A livelihood analysis of resettlement outcomes: Lessons for climate-induced relocations

Abstract

The resettlement of communities has occurred throughout time from a variety of drivers. More recently, relocation from climate change impacts has emerged in policy frameworks and on-the-ground initiatives. While there are few case studies of climate-induced relocation globally, this is expected to increase in the future. Exploring the livelihood implications of past resettlements is one way of better preparing for this. This paper reviews 203 resettlement case studies to evaluate the implications on livelihoods and extract key lessons applicable for future climate-induced relocations. Findings revealed physical outcomes as the only in which any improvement was seen while natural, social, financial, human and cultural outcomes fared worse. Key consideration for future relocations emerged surrounding: (a) land and compensation, (b) accounting for the issue of access to livelihood assets, (c) accounting for the intersections of vulnerability within a ‘community’, (d) explicit recognition and attention to the cultural dimensions of relocation, and (e) meaningful participatory planning.

Electronic supplementary material

The online version of this article (10.1007/s13280-019-01289-5) contains supplementary material, which is available to authorized users.

Introduction

With the impacts resulting from climate change, both realized and anticipated, comes a new era of population mobility with the prospect that millions will be affected (Ferris 2015). It is however important to note that mobility (be it migration, displacement, or relocation) driven by climatic and environmental change, has been occurring throughout history (Oppenheimer 2003; Nunn 2007; Turney and Brown 2007; Tsonis et al. 2010). Not only has it occurred, but mobility (in its various forms) has played an important role in the survival and livelihood resilience of populations (Barnett and McMichael 2018). Yet current climate change impacts and politicization, coupled with an increasingly globalized and urbanized world, has resulted in mobility patterns being altered in new and evolving ways. For instance, this has been experienced in the Pacific atoll nation of Kiribati where a ‘migration with dignity’ policy was aimed at developing I-Kiribati educational and vocational skills in line with neighbouring countries including Australia and New Zealand, thereby allowing residents to be better prepared for migration under an unfavourable future climate change scenario (McNamara 2015; Hermann and Kempf 2017). Across Europe and Northern Africa, there is likely to be an increased influx of migrants from parts of sub-Saharan Africa resultant from climate change impacts (Anon. 2010). Climate change-related human mobility has also occurred on a more localized scale with entire communities being displaced and resettled often to nearby locations (Bronen and Afifi 2010; Barnett and McMichael 2018).

Across the world, populations have been displaced and resettled as a consequence of a range of drivers. These include infrastructure and development, urbanization, agricultural expansion, conservation, war and conflict, and environmental disaster (Mclean and Stræde 2003; Manatunge et al. 2009; Chen et al. 2017). Displacement driven by development activities such as dam construction, mining, and tourism, known as Development-Induced Displacement and Resettlement (DIDR), is arguably the most studied in terms of its design and outcomes. It has been estimated that roughly 15 million people per year are affected by DIDR (Terminski 2013). The displacement of people from DIDR primarily occurs to aid greater economic development. For example, throughout China the building of large-scale dams is a core aspect of its national economic development strategy and goals to alleviate economic inequalities throughout the country (Galipeau et al. 2013). Alongside the benefits that DIDR can bring, an extensive amount of research has attempted to understand the complex and negative consequences that this type of population displacement and movement poses to the lives and livelihoods of those resettled (rather than downstream beneficiaries) (Terminski 2013). Such implications for affected people have been widely documented within the DIDR literature and categorized as homelessness, joblessness, landlessness, marginalization, increased mortality, food insecurity, expropriation, and social disarticulation (Cernea 1997; Kirchherr and Charles 2016).

Climate change is likely to increase the need for future relocations resulting from both current and projected climate change impacts. This new era of displacement and resettlement of communities as a result of climate change has been labelled differently in both the literature and policy platforms as: planned relocation (United Nations High Commission for Refugees, UNHCR 2017); climate-induced resettlement (Lopez-Carr and Marter-Kenyon 2015); community relocation (Campbell 2010); and climate-induced relocation (Bronen 2015; McNamara et al. 2018). Defining the specific terms ‘relocation’ and ‘resettlement’ is somewhat contentious (Campbell 2010; McAdam and Ferris 2015). Some refer to the term relocation as only the movement of people from one place to another, while resettlement is used to refer to the re-building of livelihoods in a new location (Ferris 2015). In 2010, the Conference of the Parties (COP) to the United Nations Framework Convention on Climate Change (UNFCCC) stated that parties should enhance understanding, coordination and cooperation with regard to climate change induced displacement, migration and planned relocation (UNFCCC 2011). Since then, the UNHCR has established guidelines to assist countries dealing with planned relocation. In these guidelines, the UNHCR define relocation to include the process of resettlement, stating planned relocation is when “a community is physically moved to another location and resettled there” (UNHCR 2015, p. 10). Throughout this paper, we will refer to the movement of people by climate change impacts as climate-induced relocation.

Recognizing the various ways climate change is and will continue to influence human mobility patterns, this research focuses its attention on the process of climate-induced relocation. Fiji is perhaps the most notable example of such as four iTaukei (Indigenous Fijian) communities have relocated with the assistance of the Fiji Government (Charan et al. 2017; Barnett and McMichael 2018; Martin et al. 2018; Piggott-McKellar et al. 2019). Furthermore, over 80 communities have been earmarked for future relocation (Republic of Fiji 2014), and suggestion that this could be as high as over 600 communities (Neef et al. 2018). In other parts of the world including Alaska (Bronen 2008; Bronen and Afifi 2010; Simon 2018), Mozambique (Arnall 2014), Papua New Guinea (Lipset 2013; Connell 2016), Vietnam (de Sherbinin et al. 2011b) and Solomon Islands (Simon 2018), communities have also begun undertaking the complex process of climate-induced relocation.

Given this new era of displacement, relocation and resettlement driven by climate change, along with and exacerbated by other drivers, there is potential to draw lessons from the depth of research available from DIDR, along with other forms of displacement and resettlement, to apply to future climate-induced relocations. Yet in doing so, care must be taken when attempting to compare these processes, as other forms of resettlement are distinctly different from climate-induced relocation, particularly in terms of processes such as political motivation, resources (monetary and physical), and timing (lead times and deadlines) (Ferris 2011). While acknowledging these differences, there is still significant overlap and opportunity to offer insights into climate-induced relocation, with numerous researchers already drawing these links (see de Sherbinin et al. 2011a; Ferris et al. 2011; McAdam 2014; Ferris 2015; McAdam and Ferris 2015; Wilmsen and Webber 2015; Arnall 2018). For example, a study by Wilmsen and Webber (2015) note five similarities between DIDR and climate-induced relocation as follows: (1) both result from human actions, (2) often having long lead times to plan for resettlement, (3) resulting in livelihood impacts, (4) those most affected are the least powerful, and (5) there is limited international protection. A recent study by Arnall (2018) draws lessons from state-led resettlements across Africa and Asia to apply to future climate-induced relocations, while similar links have been explored in Pacific Island countries from historical displacement and resettlement cases (McAdam 2014; McAdam and Ferris 2015; Tabe 2019). Further, de Sherbinin et al. (2011a) note there is a possibility that existing policy and legal frameworks used to determine processes of DIDR may be applied to climate-induced relocation.

The relationship between climate change and DIDR is more connected than applying lessons from previous DIDR cases to climate-induced relocation. de Sherbinin et al. (2011a) argue that greater attention must be attributed to climate change not only displacing people from direct climate-related impacts (such as sea-level rise, increased intensity of storms surges, prolonged periods of droughts) and subsequently leading to climate-induced relocation, but that climate change adaptation and mitigation programmes and initiatives themselves can result in displacement and resettlement. For example, the increased development of dams for water storage and hydropower, biofuel expansion and wind farms implemented to combat levels of greenhouse gas emissions, and large-scale seawalls and other coastal defences have all been identified as causing displacement and relocation of populations (see de Sherbinin et al. 2011a).

With this context in mind, the impetus for this study is threefold. First, climate change is likely to increase the likelihood and need for people to relocate and resettle in areas of lower environmental risk in the future (Adger et al. 2012; Ferris 2015). Second, resettlement effects people’s livelihoods (Cernea 1997; Terminski 2013) thus making it important to understand. Third, there exists a range of lessons from empirical resettlement case studies and literature that can provide insights into the livelihood futures for climate change affected communities (Wilmsen and Webber 2015). Building off previous literature which asks, what can be learned from resettlements (specifically DIDR) when planning for climate-induced relocation (Wilmsen and Webber 2015), the aim of this study is to do so with a livelihood lens, through exploring and synthesizing the livelihood outcomes of past community resettlements, as documented in the peer-reviewed literature. Two associated research questions guide this study: what has been the extent and variation in livelihood outcomes for affected communities post-resettlement? and, what lessons from these resettlements exist that can be applied to future resettlements, particularly climate-induced relocations?

Livelihood analysis

Owing to the significant impact that resettlement can have on the livelihoods of affected people and communities, a livelihood analysis is applied in this research. Livelihoods is a term used widely within the literature and defining “what is a livelihood, and how can you measure it?” can be challenging. The sustainable livelihoods framework (SLF) is a widely used tool in both development-based organizations such as OXFAM, CARE and UNDP, as well as by researchers (Ferrol-Schulte et al. 2013). The SLF was established to understand how people generate a livelihood and how to most appropriately and effectively design human development initiatives accordingly (Scoones 1998). The SLF considers livelihoods to be people’s ability to access assets across five categories (natural, social, financial, human and physical). These assets go beyond what people utilize to build a livelihood, but further represent people’s capacity to be and act (Bebbington 1999). These assets both influence and are influenced by their interaction with the external context (i.e. shocks, vulnerabilities and trends) as well as transforming structures and processes (such as organizations, cultural norms and governance structures). These structures and processes influence the livelihood strategies employed and result in livelihood outcomes (e.g. increased adaptability to climate change, food security, poverty reduction) (Morse and McNamara 2013). Despite its prominence, there are many criticisms of the SLF. These include its lack of inclusion of power dynamics which can influence access to livelihoods (De Haan and Zoomers 2005) and a lack of accountability of the wider factors and influences that lead to poverty in the first place (McDowell 2002; Biggs et al. 2015). Despite these criticisms, the SLF remains an important and useful lens through which to understand livelihoods.

The SLF’s asset pentagon (natural, social, financial, human and physical) along with the inclusion of ‘cultural’ is used to guide the analysis in this research. ‘Cultural’ has been included for two reasons. First, other researchers have included ‘cultural’ as a distinct category in previous research (Bebbington 1999; Emery and Cornelia 2006). Second, the impacts of moving people from their land, as is often the case in resettlement, has a strong cultural component and thus it is important to consider. This is especially so as in many traditional cultures, which are disproportionately impacted by resettlement (Terminski 2013), land is an integral aspect of identity, spirituality, and broader cultural significance (Campbell 2010). As such including it here allows for a more holistic understanding of the impacts and outcomes of affected people’s livelihoods.

In this paper, the SLF asset pentagon is used to aid in framing the outcomes of resettlement for affected people. By using this approach, we can understand the wider implications on livelihoods by moving further away from the simplistic financial focus of livelihood rehabilitation often employed in resettlement research and practice (Mathur 2013). This is of particular importance as the UNHCR recommends in relation to climate-induced relocations that, “States should ensure at a minimum the restoration, but ideally the improvement, of livelihoods of Relocated Persons as both a matter of right and as an essential component in preventing impoverishment” (UNHCR 2015, p. 24). It is also important to understand that while a purpose is served by extracting and defining these categories of livelihood outcomes, there is significant overlap across them and often debate about which falls into which category (Scoones 1998). This is exemplified in the literature where the same asset has been attributed differently (see categorization of livestock in Hang Bui and Schreinemachers 2011; Mallick and Sultana 2017). Table 1 provides a guide to the categorization of the livelihood assets used in this analysis.

Table 1.

Livelihood categories—definitions and examples

Capital	Definition	Examples
Natural	The natural resources used to provide resources and services for livelihoods, such as land, water and climate	Land size, land quality, access to common resources (forests, rivers, oceans, etc.), crop quality and diversity, livestock, food security
Social	The social fabric and networks through which people interact, build relationships, and share resources on a daily basis	Relationships, community cohesion, community groups, relationships with government and private sector, sense of place
Financial	The financial resources people use to achieve livelihood objectives. Accounted for by both available stocks, and flows of income	Income, savings, access to loans/borrowing, compensation, expenses, remittances
Human	The skills, knowledge, ability to work and good health that enable people to pursue different livelihood strategies	Access to and availability of services for education, skill building, health, employment and safety
Physical	The basic infrastructure and goods needed to support livelihoods	Housing, infrastructure, access to water, provision of services, facilities and amenities, transportation, access to markets
Cultural	The set of constructs and rules for constructing the world, interpreting it, and adapting to it	Connection to land, religion, language, rituals, traditional practices

Based on: Downing and Garcia-Downing (2009), Sati and Vangchhia (2017) and Scoones (1998)

Methods

The following section describes how this archival research was undertaken. The search criteria, method employed, and data analysis are first discussed, followed by some of the limitations of this process.

Defining ‘Community Resettlement’

As this research takes the position of detailing ‘community’ resettlement (i.e. the resettlement of a community) the parameters through which this research is defined must be established. Acknowledging the term ‘community’ has its limitations and ambiguities (Buggy and McNamara 2016; Titz et al. 2018), it was chosen largely because of its wide use within the literature. As such, the parameters of a ‘community’ as well as ‘community resettlement’ were established to allow for appropriate and relevant literature to be selected for the purpose of this review. In this research, a community is defined as a social group of any size residing in the same specific location and under the same government umbrella (Schabas et al. 2016). From this, a set of parameters relating to ‘community resettlement’ were selected to include and exclude case studies.

The first parameter was that resettlement projects needed to involve a significant portion of a community, resettling from one location to another. This further included: cases of two or more communities amalgamating in a new location to form a new community; one or multiple communities amalgamating with an already-established community; as well as cases of one community being resettled into a number of locations, with either new or already-established communities. Second, the resettlement projects had to involve the permanent, not temporary, resettlement of a community. Third, the resettlements, or a significant portion of, had to have been completed at the time of the research. Fourth, to align with the aim of this research, case studies needed to be empirical and include experiential information from communities along with sufficient detail on the outcomes of the resettlement project for those affected by it.

Search method

An initial search of the academic literature was undertaken in August 2017 using Scopus. This was chosen due to the heavy importance placed on social sciences within this database (McNamara and Buggy 2017) which is especially relevant to this research. The search involved two separate search terms: ‘community’ AND ‘resettl*, and ‘community’ AND ‘relocat*’. The asterisks (*) were used to ensure the suffixes ‘-ment’, ‘-ed’, and ‘-e’ were all accounted for in the search. The search term ‘community’ was added for two reasons. First, this research is aimed at cases where a ‘community’ (see “Defining ‘Community Resettlement’” section) has been resettled. Second, by applying this condition, the number of results was substantially reduced from > 40 000 which was the number identified without the use of ‘community’ in the search term. The search was also limited to the social sciences, and academic peer-reviewed papers, with no time limits imposed. A second search, using the same criteria, was undertaken in June 2019. This was done to expand the literature sources to include books and book chapters, as there are a number of seminal works on DIDR and other forms of displacement and resettlement available from these sources. Further, this additional search updated the academic peer-reviewed literature.

Once the searches had been completed, a three-step method for the final selection of sources was used. The three steps were: (1) screening the titles, (2) screening the abstracts, and (3) a full text read. The criteria used to define ‘community resettlement’, which served the basis for inclusion or exclusion, are described in “Defining ‘Community Resettlement’” section. The ‘snowballing’ method was used to identify any new references from the reference lists of the selected sources. The same criteria were used to exclude or include new papers. These steps are outlined in Fig. 1, showing how many papers passed each step. In total, 142 sources were identified for analysis (see Electronic Supplementary Material for full list of accepted sources).

Fig. 1.

Search method utilized, along with the number of articles accepted at each stage

Data analysis

Within the 142 sources identified, 36 of these analysed multiple case studies. This was through a comparison between different resettled communities, a host and resettled community, or a comparison of the same (resettled) community over 2 different time periods (for example after 2 years post-resettlement, and then again after 10 years post-resettlement). These have been included as additional case studies to account for this, bringing the total to 203 case studies. It is also important to note that numerous resettlement schemes were discussed in different sources, each offering different perspectives, having been researched at different times, or with a different focus, or by different people.

To analyse data, an Excel database was created. General information such as the year of resettlement, year of study (or publication date if not available), driver for resettlement, and location was extracted. Further, the outcomes of each case study were extracted and categorized in the database based on the livelihood outcome discussed. Once all the outcomes within each case study were disaggregated, a content analysis was undertaken. This process involved making evaluative judgements of the text and categorizing it (Leavy and Prior 2014) as negative, neutral, or positive. A negative outcome was selected if there were overwhelmingly negative outcomes across that livelihood category for the affected population. For example, taking the example of physical, significantly reduced access to electricity for the majority of the resettled population would yield a negative response. A neutral outcome was selected when there was no change since the resettlement, or a balance between negative and positive outcomes. For example, there might have been improvements in access to one service, such as water, yet reduced access to electricity. A net positive outcome was declared when there were improvements on average, for a majority of affected people. For example, if there was an improved provision of electricity, and access to water and transportation, yet a decline in the size of houses, this was considered positive overall. If the case study did not report any outcomes for physical, it was marked as such.

Limitations

There are some notable and significant limitations to this study. These exist primarily around the search method and analysis. First, the use of the search terms ‘community’ and ‘resettlement’ has limited the scope of this analysis. For example, there is literature on resettlement that is discussed under other terms such as migration and displacement. In addition, within resettlement practice and research, there is a large field outside of academia. As such, the exclusion of grey literature from this search meant several case studies were not included. Further only English literature was used, and the search was undertaken in just one search engine (Scopus) which presents a further limitation to the scope of the study. These factors could all contribute to the search method missing relevant case studies. However, given that this research is primarily aimed at garnering lessons from the literature, the potential of missing some sources is not seen as a major downfall to the rigour of this study, especially as a significant number of cases (142 sources documenting 203 case studies) were identified and analysed. Another limitation of this study was the variation in information presented in the case studies, with some sources providing an in-depth quantitative and/or qualitative analysis while some provided only a descriptive overview. As such, authors’ discretion was used to determine whether a livelihood outcome was detailed enough to be included. Further defining outcomes as positive, neutral, or negative involved the authors making value judgments, albeit driven by the experience, based on the information published, which varied significantly. The implications of these limitations for this could be that the information presented in the sources was not holistically accurate and balanced, therefore skewing the outcomes of this analysis.

Overview of resettlement case studies

This section provides a summary of the resettlement case studies (n = 203) documented in the 142 sources. Based on the case studies reviewed, the construction of dams was the primary driver for community resettlement, accounting for just under half of all cases (see Table 2). Other development-induced resettlement followed, such as tourism, mining, and infrastructure, followed closely by conservation and natural disaster-induced resettlements. There are five examples attributable to climate change, all from Asia and Oceania, with four published in the last 5 years (2015–2019).

Table 2.

Overview of the drivers, location, year of publication, year of resettlement, and time between resettlement and study

Driver	%	Location	%
Dam construction	49.3	Asia	70.5
Natural disaster	12.3	Africa	20.7
Development	15.7	North America	1.5
Conservation	13.3	Oceania	3.4
Livelihoods	6.9	Central/South America	3.4
Climate Change	2.5	Europe	0.5

Year of publication	%	Year of resettlement	%
2015–2019	25.8	2010–2019	13.7
2010–2014	27.3	2000–2009	26.7
2005–2009	31.9	1990–1999	31.1
2000–2004	8.8	1980–1989	8.7
< 2000	6.2	1970–1979	7.5
		< 1970	12.4

Time between resettlement and case study (years)	%
< 2	8.9
3–5	10.3
6–10	23.3
11–20	32.2
21–40	17.1
> 40	8.2

In terms of location, Asia had by far the highest number of examples with 70.5% coming from this region (Table 2). This can be attributed largely to Asia housing a large concentration of hydropower dams (Stanley 2011). Africa followed as the next most common region, followed by Oceania and Central/South America.

The dates of resettlements span from as early as 1945 with the highest concentration during the 1990s and 2000s, together accounting for over half (57.8%) of the total. This is due to the high economic growth during this time, with dam construction accounting for the subsequent energy needs (Terminski 2015). In terms of publications, the highest concentration came during the period 2005–2009 (31.9%). Most studies were undertaken between 11 and 20 years of when the resettlement occurred, followed by between 6 and 10 years. The least number of cases were studied over 40 years from when the resettlement occurred (8.2%).

Implications and outcomes for livelihoods

The following section explores the implications and outcomes for livelihoods using the six livelihood categories as a framing for exploration. Figure 2 shows how many case studies reported, or did not report on, livelihood outcomes across natural, social, financial, human, physical, and cultural asset categories.

Fig. 2.

The percent of case studies that described an outcome across natural, social, financial, human, physical, and cultural asset categories

Cultural outcomes were reported on at a lower extent than any other, with only 43.84% of case studies describing associated impacts. Social outcomes were the next least reported on, with 71.43% of case studies discussing them. Physical (75.37%), natural (81.28%), and human outcomes (85.71%) then followed. Financial outcomes were the most commonly reported on with 90.15% of case studies discussing them. This exemplifies the greater focus on the financial component of re-building livelihoods following resettlement.

The case studies that had either positive, negative, or neutral livelihood outcomes (based on the method described in “Search method” section) show that all, except physical, had a greater portion of negative outcomes than positive (see Fig. 3). Cultural outcomes fared the worst, with 84.27% of case studies that reported on cultural outcomes, reported it negatively. Natural followed with over three quarters of cases that described it stating worse outcomes (76.97%). Social (68.97%), financial (63.39%) and human (56.32%) closely followed with high percentages of negative outcomes. Physical saw over half (58.17%) of case studies that reported, noting positive outcomes with under a third (30.72%) reporting negative outcomes.

Fig. 3.

The percent of case studies that described an outcome as positive, neutral, or negative, across natural, social, financial, human, physical, and cultural asset categories

One key aspect to consider pertaining to livelihood outcomes post-resettlement, is their potential to improve over time. For example, modifying an earlier framework by Scudder and Colson (1982), Scudder (2005) presented a four-phase model of resettlement which considers this temporal element of livelihood restoration post-resettlement. The model begins with phase one, where a high level of stress within a resettled community exists due to the uncertainty surrounding the future, to phase four where resettled people are fully integrated into the economy and self-sufficient. This model indicates that over time livelihood outcomes will improve as people become more familiar and adjust to their new surrounds. However, if resettlement is not done properly, these stages will not occur (Scudder 2005). One example of this temporal aspect at play is a case study of social networks from Indonesia and the Philippines which revealed community leaders and household heads will eventually replace disrupted social networks after a period of time in a new community (Quetulio-Navarra et al. 2017). Yet, the degree to which this temporal component plays out across displacement and resettlement case studies is equivocal. In a seminal study on the outcomes of World Bank driven dam-induced resettlement, Scudder (2005) states there is no evidence of an improvement in livelihoods for a majority of people over time. This study reinforces this finding, with no evidence of a temporal improvement in livelihood outcomes overall, across this review.

Natural

Loss of access to common (shared) property assets was a significant issue. Negative outcomes occurred pertaining to the loss of access to supportive resources, such as rivers, forests, wild foods and animals. For example, in the case of a hydropower dam construction in Malaysia, the community lost access to the river and forest where they previously hunted, collected and foraged for food, and had access to clean drinking water. As they expressed it, they lost access to their ‘supermarket’, which impacted their local economy as they could no longer rely on unlimited/free access to natural resources (Jehom 2013). Neutral or positive outcomes were rarely identified with regards to common natural resources, only occurring when people were resettled close to their previous homes, maintaining access to these common resources.

Landlessness refers to a lack of access (or entitlement) to land (Sati and Vangchhia 2017). People experienced landlessness either through not being directly compensated with land, not being provided with secure land tenure or, as described in a case study from Nepal, having to sell land they were given owing to the difficulties of life after resettlement (Lai Ming and Saumik 2013). In a community in the Philippines resettled after a volcanic eruption, the lack of legal ownership of land resulted in people becoming fearful of forced eviction by the government (Gaillard 2008). Women can also be more vulnerable to landlessness, as described in a case study in Malaysia: when women are dispossessed from their traditional lands, their vulnerability is significantly exposed as they cannot legally acquire property (Yong Ooi Lin 2006).

In terms of land size, this predominantly decreased after resettlement. In a case study from Vietnam, a hydropower dam construction was reported to have reduced 79% of original land size for the affected community (Hang Bui and Schreinemachers 2011). An increase in land size was reported only in a few case studies. A reduction in land quality was another common theme. Roughly four times the number of case studies reported a decline in the quality of land in the new location. The common decline in both size and quality of land led to reductions in crop production and diversity of food sources, and in some cases food insecurity resulted. For example, a case study in Laos where people relied on rice prior to the resettlement, 95% had rice deficits after the move (Delang and Toro 2011). A reduction in livestock was also widely identified across case studies. This stemmed largely from the loss or reduction of land size which resulted in there not being enough space for animals to live and graze, nor to produce feed for them.

Social

Cases that involved the integration of communities into already-established communities, or bringing together two or more communities into a new location, overwhelmingly reported negative social outcomes. Resettlements involving host communities invariably resulted in disputes over land given that the host community gave up land for the incoming population with often inadequate compensation. Conflict extended beyond land to other resources such as competition over health services and other communal resources, causing hostility, exclusion, feelings of isolation as well as physical threats. Other negative outcomes due to multi-community integration were power dynamics between different communities, a breakdown of social networks and ethnic tensions. For example, a case from a dam development in Chile noted that there were severe ethnic tensions from the integration of Indigenous and non-Indigenous peoples and, despite efforts to resolve this, the tensions remain significant (Gonzalez and Simon 2008). Few case studies indicated some form of positive interactions between new village members. This included people adapting easily to this transition after a dam-induced resettlement in Indonesia (Sisinggih et al. 2013), the maintenance of strong social ties despite kin groups being dispersed, described from a resettlement in Inner Mongolia (Rodgers and Wang 2006), as well as reduced isolation, and greater collaboration and assistance amongst communities. Case studies where communities remained together had more equally distributed outcomes.

The resettlement case studies reveal a number of major changes to social life and structure due to lifestyle changes. These included a more urban and modern lifestyle, limited economic opportunities, excess free time, and a loss of livelihood resources often resulting in alcohol abuse and other antisocial issues. An example from Malaysia as a result of a dam construction described how the younger males lost access to the river, previously an essential livelihood source, which was linked to increased abuse of alcohol (Swainson and McGregor 2008). Across these examples, this rise in alcohol abuse has led to social tensions and a feeling of insecurity and lack of trust within new settlements, and has also been attributed to a decline in cultural practices (Kingston and Marino 2010).

Resettlement entails enormous change in terms of village layout, communal spaces and housing. These case studies showed how these physical or infrastructural changes impacted significantly on social networks. An example showcasing this comes from the King Island Indigenous Alaskan community where the physical closeness of houses was of traditional importance as it represented the connectedness of village members. When resettled, the new location had a completely different layout and the sense of community and closeness was lost (Kingston and Marino 2010). Similar outcomes arising from an altered layout impacting social networks are discussed in multiple case studies. Yet, in one case, positive social outcomes resulted from the design of houses. This stemmed from the incorporation of local NGOs which were well versed in designing houses for urban poor and incorporated areas for supporting social cohesion into the design (Cronin and Guthrie 2011). Another major issue was the breakdown of communal interactions. This was attributed largely to the lack of prioritization of social and communal spaces. A tourism-induced resettlement in China described the new village as a ‘bedroom community’ as interactions have been reduced because public spaces were omitted from the new design and layout of the village (Wang and Wall 2007).

There are many important benefits of strong social networks such as avenues to loan and access money from neighbours and families. The dislocation of communities often leads to these networks being broken. This is apparent in numerous cases here where there are fewer exchanges between households. In one instance in Inner Mongolia, it was noted that labour exchanges amongst community members were still strong, which was a case where households resettled and remained close to each other, with the same neighbours and subsequent strong social connections (Rodgers and Wang 2006).

Financial

The associated financial costs of resettlement often proved a significant burden for those affected. This was in part due to improvements in access to facilities and services (described in “Physical” section) that incurred a financial cost. The cost of electricity was cited as a primary issue. Resettlement due to flooding in Mozambique showed that, despite having access to electricity, most households reported not being able to use it due to its prohibitively high costs (Arnall et al. 2003). Other expenses included the daily costs of a more urbanized lifestyle where people have to buy most of their food, and catch transport to school, work, and town. This was well illustrated by a case in Inner Mongolia where people could not afford to catch the bus, and as such could not access the opportunities of markets in the city (Dickinson and Webber 2007). Prior to the resettlement, in many such instances, these activities were free.

There was an abundance of negative outcomes documented in the case studies related to compensation. Key issues included the amount of compensation received not being sufficient in the opinions of those receiving it, delays in receiving compensation, as well as not receiving the full amount promised. An example from a DIDR in India stated that female heads of households could not access compensation; only males could (Asthana 2012). Related issues included people not being compensated for direct assets such as crops that were in the early stages of growing—and therefore valueless—when the resettlement occurred. This was especially pertinent with poorer groups. On the contrary, there were some positive outcomes in terms of financial capital, which came from compensation being adequate and appropriate to the expectations of the affected group.

While tied largely to employment (discussed in “Human” section), there were negative outcomes surrounding income and savings after resettlement. This included income streams no longer being stable, reduced, or completely lost after resettlement. People were also not able to save money, or had to use what little savings they had accumulated during and after the resettlement to rebuild their livelihoods. Many people had to borrow or take out larger loans that led to indebtedness. In saying this, the availability of loans and subsidies to resettled people was seen as a positive improvement in many cases. In one case study from Turkey, the government provided free services to those resettled (Akça et al. 2013). Another positive example was from the Philippines where the provision of and access to micro-financing resulted in greater livelihood opportunities after resettlement (Usamah and Haynes 2011).

Human

Negative outcomes pertaining to loss of employment or job insecurity were common. This occurred primarily because of the move from a subsistence lifestyle to urban areas where there are limited jobs and people lost their capacity for subsistence. The migration of people in search of work, causing greater disruption at home and social disarticulation resulted from this. Difficulty attaining stable and sufficient employment was especially noted amongst the elderly population who did not have the skill set to gain employment in new areas. This is shown in a case study from China by a comparison of ages and rates of unemployment: 3.3% for the 18 to 30 age group, 5.9% for the 31 to 50 age group, 26.1% for the 51 to 60 age group, and 68.9% for the over-60 age group (Tong et al. 2017). Numerous studies noted specifically a lack of appropriate skills and training to help people enter a new stage of employment. The term ‘appropriate’ is key here as it was recognized that, while training was provided, it was neither long enough nor sufficiently relevant to what the communities needed. It is also important to note that there were instances where training and skills provision was deemed successful, and resulted in positive employment outcomes.

Substantial gains in access to, and improvements in, education were noted. Greater access to schooling through the provision of new schools in the area, or being closer to established schools through moving to more urbanized areas accounts for this. On the contrary, multiple case studies noted that access to education has become worse. A case study from Malaysia stated there was the provision and building of new schools yet no teachers to work there so the facilities remained unused (Swainson and McGregor 2008). Another case in India from a development-induced resettlement showed an 18% dropout rate of school students after displacement due to difficulty in accessing schooling (Patel et al. 2015).

In terms of health, numerous case studies reported positive outcomes in terms of access to and improvement in health-related facilities. Yet, negative health outcomes were noted due to having to travel further distances to access health facilities, while in other cases it was increased incidences of disease, which in some cases lead to increased mortality. One example from Turkey stated that the resettled community experienced higher levels of obesity due to the profound change in diet (Akça et al. 2013). There were also reports of increased incidents of stress and other negative mental health outcomes after resettlement.

Physical

One of the most commonly reported improvements to people’s livelihoods in post-resettlement contexts refers to access to services. Improvement in access to electricity was the principal improvement. While improvements were generally seen in this area, there were cases were this was reduced. For example, a case from a resettlement driven by an airport expansion in Tanzania found that prior to resettlement, 95% of people had access to electricity, this dropping to a mere 8% post-resettlement (Mteki et al. 2017). While it is evident that there is room for improvement, there is a still a large majority of cases where increased access to services occurs. This is in many cases due to the transition to more urban areas where these services already exist. Greater access to water, transport, and communications was also identified.

Public infrastructure is another aspect that invariably improved with the building of schools, health centres, shops and restaurants, and roads. While this is important, issues of delays in finishing or simply never finishing were widely raised. In terms of housing, there were more mixed outcomes. The provision of new housing was stated as a positive in many instances with improvements in size, the addition of bathrooms and kitchens, flush toilets and solar panels, and improved structure of houses. On the contrary, instances existed where the provision of housing was either incomplete or ill-suited to the lifestyle or needs of the community. An example from a tourism resettlement in China stated that while a new modern kitchen was provided, this was not the way people wanted to cook, and they also did not include a shrine room in the house design which is important for them (Wang and Wall 2007). There was generally significant improvement in access to appliances and other household goods after the resettlement, particularly vehicles, fridges, televisions, phones, and stereos. This was primarily from monetary compensation.

Cultural

The loss of connection to land and place was expressed throughout the case studies, through a sense of loss from being moved away from familiar ancestral local environments. This was shown well in a case study from Inner Mongolia where attachment to the old village site and the mountains remains strong and imbued with meaning for which there is no substitute (Rodgers and Wang 2006). The impact on religion and associated practices was also noted. For example, a lack of places of worship in the resettled site or the enforced integration of multiple religions into the new location often resulted in reduced religious practice. By comparison, a case study community in Indonesia illustrated how the process of resettlement served to strengthen religious practices and increase religious activities in the community (Nakayama 1998).

There was also a noted significant change in the type and quantity of traditional rituals and practices undertaken by people. An example from a dam development in Malaysia showed that the traditional practice of ‘ngajat’ (a traditional welcoming dance) is no longer practiced due to the preoccupation of earning money and integration into a more modernized society (Choy 2004). The change in lifestyle, in many cases away from traditional means of fishing and agriculture has resulted in both the loss of traditional practices of hunting and farming, as well as the erosion of traditional knowledge of these areas and their history. A case from the King Island community from Alaska expressed the loss of language as the village has been integrated into a more modern society with the benefits of English speaking becoming more pronounced. Now, none of the younger generation speak the Iñupiaq dialect (Kingston and Marino 2010).

Lessons for future climate-induced relocations

Climate change is ushering in an evolving era of human mobility both through direct climate-related hazards as well as the implementation of mitigation and adaptation projects that displace people from their homes and lands. Within the resettlement literature and in practice, it is commonly understood that the implications for livelihoods of affected people can be disastrous (Colson 1971; Cernea 1997; Cernea and McDowell 2000; Scudder 2005). This review serves to replicate such findings and bolster these critiques. It details how physical outcomes, accounting for tangible aspects such as provision of services, improved community infrastructure and housing, was the only area in which most resettled communities witnessed any real improvements following resettlement. Across natural, social, financial, human and cultural assets, outcomes were generally negative post-resettlement.

If the findings from this review are in any way representative of the future of climate-induced relocation, this is extremely concerning. As such, this research has the potential to contribute to existing policy frameworks for climate-induced relocation including the UNHCRs guidelines in which impetus and importance is placed on, in the least restoring, but preferably improving the livelihood outcomes for affected communities post relocation (UNHCR 2015). In addition, other guidelines which exist pertaining to disaster-induced relocation and resettlement which have strong links to climate-induced relocation, including the Inter-Agency Standing Committee’s (IASC) Operational Guidelines and World Bank guidelines are of significance. For example, the IASC Operational Guidelines, which adopt a human-rights framing, identify a key guideline that resettlement should protect the rights of affected people in relation to housing and livelihoods (Oliver-Smith and de Sherbinin 2014).

This review therefore provides lessons from the experiences of previous resettlements in an effort to move towards better livelihood outcomes for affected populations in this new era of human mobility. Owing to the differences in contexts of each case study analysed (such as geographical location, drivers of resettlement, involvement of affected people, and the framing of research, as well as the complex underlying and dynamic political, economic, and socio-cultural differences), these key themes are not intended to be prescriptive but rather draw attention to some broad lessons and considerations common across a majority of case studies. These key themes are (a) land and compensation, (b) accounting for the issue of access to livelihood assets, (c) accounting for the intersections of vulnerability within a ‘community’, (d) explicit recognition and attention to the cultural dimensions of relocation, and (e) meaningful participatory planning. Each of these themes are discussed separately below.

Land and compensation

The results from this review found only 12% of cases showed an improvement in natural capital (such as land size and productivity, and access to natural resources) post-resettlement. Current examples of climate-induced relocations have largely been with rural communities (as is the case with most DIDR cases), where a heavy reliance on subsistence lifestyles and thus access to appropriate and suitable natural capital is a core aspect of livelihoods, making this finding particularly concerning. Considering the implications of this for future climate-induced relocations, executing land compensation in the form of just and appropriate land transfers, and land-based re-establishment accounting for size and productivity potential of, as well as access to relevant and sufficient natural resources including access to common pool resources, becomes clear. Furthermore, issues of land rights must be accounted to include both those who hold informal or unrecognized land title, and to ensure that adequate provision of land rights is maintained in the new location so as to not undermine long-term land security. This is an area that may become increasingly problematic, particularly in some Pacific Island Countries, where land boundaries and tenure arrangements can be unclear and are not formally recognized (Campbell 2010).

The UNHCR Planned Relocation Guidelines aim to address concerns about land by specifying that equitable compensation must be provided to relocated persons pertaining to the loss of land and subsequent land associated assets (UNHCR 2015, p. 24). Yet despite almost all cases analysed in this review providing a form of compensation for the loss of land for affected people, compensation was often found to be inadequate and led instead to undesired outcomes such as deterioration in cultural practices and increased living expenses. This raises questions about what constitutes equitable compensation, in particular the use of monetary compensation for land, especially considering the loss of invisible assets and the ensuing implications for long-term livelihood sustainability (Witter and Satterfield 2014). One positive example of an appropriate transfer of land is that of a recent climate-induced relocation in Fiji where the community was relocated to a nearby site from which residents could maintain access to their previous land, as well as being provided with alternative livelihood provisions in the name of land-based resources (crops, cattle, and fishponds). This helped the successful land-based re-establishment as well as allowing the community to maintain spiritual ties to their land (Piggott-McKellar et al. 2019).

Accounting for access to livelihood assets

Resettlement has the potential to improve access to resources and services for affected people, as documented in this review. This is especially pertinent in some regions of the world where rural communities lack access, namely to education, health care, employment opportunities, income, and electricity. Yet, in understanding the real and important potential resettlement can play in improving these assets, a deeper exploration of what access means must be considered. Examples of when resettlement projects failed to account for ‘access’ abound. For example, Galipeau et al. (2013) identified a case from a DIDR in China that provided schools and medical clinics to the community, yet these remained unused as there was no equipment and no qualified staff. Another example is the provision of electricity and services which, while important and beneficial, can lead to increased expenses and render the services impractical and unused if people cannot afford to use them, as described from a flood-induced relocation in Mozambique (Arnall et al. 2003). Similarly, while transportation was newly available to resettled people to access markets and opportunities, in numerous cases this could not be used as the cost proved prohibitive. This is shown by a case study in Inner Mongolia where the cost to catch the bus to the city and back was almost the entire average daily wage, making it unlikely that anyone would utilize this service (Dickinson and Webber 2007). When discussing the restoration of livelihoods, it is not only meeting immediate and direct needs, such as providing a new service, but also looking at whether these services can actually be utilized by the affected population, as well as their long-term sustainability.

Accounting for the intersections of vulnerability

This review shows that the outcomes of resettlement are disproportionately felt by certain groups and individuals, making them more vulnerable to impoverishment. The disproportionate impact on women is seen through an example provided by Asthana (2012) in which, after being resettled from a dam construction, women could not access compensation solely because of their gender. This greatly affected female-headed households and widows, as in parts of India (as in many other countries) only men are recognized as official heads of households and are thus the only ones able to claim compensation (Asthana 2012). Greater risks also fall onto the elderly as they are often less educated than the younger generations, especially in rural areas from which most resettlements occur, and therefore are unable to adequately engage with the process. As described from a tourism-induced resettlement in China, the older population experienced significantly higher rates of unemployment compared to younger generations after resettlement for this reason (Tong et al. 2017). The general failure to account for intra-community inequalities and vulnerabilities throughout the resettlement process can be seen as the World Bank themselves admit the failures they have made in dealing with resettlement in respect to more marginalized groups and individuals (World Bank 2015).

Climate change has been documented as disproportionately affecting those most vulnerable. This occurs both on a global scale in that developing countries, which contribute least and are generally considered more vulnerable, are those that are most affected (Althor et al. 2016). Similar arguments apply within communities due to inequalities on the basis of gender, class, religion, and ethnicity (Yamin and Rahman 2005). Further, it has been shown that if climate change adaptation projects do not explicitly target existing intersections of vulnerability, they risk perpetuating them. When looking toward climate-induced relocation, there is an explicit need to recognize and target these intersections of intra-community vulnerability to ensure they are neither aggravated nor perpetuated. This can be achieved through considered community planning processes that explicitly aim to target an array of local voices rather than pursuing a more tokenistic level of community participation, or none at all (Mallick 2011).

Recognition of the cultural dimensions of relocation

Compared to other areas of livelihoods, this review identifies that there are limited studies focusing on the cultural implications of resettlement (with over 50% not discussing any cultural outcomes). Of the case studies that did discuss culture, it is clear that culture is impacted greatly through the process of resettlement with almost 90% of cases reporting negative outcomes. Such impacts on culture included declining use of traditional practices, loss of connection to land, languages being lost, and undesired and unanticipated impacts on religious practices (described in “Cultural” section). Owing to both the limited detailed studies focusing on cultural outcomes, and the high rate of negative outcomes from those that did, this is seen as a key aspect to explore and understand further, especially within a climate change context. This is of further significance as the Planned Relocation Guidelines for climate-induced relocations state that an overarching principle is the preservation of culture for affected populations (UNHCR 2017, p. 11–12).

Downing and Garcia-Downing (2009) argue that while it is unlikely that post-displacement culture can ever be restored to its pre-displacement condition, largely because of a detachment from place, there are mechanisms that can be put in place to mitigate cultural impoverishment. As a first step, cultural impacts need to be understood, accounted and planned for with active participation of those likely to be affected. Benefits of following such steps can be seen with a recent climate-induced relocation in Fiji, where the village was relocated away from the coast due to shoreline erosion and tidal inundation. In this case, the burial site was moved so as not to be washed away; something extremely important to the local community members (Charan et al. 2017). While the need to relocate such a site is irrefutably not ideal, it demonstrates the importance of understanding the cultural needs, impacts and outcomes of relocation and resettlement.

Meaningful participatory planning

Appropriate planning involving the affected communities should be incorporated into the climate-induced relocation process with the support of stakeholders such as relevant local NGOs and government agencies that can assist with specific needs of relocation such as human rights, land, and health. From this review, the benefits of having targeted and meaningful assistance can be seen from a case study from India where there was involvement of a local NGO that assisted with the design and layout of housing, through an understanding of social networks and preferred living arrangements (Cronin and Guthrie 2011). In a similar, yet opposing vein, negative outcomes resulted when the consideration of affected people’s cultural and social needs pertaining to physical closeness of housing were not considered, as in the case of the King Island Indigenous Alaskan community (Kingston and Marino 2010). The involvement of affected communities must go further than just the planning process of relocation itself, but also through the initial decision to relocate. McAdam and Ferris (2015) discuss this challenge of ‘consent’ in the relocation decision, and how meaningful participatory involvement is an essential precursor to gaining this. These are essential things to consider and account for moving into a new area of climate-induced relocation.

Conclusion and future directions

This research used a livelihood analysis to explore outcomes for affected people after they have gone through the process of being resettled. This was done with the aim of learning lessons to enable best practice for future climate-induced relocations. The asset pentagon of the SLF was used to analyse the results, looking specifically at livelihood outcomes across natural, social, financial, human and physical categories, with the addition of cultural assets. This analysis examined 203 case studies of empirical research exploring resettlements and the outcomes for those affected. It was found that the resettlement process overwhelmingly resulted in negative outcomes for affected people across natural, social, financial, human and cultural assets. The only area that saw, on balance, any benefit from the resettlement process was physical outcomes including the provision of services and infrastructural improvements. From this analysis, key themes that should be considered moving into a new era of climate-induced relocation are identified. These include considerations of: land and compensation, the issue of access to livelihood assets, the intersections of vulnerability within a ‘community’, the cultural dimensions of relocation, and meaningful participatory planning.

As impacts from climate change will increase into the future, and more communities and governments face the complex reality of climate-induced relocation, it is essential that we learn from and draw on examples and understandings from previous community resettlements. As this review shows, there is significant room for improvement when re-building livelihoods following resettlement. Yet this also comes with major challenges. Moving forward, governments, the private sector, community groups, researchers and those at the frontlines should be working together more closely to ensure that livelihoods are rebuilt at an absolute minimum to pre-resettlement levels, but preferably improved in the resettlement site.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Biographies

Annah E. Piggott-McKellar

is a PhD Candidate with the School of Earth and Environmental Sciences at the University of Queensland, Australia. She has an Undergraduate Degree in Environmental Management, majoring in Sustainable Development where she attained first class Honours. Her research interests centre on human–environment interactions, specifically how communities and industries around the world are impacted by and responding to a changing climate.

Jasmine Pearson

is a PhD Candidate with the School of Earth and Environmental Sciences at the University of Queensland, Australia. She holds a Bachelor of Environmental Management (Honours Class I) with a major in Sustainable Development. After completing her Honours Thesis, Jasmine developed strong research interests in the field of local and Indigenous knowledge, particularly on how coastal communities in Small Island Developing States (SIDS) interact with local ecosystems and respond to environmental changes.

Karen E. McNamara

is a Senior Lecturer in Sustainable Development and Livelihoods at the University of Queensland, Australia. She has a Bachelor of Environmental Science (Honours I) and a PhD in Human Geography, and has worked at various universities and government departments in Australia and Fiji. Karen undertakes research on broad issues relating to sustainable livelihoods, most notably community-based adaptation to climate change, environmentally induced population relocations and disaster risk management—largely in the Asia–Pacific region.

Patrick D. Nunn

After 30 years of research, mostly focused in the Asia–Pacific region, Professor Nunn has accumulated a degree of expertise in a number of different fields. His primary field is geography, once largely physical in focus, but now straddling various aspects of sustainability. Professor Nunn has worked for a number of years in climate change, mostly on sea level and on the challenges of effective adaptation in poorer countries. Since 2000, when a coup in Fiji interrupted a planned research programme, Professor Nunn became interested in myths as potential sources of information about geological hazards, particularly earthquakes and tsunamis, volcanic eruptions and abrupt coastal change, something that is the topic of his latest book, The Edge of Memory (Bloomsbury, 2018).