Sierra: Rural-to-Urban Migration and Immobility Related to Water Scarcity in Peru’s Highlands

Abstract

The research interest in the Lima Region in Peru’s Central Highlands was in longer-distance, rural-to-urban migration from two villages, both harmed by two types of gradual climate impacts: glacier retreat and rainfall changes. The villages V1 and V2 were the starting points to trace migrants in the Regional and national capitals Huancayo and Lima. This dynamic is of interest for two reasons: first, water scarcity due to glacier retreat and rainfall changes is already salient across the highlands (Sierra). Migration from the Sierra can be shaped by both such rainfall changes and glacier retreat. Second, while rainfall projections are uncertain, future glacier loss will be severe even for low emission scenarios, with likely strong impacts on water security and migration. In the first section of this chapter, I provide information on the geographical context, measured and projected climate change trends and impacts, exposure, vulnerabilities, local coping and adaptation, and hazard-related migration in Peru’s highlands. Afterwards, I describe the empirical results of the new case study, discuss them, and induce propositions on broader well-being impacts of climate (im)mobilities. The findings suggest that the climate-related migration occurred under relatively adverse structural conditions. Movements ranged from more voluntary to clearly forced instances, often with relatively limited agency. Among those remaining, age differentials influenced future migration aspirations. At the destinations in Lima and Huancayo, migrants’ well-being results were on average more net-negative than net-positive. Two factions can be identified. The first group covers most of the interviewees who had left the increasingly adverse situations in their home villages in recent years under distress conditions and arrived with limited baseline resources and skills in the oversaturated, large urban agglomerations. These migrants experienced great challenges for their development from a secure base and a space to live better, while gradually reaching similar or better social relatedness as before migration. Although they were subsisting in the cities, multiple deprivations restricted their development prospects. These migrants with multiple unmet needs subjectively suffered from deprivation in their cognitive satisfaction and emotions; only few reported partial and gradual adjustment. By contrast, the second group comprises a small number of migrants who achieved urban upward mobility, in particular those who had moved voluntarily and a longer time ago. In often-arduous processes, they have used urban education and job opportunities to improve their well-being and long-term prospects. This small group evaluated its needs as mostly fulfilled. Nonetheless, both the large number of migrants losing OWB and the few ones gaining OWB held mostly negative views of the future. Using these empirical findings from Peru’s highlands, I conclude by inducing broader propositions on the well-being of climate migrants in cities and stayers in rural source areas.

“Toda la gente se va a las ciudades, pero ya no hay espacio donde vivir, … y el agua se va a escasear, y en las ciudades serán peor seguramente …”

“Everyone goes to the cities but there is no longer space to live, … and water is going to be scarce, and in the cities, they will be certainly worse ...” (own translation, as in all the chapter).

Statement by a mestizo¹ man living in a village in the highlands that is experiencing glacier retreat. He was in his early 60s and in good health at the time of the interview and made a living as a civil servant and farmer (V1-1).

The research interest in the Lima Region in Peru’s Central Highlands was in longer-distance, rural-to-urban migration from two villages, both harmed by two types of gradual climate impacts: glacier retreat and rainfall changes. The villages V1 and V2 were the starting points to trace migrants in the Regional and national capitals Huancayo and Lima.² This dynamic is of interest for two reasons: first, water scarcity due to glacier retreat and rainfall changes is already salient across the highlands (Sierra) (Heidinger et al. 2018; Seehaus et al. 2019). Migration from the Sierra can be shaped by both such rainfall changes (e.g. Hook & Snyder 2021; Lennox 2015; Milan 2016; Milan & Ho 2014) and glacier retreat (e.g. Alata et al. 2018; Altamirano Rua 2021; Figueiredo et al. 2019; Heikkinen 2017; Wrathall et al. 2014). Second, while rainfall projections are uncertain, future glacier loss will be severe even for low emission scenarios (Adams et al. 2014; Marzeion et al. 2012; Radić et al. 2014; Schauwecker et al. 2017), with likely strong impacts on water security and migration. In the first section below, I provide information on the geographical context, measured and projected climate change trends and impacts, exposure, vulnerabilities, local coping and adaptation, and hazard-related migration in Peru’s highlands. Afterwards, I describe the empirical results of the new case study, discuss them, and induce propositions on broader well-being impacts of climate (im)mobilities.

1 Context

28% of Peru’s population resides in the Sierra (INEI 2018c), a share that has significantly decreased over the past decades, similarly as in other Andean regions and highlands worldwide (de Sherbinin et al. 2012; Valdivia et al. 2010). The study sites V1 and V2 are in one of the mountainous Provinces of the Lima Region (Figure 5.1). In 2017, one year before I collected the data, 0.91 million inhabitants lived in this Region, not counting the Metropolitan Province of Lima. V1 was home to about 200 residents (56% male / 44% female) and had 200 houses (INEI 2018c).³ It was connected to the public water network, electricity grid, and drainage system, and hosted a health post as well as small primary and secondary schools. By contrast, V2 had only 70 houses home to about 40 villagers (49% m / 51% f). It had access to electricity, but not to the water network or drainage system, and only a primary school with one pupil left. In both sites, villagers speak Spanish and consider themselves as mestizo. V1 is a district village whereas V2, as V1’s administrative annex, lacks an independent budget.

Figure 5.1

Sites for qualitative data collection in Peru’s highlands. (Note: To protect the respondents, the pins indicate approximate locations only. Created by the author, based on CIA (1970))

The Andes, home to V1 and V2, feature an alpine Tundra climate with wet and dry seasons (Beck et al. 2018; Mächtle 2016).⁴ Humid air masses from Amazonia travel westwards over the high mountain chains (Cordilleras) that are separated by deep valleys, and bring summery rainfall from December to March. Precipitation gradients and topography-related temperature differences shape vegetation and agricultural potential: western Amazonia features open mountain forests, followed by humid grassland highlands in the north, and arid desert puna in the south. Between 2,000 and 3,500 m.a.s.l., the cold land (tierra fría) and cloud forests allow for livestock farming and field crops, such as corn and quinoa. Above the timberline, in the frosty land (tierra helada), few crops such as potatoes and pastures such as alfalfa grow. Vegetation in the puna consists of grasslands that provide forage to camelids such as alpacas. Above 4,500 m.a.s.l., vegetation is scarce.

The river basin in the Cordillera Central that is home to both V1 and V2 has around 500 lagoons and 15 km² of mostly small, glaciated surfaces higher than 3,700 m.a.s.l. (Proyecto Glaciares 2018). Figure 5.2 shows the glaciated surfaces. Rainfall and glacial meltwater feed a high-altitude dam lake, which supports the water regulation in the basin; during the dry season, up to 80% of the water in the dam lake is from glaciers. 85% of its water is used for hydropower generation and 15% for agriculture.

Figure 5.2

Ecosystems in the Cordillera Central and glaciated surfaces. (Note: Cropped from INAIGEM (2018: 156) and adjusted by the author)

Within the basin, V1 and V2 are in a remote, protected natural area at an altitude of around 3,600 m.a.s.l. The next larger cities are several hours away by car. Pulgar Vidal (1972) coined this ecological floor as Suni.⁵ The cold subalpine climate here has distinct dry and rainy seasons and low average temperatures (INAIGEM 2018) (Figure 5.3). Rainfall occurs mostly from October to March and is the main source of water in the basin (Proyecto Glaciares 2018; SERNANP 2006).

Figure 5.3

Temperatures and rainfall close to V1 and V2. (Note: Average temperatures and rainfall per month from 1982-2016. Produced by Stephanie Gleixner with PISCO data, edited by the author. The station name here and in the following figures is concealed to protect the privacy of respondents)

However, for some rural zones in the Andes, glacial meltwater is key during the dry season (Buytaert et al. 2017). In the mountainous Lima Region, the focus herein, meltwater contribution to river flow can reach between 50% to 100% during the driest months of a drought year (Figure 5.4, right). Similarly, the villages V1 and V2 have an increased dependence on meltwater during dry months.

Figure 5.4

Contribution of glacier meltwater to river flow in the Lima Region. (Note: Spatial propagation of the contribution of glacier melt water [%] to river flow; yearly average for a normal year (left) and monthly maximum for a drought year (right). Cropped by the author from Buytaert et al. (2017: Supplementary Materials S4, S7))

Most highland areas in Peru are exposed to various natural hazards, climatic changes, and non-climatic stressors, such as soil degradation and overgrazing (Dourojeanni et al. 2016; MINAM 2011, 2016b; SERNANP 2016). Many inhabitants in the Sierra must confront multiple hazards, sometimes concurrently or in short subsequent periods (Cavagnoud 2018; Perez et al. 2010). Hazards often revolve around water (droughts, floods, hail, snow, glacier retreat, rainfall changes) and temperature (daily and seasonal extremes, average increase) (e.g. Alata et al. 2018; Aragón et al. 2018; Koubi et al. 2016). Highland villagers perceive that climate change has multiplied hazards (Oft 2009), especially glacier retreat and rainfall changes (Heidinger et al. 2018; Seehaus et al. 2019).

For V1 and V2 specifically, a trend analysis is possible with gridded satellite and station data from the PISCO dataset (Lavado-Casimiro et al. 2016).⁶ Figure 5.5 displays that average monthly temperatures have increased by up to 0.5 °C when comparing 1997-2016 with 1982-2001. Annual mean average, minimum, and maximum temperatures validate this warming trend. Maximum daily temperature ranges show strong interannual variability but are overall decreasing. Despite the low temperatures, cold spells of several days are rare and their length has decreased over the past years.

Figure 5.5

Temperature trends close to V1 and V2. (Note: Average monthly temperature 1982-2001 compared to 1997-2016 (top left) and annual mean, maximum, minimum temperature (bottom left); cold spell analysis (top right) and daily maximum temperature range (bottom right). Produced by Stephanie Gleixner with PISCO data, edited by the author)

Previous studies do not find significant rainfall trends in the tropical Andes in the 20th century (Rabatel et al. 2013). For V1 and V2, the data in Figure 5.6 seems to suggest an upward trend in annual precipitation, but with strong interannual variability. The rainy season has ostensibly intensified, whereas dry months have become marginally drier. While the frequency of dry spells seems to have decreased, their length might have increased. For rainfall extremes, the data displays no clear trend.

Figure 5.6

Rainfall trends close to V1 and V2. (Note: Average daily precipitation (top left) and annual precipitation (bottom left); dry spell and 95^th percentile trends. Produced by Stephanie Gleixner with PISCO data, edited by the author)

Rainfall changes and glacier retreat—the hazards in focus here—often affect people in the Sierra in parallel (e.g. Charbonneau 2008; Heikkinen 2017; Magallanes 2015). In prior studies of climate migration, people reported changes in seasonal rainfall patterns (Milan & Ho 2014), such as delays in the seasonal onset (López-i-Gelats et al. 2015), changes in seasonal length (Alata et al. 2018), and abrupt seasonal turns (Adams 2016). Altered rainfall availability was also reported, such as periods with more or less precipitation than usual (Cometti 2015a), reduced water availability (e.g. Adams 2016; Alata et al. 2018; Milan & Ho 2014), and droughts (Cavagnoud 2018; Koubi et al. 2016; Kuznar 1991; López-i-Gelats et al. 2015; Oft 2009). Studies also document changing rainfall intensities (e.g. López-i-Gelats et al. 2015), such as more frequent and extreme events (Oft 2009; Wrathall et al. 2014), which can contribute to mud- and landslides (e.g. Adams 2016; López-i-Gelats et al. 2015). The studies above conclude that rainfall changes can have widespread impacts on people’s lives and livelihoods in the highlands, including losses or reduced crop and fodder yield, crop diseases and pests, reduced quality of products, losses and diseases of animals, a decline of biodiversity, damages to infrastructure and assets, food insecurity, and other health challenges.

The severe glacier retreat in Peru is due to warming and changing El Niño events (Rabatel et al. 2013; Vicente-Serrano et al. 2018; Vuille et al. 2018). The retreat has accelerated since the mid-twentieth century (Bury et al. 2011; Georges 2004; Hastenrath & Ames 1995; Racoviteanu et al. 2008). Today, surface losses are at least 40% for all glaciers and various smaller ones are about to disappear with below 10% to 30% surface remaining (INAIGEM 2018). From 2000 to 2016 alone, close to 30% of Peru’s glacier area was lost (Seehaus et al. 2019). In the Cordillera Central, where V1 and V2 are located, data indicates that the glaciated surfaces decreased by 64% from 1962-2016 (Figure 5.7). Once a peak is crossed, glacial meltwater flows decrease, which is especially problematic for agriculture during the dry season (Orlove 2009; Wrathall et al. 2014). Glacier retreat also frees contaminants stored in the ice (Alata et al. 2018), has started to create water conflicts, and is affecting spiritual and cultural dimensions of people’s lives in the Sierra (Altamirano Rua 2014).

Figure 5.7

Loss of glaciated areas in the Cordillera Central. (Note: Reproduced and edited by the author from INAIGEM (2018: 166))

Even in a low emission scenario, future average temperatures in the Sierra could increase between 0.75 and 1.5 °C by 2050 and between 1 and 1.75 °C by 2100 compared to 1985-2005;⁷ in a high emissions pathway, they could rise by as much as 1 to 2 °C and 3.5 to 6 °C (Bergmann et al. 2021a).⁸ Specifically in the Province home to V1 and V2, in a high emissions pathway⁹, the 50^th percentile of mean air temperature would rise from ~7.5 °C in the 2010s to above 11°C in the 2080s, and maximum temperature would rise from 20.5 to 24 °C (Figure 5.8). These changes lead to ample risks for crops (Sanabria et al. 2014). 1.3 °C and 2.6 °C higher mean temperatures over the coming 35–70 years could reduce maize and potato production in Peru by more than 87% in current elevations (Tito et al. 2018).

Figure 5.8

Observed and projected temperature in the Province where V1 and V2 are sited. (Note: Mean air temperature (top) and maximum air temperature (bottom), 1975–2085. Observations/ W5E5 observations-Regional Rivalry SSP3, 7.0 W/m² / CMIP6 GCM ensemble. Created with data from Climate Impacts Online by PIK, http://kfo.pik-potsdam.de)

Rainfall projections are less certain; Peru may see fewer rainy days but with more intense rainfall (Christensen et al. 2013; Giorgi et al. 2014). For the Province home to V1 and V2, a high emissions pathway could lead to fluctuating but rising average rainfall and more wet days by 2085 (Figure 5.9).

Figure 5.9

Observed and projected rainfall for the Province where V1 and V2 are sited. (Note: Precipitation (Top) and number of wet days (Bottom), 1975–2085. Observations/ W5E5 observations-Regional Rivalry SSP3 7.0 W/m² / CMIP6 GCM ensemble. Created with data from Climate Impacts Online by PIK, http://kfo.pik-potsdam.de)

Future glacier volume losses are projected between 78% and 97% for the Central Andes, home to V1 and 2, for low and medium emission scenarios (leading to 2 ° or 3 °C warming above pre-industrial temperatures by 2100). By contrast, the loss would be close to complete with 93% to 100% in a high emissions scenario (inducing 4 °C warming) (Adams et al. 2014; Giesen & Oerlemans 2013; Marzeion et al. 2012; Radić et al. 2014). Lower-lying glaciers are most at risk (Rabatel et al. 2013). If the current trend in the Central Andes was to continue linearly, its glaciers would disappear in 2048 (INAIGEM 2018). Runoff will decrease once peak flow is crossed (Seehaus et al. 2019; Veettil 2018; Vuille et al. 2018), which is projected in 20-50 years for most tropical glaciers (Adams et al. 2014).

The projected retreat of Peru’s glaciers will heavily affect downstream ecosystems and users. Future glacier reductions would not necessarily reduce total water yield in Peru, but dry season runoff and seasonal buffering capacities may sink, especially as rainfalls may become even more seasonal at the same time (Buytaert et al. 2017). Projections show reductions of dry-season runoff and increases of wet-season discharge for the 2050s and 2080s (Andres et al. 2014; Juen et al. 2007; Lavado-Casimiro et al. 2011), including for a basin in the Lima Region close to V1 and V2 (Olsson et al. 2017). Meanwhile, rising populations, especially in cities, and more usage by industries, hydropower generation, and intensive irrigation on the coast increase the demand on already stretched resources (Buytaert et al. 2017; Buytaert & de Bièvre 2012). Moreover, meltwater can accumulate in glacial lakes that carry the risk of potentially destructive outburst floods (GLOFs) (Emmer et al. 2018). Such GLOFs threaten Huancayo, Huaraz, and other large cities in the Sierra (Frey et al. 2016; Haeberli et al. 2016; Huggel et al. 2020; Stuart-Smith et al. 2021). Rapid deglaciation will also result in water quality issues and severe non-economic losses, including the disruption of moorlands (Adams et al. 2014) alongside the loss of aesthetic and spiritual ecosystem services key to people’s identity (Adams 2016; Paerregaard 2013, 2016). Because glacier landscapes are also key destinations for tourism, their loss might threaten jobs in the tourism sector (Altamirano Girao 2012).

Vulnerabilities to the discussed climate hazards are significant in the Sierra. They depend on the extent, quality, and location of household resources, including land and livestock, and factors such as age, family size, and health (Heikkinen 2017). In many areas, poverty, insufficient property and resource rights, poor soil quality, poor basic infrastructure and services, a lack of quality education, food insecurity, as well as a lack of savings and access to credit raise vulnerabilities (Koubi et al. 2016; Oft 2009; Oliver-Smith 2014; Sperling et al. 2008)). For example, a large-scale survey across the Peruvian Sierra finds a high poverty incidence of 55% (Aragón et al. 2018). Vulnerabilities are differentiated within households; for example, illiteracy is especially high among women and malnutrition particularly affects children (Oliver-Smith 2014). Constrained livelihood options also raise vulnerabilities: many households depend heavily on few agricultural activities, with limited diversification options and a small number of income earners (Cavagnoud 2018; Oft 2009, 2010). They are often subsistence crop and livestock farmers (Oft 2009, 2010; Perez et al. 2010), and as smallholders who use traditional practices, they are mostly dependent on rain-fed crops and have limited means to irrigate (Aragón et al. 2018; Heikkinen 2017). Some villages hold lands in various ecological floors to diversify livelihoods (Crespeigne et al. 2009), yet options are more limited the higher the altitude. Crop farming becomes less viable as elevation increases (Magallanes 2015; Milan & Ho 2014; Oliver-Smith 2014), and in the highest altitudes, pastoralism is often the first and, in some cases, the only option (Alata et al. 2018; Oliver-Smith 2014; Orlove 2009). Off-farm options can decrease vulnerability; for example, some farmers work complementarily in coffee harvest areas or in urban activities, such as construction (Milan & Ho 2014). However, additional non-farm incomes are only available where spatial proximity permits (Adams 2012, 2016; Adams & Adger 2013; Cavagnoud 2018; Cometti 2015a). Remoteness, land tenure, market competition, globalization, rural population changes, renunciation of traditional practices and institutions, as well as a lack of state presence and public services further heighten people’s vulnerabilities (Lennox 2015; Lennox & Gowdy 2014; López-i-Gelats et al. 2015; Oliver-Smith 2014).

Such vulnerabilities—especially those related to poverty and constrained livelihood options—also limit people’s range of strategies to deal with hazards. Generally speaking, people in the Sierra react to variations in precipitation and glacier retreat through changes in crop management and crop varieties or changes in animal raising, care, and derived goods (Alata et al. 2018; Heikkinen 2017). To anticipate extreme heat, farmers use different plants and expand land use; once heat creates income losses, they sell livestock, make children work, and invest extra time in off-farm activities (Aragón et al. 2021). Yet, vulnerable groups usually have limited adaptation options: in one study, wealthier pastoralists accumulated livestock, less wealthy factions diversified assets, and the poorest groups reduced participation in pastoralism and sometimes completely abandoned it (López-i-Gelats et al. 2015). Moreover, climatic and non-climatic stressors often already exceed the adaptive capacity of mountain villages and threaten to create downward spirals of poverty (López-i-Gelats et al. 2015). Some practices, such as giving up traditional crop rotations, can further deteriorate soil quality and impair future production, and others may increase vulnerability, for example, if they raise dependency on single products (Lennox 2015; Lennox & Gowdy 2014). Recovery from shocks is often incomplete. In one study, surveyed households could only recover 76% of losses after droughts and floods (Oft 2009, 2010). In addition, today’s practices come with limits—such as the availability of land—and may be insufficient to meet the magnitude of future changes (Aragón et al. 2018). For example, in one study, many highland villages struggled to cope with water scarcity in the short term, mostly asking peers for help, and reportedly lacked long-term options to adapt (Oft 2009, 2010).

Several prior studies observe migration for coping or adaptation, both due to rainfall changes and glacier retreat. For contextualization, it is key to know that Peru has experienced strong demographic changes and migration patterns are long established, as I analyzed in detail elsewhere and only briefly outline here (see Bergmann et al. 2021a). Over the past decades, Peru’s population has grown at a flattening rate, moving the country toward a low transitionary state with a rising life expectancy in a population that is still young, but has a greater share of older people than before (INEI 2018c; UNDESA 2019; World Bank 2021c). This shift may have raised migration potential, as younger people tend to be more mobile (Millington 2000; Plane 1993; Rogers & Castro 1981). The focus of this dissertation is on internal migration, which outweighs cross-border migration in Peru (INEI 2018b; Sánchez Aguilar 2012a; UNDESA 2016).¹⁰ About one fifth of Peru’s population are internal migrants, with slightly more men than women (Sánchez Aguilar 2012a). They tend to be relatively young and educated above the average. Many internal migrants work usually outside of agriculture and seem to earn more than non-migrants, although data on available net income is limited. Internal migrants remit some of their income to support family members—mostly women and economically inactive relatives—in their home communities, which could contribute to climate adaptation.

This internal migration is a major driver of Peru’s population redistribution: between 1940 and 2017, the share of the population living in the highlands has dwindled from 65% to 28%, while the share of population on the coast has increased from 28% to 58%, and that in the rainforest from 7% to 14%. Prior research confirms that migration is part of the social fabric in the Sierra (Cavagnoud 2018; Skeldon 1977, 1985). Both permanent and circular migration is common.¹¹ Demography, lack of land or educational opportunities, poverty, and unequal market access are among the main drivers (Crespeigne et al. 2009; Heikkinen 2017). Increasingly, disrupted lifecycles of many rural highland families also drive migration (Alata et al. 2018), so that many Andean and other mountainous villages have become skewed toward older adults (de Sherbinin et al. 2012; Valdivia et al. 2010).

Existing internal migration patterns are deeply embedded in Peru. They involve relatively stable shares of the population and reproduce disparities between receiving and sending areas. Rural–urban migration is high but urban–urban and intra-metropolitan flows are increasing (INEI 2011, 2018c). Metropolitan Lima and Huancayo are hubs for migrants from the rural Sierra given perceived opportunities for education and jobs (Sánchez Aguilar 2012a). They are also two of the main destinations for V1 and V2. Lima is the primary hub of migration in Peru; by 2017, a net number of close to 2.8 million lifetime migrants had moved there. Both cities have doubled their inhabitants since the 1980s; Lima has become home to about a third of Peru’s population or 10.5 million people, while Huancayo has reached 0.46 million inhabitants (INEI 2018c). Both cities have witnessed attendant urban restructuring and problems related to infrastructure, basic services, social cohesion, and livelihoods (Carpenter & Quispe-Agnoli 2015; Haller & Borsdorf 2013; Ioris 2015; MML 2021).

Among other drivers, migration is also a traditional diversification strategy for subsistence farmers in the Sierra to anticipate or react to hazards or crop failures in general (Heikkinen 2017; Perez et al. 2010; Sperling et al. 2008). Such migration is especially frequent after abrupt and rapid-onset hazards, whereas gradual climatic changes, such as droughts, can make migration less likely because people tend to invest first in local adaptation efforts (Koubi et al. 2016). When affected by environmental change, it is often the young, poor, and males who leave, at least initially, because they tend to lack livelihood options in-place, sufficient quality land, and support systems or risk-sharing strategies (Crespeigne et al. 2009; Lennox 2015; Lennox & Gowdy 2014; Milan & Ho 2014; Sperling et al. 2008; Wrathall et al. 2014). In some areas, climate impacts add to but remain behind other main drivers of migration, such as demographic and aspirational changes in high-altitude pastoralist communities (Alata et al. 2018). In other zones, migration is strongly driven by frost, freezing, and hail (Cometti 2015a; Sperling et al. 2008), cold spells (Crespeigne et al. 2009), and coping with floods (Oft 2010; Sperling et al. 2008). One survey in five highland communities identifies “environmental problems” as the second-most important motivation for migration (Cavagnoud 2018). The two hazards of focus here—glacier retreat and rainfall changes—also figure among the drivers of migration. Rainfall changes, drought, and food security stressors are key drivers of migration in various studies (Altamirano Girao 2012; Oliver-Smith 2014; Sperling et al. 2008), especially for agricultural communities in the high mountains (Milan & Ho 2014). Migration is used both for short-term coping with and long-term adaptation to these hazards (Oft 2010). When glaciers retreat, many variables shape adaptation decisions. Smallholders rely more on migration in later stages of retreat, when peak flow has been crossed and water levels have sunk during the dry season (Wrathall et al. 2014). Family members—often young adults—are asked to migrate to cities to send remittances back home to support their families. Often, rainfall changes and glacier retreat jointly drive migration of small-scale farmers (Altamirano Girao 2012; Altamirano Rua 2014; Cometti 2015a, 2015b, 2018; Orlove 2009). Low stream levels due to glacier retreat also harm pastures and complicate herding for pastoralists, making migration of these already mobile people to other regions more likely; at most, some of them may be able to stay in the Sierra during the rainy season (Orlove 2009). Glacier retreat can also drive migration of tour operators and their employees who lose income opportunities due to deteriorating environments (Altamirano Girao 2012; Oliver-Smith 2014).

Simultaneously, people’s reasons for staying despite such hazards are diverse. In one survey in the highlands of the Lima Region (the home to V1 and V2), many people were strongly affected by hazards, and every second respondent had considered migration in the five preceding years but stayed (Adams 2012). Most chose to remain because of high levels of satisfaction. To a lesser extent, respondents stayed because they feared or were not interested in leaving or had obligations that tied them to their communities. The smallest share of respondents who stayed did so because of resource constraints. The study highlights that instrumental and affective bonds can bind people to places that are already imperiled; these bonds continue to be a strong link even when climate change increases the risks for these places. However, thresholds of place satisfaction may be crossed eventually.

The climate projections discussed above may have several consequences for future migration patterns in Peru’s Sierra. Emigration is a common trend from the highlands and will likely be amplified because of rising climate impacts and demographic changes that increasingly disrupt the villages lifecycles. Available models indicate that water scarcity driven by glacier retreat could continue to amplify emigration, especially for less resilient rural households and people born in the cities in Peru’s highlands (Magallanes 2015; Milan 2016).¹² The projected warming could create high crop and animal losses and food insecurity risks, which could coerce migration. Rainfall projections are uncertain, but glacier losses will certainly be extreme in all emissions scenarios and raise water scarcity especially during the dry seasons. This glacier loss could result in more seasonal migration during the dry seasons as an attempt to spread risks to livelihoods and generate remittances. Conversely, where dependency on glacial meltwater for agriculture and glacier-related tourism is high, losses of glaciers could also lead to gradual settlement abandonment and conflicts between water users. In addition, people derive a sense of well-being from the many non-economic ecosystem services where they live, yet climate impacts are threatening these services. Their degradation could drastically decrease place utility and thereby shape the migration decision processes of more people than usually assumed (Adams 2016). Finally, growing glacial lakes will pose risks of outburst floods and could spark relocation discussions in several highland cities. After this contextualization, I examine climate (im)mobilities from and in V1 and V2 in detail in the next section and explain how the affected people evaluated the well-being impacts associated with these (im)mobilities.

2 Empirical Results

In November 2018, I interviewed eleven people (4m / 7f, see Table 5.1) affected by climate hazards within the villages of origin V1 and V2. The age range of the eight migrant family members and three returned migrants was 18–81 years with an average of 55 years. Because the age structure in both villages was inclined toward older adults, I also convened a focus group of twelve pupils aged 14-16 in V1 (3m / 9f) to represent the views of the small young population appropriately. Some older dwellers had only primary, later generations usually secondary education. Having lived for some days in village V1, I also took field notes, observed participants, and engaged in less formal conversations and community meetings with various other farmers (see Figure 5.10).

Table 5.1 Basic data on interviewees in the mountain villages V1 and V2

Figure 5.10

Impressions of the field work in the highlands. (Note: Community meetings with farmers in V1 (left, photo by the author) and in V2 (right, photo by the Mountain Institute))

Table 5.1 details that most interviewees were farmers who mainly worked in smallholder crop cultivation. Some had additional livestock farming in the upper elevations, including sheep, cows, and few alpacas in V1. In the basin’s sections where the interviews took place, subsistence farmers mostly rely on rain-fed agriculture.¹³ In V1, people additionally use basic flood irrigation techniques with water from an adjacent river fed by glacial meltwater and rainfall. In V2, only a small spring was left for limited irrigation (see section 5.2.1 for more details). Farmers produced typical Andean transient crops such as corn or potatoes and cultivated pastures such as alfalfa. Their fields extended across various ecological floors and were maintained in a rotational system. Given their isolated locations, interviewees had limited income sources. Even the few ones who offered services, such as municipality staff, still sowed their crops and had a few animals, similarly as most retired villagers.

To gather background material, I also talked to several experts at the village level, including the two respective presidents of the farmers’ associations (comunidades campesinas). Further discussions were held with the mayor, the alderman (regidor), the peace justice, the health post worker, and the secondary school director in V1, as well as the teacher in V2. For further context, I visited two additional villages in higher elevations to discuss informally with farmers and local experts.

As a next step, I traced and interviewed 20 migrants (11m / 9f) from V1 and V2 mainly in Lima and Huancayo (Table 5.2), often at sports events organized by migrant hometown associations (Figure 5.11). Migrants were between 18 and 77 years old with an average age of 47. All had left as young adults, and some had migrated decades ago. Three of the migrants had returned to the Sierra after having spent time in the cities. Most interviewees provided services in the cities and remained in informal or low-wage positions, whereas a small number had entered better-paid jobs (see section 5.2.3: Decent livelihoods). All migrants identified as mestizo and spoke Spanish.

Table 5.2 Basic data on migrant interviewees originally from the mountain villages

Figure 5.11

Impressions of the field work in Lima and in Huancayo. (Note: The photo on the left displays a sport event in Lima organized by the migrant hometown association; the interviews partially took place at these regular sport events. On the right, the head office of the Regional government in Huancayo. Photos by the author)

In addition, I interviewed several experts at the municipal level in Huancayo for background information, including the responsible officials for rural development, urban development, communal kitchens, civil defense, and environmental management.

2.1 Climate Change Dimensions

In the interviews, young and old villagers alike reported a wide range of climatic changes, often talking spontaneously about them and without specific prompting. Figure 5.12 exposes that the main observed impact was water scarcity related to glacier retreat and rainfall changes, and views on these changes converged across sex and age groups. Impacts were felt on crops, pastures animal and human health, as well as on people’s lifestyles and traditions. Villagers said that climate impacts were already undercutting livelihoods in the zone and anticipated that they would continue to increase. While many farmers linked these alterations to climate change, few knew about underlying mechanisms. Among the interviewed young pupils, for example, only a small number correctly connected climate change to greenhouse gases. Especially the young people also observed additional non-climatic stressors such as contamination, deforestation, overgrazing, solid waste problems, and soil degradation.

Figure 5.12

Hazards affecting interviewees in the highlands. (Note: Depicts the percentage of interviewed affected people from the highland villages who mentioned different types of hazards affecting them at least once during the interviews. Created by the author)

When asked about changes in their environment, most dwellers first mentioned the drastic losses of their nevados (snowcapped mountains), which have caused gradually rising water stress in V1 and severe scarcity in V2. The glaciers, firn, and snowpack in the higher elevations—connected to the villages by lagoons and rivers—are key for people’s identity and for their water supply, especially in the dry period. In V1, the firn used to almost touch a river close-by the village. However, respondents claimed a surface reduction by at least half since the 1950s, which they explained by reduced rainfall and warming that prevented snowfall. As the village’s alderman observed:

The water has also gone down, don’t you see that the mountain range is already thawing, the weather is not as before, that hail no longer falls now, it rains pure water. … Yes, there is not much snow, pure water no more, that’s why the snow does not accumulate. … There is no longer much water we can use to irrigate …. (V1-1)

As a result, the once abundant meltwater has decreased. Although the glacier retreat was still manageable, it affected irrigation, production stability, and crop yields in the dry season. Moreover, water scarcity and declining pastures have made livestock raising less profitable, so that farmers have had to reduce their herd sizes and the remaining healthy pastures are increasingly overexploited.

Although adjacent to V1, V2 suffered from more severe impacts of a later stage of glacier retreat. The village was once fed by canals to artificial meltwater lagoons in the higher elevations. Yet, villagers noticed how glacier retreat starting in the 1970s had gradually reduced meltwater from the 1990s onward until ten to fifteen years ago, when the meltwater disappeared completely. All farmers openly worried about the resulting “water crisis” (V2-1). The president of the peasants observed that,

The first thing that the global warming of the Earth has eliminated was the mountain range, the snow-capped mountains… It dried up completely. [Before], the water just came from that hill, the one that is here … Well then, it was nothing left … Little by little, with the heat it melted ... The water was no longer advancing… The land has become dead and there is nowhere to put crops. (V2-3)

The loss of meltwater has gradually reduced crop yields. Villagers had to switch mostly to temporary rain-fed crops and pastures. The remaining water from a small spring was scarce even for the now much reduced village size: “As I say, we are few and for that it suffices. Of course, it is not enough, and sometimes we are struggling …” (V2-2). Farmers reported misery especially during the dry season. The glacier loss has made the land desertic and people who once had sold their produce now could only produce for subsistence. Animals suffered from thirst and sometimes died because of the lack of pastures, as for example in 2017, a sad event in people’s collective memory.

In both villages, respondents mentioned rainfall changes second-most often. Since 2000, they have seen more erratic rainfall seasons, with both unexpectedly more and less rain: “Right now, here it has changed a lot and it is varied, … when it should rain in its period, it doesn’t” (V1-1). It is “a total change …, before [the rain] had its periods, … now it rains anytime” (V2-7). Farmers reported that they sowed less because, “Sometimes the rains delay, and the food is spoiled; in previous years that was not seen” (V1-16). They also described diminishing rainfall quantities but greater intensities, which aggravated crop losses. Young people were saddened that the rain was occasionally so scarce that animals died due to a lack of fodder. Many interviewees also noticed more warming and heat periods that have become “unbearable” (V1-16). The heat affected health and crops and resulted in higher irrigation needs while water resources have kept declining. Moreover, cold waves have caused diseases, animal losses, and crop damages, and people noted more weather-related pests and diseases.

Adaptation efforts and options differed considerably between V1 and V2. In V1, where glacier retreat was still less severe, a local NGO project supported people in conservation and water harvesting techniques to optimize their resources and adapt to declining water availability. Due to the project, farmers perceived some progress and that they had learnt new agrarian skills. Yet, some of the more educated villagers worried about the lack of external help, expertise, technology, and market networks to mitigate the declining yields. Looking to the future, farmers worried that surviving only with erratic rainfall for irrigation would be difficult. For example, two well-educated men worried that their glaciers might disappear over the coming two decades. A woman in her mid-70s voiced concerns about the disappearing water and the lack of good governance, fearing that V3 may turn into a “ghost village” in twenty years: “I don’t see how it’s going to improve, how? … What will we do if there is no water? All lies. There is no water, no” (V1-3). Village leaders were also aware that they were still better off than V2 but could face a similar future. Nonetheless, some of the younger residents were hopeful that people were finally waking up to the change, as illustrated by the election of a young agronomist as the new mayor. Some put hopes for adaptation into unrealistic ideas, such as painting the glaciers with sun-reflective color; such hopes seemed to be a sign of cognitive dissonance or denial that the situation could become as severe as in V2. A few other respondents downplayed the possible effects and completely discounted the probably challenging future for V2.

Foreshadowing possible future challenges for V1, the situation of V2 demonstrates how difficult adaptation to a complete loss of glacial meltwater is for poor, remote, and marginalized villages. Farmers made some adaptation efforts, such as changing cropping techniques and cementing the water canal to maximize the decreasing streamflow to the village. Yet, the attempts failed because the farmers lacked needed finances, resources, and skills. Given their remote and marginal location, they had no market access or diversification options; the only attainable adaptation was reliance on rain-fed agriculture and the limited water provided by a local spring, which offered less cropping and pasture potential for substantially fewer people. Bad governance and a lack of voice also limited adaptation possibilities. As an administrative annex, farmers lacked a proper budget and needed to ask the district authorities in V1 and the Ministry of Agriculture in Lima for help. However, the promised support never arrived, and interviewees suspected that the authorities did not care enough about their misery or did not possess enough resources. Several villagers perceived themselves “in a system of abandonment by the government”; they alleged corruption and that the authorities “cheat the people” and “only want the money” (V2-1). Eventually, farmers resigned, “praying to God for rain” and hope (V2-7). Several migrants, discerning the situation from afar, hoped that adaptation might still be possible. Yet farmers in V2 did not seem to hold any hope for a better future absent water and resources or state support to address the crisis. For example, a woman almost 70 years old asked:

What state have we come to? To a very sad life, a water crisis. Why do we lose both our youths and our children? Because there is nothing good here for how to survive. There is no work, there is no water! We, the older adults, are the only ones to stay in this village. There are no more young adults, we cannot sow anymore given the lack of water. We only have one life but are in a time of crisis in [V2]. (V2-1)

In both villages, farmers observed migration related to these changes (see next section 5.2.2). Farmers also reported more quarrels around water than in the past. Some considered more conflict possible in a future with less water, but beyond the call for more climate adaptation measures, local-level institutions or mechanism preparing for such conflicts were missing. In the depopulating village V2, the community seemed to partially disintegrate into disunity and withdrawal with quarrels among themselves, which were driven by the emigration related to the water scarcity (see section 5.2.4).

2.2 Migration Dimensions

The cases of V1 and V2 offer a detailed view of migration decisions in the context of slow-onset hazards and their interplay with other drivers. Interviews indicate that temporary and permanent migration have been common for long in the villages, whereas transhumance or commuting to urban areas is rare to the spatial setting. In the second half of the 20^th century, agrarian crises and violent conflict triggered substantial emigration. Later, some migration continued for social reasons, to establish a family or join relatives. In addition, more villagers have left yearning for a new urban lifestyle instead of rural agricultural lives. Yet, the major driver of migration has been the lack of jobs and income possibilities; traditionally, many youths have left to find urban jobs, aiming for higher and more stable incomes to support their families, and sometimes returning later to their birthplaces.

However, over the past two decades, climate impacts have exacerbated the economic situation in V1 and V2 and raised migration rates. In V1, the village alderman observed that more peasants have migrated for work since the glacial meltwater has decreased and the rains have started to be more erratic ten to 15 years ago, because there is “not enough to provide for yourself anymore” and farmers cannot “produce as much anymore, so people leave to search for jobs” (V1-1). Around five families were leaving the village every year because they could only sow few crops. After V2 had completely lost its glacial meltwater for irrigation, agricultural production went down drastically, and much more young adults than usually had to leave to the cities for work, while fewer returned. A woman in her late 60s mourned that they “lost their youths and children” who “cannot survive here anymore - there is no work or water to sow anymore” (V2-1). Migrants remarked that they had had to leave as there was “no more life” or work (L-25), and, “Farming did not render enough, there is nowhere to harvest, and livestock does not give good results anymore” (H14). A migrant in Huancayo explained:

People left as the animals… had nothing to eat, they got thinner, and died… The village is depopulated, there are no more people, a few no more. All are leaving as there is no more water. A lot of dryness, for this reason the young people have left… In some rooms are still the older adults, who are now dying. (H13)

Increasingly, water has become a main “motivation for people to leave … since the land is all powder” (L23). The president of the migrants’ hometown association in Lima indicated:

Principally, people left because there is no water anymore, there is no more future in the Andes… They cannot produce anymore… Mainly, they came for work because in the village, there is no income, there is no work, and there is also no water. (L15)

These climatic drivers interacted with other economic, educational, and lifestyle drivers. For example, a final economic push for migration was when a small mining site near V2, in which ten to fifteen young dwellers used to work, closed some years ago. The general lack of jobs also combined with a lack of educational opportunities, the second-most reported driver. While V1 had a secondary school, higher education was only possible in the cities and continued to pull migration. Some feared that climate impacts might indirectly exacerbate this pull: as more people leave the villages (partially due to climate impacts), schools might close soon, which could propel further migration of families with children. Several other feedback mechanisms also accelerated migration. First, urban migrant networks have reduced the social and financial costs of moving for migrants, while remittances and stories of urban lifestyles provided additional pulls. Second, migration has deprived the village of workers required for manual agriculture; as a result, rising salaries for the limited workforce have reduced agricultural profitability, and thus made migration more likely. Third, depopulation removed social resources that were key for place satisfaction. A migrant from V2 said that he had wanted to stay but moved for the good of his son, who aspired to leave as most of his peers had already gone.

While advancing glacier retreat has thus been a mostly emerging driver in V1, in V2, complete glacier loss has been a major driver of an exodus. Among the remaining villagers, migration aspirations differed between generations. Most of the older remaining dwellers in V1 wanted to stay. Some were voluntarily immobile and stayed as they enjoyed the calm natural surroundings and their customs. Others were acquiescently immobile, as they worried to fall sick in the cities or could not grow used to city lives, which they linked to physical insecurity. By contrast, a few villagers felt forced into immobility: for example, a poor woman in her 80s suffered from serious health issues and aspired to leave V2, but her two migrated children in Huancayo and Lima refused to take her in. Conversely, young villagers in V1 and V2 usually strongly aspired to leave and considered migration as their best option, although they liked the quieter highlands and were aware of the urban stressors. The construction of a road, better access to markets, internet, media, and mobile phones have raised the knowledge and attraction of city life. This city life pulled those who preferred jobs outside agriculture, especially “The youths who do not want to live off the fields anymore, they dislike it” (H-13). In the past, pupils used to work in the fields after secondary school, but nowadays, all interviewed pupils who were about to graduate desired to migrate to continue their studies, find off-farm jobs, and move out of poverty (voluntary improvement migration). Many established migrants also confirmed retrospectively that they had had strong aspirations to leave. However, other recent migrants from V2 had moved against their will to survive (anticipatory, forced survival migration); they missed their homes and still felt the pain of leaving and seeing their village decaying.

Few people from V2 had sufficient resources to implement their migration decisions fast. Once the glacial meltwater had disappeared, the few relatively better-off households with resources to settle elsewhere left first, whereas most poorer households faced limitations and depended on the support of relatives. Family networks helped with information, finances, food, and housing to the extent they could. However, poor families could only offer limited support for migrant members. Related, initial settlement often depended on where relatives lived who could host newcomers and was therefore dispersed. Gender and age also played a role: when V2 lost its glacial meltwater, young men tended to leave first to the cities, while their wives and children, if applicable, joined later. Overall, many migrated at a young age. The main destinations for migrants from V1 and V2 were Huancayo and Lima.¹⁴ While many of them went directly to these cities, some first moved to the surrounding, larger villages or provincial towns, and then eventually on to the Regional and national capitals. Since V1 and V2 are in remote locations with difficult access, migration was often permanent.¹⁵

Few migrants returned permanently; only some younger migrants aimed at gaining new skills in the cities, hoping that they would enable them to return and help their home villages. Nevertheless, close to all migrants preserved emotional, social, cultural, and material links to their home villages. They routinely visited for traditional celebrations or to see relatives. For example, one woman in her late 70s had moved to Lima at the age of 16 to study for two years, then went back to V2 to work, and eventually moved to Huancayo when she was 50, where she opened a small shop, but still returned sporadically to V2. Visits were a two-way street: people staying in the villages also visited their migrant children in the cities, for example to skip the rainiest highland months. Many migrants kept their highland properties, and a few admitted that not selling them to the young adults in their birthplaces may break the villages’ life cycle and render agriculture more difficult. Some soils were already eroding. Moreover, migrants in the cities sent a limited amount of remittances (see section 5.2.4). Finally, translocal links were also institutionalized through hometown associations (HTAs). Both V1 and V2 had founded associations decades ago, with more than 200 migrant members from V1 in Lima and 80 in Huancayo. The much smaller village V2 has around 50 members in these cities. The organizations provide mostly non-material support to newcomers, who can participate in sports games, celebrations of Andean traditions, and social media communication. The presidents of the HTAs reported that they had tried to help V1 and V2 economically, providing technical support, infrastructure investments, and food and income support. While their help has not alleviated the impacts of glacier loss in V2, it might back adaptation to future impacts in V1, if channeled suitably.

2.3 Well-Being Dimensions: Migrants at Their Destination

In this section, I explore how migration from the two highland villages threatened by water scarcity has affected people’s well-being. I apply the well-being axes explained in chapter 2: development from a secure base, a space to live better, and social relatedness (objective well-being, OWB), as well as subjective well-being (SWB). The analysis is based on migrants’ retrospective assessments of how their situations have evolved after arrival in the cities. Future work on intergenerational processes would be desirable, which could not be examined in detail due to due to time and resource constraints.

2.3.1 Development from a Secure base

2.3.1.1 Decent Livelihoods

Settlement in the cities affected people’s livelihoods differently, although many migrants from V1 and V2 had had similar starting conditions. Most arrived after leaving school or with agricultural skills not in demand in cities. On one end of the spectrum, the majority and especially young migrants struggled to make a decent living. They worked long hours in informal, unstable, and low-paid jobs to survive—often in addition to family and housekeeping obligations—but were unable to advance. In some deprived situations, children had to work to support the family. Typical work included ambulatory sales, construction, day labor, manufactories, or small trading. Competition was high. For example, a young woman had left V2 after the glacial meltwater had disappeared and experienced a “very shocking start” in Lima; only after years, she had gotten “used to it … more than anything else out of necessity” (L-25). She and her husband worked in dressmaking and transportation, noting that

… [Y]ou must work here, both spouses have to work, otherwise the money is not enough… Yes, all day, so much sun, and sometimes all night, and there is a lot of competition here now… Yes, it is not easy, we must continue working, that’s it. You must work hard, to survive more than anything else…. (L-25)

According to the presidents of both hometown associations, migrants coming to Lima and Huancayo ended up two times more often in precarious than in decent jobs:

Look, here the people who arrive have mostly no income, it is just enough for getting by… [T]hey do not have good jobs… Of course, the vast majority are small businessmen, as well as street vendors, they work in factories ... Dad and mom both work and with that they sustain their household…. (L-15)

Municipal and NGO experts agreed that although a few migrants improve after moving to Lima and Huancayo, many suffer from multiple deprivations and frequently fail to escape poverty.

On the other end of the spectrum, a few migrants did profit from their transition. Especially some of the older interviewees who had arrived decades ago had tapped into urban education opportunities to switch into higher wage positions in services, such as accounting, administrating, engineering, and teaching. For example, a migrant now in his early 70s reported about his strenuous but eventually successful journey in Lima, and the dual burden of working precariously while studying:

I left [to Lima] at the age of ten ... and finished my primary school, but as my father had no financial solvency, I could not study, ... as my father was a farmer ... I stopped studying for at least six years… I worked in the field, I even worked in construction… My life has been work ... Ever since I use reason, I thought I cannot be like this, I go to Lima. I had a brother who was here ... When I was in Lima, I first started working as a waiter ... The next year, I started studying, ... for six years I studied at night ... I worked during the day and at night I studied… I worked in factories, later as a baker, that’s how it was, tough sacrifice… [Now], look, I am a professional, I studied engineering…. (L-15)

Some also improved their incomes without studies. For example, one female migrant first was a street vendor, then took a credit from relatives to start a small market stall, and eventually gained well. Yet, income gains compared to prior rural lives did not automatically make for a decent live. For example, a man said that working with his disability had been impossible in V2, whereas in Lima he “dedicated myself to sewing, and with that I am living more or less because I do not want to be a burden on the family”. Despite the improvement, his livelihood in Lima was still “very tiring, a lot of stress because we really can’t stop working, as we have to survive in some way” (L-24). Intersectional factors, such as health status, thus also shaped livelihoods. As another example, most male migrants provided services, while women were said to work more often in housekeeping and lack individual incomes.

2.3.1.2 Health and Food Security

On the negative end, financial resources were the major constraint to access health services. Various poor migrants became more food insecure in the cities. One claimed, “Everybody wants to go … [to the] invasiones and cerros [human settlements on the city hills] although they do not have enough to eat [there]” (V1-4), whereas in the Sierra, people had produced their own food. A few better educated migrants were concerned that indirectly, climate impacts also affected the prices for and quality of agricultural products, which could in turn worsen urban food security. The alderman in V1 stated:

Everyone goes to the cities but there is no longer space to live, … and water is going to be scarce, and in the cities, they will be certainly worse. I think that rather with time, people will return to their land… in Lima, on a daily basis, [basic necessities] have to be bought. Instead, here we sow a piece of land, and we already have to eat all year round. (V1-1)

Especially recent migrants suffered from a lack of basic services that might contribute to healthy lives, such as water, sanitation, and drainage systems. These gaps were often due to the lack of landownership or housing titles. Moreover, urban pollution and contamination worried many migrants and caused diseases for some. For example, one migrant reported that “we are crying because of the contamination here” (H-14). Another young migrant strongly suffered from the economic and social challenges in Lima, exhibited mental health issues, and, after getting asthma in the humid and polluted city, eventually had to return home. Similarly, various migrants reported high stress levels due to long working hours, intense traffic, and multiple parallel burdens through jobs, education, or caregiving. Younger migrants also observed ill-being due to the separation from their families. On the positive end, a few migrants improved their health in the cities, if they could afford to access the better health system and thus to receive the specialized attendance they required.

2.3.1.3 Educational Opportunities

Migrants’ educational opportunities frequently improved in the cities, but sometimes at a high cost. The cities offered more opportunities to study, which, constituted strong motives for migrants to stay, according to parents who hoped their children could become professionals one day. Vocational training and evening schools enabled migrants to work during the day and still improve their skills. For many, this setup was a necessary but challenging one, given the relatively high costs for urban education. They described the triple burden of working during the day, going to school in the evening, and studying during the night, which led to health problems and self-exploitation that hindered learning. For example, one young man had moved to Lima for education, but pollution and the stress of studying while working in a gas station to pay for his expenses forced him to return:

Unfortunately, … we have to work to start studying… The truth is that at first it was shocking… [S]ometimes, I had to work at night and study during the day, in class I fell asleep. Sometimes, I attended and sometimes I didn’t. This way, I suffered a couple of months… [Studying] is costly, life becomes difficult for a poor person to progress… Then, because my finances were not enough, I had to leave my second semester… The minimum salary at that time was 850 soles and my institute costed 500. Paying for food and tickets, everything was gone…. (V1-4)

He also felt that the generally worse education situation in rural areas led to a disadvantaged start for migrants from the highlands compared to city natives.

Other migrants spoke of tough sacrifice, but a worthwhile one. As one example, a male migrant who had gone to Lima at the age of ten to study had to return as his parents could not afford the costs. Several years later, he went to Lima again, working precariously during the day and studying at night, but through “tough sacrifice”, he progressed: “[Now] look I am professional, I studied engineering”. While he felt content, these taxing years also “have marked” him (L-15). Another female migrant remembered the “hard” and “sad” separation from her family when she came to Lima to study, but in retrospect, saw the greater good of her studies, which enabled her to become a teacher (L-20).

2.3.2 A Space to Live Better

2.3.2.1 Adequate Housing

Migrants’ quality of housing depended on their available resources and social contacts. A few of them were fortunate to move in with relatives who lived in decent houses. However, most of the poorer migrants needed to settle in the outskirts of the cities, in new and frequently irregular settlements:

It was not easy, if you don’t have a place to stay here, you don’t have an income or you don’t work… Over there, they have gone, outside of Lima because here in Lima, they have not been able to get homes. Further on the side of the hills, in asentamientos humanos [new or irregular settlements]. (L-15)

Many needed to move into crowded homes of relatives or rent precarious houses at a high price: “It is not easy in the city, [even] to have a small house here, it costs much to rent it” (H-13). Others had to build their houses from nothing, and some achieved gradual upgrades. For example, a woman in her 70s reported how today, she felt content with her home, but it had required years to obtain a small terrain and only then, she said, “I was able to make my little house little by little … Years, years, and years we have built, until finally now, we have the little house finished” (L-21). Financial resources, networks, and titling efforts were key for achieving progress regarding housing and landownership.

2.3.2.2 Basic Services

The principal determinant of migrants’ access to basic services was where they could afford to settle in the city. The president of the hometown association of V2 described how fellow migrants in Lima

are quite dispersed, they are in different places, some of them who earn well are in urbanizations… Many live in places, asentamientos humanos, invasiones [irregular settlements], well away from the city center ... You do suffer there, it takes several years for them to have electricity, water. (L-15)

The many poorer migrants who moved to new and irregular settlements lacked access to basic services, and many failed to receive titles or to have water, sanitation, and drainage systems installed for years. In the meantime, migrants had to buy water from car distributers or walk long distances to distribution points. For example, one female migrant in her early 40s moved to an invasion that she described as a “desert”; lacking a property title, she had to petition and wait for fifteen years until the settlement was eventually regularized and she and her family received access to water and other services, whereas “those who live on the hill still receive water from the aguateros [small-scale water providers]” (L-19). Another woman now in her late 70s described the struggle upon arrival:

[My children] looked at me as I suffered and there was no water. We carried water from afar, from the park, far, far from the house, to cook at home here… In a wheelbarrow I have carried it, with a bucket. (L-21)

In Huancayo, unlike years ago, even the few better-off migrants had only water by the hour and worried that climate change would increase water stress: “There is no water, here we suffer” (H-13).

2.3.2.3 Pleasant Surroundings

Most migrants depicted their urban surroundings as dissatisfying. They perceived great differences between their prior rural villages, with fresh air, clean water, appealing natural landscapes, and recreational areas, and their current arid, hectic, and congested city neighborhoods: “We lived there, where my sister was, it was like a desert … all just gasoline, smell of smoke” (L-19). With the rising saturation of the city, pleasant areas have become less accessible for poor migrants. This view held across age groups, but some older migrants seemed especially dissatisfied; they described how Lima had become more polluted and congested with traffic. Even those who enjoyed urban commodities stated they missed the tranquility of the countryside, including the farm life with physical outdoor work. Only few migrants in relatively wealthier areas of Lima were more content. Some also felt more satisfied with their urban surroundings since the climate was warmer than in the highlands.

2.3.2.4 Safety from Hazards

While migrants had left rural areas that faced climate hazards, such hazards also threatened Peru’s cities. In Huancayo, cited concerns included flooding from torrential rainfall and from glacial lake outbursts. Migrants in Lima struggled with the temperature difference to the highlandy and the city’s seasonal humidity. Water quality and quantity—one driver of migration from V1 and V2 to the cities—also constituted a main urban concern. Migrants worried that Huancayo and Lima partially depended on glacial meltwater that would eventually disappear, thus exacerbating water stress. In Huancayo, unlike years ago, many migrants had only water by the hour. Experts explained that diminishing water resources, the slow exploration of new sources, and simultaneously increasing demand made higher water prices and more rationing likely. Given this situation, a migrant asked:

Eventually, where will people go? It will start in the small villages, but from Huancayo, where will they continue to go, as there will be no water? From what will we live if agriculture decays? (H-14).

2.3.2.5 Security

Most migrants lived in “movidos [dangerous]” zones of Lima (L-15), which are “very hectic, with a lot of crime” (L-17). As one migrant said, “There are so many thieves and criminals and all. They all come to Lima and the big cities” (H-13). Established migrants described that insecurity had increased. Similarly, a returned migrant in V1 worried about her migrated children because

In the year 1980, Lima was not like now, it was calmer. At least for me it was nice to be in Lima, there was not much violence, there was no robbery like now… [N]owadays, there is too much theft, it is already too much, too much in Lima. (V1-12)

Few migrants were able to move to more secure areas or described that the security in their neighborhoods was improving. For example, two migrants felt lucky to live in a place with “sufficient security”, “a little bit more peaceful than other places” (L-27; L-25). Similarly, a carpenter in his early 40s, who had come to Lima in the 1990s, observed that, “Where I live it is quiet, I hardly see any theft, or it goes unheard, but in the nearby neighborhoods, yes there is a lot of crime” (L-26).

2.3.3 Social Relatedness

Several migrants struggled with the separation from their relatives. Especially young migrants often lived with urban relatives they seldom knew well; one of them lamented that, “Indeed, a separation from parents is always sad” (L-16). The initial lack of support networks caused hardship:

There [in Lima], I lived alone. Sometimes with my uncle, but it’s not the same, it’s not like mom or dad ... [I felt] alone, alone, I felt that everything was coming over me. Away from home, away from grandparents ... It is difficult to be alone and not have the support, no one to talk to … I thought of leaving everything. I was alone ... When you’re away from home, it’s shocking ... There, no one gives you a glass of water if you are ill, if you are sick. I don’t know, they don’t care, you must look after yourself. Even family members do not look after you, much less strangers. The change was quite hard. (V1-4)

Some migrants felt that unfamiliar social codes in the cities also hampered relationships initially.

However, with time, many migrants had gotten to know partners and spouses in the cities, founded families, and started to raise their own children, with whom they usually described good relations. For example, a now retired migrant in his late 50s said his social relations were “slightly better” in Huancayo than in V2 because he had married and founded a family (H-8). Many reported “good connections with neighbors” (L-26) and peers, partially facilitated by hometown associations.

2.3.4 SWB Dimensions

2.3.4.1 Emotional Balance and Cognitive Satisfaction

Many migrants described sadness and stress especially at their arrival, due to the separation from loved ones and their cherished highland environments. In addition, the dual burden of working and studying in the cities to fulfill migrants’ own and their family’s expectations caused anxiety in several respondents. Migrants were anxious in some cases despite improvements in material conditions. For example, a migrant in his 50s described the past sacrifices of working as a mechanic while studying during the night. Although he had met his wife in the city, had children, had tapped into educational opportunities, and has improved living standards, he concluded that, “Well, I wasn’t happy… [but] a little more anxious” both at arrival and later in his life (H-8). He did not grow accustomed and considered return, claiming that various other migrants were in a similar state. In addition, many described tension and stress due to precarious livelihoods. Various migrants were also frightened due to urban insecurity. Moreover, some felt fatigue due to the city environment: “Life is more tiring now ... very tiring indeed” (H-8). A few expressed feelings of being overwhelmed, saying that they liked “almost nothing” in Lima (L-25). For example, a young student had strongly aspired to leave the highlands and had considered migration as his best option at the time but regretted his decision to migrate in retrospect. In Lima, without sufficient networks and finances, his income “improved a bit”, but he faced higher costs of living. Although he gained valuable skills, he “suffered”: “It was bad for me, not only in the part of trying to progress, but also in health ... To be honest, almost nothing makes me happy because I am full of problems”. He retrospectively assessed his urban situation with dissatisfaction, “Because I did not have the emotional support of my family”, and also felt “more anxious … much more, because of the surroundings and the social change” (V1-4). Migrants were also angry and frustrated due to corruption and governance failures, and stated, “Over there [in Lima], you must learn the reality of your country, where you live, the corruption” (V1-4).

On the positive end, the fewer migrants who had succeeded in leveraging educational and income opportunities were satisfied with their urban lives. A retired engineer reported that, “The life I live, well, I like it, it does not preoccupy me, I like having a cheerful life …” (L-15). Others in less well-paid but still formal jobs or their own businesses were also more satisfied. They were grateful for the chance to advance economically and that their children would be better off. In addition, those living in quieter areas reported more positive feelings. Moreover, new family ties and lifestyles influenced satisfaction; for example, when asked if she had gotten used to Lima, a female teacher replied:

Uy! We no longer want to know about my land [in the Sierra] … Due to the cold there, and we also don’t want agriculture anymore. We also don’t have family there anymore, everyone is here. (L-20)

2.3.4.2 Outlook on the Future

Many migrants who answered questions about their future expectations were more anxious than hopeful. The challenges back home in the highlands were one key concern. As one migrant said exemplarily: “I suffer for my land, for my town, I would like to have water so that my countrymen and the youth can return [to the village of origin] ... That makes me worry a lot” (H-13). Migrants “really worried” and felt “much more fear [than hope]” as “the water does not come anymore, and the mountains are drying” in the Andes, which, in their view, foreshadowed future problems in the cities: “The hills dry up, well, in some time. Where will we get water from?” (H-8). One worried, “From what will we live if agriculture decays?” (H-14). Others in the city seemed resigned given their own deprivation and economic fears, especially if in low-wage temporary jobs or informal positions. Partially due to such resignation, various migrants did not hold hope for themselves but only for their children’s advancement. This hope for the next generation was a main driver and motivation for them to endure hard work. By contrast, others expressed anxiety that the younger generation had few perspectives in the cities. For example, a migrant in Huancayo stated, “My daughter is 32 years old and so I say to her, ‘You don’t want to have a son?’ She replies, ‘For what, only so my children suffer?’” (H-8).

A small number of migrants who had objectively good prospects, religious faith, and personal resilience in the face of adversity were more hopeful. For example, a migrant in his mid-twenties worked as an administrative assistant in Lima and saw more possibilities in the city than in V2, where the glacier was lost: “Well, yes, there are more opportunities to have something, perhaps to progress more” (L-27). Some migrants retained hope despite “so many obstacles”, like a young man who had to return from Lima to his village after he had gotten sick and could not finance his studies anymore: “I do not lose hope that there is more, … that I can achieve my goals ... Faith and hope is the last thing I will lose” (V1-4). Another returned migrant recognized “the deprivation of the people” in V1, but still felt good about the future, “Because we talked and agreed that we must make a change among the youth … There is much potential here … [W]e’d like to return and bring in ideas” (V1-5).

2.4 Well-Being Dimensions: Stayers in Source Areas

While not the focus of this study, I briefly explore the well-being of families staying in V1 and V2 in this section. Migration affected the home villages by changing their sizes and compositions, alongside people’s economic, health, educational, social, and cultural situations, as well as their SWB.

2.4.1 Demographic changes

Family planning and rising emigration have decreased population sizes in V1 and V2 in a mutually reinforcing process, leaving many homes abandoned (Figure 5.13). Because families have significantly fewer children today than some decades before, the migration of teenagers more strongly affects population sizes. V1 had had reportedly about ten births per year until the 1990s but now, only one child was born per year, so that the migration of the fewer young adults could not be offset in the same way as before. At the same time, climate impacts (together with other drivers, see 5.2.2) have increased the flows from V1. As a result, V1 has gradually shrunk: between the 1960s and 80s, it had been allegedly home to more than 150 families yet in 2018, only 40 families with mainly older adults were left, who worried that the village might become depopulated. V2 was reportedly home to 60 to 80 families in 2000. Yet, as the glacial meltwater disappeared, more and more youths left. In 2010, only 25 families remained and by 2018, fifteen more had left. In contrast to previous decades, most migrants left permanently since they lacked water and jobs in V2. Because mostly older adults stayed behind, no more new children were born and the life cycle in V2 was disrupted.

Figure 5.13

Situation in the highland villages of origin. (Note: The photos present abandoned houses of migrants in V1 (left) and V2 (right). Photos by the author)

2.4.2 Economic, Health and Educational Changes

Economically speaking, emigration subtracted labor force from the villages but also generated some remittances. For V1, the economic effect was noticeable but still less severe than for V2. For example, one father in V1 reported how his children had migrated to Lima and how he struggled to compensate for the loss of their labor support on his fields. Conversely, in V2, the economic situation deteriorated severely because many residents had left. Mostly older adults stayed behind, and they had limited resources or physical abilities to make a decent living. These stayers observed that glacier loss and ensuing migration had ruined their economy. For example, a woman in her early 80s remained back alone in V2 in deprived conditions and with serious health problems, mostly dependent on the few remittances sent by her children in the cities. Migrants from V1 and V2 now living in cities also worried about the economic downturn, constrained resources, and decaying houses in their home villages. The migrants and hometown associations from V1 and V2 would send support to their relatives at home, but not enough to buffer the lack of water and labor force.

In V2, the remaining mostly older residents suffered from typical health problems of their age, yet climate impacts and emigration have aggravated these problems. Since the village was shrinking, health workers would visit less regularly so that inhabitants lacked needed medical care. Hunger and despair, loneliness, and social disintegration further threatened people’s health. Conversely, in V1, emigration had not yet had the same extent of adverse health effects. Because sufficient dwellers have remained in the village, it still disposed of a medical post with semi-regular attendance.

Emigration—in combination with other factors that have reduced population sizes in V1 and V2—has also affected education. In V1, the director of the secondary school worried that the service might close soon due to the low birth rate and ever fewer incoming pupils. For the few remaining younger families, such a closure would create additional emigration pressure, as the next schools are too far for daily commuting. In V2, the school was about to close with only one pupil left. The vanishing educational opportunities have exacerbated the already difficult prospects for the village.

2.4.3 Social Changes

Emigration also has affected social relationships at home. The effects were especially severe in V2, where the president of the farmers’ association worried about social disintegration. He observed increasing fights, grudges, and egoism among the stayers, which resulted, for example, in one episode where they failed to take the needed logistical steps to accept help offered by a local NGO. A female farmer in her late 60s remarked that the village had disintegrated due to the emigration and resultant losses of markets and community organization. People used to be caring and compassionate but now mostly remained within their houses. Remaining connected with migrated relatives was difficult because the remote village had limited mobile reception. Another woman in her early 80s deplored that her children would only rarely come to visit. She felt on her own and abandoned, missing the unity and mutual help she had experienced before the exodus. Migrants in the cities also worried about the decay of social relationships in V2 and the resulting loneliness of the remaining dwellers.

In V1, social ties still seemed more intact thanks to the larger remaining population. Village associations and religious celebrations continued to bring people together. Families stated that their children’s migration was usually emotionally challenging in the beginning, especially when the first child left. However, they also said that they had tended to grow accustomed to emigration and were able to adjust slightly easier to subsequent departures of further children. Many stayed in constant contact through mutual visits and their phones, the latter facilitated by a decent mobile reception.

In both villages, older household members lamented that migration had resulted in losses of traditional knowledge and customs. However, these changes may be part of a larger social change that migration has only accelerated; a focus group with pupils who still lived in the villages but aspired to migrate revealed that many were disinterested in maintaining such customs and traditions.

2.4.4 Changes in SWB

People in V2 felt depressed and desperate due to the decay of the village, the exodus of relatives, and the lack of social contacts. One stark example was a woman in her early 80s: her husband had died decades ago, and her migrated children were not taking care of her regularly. She struggled to walk, was blind and almost deaf, and lived by herself. She suffered hunger, seemed distressed, fearful, and exhausted, and lacked medical care. Moreover, most villagers were resigned and tired due to the loss of water and jobs. For example, a woman in her late 60s saw, “No future; rather we need, we live in great need… I will die this way, without seeing the water [again]” (V2-4). Another old woman asked,

So, how will I spend my life now, how will I eat when my belly asks for food when I go to sleep, as a poor person? ... I don’t know how I will spend my life, young man [sighs]… I’m already [in my early 80s], I can’t anymore, I can’t anymore. (V2-5)

Similarly, the president of the farmers’ association asked,

[W]hat are we supposed to do without water? The water is essential. If we do not have water, we will remain the same here, more and more people of the few ones left will migrate ... Even I think, ‘what to do if there are no resources to make a living?’. (V2-3)

In their misery, a small number of villagers held on to religious faith. For example, a farmer asked, “Do you see how the town is? A desert, it is a desert … There is no future, with nothing but the grace of God we live” (V2-1). Some farmers kept hopes that their children would be able to advance, such as one young man who stated that, “Well, for the future I only think about my son, to educate him, give him the best possible” (V2-2). One woman in her mid-60s, when asked about the next five to ten years, expressed a grim sense of humor by noting, “I will have died by then [laughs]” (V2-7).

In the less-affected V1, especially some of the older villagers perceived that migration had caused partial but limited social deprivation. They also worried about the well-being of their migrated relatives due to the insecurity and criminality in the cities. For example, a woman in her mid-70s felt saddened and alone after her children had migrated. After their migration, she was by herself, lacked support, and felt less satisfied with life, a situation captured in a traditional song (huayno) she recited:

[The relatives] only come a little while, and then they leave again; we remain behind, sad. What will we do? … As I say, I keep crying. There is a huayno that goes like this [sings]: ‘whoever leaves is happy seeing new horizons, whoever stays is sad, heartbroken [crying].’ It is sad, sad [crying]. (V1-3)

Yet, others described that they were more contented knowing that their children were advancing and becoming professionals in the cities. A few held hopes that adaptation to future climate impacts would be possible (see section 5.2.1), but such views were mixed with strong concerns about more hazards, deprivation, and economic insecurity. For example, a woman worried that in the long term, V1 would become a “ghost town, because there are no people” (V1-3). The focus group of young pupils, many of whom aspired to leave, were afraid that climate change would imperil their futures. One pupil was

Sad for what has happened. Our children will not have the privilege of living everything we have lived, in a healthy environment, the river by your side. It seems that our children will no longer have water, they will suffer many things. It is very sad, disappointing. (V1-16)

3 Discussion

These findings in the Sierra case study can contribute to the understanding of the nexus between climate (im)mobilities and well-being in several ways. In this section, I interpret the observed hazard-(im)mobility links and ensuing effects along the studied four well-being axes, situate them in the broader literature, and analyze relevant mechanisms and structural conditions.

3.1 Hazard-(Im)mobility Links and Pathways

The poor agricultural villages V1 and V2 have long histories of migration to cities driven by multiple reasons, similar to the historical trends in mountain areas worldwide (de Sherbinin et al. 2012). Traditionally, mainly younger people have left the study sites seeking jobs, education, and lifestyle changes, an age-specific pattern known worldwide (Millington 2000). Changes in family planning and family sizes have further reduced populations, as statistics indicate (INAIGEM 2018; INEI 2018c). Hazard-related migration specifically is also common in Peru’s highlands (Milan & Ho 2014; Vidal Merino et al. 2020). In the study sites, strong glacier retreat and rainfall changes have caused water stress and thereby gradually gained importance as migration drivers over the past thirty years. Simultaneously, these gradual climate impacts have made migration conditions more difficult and return less likely. The most direct climate effect on migration from the mainly ecosystem-dependent V1 and V2 was the loss of agricultural productivity; similarly, climate effects on economic migration drivers are also most salient worldwide (Foresight 2011). In V1, the mid-stage glacier retreat and rainfall changes have only gradually intensified preexisting migration because an NGO helped to mitigate the water stress to some extent. Likewise, many rural Peruvians tend to stay at the beginning of slow-onset climate impacts to attempt local adaptation (Koubi et al. 2016). By contrast, in V2, the fast and near-complete loss of meltwater caused an exodus because remoteness, poverty, lack of diversification options or necessary skills, and an absent state apparatus strongly limited capacities for adaptation in place. Increasingly, whole families have left for good, mainly to Huancayo as the Regional and Lima as the national capital. Such accelerating effects of advancing stages of glacier retreat on migration are also described in other Andean villages in Peru (Wrathall et al. 2014). In the future—when glacier loss will intensify, and adaptation limits may be reached due to insufficient state support, skills, technology, and market networks—V1 will therefore likely be at risk of a similar exodus as V2. Other studies confirm that because many people in highlands worldwide are poor and depend on ecosystem-based livelihoods, they frequently move to smooth income shocks, and degrading ecosystems make emigration from such areas more likely (Kollmair & Banerjee 2011). The cases here also corroborate that feedback mechanisms can enlarge emigration; such mechanisms include shortage of labor force, closure of critical infrastructure, and removal of social resources in villages of origin as well as expanding migration networks in cities and chain migration effects. When tipping points are crossed and life cycles are compromised—as seen in other Peruvian pastoralist communities (Alata et al. 2018)—gradual settlement abandonment can become a real threat. Such abandonment processes are known from historical cases (McLeman 2011). Simultaneously, the case of V2 emphasizes that gradual hazards can reduce the ability of people with limited prior resources to move out of harm’s way and thus forcibly trap them, as suspected by other studies (Kollmair & Banerjee 2011). Even more so, place attachment and health limitations can make people stay despite rising climate hazards, a dynamic also observed in other Peruvian highland zones (Adams 2016).

Among those who left, agency varied (de Haas 2021). Most migrants disposed of limited, but sufficient capabilities to implement their choices to move. Men and wealthier families left first, which validates that gradual climate migration is often gendered in Peru (Milan 2016; Sperling et al. 2008), and that vulnerability and capacities to move are inversely correlated (Adger et al. 2014). Conversely, the findings here provide nuances to another study of Peru which holds that gradual hazards first push those with least assets and skills to migrate (Koubi et al. 2016); such a dynamic is probable when impacts are still more manageable in place (V1) but once habitability is threatened (V2), most affected people will be compelled to leave, and the wealthier have better chances for doing so faster. Migration ranged from more voluntary to more forced instances, as some people had aspired to leave but others hoped to stay. In Kunz’s terms (1981), migration was anticipatory at first, as people increasingly worried about the climate-related, gradual loss of jobs and educational opportunities; yet, it had acute elements later on, especially once the meltwater was entirely lost in V2 and people were forced to react. This underscores that the lens of mixed migration, which has recently gained traction in migration policy debates (Sharpe 2018), is also relevant for climate migration (e.g. Sow et al. 2016).

While much fewer migrants from V1 have opted to return permanently than used to be common decades ago, all migrants have maintained translocal links with their birthplace, which underlines the merit of studying climate migration through translocality lenses (Greiner & Sakdapolrak 2016; Rockenbauch & Sakdapolrak 2017). Due to V2’s remote location, limited mobile reception, and the few stayers left, translocal ties to the village were weaker than to V1. In the cities, hometown associations were key actors for migrants’ new pursuits (Pairama & Le Dé 2018), yet such HTAs and other collective actors, such as faith-based organizations or NGOs, have still received limited attention in the literature on internal climate migration.

3.2 Well-Being Effects, Structural Conditions, and Mechanisms

3.2.1 Migrants’ Development from a Secure base

Mechanisms conducive to decent livelihoods included vocational trainings or studies in the cities that enabled a small number of migrants to take up better positions; networks permitting access to credit and jobs; entrepreneurship and innovation to find niches or make use of urban opportunities; as well as sacrifice and hard work, often at the cost of other well-being elements, such as health. Conversely, especially the case of V2 corroborates findings from other highland studies that migration becomes “more of an imperative to maintain livelihoods than a way to enhance well-being” when climate impacts cause severe harm and exacerbate unequal access to resources or power (Lennox 2015: 793; Wrathall et al. 2014). This dynamic was evident in the heavily harmed, spatially and administratively isolated V2. When farmers from V2 had to migrate to the cities, they lacked transferable skills and often ended in precarious jobs, which confirms a concern raised before in Peru (Sperling et al. 2008). Interviewees with prior vulnerabilities suffered from urban livelihood stress, which inflicted further harm such as food insecurity, in line with the larger internal migration literature (Lagakos 2020; Selod & Shilpi 2021), other climate migration studies (e.g. Melde et al. 2017), and a study in Peru’s Selva (Sherman et al. 2015). Most migrants’ urban occupational quality did not greatly improve, contrary to global findings (World Bank 2018). This observation substantiates the warning that distress migrants are at risk in labor markets split into high- and low-wage jobs (Waters et al. 2010), as is the case in Peru (Carranza 2016). The lack of skills transferable to the cities or the inability to acquire new skillsets, for example due to the high costs of education, were a first major obstacle for decent job and income situations. Second, additional barriers included baseline health problems or worsening health conditions due to migrants’ new urban lives. Third, limited social networks posed further obstacles. Fourth, structural factors also obstructed advances: due to population growth, labor market segments of interest have gotten more saturated and highly competitive while occasionally, the rising costs of urban living nullified relative income gains compared to migrants’ previous lives. Furthermore, weak governance and alleged corruption issues as well as a lack of adequate resources and personnel in municipal service provision hindered economic advances.

For education, several mechanisms worked in different directions. Migrants started with disadvantages in the cities because education in the Sierra lagged. Nevertheless, many appreciated the better availability of educational institutions in the cities with more programs, better staff, and greater market recognition. Flexible options such as evening schools were key resources for migrants. However, these options were also expensive; migrants had to study, learn, and work simultaneously to pay the costs, and this burden prevented several of them from accessing or progressing in education. Hence, few migrants had access to this educational gateway to better jobs and health, which echoes UNESCO’s (2020) concern about educational vulnerabilities in climate migration, but contradicts the gains found by prior studies on climate (Melde et al. 2017) and general migration (UNDP 2020).

Additionally, many interviewees suffered from several of the health challenges common in climate migration (see review in Schwerdtle et al. 2020), which contradicts the more positive results by Melde et al. (2017). A first, major constraint for food security and access to the health care system was migrants’ lack of income. Second, deficient basic services, often due to a lack of land- and homeownership, aggravated health problems. Third, stressful lives, pollution and contamination, and precarious livelihoods also affected health, as did family separation. Finally, partial health gains were often offset by higher costs of living and adverse living conditions, as can be the case for internal migrants worldwide (Selod & Shilpi 2021). Conversely, the few migrants with the necessary resources to access urban health services, staff, and infrastructure tended to profit from moving.

All things considered, the new cases substantiate results of prior studies in Peru’s highlands (Afifi et al. 2016) and the general literature (see chapter 4), which stress that well-being effects of rural-to-urban climate migration vary: the higher prior vulnerabilities are, the more likely adverse results. More voluntary, improvement migration under tolerable conditions made it possible for a small number of migrants to advance in their development from a secure base. Conversely, those with little prior resources moving forcibly and under distress conditions experienced hardship.

3.2.2 Migrants’ Space to Live Better

Achieving a space to live better depended on migrants’ resources, social networks, land availability, and administrative decisions regarding titles and basic services. A small number of migrants was privileged to move in with established relatives who lived in decent houses, which stresses the key role of family networks to mitigate some of the possible hardship in cities (OECD & EU 2015). Better situated migrants were generally content in the nicer and safer areas of Lima, and some preferred the warmer climate compared to the Sierra. By contrast, for most of the poorer and more recent migrants, the oversaturation of the rapidly growing destinations Lima and Huancayo has created challenges. Vertical, horizontal, and informal growth has led to problems related to land acquisition, infrastructure, basic services, social cohesion, and urban livelihoods (Carpenter & Quispe-Agnoli 2015; Haller & Borsdorf 2013; Ioris 2015). Many interviewees had to move gradually more toward the outskirts, often to precarious and unsafe housing in new settlements. Because few of the attainable settlements were legal, many migrants lacked titles and therefore access to basic services for long and some for life. Most poorer migrants also described their new urban surroundings as arid, hectic, and congested, and thus as dissatisfying compared to their previous rural environments, confirming that the lack of green spaces and connections with nature can heavily reduce well-being (Wolsko & Marino 2016). The observed loss of place satisfaction compared to rural homes also supports the prediction by Adams (2016) that climate migrants who suffered non-provisional losses in their birthplaces face well-being risks in the cities, because they cannot recover these losses. A growing literature on distress caused by environmental change refers to this phenomenon as solastalgia (Albrecht et al. 2007; Tschakert & Tutu 2010; Warsini et al. 2014). Most poor migrants also had to live in urban zones that were insecure and dangerous, such as riverbanks or hills surrounding Huancayo, and the arid, hilly outskirts of Lima. While they managed to escape from primary hazards in the Sierra, many suffered from secondary hazards in the cities, a concern also stressed by the climate migration literature (Adger et al. 2014; de Sherbinin et al. 2012; Foresight 2011). In Peru specifically, studies indicate that rural-to-urban climate migrants from the Sierra (Oliver-Smith 2014), Selva (Langill 2018), and Costa (Rubiños & Anderies 2020) often end up in hazard-exposed zones due to structural and intersectional factors (Erwin et al. 2021). In both Huancayo and Lima, numerous interviewees worried about water scarcity, and the literature substantiates this concern. Water is already scarce, but demand is increasing in many growing cities in Peru; simultaneously, the projected vast glacier loss further threatens highland cities that depend on meltwaters and coastal cities’ inter-annual water buffering systems, and thereby exacerbates the competition for water between human use and agriculture (Buytaert et al. 2017; Buytaert & de Bièvre 2012).

To summarize, a space to live better was unreachable for most poor climate migrants who had moved under distress to disadvantaged urban zones characterized by tenure insecurity, deficient housing, poor basic services or infrastructure, as well as insecurity, pollution, and hazards. These results substantiate the risks for climate migrants in rapidly urbanizing areas with adverse structures identified in other climate migration studies (e.g. Adger et al. 2020; Melde et al. 2017).

3.2.3 Migrants’ Social Relatedness

The review in chapter 4 led to mixed expectations regarding the effects of rural-to-urban climate migration on social relatedness, ranging from gains (Melde et al. 2017) to losses, especially for forced migrants (Schwerdtle et al. 2020). The limited information for Peru also suggested negative impacts (Sherman et al. 2015; Sperling et al. 2008). This study adds nuances to this literature by demonstrating that the quality of migrants’ social relatedness was in part a function of time. Initially, different social systems and codes, family separation, and losses of strong prior ties often reduced migrants’ social relatedness, in line with results of general migration studies (Selod & Shilpi 2021). However, consolidated migrant structures in the large cities reduced social costs for newcomers by allowing them to acculturate selectively, namely, to gradually adopt new ways while nurturing partial roots in their co-ethnic communities (Waters et al. 2010). Over time, migrant networks therefore created social gains. The more established migrants became, the more often they also met partners and spouses in the cities, founded families, and started to raise own children, with whom they usually described good relations. Many also experienced positive social cohesion in their neighborhoods. This value of primary relationships and community support in destinations is consistent with findings of general migration reviews (Munshi 2020; Selod & Shilpi 2021).

3.2.4 Migrants’ SWB

No previous study had explored the SWB of climate rural-to-urban migrants in Peru.¹⁶ On the negative end, the findings here reveal that most of those moving under structural distress conditions and settling in places with relatively hostile modes of incorporation suffered from deprivation¹⁷ due to multiple unfulfilled needs and limited development prospects. Many migrants described negative feelings especially at the beginning of their time in the challenging urban areas. A first burden was sadness due family separation. Second, anxiety, stress, and tensions frequently persisted due to the demanding city life, precarious livelihoods, the dual burden of working and studying, and insecurity. Various migrants were overwhelmed by these dynamics, exhausted, and dissatisfied with life. Third, migrants were also angry and frustrated due to corruption and governance failures. Fourth, negative feelings often related to adverse material conditions, and even some migrants whose conditions had improved reported negative emotions. Nevertheless, other than predicted before (Koubi et al. 2018; Magallanes 2015), the interviewees’ experiences of hazards, grievances, and deprivation did not (yet) seem to make conflict more likely in the cities.¹⁸ Over time, several but far from all migrants experienced partial hedonic adaptation to the surrounding urban risks. These results corroborate that migration voluntariness, resources, and conditions strongly determine SWB (Bartram et al. 2013; Hendriks 2015) and that most rural-to-urban migrants in poorer countries stand to lose SWB (e.g. Chen et al. 2019; Mulcahy & Kollamparambil 2016). Various of the mechanisms identified in the review chapter 4 are plausible. First, it is known that migrants’ SWB partially converges to the average SWB of locals (Helliwell et al. 2018b), and most interviewees lived in poor zones with expectedly low SWB. Second, relative deprivation compared with better situated segments of the cities may have further reduced SWB because the more migrants integrate, the more they tend to take residents in their new homes as reference points (Melzer & Muffels 2017). Third, the difficult macro conditions in the cities may have further decreased SWB (e.g. Hendriks & Bartram 2016). And finally, a part of migrants’ low SWB in the cities may relate to footprint effects from source areas (Helliwell et al. 2020), especially for migrants from the SWB-deprived village V2.

On the positive end, the small number of improvement migrants who had achieved urban upward mobility seemed to experience true well-being. This result contradicts most other studies on internal migrants in poorer countries (for whom SWB losses are common even despite high OWB gains (e.g. Chen et al. 2019; Czaika & Vothknecht 2014)), but is closer to results in a few richer areas (where some male migrants make lasting SWB gains (e.g. Ek et al. 2008; Kratz 2020)). Advantageous mechanisms included successfully leveraging educational and income opportunities, working in formal jobs or in own businesses, living in quieter areas, and having good family ties. Therefore, the reasons behind these rare SWB gains in this study may be large relative OWB improvements compared to migrants’ own prior lives, combined with relative improvement compared to their many often-poor urban peers. Prior studies also find that OWB gains (especially in income) compared to migrants’ previous lives can drive lasting SWB increases beyond hedonic adaptation (Melzer & Muffels 2017). By contrast, the mechanisms most commonly stressed in the field—rising or unrealized aspirations, lacking information, relative deprivation, and migration costs (Haindorfer 2019a)—ostensibly had limited influence on this small number of migrants. Even so, the reviewed literature suggests that most interviewees will likely remain unable to catch up with the SWB of city natives (e.g. Hendriks et al. 2016), a prediction that future studies could assess.

Beyond their emotional balance and cognitive satisfaction in the present, most migrants’ views of the future were negative (enforced and dramatized fear)¹⁹ due to various mechanisms. First, migrants in low-wage temporary positions in the city suffered from deprivation and economic fears and had often resigned. This result contrasts other cases where hope aided low-wage migrants to endure present hardship (Pine 2014) but is in line with cases where migrants’ hopes wane over time (Boccagni 2017). Second, several migrants worried about the observed challenges in their highland villages of origin and third, various of them felt anxious for the coming generations due to their lack of development prospects and the rising magnitude of climate impacts. These findings stress how salient climate anxiety can be for climate migrants (Clayton 2020); they also suggest that migrants may not be able to escape this anxiety by moving, since they are linked to stayers in still-exposed hometowns and often remain affected themselves by hazards out of their control in cities. Conversely, those who held hopes did so due to relatively good economic prospects, religious faith, or personal resilience in the face of adversity, occasionally despite deep poverty, which reveals that circumstantial and personal factors interact to create hope (see chapter 4). Other migrants only held hope for their children’s progress, a finding consistent with other studies (e.g. Boccagni 2016). However, as noted above, still other migrants despaired over impending challenges for their children.

3.2.5 Stayers’ Well-Being

As a secondary interest, this study also led to insights into several mechanisms through which emigration affected the social systems and well-being of stayers in the Sierra. Migration together with changing family planning have reduced population sizes in migrants’ birthplaces toward (V1) or below (V2) critical thresholds that have started to threaten village life. Economically speaking, emigration subtracted labor force from the shrinking villages, which complicated labor-intensive agriculture. It also generated remittances, but their extent depended heavily on the urban jobs attainable for migrants. Many of the interviewed poor migrants in precarious and informal jobs had limited resources for remittances, similar as in Peru’s Selva (Sherman et al. 2015) and other contexts (Le Dé et al. 2013; Schade et al. 2016). Adding insights to a still-open debate (Banerjee et al. 2017; Bendandi & Pauw 2016), in this study, financial remittances did not seem to buffer the loss of water and labor force and thus failed to support climate adaptation. As a result, in V2, many economically inactive and physically limited older adults suffered economic hardship due to emigration. Interviewees also did not report large benefits from remittances on health and education, contrary to a study of climate migration on Peru’s coast (Badjeck 2008) and to general reviews on migration (Obi et al. 2020; Ratha et al. 2011). Social remittances, another possible development catalyst (de Haas 2009), were also limited due to migrants’ economic struggles and the remoteness of their birthplaces. Hometown associations from V1 and V2—another potential supporting actor for climate adaptation in source areas (ADB 2012; Webber & Barnett 2010)—mainly focused on sociocultural events and support for migrants in the cities. Nonetheless, general statistics indicate that internal remittances hold potential to support at-risk stayers in rural source areas in Peru, as almost one third of the recipients are women, 65 years or older, and 47% of them are economically inactive (Sánchez Aguilar 2012a). However, this study suggests that limits apply; once thresholds toward settlement abandonment are crossed, remittances cannot counterbalance the ensuing threats to the survival of critical infrastructure, services, and social stability. Similarly, extended family networks can help mitigate some of the social effects of emigration (Bedford et al. 2009), but only to a certain degree. These networks gradually failed in V2, where too many entire families had already left, and the few remaining villagers suffered from a disrupted social system. Overall, the socio-cultural effects of emigration depended on who left, their numbers, the pace of departures, and the possibilities to maintain contact. In V1—with its still lower pace of emigration and better mobile reception—social relationships were less affected than in V2. The exodus from V2 also aggravated existing health issues by deteriorating health service provision, food insecurity, and psychosocial issues, which confirms the concern by general reviews that emigration can raise morbidity for stayers (Abubakar et al. 2018; Paudyal & Tunprasert 2018), especially for older adults (Ao et al. 2016). By contrast, increasing emigration has had few effects on health in V1 so far, where more people have remained in place. Regarding education, the increasing departure of the youth and young families has led to ever fewer pupils in both villages, which has threatened the continuance of the educational infrastructure. In summary, the findings on the OWB of stayers stress that emigration can induce losses of labor force, critical infrastructure or services, social networks, and traditional knowledge, which corroborates findings of prior studies of climate migration in Peru’s highlands (Altamirano Rua 2021; Lennox & Gowdy 2014; López-i-Gelats et al. 2015) and the general literature (Obi et al. 2020).

Finally, emigration also affected SWB in the villages. Key mechanisms were the severity of climate impacts and related magnitudes and conditions of emigration; the well-being of migrated relatives; the possibility to maintain contact; and stayers’ social and material prospects. The entirely negative SWB effects of emigration in V2 are consistent with results of a prior review study (Paudyal & Tunprasert 2018), and the cases suggest that the reasons behind this decline are losses of identity, community organization, and critical infrastructure due to emigration. By contrast, migration has had more mixed SWB effects in V1, confirming that gains in some SWB factors can occur simultaneously with losses in others (Graham & Nikolova 2018). Future outlooks in V1 were mixed given concerns about livelihoods and climate change, yet still more positive than in V2, where personal coping mechanisms like faith or humor could not offset the salient stressors. These findings confirm the expectations from the review in chapter 4 that climate change can worsen views of the future through various pathways, including past and expected climate impacts, and that rising despair can deepen perceptions of vulnerability, uncontrollability, and unpredictability, which may ultimately block action or engagement (e.g. Carver & Scheier 2014; Forgeard & Seligman 2012).

4 Summary and Induction of Propositions

In this section, I summarize the key structural conditions, mechanisms of action, and well-being dynamics in this case study of rural-to-urban migration related to gradual hazards in the Sierra. In a final step, I then induce broader propositions on the well-being impacts of climate (im)mobilities.

Figure 5.14 displays the identified well-being outcomes for climate migrants as well as significant structural conditions and mechanisms shaping these outcomes. Stayers are not depicted. The figure illustrates that the migration in the context of gradual glacier loss and rainfall changes occurred with average to low structural opportunities, including freedom of movement within Peru and established translocal networks. However, structural constraints were larger and included severe climate risks and limited adaption possibilities, poverty and inequality, limited livelihood options, spatial insularity, tenure insecurity, poor basic infrastructure and services, weak governance, lack of political voice, and population growth. Given these structural conditions, the flows from the rural areas to the large cities were situated on a spectrum between improvement and distress migration. Regarding agency, most migrants had limited but sufficient migration capabilities to implement their decisions, yet aspirations to leave varied, and movements thus ranged from more voluntary to clearly forced instances. Most remaining adolescents aspired to depart to the cities. Among those staying in V1, many older adults were voluntarily immobile, whereas others, especially those in V2, were trapped.

Migrants’ well-being results in Lima and Huancayo were on average more net-negative than net-positive. For finer analysis, the figure below differentiates between two factions. The first group covers most of the interviewees who had left the increasingly adverse situations in their home villages in recent years under distress conditions and arrived with limited baseline resources and skills in the oversaturated, large urban agglomerations. These migrants experienced great challenges for their development from a secure base and a space to live better, while gradually reaching similar or better social relatedness as before migration. Although they were subsisting in the cities, multiple deprivations restricted their development prospects. These migrants with multiple unmet needs subjectively suffered from deprivation in their cognitive satisfaction and emotions; only few reported partial and gradual adjustment.²⁰ In the figure below, they are marked in red as the “losing” group. By contrast, the second group comprises a small number of migrants who achieved urban upward mobility, in particular those who had moved voluntarily and a longer time ago. In often-arduous processes, they have used urban education and job opportunities to improve their well-being and long-term prospects. This small group evaluated its needs as mostly fulfilled (true well-being). In Figure 5.14, this faction is highlighted in green as the “gaining” group. Nonetheless, both the large number of migrants losing OWB (enforced fear) and the few ones gaining OWB (dramatized fear) held mostly negative views of the future. The light gray boxes in the figure synthesize the key structural conditions and well-being mechanisms explained above that may be applicable to other contexts.

Figure 5.14

Main well-being effects for migrants from V1 and V2, structural conditions, and identified mechanisms of action. (Note: Created by the author)

These empirical findings on improvement and distress climate migration from Peru’s highlands, ranging from voluntary to forced cases, make it possible to induce broader propositions on the well-being of climate migrants in cities and stayers in rural source areas.

Foremost, when severe and irrevocable climate impacts threaten the means of existence of remote and poor agricultural communities with constrained on-site adaptation options, migration likely gradually becomes an imperative. Self-accelerating migratory feedback mechanisms then can reduce population size at an exponential pace, pushing settlements to points of no return and toward eventual abandonment. This is particularly likely when village life cycles are disrupted because the youth and young families leave, as is common in many small mountain towns worldwide with deeply rooted cultures of migration, which can raise the pace and volume of departures. If land and livelihoods are not entirely lost, increasingly smaller shares of the residents may be able to apply the limited in-place diversification options to carve out a living in the original site. Yet, absent strong support from the state or translocal networks, these remaining villagers are at risk of progressively more severe OWB and SWB declines due to the losses of labor force, knowledge, social networks, and key infrastructure or services, which can endanger their self-organization, identity, and survival.

If enough freedom of movement and migration networks exist, two internal migration paths are possible for such former subsistence farmers who leave areas with progressively worsening slow-onset impacts. First, if impacts are still more manageable in place, they are likely to emerge as one factor among many that jointly drive gradual, anticipatory improvement migration. People engaging in this type of migration have a chance to experience improved OWB in the cities, albeit with taxing and laborious processes. Second, OWB is likely at risk if climate impacts reach critical thresholds and increasingly force distress migration, and if such migrants arrive with limited baseline resources and untransferable rural skills in overcrowded zones of large cities. Such migrants may subsist in their new urban lives, but multiple deprivations will significantly restrict their development prospects. The well-being results of these two groups will likely diverge strongly in the four studied dimensions:

• (1) Most of the poor farmers who must move to irregular, peripheral zones marked by high job competition, limited state presence, and deficient basic services will have almost no chance to transform their skillsets to meet urban demands. They will likely continue with limited opportunities for development from a secure base while they struggle to achieve decent livelihoods, education, or health and food security. A few migrants may see progress if they arrive in the cities during periods with more favorable labor, land, and housing markets as well as receptive policies. However, making such advances additionally requires strong educational and social entry points and migrants’ entrepreneurship, hard work, or sacrifice.

• (2) A space to live better is likely out of reach for many poor climate migrants who move under distress conditions to irregular urban settlements at the margins of oversaturated cities, where they lack state support, tenure security, as well as basic infrastructure and services, and may be exposed to insecurity, pollution, and unfamiliar climate hazards. In the rarer cases where positive conditions work together—such as sufficient migration resources, strong social networks, state support, and urban livelihoods gains—migrants may start with or gradually achieve more decent living conditions.

• (3) If a strong diaspora provides climate migrants with time and support so they can gradually blend into segmented, co-ethnic urban social systems and learn new social codes, they may, gradually, reach similar social relatedness as before, especially as they build new ties and found families.

• (4) As a result of multiple unfulfilled needs, many climate migrants are likely to suffer from subjective deprivation. Such deprivation threatens to be especially salient when migrants start to settle in, a time when several negative factors such as family separation, parallel burdens, insecurity, weak governance, and unfamiliarity with the new social system are common. If people’s attained human development state as well as the quality of their urban locations and security remain low, even expectable social gains made over time and personal coping mechanisms are unlikely to raise their evaluation of need fulfillment in the present. Few of these migrants may gradually and partially experience adjustment through hedonic adaptation. By contrast, a smaller group of migrants with urban upward mobility can experience true well-being. Nevertheless, most of the internal climate migrants moving, settling, and living under the conditions described above are likely to experience enforced (and some dramatized) fear regarding the future, given economic problems and lack of progress, anxiety regarding uncontrollable climate impacts, concerns about the future of their children, and worries about their rural hometowns, which personal coping mechanisms may moderate only to some degree.