1 Introduction

Bhutan is a small country, approximately the size of Switzerland, secluded in the foothills of the Himalayas, nestled between China to the north and India to the south (Fig. 8.1). Administratively, Bhutan is subdivided into 20 dzongkhags, which are structurally analogous to states in the United States and, at present, 205 gewogs, which are structurally analogous to counties in the United States.

Fig. 8.1
A map of Bhutan. Some of the primary subdivisions of Bhutan are Gasa, Wangdue, Punakha, Thimphu, Bumthang, Lhuentse, Trongsa, Monggar, Trashi Yangtse, Trashigang, Samdrup Jongkhar, Zhemgang, and Sarpang.

Administrative boundaries of Bhutan

Full size image

The Population Housing Census of Bhutan (PHCB) has conducted two 100% censuses since 2000, one in 2005 and one in 2017. The 2017 census found a total of 727,145 people residing in Bhutan, an increase of 54,720 people from the 672,425 people enumerated in the 2005 census. The most populous Dzongkhag is Thimphu, which houses the capital city of Thimphu, with an overall population of 138,736, accounting for almost 20% of the country’s total population. The least populous is Gasa, the northernmost Dzongkhag, with a population of 3952.

2 Literature on Migration Drivers

Characterizing and understanding migration and its drivers has been a line of research in geography and demography since Ravenstein’s pivotal work in the 1885s providing the “Laws of Migration.” For its strengths and weaknesses, this work helped set the tone for a variety of related research.

Series of factors are taken into considerations regarding internal migration: factors at the origin, factors at the destination, intervening obstacles in between, and personal considerations (Lee, 1966). Researchers may be intimately aware of the factors at the origin, both positive and negative, attracting and detracting, that can urge residents to stay or encourage them to migrate. The factors at the destination are not as familiar to researchers, thus the migrations may be influenced by a perception of these factors. While origin and destination factors certainly influence decisions to migrate, there are also a variety of steps in between the origin and destination, such as mobility, and of course personal considerations that also weigh into the decision to migrate. Migration thus is part perception and part empirical evidence. Human capital models also seek to explain migration drivers, suggesting that migrants experience tradeoffs of costs and benefits associated with moving from one place to another, with the expectation of a net return in moving to their destination. These models mostly indicate that the chief motivator of migration are real and perceived economic advantages (Sjaastad, 1962; Todaro, 1980).

Dorigo and Tobler (1983) took Lee and Ravenstein’s (1885) work further by quantifying and proving a formulaic representation of push and pull factors, using distance as a proxy for obstacles between the origin and destination. Even with this quantification, Dorigo and Tobler recognized that true estimates of push and pull factors would be impossible. These pull (attracting) and push (detracting) forces are intimately intertwined; however, impact migrants differently. While employment and economic factors have a large impact when considering migrating, there are also non-economic factors. Some migrants may be more interested in a particular destination due to the attracting forces, and thus may not necessarily be responding to detracting forces at their origin. Conversely, there are migrants who are motivated to migrate more by the detracting forces at their point of origin than by the attracting forces at their destination. This dichotomy represents a highly nuanced and individual preferences throughout the decision-making process for migrants.

More recent migration literature explores some of these non-economic factors; however, majority of these studies focus on motivations in the developed world. Despite this focus on the developed world, many insights can be gleaned, such as Halfacree’s (2004) call for placing migration more heavily in a cultural context and exploring the multiple motivations for migration. Halfacree explores the idea that questionnaires that only allow one reason for migration overemphasize the economic reasons for migration instead of allowing for the highly intertwined nature of rationales surrounding migration. Morrison and Clarke (2011) sought out micro motives for migration, looking at social, education, housing cost, housing size, environment, and other reasons for internal migration in New Zealand, and found most internal migrant’s primary motives for moving were not primarily employment. While the reasons for migration in a developed nation like New Zealand may differ from developing countries, this research incorporates these cultural motives together with employment-related motives.

Other researchers have examined the drivers of migration by comparing direct and indirect connections. These connections may represent complex multidimensional composition of economic, political, social, and other inequalities and events that dynamically change migration motivations and options for different groups of people. By examining empirical studies Czaika and Reinprecht (2020) identified 24 categories of factors influencing migration processes and decision-making, they outlined the drivers of migration by developing a simple classification scheme consisting of 24 driving factors and grouped them into nine categories: demographic, economic, environmental, human development, individual, politico-institutional, security, socio-cultural, and supranational. Of the nine categories they found economic and socio-cultural as the most cited drivers of migration, the two categories accounting for 47%. The economic drivers of migration included two major categories that were cited the most, labour market and employment conditions, which coincides with other historical studies by Sjaastad (1962) and Lee (1966). This may support the claim that some attracting and detracting forces are historic in nature however, an in-depth review of the geography of migration in current literature is needed to understand if the phenomena is historic or contemporary.

2.1 Internal Migration in Asia

Internal migration is often characterized as a major flow of people in Asia, especially South Asia (Deshingkar, 2006). Recently, one type of internal migration “circular migration” has emerged as a trend in internal migration in Asia, where migrants from rural areas go to urban areas to find jobs in the informal sector (Deshingkar, 2006). Efforts to compare internal migration patterns within the countries of Asia have often met with difficulties in data collection and comparison (Charles-Edwards et al., 2016). One effort to compare the internal migration in Asian countries, Charles-Edwards et al. (2016) compared the Migration Effectiveness Index and Aggregate Net Migration Rate across the countries of Asia, finding variations within the countries, but also some similarities. Countries such as Bhutan, Armenia, Mongolia, Nepal, Timor Leste, and Turkey appear to exhibit characteristics of rural-urban migration. They also found that countries such as Cambodia, China, and Thailand have migration to urban areas as well as to rural areas representing a more complex flow of movement. Other Asian countries, such as India, have lost populations in the densest areas. Other research has identified the rural-urban migration trends in Bangladesh, Pakistan, and India, while also recognizing the urban-urban migration that also occurs in India (Haque, 2005). The IMAGE project also sought to conduct both macro and micro-levels of analysis in internal migration in Asia, comparing the countries among each other, but also providing micro-level analyses detailing patterns unique to each country (Bell et al., 2020). Each research recognizes the nuances within each country, making general trends amongst the countries difficult to associate.

2.2 Internal Migration in Bhutan

Bhutan has widely been cited as one of the countries with the highest rates of internal migration in Asia (Choda, 2012). Studies of internal migration in Bhutan have mostly examined data within the administrative boundaries of Bhutan at the dzongkhag (Gosai, 2009; Gosai & Sulewski, 2020; Ura, 2013) and gewog levels (Choda, 2012; Gosai & Sulewski, 2014), with few focusing on individual communities.

Migration in Bhutan can also be characterized as either permanent or temporary. Literature on temporary migration in Bhutan is sparse, as data on these patterns tends to be difficult to obtain. Temporary migration in Bhutan includes seasonal, circular migration, such as winter labor migration in the rural areas of Bhutan (Chand, 2009). Chand (2009) studied seasonal labor migration in a rural gewog: Lauri, where 80% of those studied migrated during the winter to other areas of Bhutan. Yak herders are another group in Bhutan who migrate seasonally, including alpine meadow grazing in the summer and winter grazing near villages (Wangda, 2016).

Permanent internal migration in Bhutan has more data, and thus is studied more widely (Gosai, 2009, Ura, 2013, Gosai & Sulewski, 2014, NSB, 2018a). There is generally an east-west dichotomy with regards to migration in Bhutan, with larger net out-migration from the eastern part of the country and larger net in-migration to the more populous eastern part of the country (Fig. 8.2). Eleven dzongkhags (55%) exhibit net negative lifetime migration rates, with Trashigang on the far east of the country experiencing the largest net out-migration. Of the 11 dzongkhags exhibiting net negative lifetime migration rates, seven (63.6%) are in the eastern part of the country. Nine dzongkhags exhibit net positive lifetime migration rates, with Thimphu dzongkhag (the location of the country’s capital city) exhibiting the largest net positive migration rates. All nine dzongkhags are located in the central or eastern part of the country. In addition, the age profile of migrants tended to be around the working ages (Choda, 2012; Ura, 2013).

Fig. 8.2
A map of Bhutan. The central from north to south except Ha, Samise, Dagana, and Tsirang have net positive lifetime migration. The remaining dzongkhags toward the east are net negative lifetime migration.

Net positive and net negative lifetime migration by dzongkhag

Full size image

A similar pattern emerges when examining the recent (within five years of the 2017 census) gewog migration rates. Of the 205 gewogs, 138 (64%) are experiencing net negative migration rates for recent migrants (Fig. 8.3). A notable difference is that only approximately 42% of the gewogs with net negative recent migration rates fall within the eastern part of the country. In the central part of the country, approximately 38% of the gewogs have a net negative recent migration rate, leaving only 20% of the gewogs in the western part of the country having a net negative recent migration rate. While different from the dzongkhag patterns, the gewog data for recent migration exhibits a similar east-west dichotomy, with the greatest percentage of gewogs with net negative recent migration rates being in the eastern part of the country, less so in the central part of country, and the least in the western part of the country. Those gewogs with a net positive recent migration rate exhibit a different pattern. Only 28% of the gewogs in the eastern part of the country exhibited net positive recent migration rates, 38% of the gewogs in the central part of the country, and 32% in the western part of the country.

Fig. 8.3
A map of Bhutan. Some minor parts of the western, central, and eastern parts of Gewog have net positive lifetime migration. The remaining majority of Gewog is net negative lifetime migration.

Net positive and net negative lifetime migration by gewog

Full size image

The 2017 PHCB sought to explain these trends by asking migrants their reasons for migrating. The top three reasons in 2017 included familial moves, employment, and education (NSB, 2018b). Other less prominent reasons for migration include resettlement, natural disasters, and security. This research seeks to identify the geographic variables that could help explain the attracting and detracting forces in different regions of Bhutan to help further explain and validate the cited rationales for internal migration in Bhutan.

3 Discussion on Migration Drivers

3.1 Attracting Forces

Historically, economic declines in rural areas have led to internal migration to urban areas while national economic downturns have resulted in regional and seasonal migration from countries surrounding Bhutan. There are a variety of forces that would attract migrants from point of origin to move their residence. The PHCB 2017 identified several reasons for migration, education and employment being most predominant. Those locations with educational and employment opportunities would be more attractive areas for migration than areas that lack such opportunities. Conditions such as educational and employment opportunities at the current residence might be deteriorating making these the attracting forces if in receiving areas show promise and pull people to migrate to urban centers.

3.2 Educational Opportunities

With approximately 8 percent of migrants identifying educational opportunities as the rationale for migration, one might postulate that areas with advanced educational opportunities would receive more inward migration than outward migration (Fig. 8.4). There are sixteen known institutions of higher education spread across the country of Bhutan, with the majority (11) located in the more populous western part of the country. Most of these institutions (70%) are in geogs that have experienced positive net rates of migration within the last five years. This includes geogs in the eastern part of the country that are generally surrounded by areas of high rates of outward migration. For example, Kanglunggeog, located in the far east of Bhutan in Trashigang Dzongkhag, is home to two colleges, Sherubtse College and Yonphula Centenary College, and has the highest positive net migration rates for Dzongkhag. For context, Trashigang Dzongkhag has the highest levels of lifetime net out migration in the entire country. While the PHCB does not have data on international moves for educational opportunities, the age (primarily young adults between the ages of 20 and 29) of emigrants and 2016 Australian immigration data suggests that employment and educational opportunities may be the rationale for emigration (NSB, 2018b).

Fig. 8.4
A map of Bhutan. The maximum number of higher education institutions are in the western regions, one in the central and 4 in the eastern regions, and most of them are in gewogs with a net positive rate of migration.

Educational opportunities with net recent gewog migration

Full size image

3.3 Employment Opportunities

Employment opportunities are often cited as a major reason for moving for Bhutanese internal migrants, and this is echoed throughout the literature on internal migrants, especially in developing countries (Todaro, 1980). Quantifying this motivating factor can be difficult, as this is partly based on the perception of available employment opportunities and not necessarily actual employment opportunities. Examining unemployment data may also not provide the whole picture, since recent migrants may be within the unemployed category while seeking out employment in their destination. One proxy for employment opportunities could be a subset of points of interest from the OpenStreetMap (OSM) database. This spatial database is created mostly by volunteers, and thus is not a complete selection of potential employment opportunities. Despite its shortcomings, this dataset could be indicative of areas of interest to individuals for a variety of motivations.

The points of interest layer in OSM includes several features not necessarily of interest when considering employment opportunities, so the data was subset to include such features as hotels, restaurants, convenience stores, and similar establishments. Features, such as benches and viewpoints, were removed from the analysis, as their purpose is likely not employment related.

The points were analyzed at both the dzongkhag and gewog level, both yielding different results. At the dzongkhag level, 87% of the selected OSM points of interest were in dzongkhags with a greater number of lifetime in-migrants than out-migrants and were mostly located in the western and central part of the country. Additionally, the dzongkhags with over 20 selected features, 80% of them are also in net positive lifetime migrant areas. Thimphu dzongkhag, which has the highest number of net in-migrants in the country, also overwhelmingly has the largest number of employment related features.

While this is also true at the gewog level, one of the gewogs containing Thimphu also has overwhelmingly the largest number of features, a different pattern emerges when examining the data at the gewog level. Of the gewogs with more than 20 employment related features only 40% were in gewogs with a net positive recent migration rate. However, when examining those gewogs without any employment related features from OSM, 78% are located in net negative migration rate gewogs.

While the data at the dzongkhag level may indicate that migrants are indeed seeking out areas where employment opportunities seem more plentiful, the gewog level patterns may actually indicate that the lack of employment opportunities at the local level may be a more powerful driver to explain recent out-migration.

3.4 Detracting Forces

Neoclassical migration theory, based on Sjaastad’s (1962) cost-benefit model, suggests that migrants evaluate the costs and benefits of moving to alternative locations. Similarly, Lee’s (1966) push-pull model of migration suggests that individuals migrate due to economic opportunities, such as employment, at the destination and/or lack thereof at the origin in hopes to find better conditions. There are a several reasons that would detract migrants from leaving their residence. Lack of market access, food insecurity, lack of water, agricultural, and wildlife have been identified by the PHCB 2017 as reasons why people move. Naturally locations that lack market, water, and food access would be less attractive and considered detracting forces for migrants seeking better opportunities.

3.5 Lack of Market Access

One detractive force in various areas of Bhutan is lack of market access, as described in surveys of those in agricultural professions, including yak rearing (NSB, 2018b; Wangda, 2016). The PHCB in 2017 collected data on how far a household must travel to get to a main road, which is available at the dzongkhag level (NSB, 2018a). In most dzongkhags, most households stated that they were within 30 minutes of a road. However, the dzongkhags with the greatest number of households that were greater than 30 minutes away from the nearest road were predominantly within dzongkhags where the net lifetime migration was negative (Fig. 8.5). Eleven dzongkhags had greater than 9 percent of households indicate that they had to travel more than 30 minutes to the nearest road. Of those 11 dzongkhags, 9 (82%) were dzongkhags where net lifetime migration was negative. The dzongkhags where the greatest percentage of respondents indicated that their household was less than 30 minutes away from the nearest road were Bumthang (98%), Thimphu (98%), and Paro (96%) dzongkhags. Thimphu and Paro dzongkhags are in the populous western part of the country, the location of capital city Thimphu and the nation’s only international airport in Paro. Bumthang dzongkhag, located in central Bhutan, is one of the most historically significant dzongkhags in Bhutan, with some of the nation’s oldest temples.

Fig. 8.5
A map of Bhutan with roadways. The roadways are maximum from the western central to the eastern central with some roadways extending toward the south extreme and little toward the north.

Known roads with net recent gewog migration

Full size image

Another potential proxy for market access is the length of roads in each gewog. Of the 205 gewogs, 164 gewogs have documented roads within them, according to the International Steering Committee for Global Mapping data from 2016, obtained from the Bhutan Land Commission (ISCGM 2016). Of the 41 gewogs without documented major roads in them, 35 (85%) have a net negative migration rate for recent migrants. While the presence of roads does not necessarily portend net positive recent migration rates, fewer roads does appear to coincide with net negative recent migration rates. Of the gewogs that have less than 20 kilometers of roads within their borders, approximately 78% of them exhibited net negative recent migration rates. Of the gewogs with less than 30 kilometers of roads within their borders, approximately 83% exhibited net negative migration rates. The reverse was not necessarily true. Of the gewogs with greater than 60 kilometers of roads within their borders, approximately half (54%) exhibited net positive recent migration rates. This pattern could indicate that lack of roads is a proxy for reduced access to markets, thereby considered a detracting force; however, it is important to note that the presence of lengthy road networks is not necessarily an attracting force for migrants.

3.6 Food Insecurity

Food insecurity may also be indicative of a lack of market access. PHCB collected data preceding the 2017 census on the number of households in each dzongkhag who have experienced food security within the 12 months. The dzongkhags that experienced the highest food insecurity are also identified as dzongkhags with the greatest number of lifetime out-migrants. Of the ten dzongkhags that have greater than 6% of households indicating that they have experienced food insecurity in the 12 months prior to the census, 70% have experienced net negative lifetime migration patterns. The dzongkhag with the lowest percentage of households reported having experienced food insecurity is Thimphu, with less than 3% of households. As noted earlier the dzongkhags where households are less than 30 min away from the nearest road were Bumthang (98%), Thimphu (98%), and Paro (96%) dzongkhags.

3.7 Lack of Access to Water

Another detracting force that survey respondents cited was a lack of access to water in their origins. The PHCB in 2017 collected data on several aspects related to access to water including what the main source of drinking for households, how long it takes for households to reach the nearest water source, and whether the water source was reliable. This data was provided by the PHCB aggregated to the dzongkhag level.

There are a variety of different water sources described by the PHCB, but for the purposes of this study, they are divided into water that is piped into the dwelling and other water sources (including water that is piped outside of the dwelling). Dzongkhags where greater than 60% of the households do not have piped water inside their dwelling exhibited predominantly net negative lifetime migration patterns (77%). Access to piped water inside the dwelling may be considered an attracting force for migrations, as the dzongkhag with highest net positive lifetime migration rate is also the dzongkhag with the largest percentage of households reporting piped water inside their dwellings: Thimphu (76%). In addition, of those seven dzongkhags with greater than 40% of the households reporting piped water inside the home, only one exhibited net negative migration rates.

While the water sources appeared to exhibit patterns that indicated that it could be a detracting force, possibly encouraging out migration, how long it takes for households to reach the nearest water source does not exhibit the same patterns. The majority of households in all of the dzongkhags (97% or greater) are within 30 minutes or less of the nearest water source. The dzongkhags with the highest percentage of households that have to travel greater than 30 minutes to reach the nearest water source did not exhibit an overall net negative migration pattern; it was a mixture of dzongkhags with both net positive and net negative lifetime migration rates. The reliability of the water source exhibited similar patterns to the considerations for the time it takes to reach the nearest water source. Households in all dzongkhags indicated that the water sources are between 73% and 88% reliable. Seven dzongkhags exhibited less than 80% of households describing their water source as reliable, and of those seven five (71%) exhibited net negative lifetime migration rates. Those dzongkhags where greater than 80% of households described their water source as reliable were relatively evenly dispersed between net in and net out lifetime migrants.

The examination of various water access related variables indicated that perhaps reliability and distance are not as large of a detracting factor as perhaps the type of access that migrants have at their origins. Data collected at the gewog level may demonstrate different patterns; however, data at gewog administrative level was not available for this study.

3.8 Agricultural/Wild Interface

A survey of migrants described animals destroying crops as a reason to seek employment opportunities outside of rural agricultural areas, and perhaps even migrate to seek better conditions and opportunities (NSB, 2018b). This agricultural-wild conflict could be a detracting force, urging migrants to leave their residence and jobs in search of better opportunities.

The presence of national parks and wildlife corridors could be an indicator of potential conflict between wildlife and agricultural areas. Out of the 205 gewogs in Bhutan, 87 have National Park or wildlife corridors running completely or partially through them (Fig. 8.6). Approximately 66% of all of the gewogs that have national parks or wildlife corridors within them have a negative net migration rate for recent migrants. Thirteen gewogs have greater than 90% of their area covered by national parks, of those nine (69%) have net negative recent migration rates. Similarly, 47 gewogs have over half of their area covered by protected land, and 33 of those (70%) have recent net negative migration rates.

Fig. 8.6
A map of Bhutan. The maximum protected natural land is toward the north and central parts of Bhutan, and small parts of eastern and western regions.

Dzongkhags with protected natural land

Full size image

3.9 Other Considerations

While many of the reasons for migration can be characterized as attracting and detracting forces, there may be other considerations that migrations must weigh. For example, some migrants are relocating as dependents. Those migrating as dependents are predominantly children and female migrants (NSB, 2018a). Another consideration in attracting and detracting forces of migrations are due to changes in marital status. These include migrating to get married, migrating due to a divorce, or migrating due to being widowed (NSB, 2018a).

4 Impact of COVID-19 to the Study Area

The global pandemic caused by COVID-19 (coronavirus disease 2019) was discovered in December 2019 in Wuhan, China according to the Centers for Disease Control and Prevention in the United States. It is very contagious and quickly spread around the world including Bhutan. As of early 2021 there were only 337 physicians and less than 3000 health care professionals serving a population 760,000 however, Bhutan weathered the COVID-19 pandemic far better than wealthier developed nations (Drexler, 2021). As of October 2022, there have been 62,380 infections and 21 coronavirus-related deaths reported in the country since the pandemic began according to Johns Hopkins’s Coronavirus Resource Center. The Ministry of Health has administered approximately two million doses of COVID vaccines so far with at least 2 doses per person (MoH, 2022a).

Research is just now emerging on how Bhutan was able to better manage the global pandemic, through diligent and conscientious leadership that provided provisions and financial means to their citizens on health guidance and a shared responsibility Bhutan was able to achieve and maintain low death rates throughout the pandemic (MoH, 2022b). As the World Health Organization was announcing “a pneumonia outbreak of unknown cause” in December 2019 Bhutan was drafting its National Preparedness and Response Plan and began a strict screening policy at various points of entry affecting the movement of people both internationally and internally. There were other measures in place that restricted the migration of people such as a national mandate to quarantine for 21 days and the establishment of disaster-relief zones that relied on local population for support.

5 Conclusions

Census data can only go so far in helping to explain the drivers of migration, as they often only collect limited information, and thus may conceal more nuanced drivers of migration (Todaro, 1980). The 2017 PHCB identified familial moves, employment, and education as the top three reasons for migrating (NSB, 2018a). While other data collected by the PHCB in 2017 combined with data explored from a variety of different sources including OSM generally reflect these rationales. However, surveys conducted with regards to migration in Bhutan by the NSB (2018b) indicate some more nuanced rationales, including lack of amenities at the point of origin. While the results for these amenities (lack of market access, lack of water access, and the interaction between the rural communities and wildlife) vary, it is clear that these variables may play a role in the decision of a Bhutanese person to migrate. It is also clear from this study that the geographic scale of analysis plays a role; while trends at the dzongkhag and gewog level are similar, there are also some distinct differences. As programs to increase amenity access and create employment opportunities across the country increase and are completed, it is likely that the migration patterns in Bhutan will continue to become increasingly more complex with rural-urban, rural-rural and other patterns of migrants becoming more prevalent. Greater availability of gewog level data would likely provide valuable insights to migration related research in the future as these complex migration patterns may make dzongkhag level and regional analyses less relevant.