2. Site selection and research methods

This study was carried out in four villages located in the Ban Puang sub-district in Lamphun province (see Figure 1). The villages, which represent typical rural upland settlements of Northern Thailand, are inhabited by the Thai and Karen communities. The area is hilly and forested. The villages are located on the slopes along small rivers that drain into the Li River. The selection of the study sites was a collaborative process between the Raks Thai Foundation which supported the research, and the United Nations University Institute for Environment and Human Security (UNU-EHS) which was responsible for the scientific side of the project. The study site is therefore purposely selected guided by a mix of scientific, practical and institutional considerations (see Promburom & Sakdapolrak, 2012, pp. 28–29).

Why successful in situ adaptation with environmental stress does not prevent people from migrating? Empirical evidence from Northern Thailand

All authors

Patrick Sakdapolrak, Panomsak Promburom & Alexander Reif

https://doi.org/10.1080/17565529.2013.826129

Published online:

12 March 2014

Figure 1. Research sites. Source: OAE (2006) and Centre for International Earth Science Information Network (2011).

Display full size

Based on a livelihoods vulnerability framework (see Warner & Afifi, 2013, in this special issue), we have addressed the relationship between rainfall-related events and trends, livelihood security and migration in the study area with multi-method field-based research design, consisting of a household survey, participatory rural appraisal (PRA) as well as expert interviews (for a detailed description of the methodology, see Rademacher-Schulz et al., 2012): (a) the household survey was carried out with 30% of households in the study villages, which were selected by a proportional cluster sampling strategy using village household maps as a sample frame. In total, information was collected from 206 households; (b) various PRAs (including among others livelihoods risks ranking, time lines, impact diagrams, mobility maps and Venn diagrams on migration) were carried out in 35 PRA sessions. All in all 148 villagers, whose selection has been based on gender, age, livelihood type and social status, have participated in these sessions; (c) 13 expert interviews with key informants from the study sites, government officers as well as members from non-governmental organizations have been conducted. The main field work was carried out by a research team in October 2011 (Promburom & Sakdapolrak, 2012) and was followed up by an additional in-depth village study between March and June 2012, during which 31 problem-centered interviews and 9 PRA sessions were carried out (Reif, 2012).

3. Rainfall-related stresses – numbers and perceptions

Being under the influence of the south-west monsoon, the study site exhibits the typical seasonal pattern of rainfall of the Northern Thai region. The average rainfall peaks twice a year in May (156.7 mm) and September (209.1 mm). November through February are dry months with almost no precipitation. The average total annual rainfall is 1017.03 mm. The analysis of the total annual rainfall between 1981 and 2011 reveals that it has slightly increased over the period of 30 years. Figure 2 shows the average monthly rainfall between 1981 and 2011, as well as the average monthly rainfall in 2011. It can be clearly seen that the rainfall in 2011 was exceptional in many regards: rain started to fall in considerable amounts already in March. Each month from March until September had significantly above average rainfall. The total rainfall in 2011 was 1706.7 mm or 67.8% above the 30-year average and led to damage and destruction not only in the study sites but also to heavy flooding in Central Thailand and Bangkok (World Bank, 2012).

Why successful in situ adaptation with environmental stress does not prevent people from migrating? Empirical evidence from Northern Thailand

All authors

Patrick Sakdapolrak, Panomsak Promburom & Alexander Reif

https://doi.org/10.1080/17565529.2013.826129

Published online:

12 March 2014

Figure 2. The average monthly rainfall in Lamphun (mm). Source: Lamphun Meteorological Station (1981–2011).

Display full size

The long rainy season in 2011 was also a prominent feature in the narratives of the study population about climate, changing climate and rainfall. Participants in the PRA sessions mentioned it to be a remarkable event. Older villagers reported during these sessions that they had never experienced a year with similar rainfall before. Furthermore, the participants in the PRA sessions stated that generally it is currently raining more than it used to when they were younger. The results of the household survey also affirm statements during the PRA sessions (Table 1): 87% of interviewees stated that over the last 10–20 years (1991–2011) heavy rainfall events occurred more often. Floods that are closely related to rainfall events in the study area were also perceived to have occurred more often. A closer examination of the perceived changes in rainfall-related events exhibits the following results (not in table): 73.3% of households stated that it is raining more; and 48.5% have noticed that the rainy season has become longer. The majority of the respondents also stated that flood events occurred a lot more (20.3%) and more often (29.7%) during the last 10–20 years compared with the period before. People in the central plains experienced slow-onset but long-lasting and geographically widespread flooding in 2011. In contrast to this, inhabitants of the research sites were affected by flash floods with a rapid rising water level of the streams that flow through the upland valleys where the villages are located.

With regard to droughts and the occurrence of dry spells, over half of the respondents (58%) reported that these had occurred less often during the past 10–20 years (1991–2011) (Table 1). Changes in the characteristics of dry spells have been noticed by a minority of respondents only. In the local context, as the PRAs revealed, drought or laeng in Thai is understood as any situation where the farmers do not have access to a sufficient amount of water to pursue their normal agricultural activities. Droughts in the study area are not related to complete failing of rain, but rather to untimely rainfall as well as the occurrence of dry spells that have a negative effect on agricultural production. Drought, in people's understanding, is closely linked to agricultural production. Therefore, the perception of the occurrence of droughts is linked to the amount and pattern of rainfall as well as to technological and infrastructural changes related to agricultural production, for example, the construction of water reservoirs and the development of irrigation infrastructure. The same amount of rainfall therefore might have caused drought in the past but might not be considered as drought today. In general, a majority (86%) of the respondents of the household survey said that during the past 10–20 years (1991–2011) extreme weather events occurred more often (Table 1).

4. How are people affected and how do they cope?

The households’ livelihood in the research area is characterized by a high level of on- and off-farm diversification and integration into national and global market structures (Table 2): Agriculture is the most widespread income generating activity. 85.9% of the households reported it as an income source. Most farmers pursue semi-subsistence agriculture. Rice is grown for domestic consumption and maize is the main cash crop in the area. In recent years, a wide range of additional cash crops have been introduced in the study area, including potato, bamboo and papaya. For 45.6% of the households wage labour, mainly in the agricultural sector, is also an income source. The most widespread off-farm income generating activity is weaving, which is carried out by 35.9% of the households. 10.2% of the households are conducting small-scale trade. Only 5.3% reported having access to income from regular salaried employment, while 20.9% regularly receive financial remittances from migrant family members. Within this livelihoods system, which is highly dependent on agricultural activities, rainfall-related events and trends, such as droughts and dry spells or heavy rainfall, are considered by the villagers to be a potential threat to livelihood security.

According to the household survey, 15.5% of households have experienced drought and dry spells in the past, which had negative effects on maize and rice production. Once the farmers put the maize seed in the earth and the rain doesn't start on time, the seeds do not germinate and then the farmers have to repeat the procedure again. Rice plants and maize plants also die if they do not get enough water during the period water is needed for their growth. The participants in the PRAs said that they cannot do anything to prevent drought and dry spells, but they can adapt to it by: adjusting the start of the maize seed planting; getting credit and planting again in cases when the first planting attempt failed; and pumping water from the streams into the rice field, which can lead to arguments and conflicts within the village.

The effect of heavy rain was mostly discussed in relation to the heavy rainfall of 2011. More than half of the households were affected by heavy rainfall and floods during the previous 12 months. The participants in the PRA explained that this heavy and continuous rainfall had a negative effect in several ways. Infrastructure such as roads and dams were damaged through heavy rain. The flash floods that occurred in the course of the rains also led to damage to the rice fields, carrying sand into and burying rice in paddy fields. The continuous rain and the high moisture also had a negative effect on maize quality through plant diseases. The participants reported that the fertilizer they had applied was washed away by the rain. All these factors resulted in reduced agricultural income. In the case of damaged infrastructure, the villages can approach the local government for help (monetary as well as equipment) to repair the road and the dam, and 22.3% of households reported that they approached the local government for help during the last 12 months. The farmers also have to get credit from banks to overcome lost investments, and 16.5% reported that they had to get credit in order to cope with the effect of environmental stress. With regard to the maize harvest, which was taking place while the PRA sessions were being conducted, the farmers said that they could either wait and hope that the rain will stop soon, or harvest and sell it anyway even if the price will be low. Comparing the effect of drought with heavy rainfall, the participants in the PRA said that the effect of droughts is more severe. Death of maize plants due to drought, for example, is worse than diseased maize caused by heavy rains. Diseased plants can still be treated with herbicides and the harvest is not completely lost.

Three-quarters of the households reported that they suffered from declining income as a result of the exposure to environmental stress due to declining crop yields and deceasing income from agriculture. At the same time, they had to struggle with rising commodity prices. During the PRA sessions, it was discussed how the villagers cope with declining income. The participants said that they try to increase income from additional work as agricultural labourers, through off-farm activities such as weaving, and by diversifying agricultural products. In the context of exposure to rainfall-related stressors, the participants in the PRA sessions have not mentioned that migration is a significant strategy they pursue to cope with and adapt to these adversities.

5. What is the role of migration and why does it matter?

Even though people in the research areas did not consider migration as a primary strategy to cope with the environmental stressors they are exposed to, migration – internal as well as international – is a widespread livelihood strategy in the area. Several PRA sessions where migration was discussed revealed that the history of migration in the study area reaches back several decades. Out-migration started around 50 years ago (early 1960s) when male members of the villages were recruited for logging and mining activities in nearby areas within the region. Some went for a few years, and other members migrated permanently. While these first migrants were mainly going to other rural areas, the subsequent generation of migrants – by this time males as well as females were migrating – searched for employment in urban areas. They took up work at petrol stations, worked as housekeepers and child-carers, in construction and as factory workers. More recently, migration for educational purposes has played an important role (a similar pattern can be observed in Vietnam; see van der Geest, Nguyen, & Nguyen, 2013, in this issue). People from the study area also migrated internationally. This movement of people started 30 years ago (late 1970s) and continues to play a role today. The household survey confirms that migration is an important feature in the life of the people in the study villages; 67% of households reported that they have a member who has migration experience – which means a member who is currently away or who has migrated in the past and returned. The total number of people with migration experience is 224. All in all, more men than women have migration experience: 61.6% of the migrants were male and 38.4% were female. Among those with migration experience, the majority (62.9%) have migrated in the past and returned. The number of current internal migrants is 83, and current international migrants are 10.

Focusing on the role of migration as an adaptation strategy with regard to the effects of environmental stress, participants said that they would not migrate as a response to this stress, as the effect was still within the acceptable range that they could cope with through a number of strategies other than migration. Rainfall-related events do not seem to have a direct effect on migration decisions from the study area. In the household survey, the respondents were asked to indicate the importance of potential causes of migration (Table 3). The three most frequent reasons that were reported to be ‘very important’ are: (1) ‘I want to build my own life in the city’ (frequency: 19); (2) the ‘bright lights’ of the city attract me (frequency: 17); and (3) ‘no relatives and friends in the village’ (frequency: 16). The three most frequent reasons reported to be ‘important’ are: (1) ‘not satisfied with my livelihood’ (frequency: 34); (2) ‘better job opportunity in the city’ (frequency: 34); and (3) ‘less crop production for sale’ (frequency: 32). The six reasons for migration with the highest cumulative score² are: (1) ‘work related to my skills is not available’ (score: 57); and five reasons with a score of 52: ‘want to build up my own life in the city’, ‘the bright lights of the city attract me’, ‘the living quality in the city is better’, ‘unemployment’ and ‘not satisfied with my livelihood’.

The results in Table 3 indicate that environmental factors as drivers of out-migration play a subordinate role. This result was then cross-checked in the PRA sessions, which confirmed the identified pattern (the need for triangulation to decipher migration-environment linkages is discussed by van der Geest, 2011, pp. 70–71). At the end of the Venn diagram sessions on migration, the relationship between rainfall variability and migration was raised as it had not been mentioned before. The participants stated that there is ‘no relationship’ and that rainfall has ‘very little influence’ on a decision to migrate. Furthermore, most participants mentioned that they can pursue a good life without moving outside as resources are available and cultivation is quite productive. Behind this background two questions arise: (a) what are the reasons for migration despite successful in situ adaptation to environmental stress? (b) In a context where environmental risks seems to play only a marginal role in migration decision, is it necessary to consider migration in the analysis at all?

To start with the first question, the results of the empirical case study show that migration is a common feature of livelihoods for at least three reasons. Firstly, from a historical-structural perspective on migration (see Portes & Walton, 1981) the historical analysis (PRA – timeline) has clearly displayed the relationship between market integration, livelihood diversification and migration. The study villages, which are situated in a peripheral area in Thailand, were successively connected to and integrated with the national and global economies since the late 1960s (the role of market access on migration pattern is also highlighted for the case of Peru by Milan & Ho, 2013, in this issue). With this integration, farmers who were engaged in non-monetary subsistence agriculture before, now needed to generate cash income in order to meet needs (e.g. electricity, education and agricultural investments) as well as consumption desires (e.g. TV, fridge and car). Migration was one of the main strategies to generate cash income. Similar processes of agricultural change are well documented (e.g. Rigg, Veeravongs, Veeravongs, & Rohitarachoon, 2008). Secondly, the pattern of international migration shows that the timing and direction of migration is determined largely by external factors such as labour demand in the Middle East as well as South-East Asia (see Piore, 1979, on the effects of dual labour markets on migration) and facilitated though commercial networks, the so-called migration industry (see Massey et al., 1993, pp. 448–451, on the role of networks and social capital). The pattern furthermore indicates, as highlighted by Massey's (1990) theory of cumulative causation, that once the migration flow has been initiated, the feedback processes of migration lead to more migration happening. Reasons for this include better availability of information, the changing social structure and the need for those left behind to catch up. Finally, migration – especially of younger segments of the population – is closely related to a process of structural change in the economy and the labour market. The results in Table 3 illustrate that personal, social and economic reasons are considered important and migration decisions reflect the interplay between push factors (e.g. lack of adequate work) and pull factors (e.g. better quality of living in the city). This is particularly the case for the better-educated younger population which is seeking employment in the secondary and tertiary sector in the cities.

As migration in the study area is not related to rainfall-related stressors and other environmental risks, one might ask if it is necessary to consider it in the context of adaptation. The results of this case study clearly show that migration is a crucial issue. Migration, regardless of exposure to environmental risks, is occurring and will continue to be a major dynamic of social transformation in the area. Migration, as the in-depth follow-up villages study shows, is facilitating flows of financial and social remittances, which is enhancing the ability of households and communities to act in the context of risks (Reif, 2012). The flows of financial remittances are considered by the villagers to strengthen their adaptive capacities as these flows diversify the risk of income losses, create redundancies and enable proactive investments which reduce vulnerability against environmental risks (e.g. agricultural diversification). More importantly, social remittances initiate learning processes and facilitate innovation. Examples include the adoption of new irrigation technologies and innovative agricultural practices, the introduction of non-agricultural livelihoods activities and the initiation of participatory decision-making processes (Reif, 2012, pp. 76–78). These examples highlight that through an intensifying translocal connectedness, the ability of households and communities to respond to environmental risks and sustain their livelihoods and well-being – that is, their social resilience (see Keck & Sakdapolrak, 2013) – has the potential to be strengthened.

6. Conclusion

The objective of this study is to gain a better understanding of the relationship between rainfall-related events and trends, livelihoods security and migration. It is of particular interest to explore under what circumstances households use migration as a risk management strategy in relation to increasing rainfall variability and food insecurity. The empirical results show that despite exposure to environmental stresses people in the study area clearly stated that they did not consider them as a factor in their migration decisions. Under both past and current situations, households do not perceive migration as a primary strategy to manage environmental risks. For the ‘Where the Rain Falls’ project, the key question that arises is why these communities do not migrate in spite of the exposure to environmental stress. There are a number of factors that might explain the situation. Firstly, the severity of the stress events at present does not exceed the threshold where the basic survival needs of the affected people are threatened. This clearly differentiates the Thai case from other case studies within the framework of the project where environmental stress mediated through food insecurity is closly linked to migration decisions (see Afifi, Liwenga, & Kwezi, 2013 for Tanzania; Etzold, Ahmed, Hassan, & Neelorm, 2013 for Bangladesh and Rademacher-Schulz, Schraven, & Mahama, 2013, for Ghana in this issue) Secondly, the affected semi-subsistence farmers are able to draw on a wide range of resources that enables them to cope with, adapt to and recover from these events without the necessity to migrate. At the household level a high level of livelihood diversification – on-farm as well as off-farm – enables households to buffer losses. As members of the larger community, households have access to credit from community funds and therefore have the possibility to access cash to compensate for losses or invest in recovery actions. Infrastructural investments by the state in the past, such as the building of a water reservoir, not only enable agricultural diversification but can also be considered as a drought risk mitigation infrastructure. This case study has also highlighted that, even if the decision to migrate cannot be directly related to environmental factors, migration is a significant feature of rural livelihoods which must be taken into account for the analysis. Even though migration decisions might not be taken to directly respond to environmental risks, migration is important factor that has to be considered. It is a means by which the rural household's diversity risks of income losses. Through the flow of social remittances (e.g. knowledge, technology) it can facilitate innovation and in doing so strengthen social resilience of households and communities.

As a clear, direct link between migration and rainfall-related factors cannot be established, one might be tempted to conclude that the case study does not add any insight to the rainfall-related events and trends, livelihoods security and migration. In our view, the importance of the case study – as one of eight case studies conducted in three continents – is that it forces those in the field of environmental migration to recognize the complex relationship between environmental factors and migration. This case study calls into question simple and deterministic statements about the relationship between environmental factors and migration, while enabling – in comparison with other case studies – nuanced statements about the nature of these relationships. Finally, the presented results point to the need to consider migration and the translocal connectedness induced by it as a source for social resilience building in the context of climate change (Greiner & Sakdapolrak, 2013). This is an issue which has been widely neglected by climate change adaptation researchers and policy makers.