primary industries on which migrants rely and, therefore, will directly threaten the livelihood of those migrant workers and their families, as well as compromising their ability to send remittances to their home countries. Furthermore, if Bangkok is threatened due to one or a combination of the climate change impacts, then that channel for human mobility could be fundamentally disrupted. If the in situ adaptation measures prove insufficient or ineffective, a key non-linear impact on migration here would be a re-shaping of the corridor for migration in South-east Asia, with the role of Bangkok being undermined, both as a sink destination for internal Thai immigrants and a stop-over point for future international migrants. Therefore, while much of the focus in the debate regarding climate change and migration is on increased pressure to migrate from origin areas, in the Thai case, as in many others, climate change will also influence the destination areas and sectors of immigrants from neighbouring countries.

Conclusion and implications for migration policy

While it is of course very important that there are policies and programmes in place that give local people the choice to adapt in situ to climate change, equally it needs to be recognised that in some cases, the facilitation of effective migration, including resettlement, will become necessary. Much of the discussion on climate change migration has exaggerated the likely extent of direct impacts of future climate on displacement; however as discussed, the risk of non-linear increases in migration due to climate change is real and considerable. Given the enormous uncertainty but significant risks if systems are not established to manage changes in migration, which may involve new orders of magnitude of internal or external migration, the Precautionary Principle should apply (defined under Article 15 of the Rio declaration, 1992). It is important to put in place strategic mechanisms at national and regional levels, which have the capacity to facilitate the processes of changed linear and non-linear migration patterns, including both encouraging and managing temporary migration and selective permanent migration, as well as supporting the potential resettlement of individuals, families and in some cases entire communities displaced by climate change.

The futuring approach described here outlines how researchers can and do (see Hugo et al. 2009) apply the concept of thresholds of fundamental change in migratory patterns to highlight key vulnerabilities in migration policy frameworks. To better inform projections of responses to climate change, Kniveton et al. (2008) rightly argue that considerably more empirical information should be collected on how people make decisions in relation to future risk and behave subsequently, but to date, this information is largely absent. In the short term, by emphasising the key risks of non-linear changes in migration in particular local contexts, vulnerabilities in current systems can be responded to directly. Arguments are being made for similar focus of research in relation to physical environmental and infrastructural vulnerabilities in relation to future climate change impacts (Zahran et al. 2006; Bardsley and Sweeney 2010). The specific value of such an approach is that government and non-government decision makers are empowered to develop important policy that leads to long-term adaptation solutions in response to specific issues of heightened risk within their sphere of influence. Where large populations will be highly vulnerable to environmental change and provided with few adaption options away from increasing mobility, migration policy would begin to develop appropriate adaptation interventions.

If this intervention is carried out effectively and in a timely manner, it will be possible for states to not only cope with the expected changes in migration, but also to harness population mobility to reduce poverty and enhance economic and social development in both source and destination regions. Strategic policy responses to risks that some people in Nepal and Thailand face would need to ensure that people can move and settle in a way that assists all parties and may involve avoiding moving to other areas of high risk, especially along vulnerable coastlines. To do this, however, there will be a need for major improvements in many areas of migration governance, both internally and internationally, including: more effective migration management; strengthening governance; developing appropriate funding mechanisms to facilitate adaption to climate change; enhancing international cooperation on climate change issues; expanding and improving development assistance mechanisms; and developing sound economic development policy and practice throughout the region. Moreover, these changes are urgent for two reasons: (1) some of the climate change impacts are already in evidence and (2) because the changes required involve substantial institutional, structural and cultural change, it will take considerable lead time to successfully operationalise them. This is a major intergenerational issue, because many of the impacts of climate change are likely to be most severely felt several decades into the future and yet the societal adaptation responses to environmental change need to be put in place by the current generation. To avoid the poverty and suffering that could arise from inaction, and to help reduce inequality and improve the general well-being of Asia-Pacific residents, multilevel policy interventions are required now.

The bulk of climate change–induced forced migration will involve poor people since they will be the most vulnerable to climate change impacts and could be impoverished further by deteriorating local situations. Moreover, most of the resettlement will occur within low-income countries. One of the clear lessons from decades of resettlement experience (e.g. Cernea 1990) is that successful resettlement is not cheap. If resettlement is not to lead to further impoverishment, morbidity or mortality of those displaced, there will need to be a significant investment of resources made to successfully establish those displaced in new locations and provide them with security and a sustainable livelihood opportunities in these destination areas. Therefore, we recommend that policy initiatives focus on the development of governance mechanisms to resource both internal and international programmes when appropriate and encourage the safe transfer and effective use of the remittances (Hugo 2009). In particular, within the Asia and the Pacific, as elsewhere, there are a number of barriers to migration, which means that even where there are manifest shortages of labour, political and social barriers prevent migration. Hence, various systems that facilitate migration at minimum cost and difficulty between areas of excess labour to those of labour shortage need to be developed. If such developments in the governance of migration could be implemented and the vested interests marginalised, migration represents not only a coping strategy for communities confronting climate change but a possible solution to socio-economic inequalities across the Asia-Pacific.

Clearly, effective migration policy will require an increase in the capacity of nations to administer their own processes, particularly in the area of development and implementation of labour migration policy. Well-trained professionals across a range of migration-related skills will require proper resources, including software and hardware to manage migration, applying equally to origin and destination officials (Hugo 2008b). There is also a pressing need to improve migration data collection systems; accurate, comprehensive and timely information on demand and supply of migrant workers is a fundamental prerequisite of effective management of migration. Good governance will require increasing cooperation between authorities at origin and destination. Improving governance of international labour migration systems will involve confrontations with well-established vested interests which operate inside and outside existing legal systems, and often not in the interests of the migrant or of development in origin communities.

At the international level, Warner et al. (2009) have pointed to the general problems of diversity and fragmentation of current institutional arrangements to address the challenges of global environmental change. This is the case for migration as both a reactive and anticipatory adaptation to climate change, including potential resettlement responses. At the international migration level, effective institutions and legal instruments are still extremely limited, and poor governance of contemporary international mobility remains one of the major stumbling blocks to achieving the potential of development gains from migration in Asia and the Pacific (Hugo 2008a). As a first step, an appropriate international policy approach would be to strengthen and enhance the United Nations’ systems disaster response strategies and framework (Warner et al. 2008; UNHCR 2008). Another important strategy may be to establish, rather than set up a dedicated special fund to resource climate change migration, a method for migration costs to be funded along with other adaptation strategies through funds set up to resource all forms of adaptation like the United Nations Fund on Climate Change Kyoto-Copenhagen Adaptation fund (UNFCCC 2007). Beyond these steps, there are important future questions for migration governance emerging: Is there a need for a new international organisation with a specific mandate for the category of environmentally forced migrations? Does there need to be a unique separate categorisation of climate change–induced migration?

Climate change needs to be owned as a reality by researchers and governors of migration now, so that planning and policy for the future can be framed through the lens of climate change projections (Warner et al. 2008; Hugo et al. 2009). That change will need to be guided by good integrative research that focuses on engaging and working with decision makers and, as has been suggested here, by highlighting the emerging risks to current policy (Bardsley and Rogers 2011). The projections of uncertain futures are rarely going to be precise, but they provide us with the best guidance available on future environmental condition. However, from an understanding of the uncertain but significant risks of future climate change, it is possible to examine how people living in vulnerable situations will be affected and review, as we have described here, whether the changes to migration are likely to occur in a linear manner or involve a fundamental, non-linear transformation in the scale or type of migration. Examples from Nepal and Thailand suggest that it is in the non-linear responses that the serious risks to social welfare, migration policy and societal stability lie. It could even be argued that these thresholds of environmental change have been reached in certain contexts, such as the Ganges–Brahmaputra delta or some small island states. Future migration policy developments must begin to take these thresholds of social mobility into account to ensure that migration can become an increasingly effective tool for societal adaptation to climate change.


The authors wish to acknowledge the support of the Asian Development Bank’s Climate Change Program. This article draws in part from the report, “Hugo G., Bardsley D. K., Tan Y., Sharma V., Williams M. and Bedford R. (2009). Climate change and migration in the Asia-Pacific region. Unpublished Report to the Asian Development Bank, Asian Development Bank, Manila.”

Copyright information

© Springer Science+Business Media, LLC 2010