Abstract
For decades, rural Thais have been employed as “guestworkers” in the Middle East and Asia. Once the predominate source of guest labor in Taiwan, Thais are being replaced by migrants from other countries. Building on the migration systems literature, we examine this transition. We find that Thais are increasingly deciding against work in Taiwan because of a diminishing wage gap between the two states. We argue that one outcome of the maturation of guestworker programs in Asia and the Middle East is the succession of migrant groups, especially a decline among those from relatively prosperous labor-sending states.
Introduction
Over the past three and a half decades, hundreds of thousands of men from rural northern and northeastern Thailand have been employed as “guestworkers” — migrants admitted on a temporary basis for employment purposes — in factories, construction sites, and farms throughout the Middle East and Asia. Since 1993, Taiwan has been the most important destination for Thai guestworkers, in all years for which we have data (1993–2000 and 2005–present), hosting between one-third and two-thirds of the entire Thai guest workforce (Tsay 2002; TOEA 2016). More recently, however, the number of Thais employed in Taiwan has fallen precipitously from a peak of 149,600 in 1996 to 58,569 in late 2016 (Xing 1997, 2016). Taiwan manufacturing jobs that were once the purview of Thais are increasingly being filled by Indonesian and Vietnamese migrants.
We examine the rise and decline of the guestworker migration system specific to Thailand and Taiwan, embedding the analysis in the broader context of guestworker migration flows both between Thailand and several labor-receiving countries in the Middle East and East Asia and between Taiwan and several labor-sending countries in Southeast Asia. Beginning in the 1970s, Thailand and a few other Southeast Asian states — most notably, the Philippines and Indonesia — began developing labor export programs as a means of earning foreign exchange currency and alleviating domestic unemployment (Arnold and Shah 1984). These nations first exported workers to countries that were wealthy from oil revenue in the Middle East and North Africa; then in the 1980s, Japan and the Asian Tiger states (including Taiwan) became important destinations for Southeast Asian migrants (Hugo 2012).
While a voluminous literature examines the origins of labor export and import programs in the Asian context (e.g., Chan 1999; Jones and Pardthaisong 1999), comparatively little research exists on the contraction of Asian guestworker migration systems. This is in marked contrast to the literature (e.g., Chin 2007; Soysal 1994) on the origins and subsequent decline of the Western European guestworker systems that emerged following the Second World War, studies that include Castles’s (1986) oft-cited “obituary” of those systems. In the European context, guestworker programs were terminated when they were discovered to result in settlement, a consequence of liberal democratic governance systems that recognized migrants’ rights to residence and family reunification (Castles 2006). 1
Implemented following the closure of European guestworker programs and seeking not to replicate the European experience, some Asian states designed their guestworker programs to ensure that migrants would leave upon completion of their contracts (Lan 2007; Tseng and Wang 2013). The Taiwan state — which Tseng and Wang (2013, 16) argue governs with only a “liberal pretense” — has largely succeeded in this task by delegating deportation to migrant brokers. While Taiwan’s guestworker program remains vibrant (Tseng and Wang 2013), its migrant composition has shifted considerably, with the retrenchment of Thai workers being particularly dramatic.
Our analysis of the rise and decline of the guestworker system that once flourished between Thailand and Taiwan suggests that succession of specific migrant groups has been one consequence of the maturation of that program and is likely to characterize the maturation of other guestworker programs in Asia and the Middle East. Unlike the Western European or US experience in which guestworker programs were temporary, Taiwan’s guestworker program shows no sign of abatement, although succession is taking place among the migrant groups participating in the program. In this context, the precipitous decline of Thai labor exports to Taiwan, the most important (in terms of volume of workers) receiving state for Thai workers, raises questions about the long-term viability of Thailand’s labor export program itself.
To address some of these questions, this paper is divided into five sections. The first argues that a migration systems approach may offer the appropriate conceptual tools to analyze the formation of the Thailand-Taiwan migration system. We discuss the emergence of migration systems theory and recent attempts to solidify its theoretical foundation and test it empirically. The second section discusses our methods. The third section describes Thailand’s emergence as a key labor-sending state in Southeast Asia and the short-lived migration system that developed between Thailand and Saudi Arabia. The fourth section analyzes the initial surge and subsequent reduction in the numbers of Thai guestworkers migrating to Taiwan. The final section discusses the significance of our findings for analyzing changes in the national composition of guestworker migration systems in general as well as for Thailand’s future as a labor-sending state in particular.
Background: Migration Systems Literature
A migration system was originally conceived as “a self-regulating apparatus within a wider social system” (Bakewell 2014, 302). Citing work by Arrighi and Saul (1968), Gutkind (1962), Portes and Böröcz (1989), and Richards (1952), Bakewell (2014) has observed that the term migration system was originally used to describe guestworker programs and labor migration flows in Africa. Building on world systems theory, the scholars who initially adopted the concept theorized migration systems at a level of abstraction that did not provide much understanding of those who were actually migrating (ibid.). The succeeding version of migration systems theory, however, put forward by Mabogunje (1970), linked “people, families, and communities over space in what today might be called transnational or translocal communities” and contributed the notion of feedback from destination to origin in influencing subsequent migration behavior (Bakewell 2014, 303).
In the 1980s and 1990s, research efforts to understand the complexity of international migration systems (e.g., Fawcett 1989) drew on the foundational work just cited. Kritz and Zlotink (1992, 2), for instance, argued that to analyze a migration system, it was necessary to attend to “the dynamics associated with the flow, from its origins, through the shifts in its composition and volume as it matures, taking account of return migration and remittances, and of the policy and structural conditions…that shape migration.” They also identified certain attributes common to international migration systems as a whole. First, at least two, but often more, interconnected countries are generally involved. Second, although people move in multiple directions, most movement is to countries with higher wages and welfare standards and away from less advantaged ones, with the bulk of remittance income flowing in the opposite direction. Third, migration volume and direction are influenced by changing social, economic, demographic, and policy conditions, all of which must be understood in historical perspective. Fourth, over time, governments and firms can become engaged in the targeted recruiting of labor, typically in expanding labor markets. Finally, changing socioeconomic and political conditions in sending and receiving states affect both the supply of and demand for labor over time (Kritz and Zlotnik 1992). Each of these points is important in the story we tell.
In the year following the release of the volume International Migration Systems: A Global Approach (Kritz, Lim, and Zlotnik 1992), Massey and colleagues published their seminal article summarizing contemporary theories on international migration and identifying the underpinnings of each one (Massey et al. 1993). Their article was oriented around theories that are important for the initiation of migration flows (macro- and micro-neoclassical economics, new economics of migration, dual labor market theory, and world systems theory) and those that concern the perpetuation of migration (network theory, institutional theory, cumulative causation, and migration systems theory). They argue that the migration systems perspective is less a theory in its own right and more a generalization and extension of the other mentioned theories. Nonetheless, several propositions are associated with migration systems, including that systems are apt to form because of economic and political ties and do not require geographical proximity to function. Equally important, systems may be multipolar “whereby a set of dispersed core countries receive immigrants from a set of overlapping sending nations” (ibid., 454), although membership in multiple migration systems is less typical among receiving than sending countries. Over time, given changing conditions, systems may become less stable, and nations “may join or drop out of a system in response to social change, economic fluctuations, or political upheaval” (ibid., 454).
A number of efforts have since been made to improve or “relaunch” migration systems theory (Jennissen 2007; Bakewell 2014) and test it empirically (van der Erf and van der Gaag 2007; DeWaard, Kim, and Raymer 2012). Nonetheless, our research draws most heavily on the conclusions reached by Massey (2013, 2015) following his evaluation of established migration theories in relation to the “empirical evidence from international migration systems around the world” (Massey 2015, 279). His assessments resulted in the identification of a framework with five characteristics deemed critical to understanding the rise and decline of migration systems. They include (1) the structural conditions in sending countries that produce a population open to the prospect of migration, (2) the structural conditions in receiving countries that create the need for migrant labor, (3) the motivation to migrate internationally among those affected by conditions in both sending and receiving countries, (4) the institutions and social structures that develop along with globalization and perpetuate migration flows, and (5) the government policies devised to respond to changing conditions and the ways they affect the volume and composition of migration flows (Massey 2015).
We argue that these five characteristics are critical to understanding the advent of the migration system that developed between Thailand and Taiwan. We argue further that the origin and contraction of the Thailand-Taiwan guestworker migration system cannot be understood in isolation from the other migration systems in which Thailand and Taiwan are embedded. We begin, therefore, with an analysis of Thailand’s emergence as a labor-sending state (and the structural conditions that set the stage for that development) and a look at the ways in which Thailand’s experience of exporting labor has compared with two of its neighbors, the Philippines and Indonesia. All three countries have exported labor to Saudi Arabia (the most important destination for Thai workers in the 1980s) and Taiwan (the most important destination for Thailand’s workers subsequently).
Method and Sources
We construct our analysis primarily from published sources, including academic studies, newspaper articles, government statistics, and published reports. 2 We also reference results of two months of ethnographic fieldwork conducted in Thailand by the first author in July 2013 and August 2016. 3 The portion of the analysis that focuses on that fieldwork draws on (1) interviews with employees at three labor recruitment companies and with government bureaucrats in the Ministry of Labor in urban Udon Thani, Thailand; (2) interviews in two rural labor-sending communities in Udon Thani that were recommended by bureaucrats as important sources of Taiwan migrant labor 4 ; and (3) an interview with a bureaucrat from the Ministry of Labor in Bangkok.
Guestworker Programs and Thailand’s Advent as a Labor Brokerage State
Guestworker programs are characterized by the admission of migrants on a temporary basis for employment purposes (Castles 1986, 2006; Surak 2013; Tseng and Wang 2013). Hahamovitch (2003, 72–73) contends that historically, guestworker programs have often been implemented in contexts of organized opposition to immigration, where they serve as “state-brokered compromises…to maintain high levels of migration while placating anti-immigrant movements.” In a survey of twentieth-century guestworker programs across the globe, Surak (2013, 86) similarly found that such programs are designed to achieve “malleable labor” without incurring “unwanted members” in the nation-state, an objective they satisfy by “discarding” workers “before they can sink roots in national soil.”
“Labor brokerage states” — those that mobilize their citizens for work abroad to generate foreign exchange currency from remittances — often emerge in tandem with guestworker programs (Rodriguez 2010, v). Thailand’s advent as a labor brokerage state, for example, largely coincided with the implementation of guestworker programs in the Middle East and North Africa in the mid-1970s. Following the massive increase in oil revenue ensuing from the 1974 Arab Oil Embargo, states in the Middle East sought guestworkers to build infrastructure (Woodward 1988). The first Thais employed on these oil-funded development projects were recruited from villages surrounding the American-built Air Force bases in Thailand’s once remote and economically marginalized northeastern and northern regions. During the war with Vietnam, the US military had engaged American and European construction contractors to build bases and roads, usually hiring local labor. In the early 1970s, many of these same contractors procured construction tenders for projects in the Middle East and returned to Thailand to search for employees (Chiengkul 1986; Roongshivin 1986). There, they found large numbers of men whose options in the Thai labor market were extremely limited and were willing to seek higher wages abroad. Consequently, from 1975 to 1990, more than 650,000 Thais traveled to the Middle East and North Africa to work, mostly in construction, as unskilled and semi-skilled contract laborers (Krom Raengngan 1978–1992).
Migrants’ remittances quickly attracted the attention of the Thai state, which came to perceive the growth in transnational migration as a means of reducing domestic unemployment and infusing foreign exchange currency into an economy staggering under the weight of increased oil prices and the American military’s withdrawal (Bank of Thailand 1978). In 1978, bureaucrats from several Thai ministries formed the Committee for Promoting the Employment and Contracting of Thais in Foreign Countries (“Chuai” 1978). Several years later, when Thailand drafted its Fifth National Economic and Social Development Plan (1982–1986), the promotion of “exports of excessive Thai labor surplus to work in foreign countries, especially to the Middle East,” became enshrined as a development objective (Roongshivin 1986, 145). A new office in the Department of Employment was charged with managing the affairs of both overseas workers and the dozens of private employment agencies that had sprung up to capitalize on the demand for Thai workers.
Expanding Programs in Neighbor States and the Turn toward Asia
Thailand’s export of labor to the Middle East occurred alongside the expansion of the labor export program from the Philippines, with Indonesia soon following (Hugo 2012). In the early 1980s, 97 percent of guestworkers departing from Thailand headed to the Middle East, as did 84 percent from the Philippines and 74 percent from Indonesia (Hugo 1998; Skeldon 1998). In the mid-1980s, though, a portion of the migration flows from all three states was directed away from the Middle East, where development projects were delayed following a decline in oil prices (Chaiwongkiat 1985; Woodward 1988), and toward newly initiated guestworker programs in the growing economies of Asian states (Hugo 2012). By the mid-1990s, Asian nations had become the destination for 89 percent of Thai overseas workers and 36 percent of overseas workers from both Indonesia and the Philippines (up from 8% and 11% in the early 1980s, respectively) (Hugo 1998; Skeldon 1998). As we describe in the following, a major diplomatic dispute between Thailand and Saudi Arabia in 1990 was a triggering event initially responsible for the sudden shift of Thai workers toward Asian destinations. We argue, however, that the abiding reason for the more intense Asianization of Thailand’s migration stream relative to its neighbors is that the appeal that jobs in the Middle East held for Thai workers was reduced by a decline in real earning potential, itself a consequence of the willingness of migrants from less prosperous labor-sending states to accept terms of employment that were not satisfactory to Thais.
Although by 1990 Saudi Arabia had become the largest recipient of Asian guest labor during the Middle Eastern development boom (Abella 1995) and the most important destination country for Thai guestworkers (Krom Raengngan 1981–1991), in the same year, Saudi Arabia banned Thai labor imports, following a gem heist by a Thai worker and the subsequent murder of three Saudi diplomats in Bangkok (McClincy 2012). The Thai state made efforts to reestablish itself in Saudi Arabia’s good graces immediately after the murders (Reuters News 1990; Kanwerayotin 1991) and again in 2010 (McClincy 2012), but relations remained strained as the Saudi government continued to express dissatisfaction with the handling of the murder investigation ( Asian Correspondent 2016). The failure to find a timely resolution for the dispute had dramatic consequences for Thai workers intent on working in Saudi Arabia. Whereas in 1989, 41,111 Thais — or 45 percent of the total Thai guest workforce — had been deployed to Saudi Arabia, in 1991, only 85 workers were sent to the kingdom (Krom Raengngan 1990, 1992).
While the gem heist and its aftermath were responsible for the initial precipitous fall in Thai labor exports to Saudi Arabia, the Thailand–Saudi Arabia migration system would likely have declined in the 1990s even if the labor ban had not been implemented. As the 1980s progressed, Thai observers were already noting that the benefits for Thais of migration to the Middle East were diminishing because of rising employment agency fees and declining wages, a development attributed to the steepening competition among migrant sending states (Naithuean 1983; Angsuthanasombat 2001; Suriyon n.d.).
Furthermore, better-paying jobs were opening for Asian guestworkers within Japan and the Asian Tiger economies. In the early 1990s, the overall export of labor from Thailand increased substantially, growing from approximately 63,200 workers deployed in 1990 to roughly 202,300 in 1995, when Thai overseas labor migration reached its peak. 5 When interviewed in 1994, just a few years after Saudi Arabia’s ban on Thai workers, a Thai Department of Employment official boasted, “We’ve secured new and better paid jobs for our people…. Frankly, if Saudi Arabia lifted the visa ban tomorrow and asked for many Thai workers I don’t know if we could find some for them” (Bangkok Post 1994).
Today, Asia continues to be the most important destination region for Thai guestworkers: 67 percent of the 117,291 Thai guestworkers deployed in 2015 were bound for Asian countries (Taiwan, followed by South Korea, Japan, and Singapore). In contrast, in 2015, only 16 percent of departing Thai guestworkers were bound for the Middle East. Of those migrants, nearly half were destined for Israel’s agriculture sector, where wages are relatively high for the region and there is a government-to-government brokered labor agreement that minimizes Thai workers’ travel expenses (IOM n.d.).
Outside Israel, however, Thais now find few employment opportunities in the Middle East. Although there has been a recent upswing in construction activity in the region, a Thai Department of Employment (DOE) official conceded that this upswing is unlikely to translate into jobs for Thai workers: “Countries in this region haven’t set a minimum wage for foreign workers, and there is a high competition to export labor to this region especially with South Asian laborers without skill who are willing to accept a lower wage than Thai workers” (TOEA 2014, 40). In contrast to Thais, Indonesian and Filipino workers continue to compete for jobs in the Middle East. Saudi Arabia remains the second most popular destination for Indonesian workers (after Malaysia), and the UAE and Kuwait continue to receive large numbers of Indonesian migrants (Asian Migrant Center 2017). Similarly, as of 2014, 62 percent of overseas Filipino workers were headed for destinations in the Middle East, and Saudi Arabia remained the top destination of new hires and rehires from the Philippines (POEA 2014, 8).
The Rise and Decline of the Thailand-Taiwan Migration System
The early migration of foreign workers to Taiwan predates the development of a formal foreign labor policy or program in that country. By the late 1980s, however, conditions in Taiwan were such that the move to import low-wage labor was widely seen as necessary. Fertility had recently fallen to below the replacement level, and population aging was becoming a serious concern (Sciubba and Chen 2017). Taiwan’s labor-intensive, export-oriented growth strategies had been deemed a success (Cheng 2002), and the newly educated Taiwan domestic workforce 6 was increasingly moving into more capital- and human capital–intensive industries (Selya 1992). Those who remained in labor-intensive jobs took advantage of the constricted labor supply and the ruling Kuomintang Party’s 1987 revocation of four decades of martial law to push for higher wages (Cheng 2002). As a result, strikes were frequent, and from 1985 to 1989, manufacturing wages rose 60 percent (Tierney 2007).
Some manufacturers sought to circumvent the rising domestic wages and increasingly militant domestic workforce by relocating operations overseas, while others remained in Taiwan but began to hire foreigners who entered the country on tourist visas. Estimates vary widely, as data on clandestine migration are extremely limited, but suggest that by 1990, between 40,000 and 200,000 undocumented migrant workers had entered Taiwan from Thailand, Malaysia, the Philippines, Indonesia, Sri Lanka, and South Korea (Selya 1992; Chan 1999).
Officially, the Thai Department of Employment (DOE) viewed unlawful migration of its citizens to Taiwan with trepidation and urged foreign diplomats to negotiate to reopen the Saudi Arabian market to again channel Thai jobseekers abroad legally (Tunsarawuth 1991). In fact, however, the Thai DOE had long had its sights set on the lucrative Taiwan labor market. Whereas before the Saudi labor ban wages in Saudi Arabia had fallen to the equivalent of 5,000 baht per month or less, undocumented Thai workers in Taiwan could expect to send the equivalent of 15,000 baht per month home (Tourngern 1990). 7 In September 1989, two Thai DOE officials traveled to Taiwan to assess the welfare of undocumented Thai workers and position Thailand for the anticipated opening of Taiwan’s market for 210,000 workers ( Bangkok Post 1989). In 1990, only days after the Saudi labor ban, the Taiwan government announced its intention to authorize a one-time importation of Thai workers to alleviate labor shortages on fourteen major construction projects ( Straits Times 1990). This stopgap measure foreshadowed the substantial transfer of Thai guestworkers that would follow with the passage of Taiwan’s 1992 Employment Services Act.
In 1992, to manage and regularize a foreign workforce that had come to be seen by lawmakers as necessary for sustained economic growth, Taiwan legalized the hiring of foreign workers in the construction, manufacturing, and domestic service sectors (Liu 1996). Guestworker programs in Singapore and Hong Kong, which had been designed to provide employers with cheap labor while minimizing the social disruptions that permanent immigration was perceived to entail, were used as models for Taiwan’s program (Lan 2007). As part of Taiwan’s Employment Services Act, the Council of Labor Affairs (CLA), a bureaucracy charged with handling migrant worker issues, was given the authority to limit the number of foreigners permitted to work in Taiwan and determine their national origin. For the construction and manufacturing sectors, the CLA was responsible for determining the total number of foreigners required (measured as a percent of the total workforce in each sector) and the number of foreign workers that each qualified employer could hire (measured as a percentage of total employees). Limitations on the total amount of time that foreigners were permitted to reside and work in Taiwan (up to three years, later extended to nine years) were established to preclude foreigners’ permanent integration into Taiwan society (Chan 1999).
At the program’s inception in 1992, the CLA allowed employers to source workers from Indonesia, the Philippines, Malaysia, and Thailand. In identifying the states from which to import labor, “the labor quality of workers” and “whether…[the sending government] is willing to assist the migration process and to work within the framework of Taiwan’s labor policy” were two primary considerations (Tseng 2004, 213). In addition, the CLA ascribed importance to “whether importing foreign workers from a specific country would enhance foreign relations between the two states” and “how friendly that specific country is towards Taiwan’s state and people” (ibid.). The CLA’s initial approval of Thailand and other Southeast Asian states was intended to further the Kuomintang government’s “Go South” policy of strengthening relations with its southern neighbors. Vietnam, also an initial target of the policy, was added belatedly (in 1999) to the list of approved source states, perhaps reflecting the influence of China, which monitors Vietnamese foreign relations closely (Ku 1999). 8
Taiwan saw improved diplomatic relations with the large developing states to its south as requisite to its sustained economic growth and as a way to leverage recognition of its national sovereignty. Since the United Nations’s recognition of China in 1972, Taiwan’s national sovereignty has been acknowledged by only about 30 mostly small African and Latin American states (all recipients of Taiwan aid) (Tseng 2004). Shortly before implementing its foreign worker program, Taiwan shifted its international relations tactics from attempting to establish direct state-to-state relations to the more accessible goal of “creeping officiality,” or the acceptance of a partial recognition of its sovereignty through upgraded diplomatic relations, international agreements, and official visits (Chen 2005). Before their approval as labor source states, all four original Southeast Asian sending states responded positively to Taiwan’s overtures for closer diplomatic relations. Each changed the name of Taiwan’s representative office to more closely align with Taiwan preferences. (Taiwan was prevented from establishing an embassy in any of the states because all adhered to the One China policy.) In addition, each state extended most of the twelve privileges typically conferred to foreign diplomats to representatives of the Taiwan government (ibid.). With the creation of its guestworker program, then, Taiwan parlayed its labor shortage into an opportunity to strengthen relations with its southern neighbors, which, for their part, welcomed the opportunity to buttress their labor export programs. The temporary nature of the Taiwan employment contracts served the interests of both Taiwan, which looked to import labor while minimizing social disruption and preventing workers’ permanent integration, and labor sending-states like Thailand, which sought high-earning foreign employment positions for its nationals but wished to prevent their permanent settlement abroad.
The Early Predominance of Thai Workers in the Taiwan Labor Market
Given that Thailand, Indonesia, and the Philippines all appeared to take comparable measures to gain access to the Taiwan labor market, it is important to understand why Thai workers initially commanded such a comparatively large market share. After the implementation of the 1992 Employment Services Act, Thais quickly became Taiwan’s predominate source of guestworker labor. Of the 304,528 work permits issued as of May 1995, for example, approximately two-thirds went to Thais, one-third to Filipinos, and 3 percent to Indonesians and Malaysians (Tsay 1995). Although poorly represented in the initially small domestic service sector (2% of all domestic workers in 1996), Thais composed the vast majority of all foreign construction workers (85% in 1996) and foreign manufacturing workers (65% in 1996) in Taiwan (Xing 1996). 9
The reasons for Thais’ early predominance in Taiwan’s construction and manufacturing sectors requires further scrutiny. One explanation is that Taiwan employers, who became acquainted with labor from the various national groups in the 1980s and early 1990s when hiring undocumented workers, developed a preference for Thai workers. Media reports from the early 1990s, for example, suggest that Taiwan manufacturing employers tended to compare Thai workers favorably to workers from other national groups. Hiring Thai workers was said to be desirable because of their “patience and [the] quality of [their] work” (Tournngern 1990) and their “relative diligence” ( Taiwan Economic News 1994b). A 1993 survey conducted by the Taiwan Council of Labor Affairs found that a majority of employers considered Thais to outperform Indonesians, Malaysians, and Filipinos ( Taiwan Economic News 1994a).
Thai workers’ early popularity was probably also related to official and unofficial dealings between Thai and Taiwan government officials. The first documented Thai workers in Taiwan were recruited for major government construction projects ( Straits Times 1990), and in 1992, Thailand was reported to have been supporting Taiwan’s bid to join GATT in exchange for a greater share of the labor quota ( Taiwan Economic News 1991). Furthermore, Taiwan government officials are known to engage in the practice of “personal diplomacy,” or utilizing private business relationships to further state goals and enable government officials to accumulate personal capital (Chen 2005). Major political scandals in both Taiwan and Thailand, including revelations following a large 2005 riot by guestworkers building the Kaohsiung metro line, exposed the ways in which both Thai and Taiwan government officials had profited from their engagement in the labor trade (Lee 2006; Tierney 2007). 10 The initial predominance of Thais in Taiwan’s workforce, therefore, appears to have been related to a combination of factors: preferences on the part of Taiwan employers for Thai workers, official dealings between Thai and Taiwan government officials, and personal relationships forged by Thai and Taiwan state officials for business purposes.
The Decline of Thai Labor Exports to Taiwan
Despite the early popularity of Thai workers in Taiwan, from 1996 to 2016, the number of Thai workers in Taiwan fell from a peak of 149,600 to 58,569 (Xing 1996; Ministry of Labor 2016). The predominate source of foreign labor in Taiwan until 2007, Thai workers have since been surpassed by Vietnamese, Indonesians, and Filipinos in number (Figure 1) (Xing 1996–2016). This is partially a consequence of the reconfiguration of the migrant labor market (Figure 2), including a great upsurge in the number of workers hired for the domestic service sector, where Thais have always been underrepresented, and a gradual decline in the number of foreign workers hired for the construction sector, where Thais were once the majority (ibid.).

Taiwan’s foreign workforce (1996–2015) aggregated by nationality.

Composition of Taiwan’s foreign workforce (1996–2015) aggregated by sector.
At the same time, however, the manufacturing sector, always the primary employer of Thais in Taiwan, has continued to hire foreign workers in increasing numbers. In 1996, 166,107 foreign workers were part of Taiwan’s manufacturing workforce, and by November 2016, that number had more than doubled to 366,127 (Xing 1996; Ministry of Labor 2016). Yet in this sector, the proportion of Thais declined from 65 percent in 1996 to about 15 percent in 2016 (Figure 3).

Taiwan’s foreign manufacturing workforce (1996–2015) aggregated by nationality.
The proportion of Filipinos in the manufacturing workforce has also declined slightly since 1996 (from 31% in 1996 to 28% in November 2016), while the proportion of the workforce that is Indonesian has increased moderately (from 4% in 1996 to 14% in November 2016) (Xing 1996; Ministry of Labor 2016). The major change in that sector, however, has been the increasing numbers of Vietnamese workers, who in 2011 surpassed Thais to become the largest national group of guestworkers in Taiwan’s manufacturing workforce. As of November 2016, Vietnamese workers constituted about 42.5 percent of all foreign workers in the manufacturing sector (Ministry of Labor 2016). One possible explanation for the rapid ascendance of Vietnamese workers, especially in relation to Indonesians and Filipinos, is that they are willing to pay substantially higher employment agency fees (Hoang 2017). If true, this would be a powerful incentive for recruiters to hire workers from Vietnam over those from other sending states.
Nonetheless, Thai workers appear to remain very much in demand in Taiwan. As the Thai Labor Officer in the Thai Consulate in Taiwan recently explained to a reporter, Thais have not fallen out of favor with Taiwan employers; Taiwan employers are hiring workers of other nationalities because they cannot recruit enough Thais (Tanikuchi 2014). This sentiment was echoed during an interview with a manager of an employment agency in Udon Thani, Thailand. He explained that while he had recruited 600 or so workers per year in the early 2000s, in recent years he had been unable to recruit more than 200 workers.
Interviews conducted in Thai labor-sending communities revealed that going to Taiwan is considered “expensive” (employment agencies are reported to charge 100,000 to 120,000 baht), “the wage isn’t good,” and there are “many [salary] deductions.” According to the head of the Labor Office in Udon Thani, Thailand’s largest labor-sending province, Thai workers have “learned from experience” that employment in Taiwan is no longer profitable. As an officer at the Department of Overseas Employment in Bangkok explained, “Taiwan employers complain, ‘Where are the Thai welders?’ They want our workers, but they need to offer a salary of at least 30,000 baht per month to motivate Thai people to come back to their country.”
A number of factors have contributed to the unattractiveness of Taiwan’s wages in the eyes of Thai workers. Over the past two and a half decades, Taiwan’s minimum wage has been stagnant for long periods. From 1997 to 2007, for instance, it remained at NT$15,840 11 (Hwang, Wang, and Chung 2011), and increases in the minimum wage for Taiwan workers have not been on par with gains in productivity (Yage 2014). Furthermore, in 2001, amid an economic recession, Taiwan’s CLA began to permit employers to make deductions from foreign workers’ salaries for food and accommodation (previously provided free of charge) and services provided by employment agencies (Tierney 2007). Foreign workers paid 2016’s minimum wage (NT$20,008) (Hsiao-han and Hou 2016) only receive NT$13,508 after permitted salary deductions of NT$6,500 (National Immigration Agency 2016). With those salary deductions, the minimum rate at which employers were permitted to pay foreign workers in 2016 (NT$13,508) is less than the minimum wage nearly two decades earlier in 1997 (NT$15,840).
Neither migration institutions (e.g., employment agencies) nor recent Thai government policy initiatives have managed to offset the decline in real wages. Intergovernmental organizations working to maximize the benefits of migration to socioeconomic development continue to advise labor-sending states to implement policies to protect the interests of their nationals. For example, the International Labor Organization (ILO; 2004, 123) recommends that labor-sending states “establish regulations on minimum standards for employment — covering wages, hours of work…. They can also supervise the issue of employment contracts [and] prepare model contracts.” Nonetheless, and despite the Thai government’s attempt to support Thai workers through such provisions, most have been unsuccessful because of Taiwan’s comparative strength in the partnership. In 2012, for example, the Thai DOE’s attempt to lobby the Taiwan CLA for a reduction in food and accommodation deductions failed (Chen 2012). Further, although the Thai DOE has come to two agreements with the Taiwan CLA on reducing employment fees (in 2001 and again in 2012) ( Taiwan Economic News 2001; Chen 2012), fieldwork in labor-sending communities in Udon Thani revealed that all known agencies were charging more than the legally stipulated limit. That the Thai DOE has not made good on its repeated promises to crack down on employment agencies that overcharge ( Bangkok Post 1999) is not surprising. Taiwan employment agencies reportedly sell employment positions to agencies in labor-sending states, and this cost constitutes the largest component of the fee that workers are charged (Khongchuai 2001).
In addition, while wages were declining in Taiwan, wages in Thailand were moving in the opposite direction. The Thai 2016 minimum wage (7,800 baht per month) represents an increase of over 300 percent from the 1993 rate (2,530 baht per month). Moreover, in Thailand’s industrial sector, minimum-wage jobs (those paying 7,800 baht per month) are now largely filled by foreign workers from Myanmar. According to one source, Thai workers now expect to earn at least 10,400 to 11,700 baht per month ( Thansetthakit 2016b). The rise in Thailand’s wage rates over this 24-year period included adjustments to inflation as well as two periods of dramatic increases, the first forced by Thailand’s labor movement, which through strikes and protests achieved an 8 percent annual increase in real wages from 1990 to 1996 (Campbell 2016), and the second through the 2011 fulfillment of a campaign promise by former Prime Minister Yingluck Shinawatra, who cultivated electoral support among the working class (Brown 2012).
It is important to note, though, that aggregated wage data hide unevenness in Thailand’s development. The motivation to migrate internationally — one of the five factors leading to the rise of migration systems, according to Massey (2015) — has not disappeared in Thailand’s labor-sending communities, and this motivation has deep roots. Jones and Pardthaisong (1999, 40), in fact, suggest that a “migration culture” emerged in a number of rural northern and northeastern Thai communities when the first generation of migrants returned from the Middle East with newfound prosperity. Even this migration culture, however, was an extension of a practice increasingly common after World War II, whereby rice farmers in the north and northeast migrated to Bangkok or the Thai Central Plains to find wage labor to supplement their income from farming (Kirsch 1966; Klausner 1987). Today, much of the rural population of northern and northeastern Thailand continues to rely on a mix of farming and off-farm wage earning, much of which continues to require migration, for their livelihoods (Rigg, Salamanca, and Thompson 2016).
As the lead author found during fieldwork in Thai labor-sending communities in 2013 and a 2016 interview with a DOE bureaucrat, employment positions in agriculture in Israel and manufacturing in South Korea — destinations where recruitment fees are low and wages are relatively high — are limited in number but still much sought after by inhabitants of Thai migrant-sending communities. At the same time, potential migrants also appear to be increasingly selective about the destinations to which they are willing to travel. One employment agent interviewed by a journalist summarized this development as follows: Thai workers would fight with one another for international employment positions that didn’t require educational qualifications, that required only that prospective migrants be hard-working. Whatever they were required to pay, they would go…. Yes, maybe in the past [jobseekers had this mindset]. At present, workers are already knowledgeable. If it’s [an unprofitable migration experience] no one will want to go. (
Mathichon Raiwan 2012)
Conclusion
Our primary objective in this research has been to document, in historical context, the rise and decline of the guestworker migration system that developed between Thailand and Taiwan. Based largely on European experience, scholars sometimes assume that the “universal tendency” of guestworker programs is “to end by creating what they set out to avoid: permanent immigrant communities” (Surak 2013, 101). Yet as Castles (2006, 746) has indicated, it is possible that temporary workers may indeed remain temporary in “less democratic states, which deny rights to foreign workers, restrict access to the legal system, and make draconian use of deportation.” Taiwan’s guestworker program, embedded in a less democratic context, has not been cancelled or reduced despite the importation of large numbers of workers. Instead, as our research demonstrates, as the program has developed, succession among migrant groups, particularly of Thai workers, has been occurring. Our findings on the decline of the Thailand-Taiwan migration system suggest that similar guestworker programs in illiberal states in Asia and the Middle East may also mature in ways that allow their programs to continue to flourish while promoting succession among migrant groups, typically involving a decline among those from more prosperous labor-sending states. This decline attests to the tendency of real wages in illiberal guestworker programs to remain stagnant over time or increase at rates insufficient to entice jobseekers from more prosperous labor-sending states to continue to migrate.
To construct our analysis, we have drawn on Massey’s (2015) five features of international migration systems: (1) the structural conditions in sending countries, (2) the structural conditions in receiving countries, (3) the motivation to migrate internationally, (4) the institutions and social structures, and (5) the government policies devised to respond to changing conditions. While this analytic is highly useful for elucidating the origins of the Thailand-Taiwan guestworker system, it may be less useful in predicting a system’s contraction. Today, each of the features that led to the Thailand-Taiwan migration system’s origins remains in place. There are insufficient employment opportunities in rural northeastern Thailand; Taiwan’s workforce is aging and reluctant to engage in manual labor; rural Thais, particularly in disadvantaged parts of the country, continue to seek employment internationally; employment agencies still facilitate the movement of workers; and Taiwan continues to encourage the importation of guestworkers while Thailand continues to seek to export its labor. That each feature contributing to the origin of the system endures reinforces the necessity of analyzing it and other systems in the larger context of the interstate guestworker labor markets in which they are embedded if we are to understand the creation, perpetuation, and transformations of such systems.
As the Thailand-Taiwan guestworker system evolved — particularly when the Taiwan state introduced legislation enabling employers to make deductions from worker wages for room and board — real wages attainable by guestworkers in Taiwan lost ground against rising wages in Thailand. The Thai government attempted to counteract this development by protesting Taiwan’s deduction policy and attempting to reduce employment agency charges. Nevertheless, its failure is illustrative of the constricted position of labor-sending states within guestworker labor markets more broadly. As long as the nationals and governments of other labor-sending states are willing to pay employment agency fees as demanded and accept wages as offered for work in Taiwan and other labor-receiving states, sending governments, even middle-income countries hoping to export workers, will have little bargaining power.
Relatively prosperous among Asian states seeking to export low- and semi-skilled labor, Thailand’s position as a labor-sending state seems poised to decline. While the number of Thai guestworkers sent abroad in any given year continues to fluctuate in response to political and economic events, guestworker emigration from Thailand is experiencing an overall reduction, a trend that we contend is likely to continue for the foreseeable future. Like Taiwan, many other states that receive guestworkers either do not have minimum wage provisions or allow employers to pay guestworkers lower wages than local workers. In such contexts, wages may be set at rates that entice workers from less prosperous nations but that are no longer sufficiently high to entice workers from comparatively affluent labor-sending states like Thailand.
We argue that the case of the Thailand-Taiwan migration system’s rise and fall is useful in understanding the maturation of other guestworker migration systems in Asia and the Middle East and perhaps more broadly. In contexts in which labor-exporting states of varying economic strengths vie for limited employment positions in guestworker labor markets, remuneration will fail to suffice for workers from more prosperous states. Consequently, opportunities will open for migrants from less affluent ones, and the race to the bottom in terms of wages will continue. Like other guestworker programs, Taiwan’s began as a concession to employers. Implemented following years of unrest that brought about a significant rise in wages, it was widely seen as an affront to the labor movement. Ultimately, the succession we document is a victory for employers who can continue to depress wages at the expense of all workers.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
