Uneven development and displaced care in Sri Lanka

Introduction: Displaced care

The last decade has seen a marked change in the global discussion of international labour migration and its link to development, most notably with the promotion of temporary migration regimes, which orchestrate the flow of labour and care from underdeveloped regions to centres of accumulation. Framed as a ‘triple win’ scenario, this flow – through the transfer of social and monetary remittances – is supposed to produce benefits for migrant workers along with their countries of origin and destination. Migration has been seen as a pathway to economic development (World Bank, 2006). The many contradictions of remittance-driven development, and accompanying expectations of ‘self-help development’ implied by the rhetoric of migrant entrepreneurship, have been subject to critique elsewhere (see Weekley, 2006). Despite numerous analyses of migrant domestic work in global care chains literature (see Ehrenreich and Hochschild, 2002; Gündüz, 2013; Lutz and Palenga-Mollenbeck, 2012), less attention has been paid to the gendered dynamics of migration and development within countries of origin, although valuable contributions have been made by recent research on the social implications for families left behind (see Hoang et al., 2015). The commodification of care at the base of global care chains, in the form of migrant domestic labour, has far reaching socio-economic implications that extend beyond macroeconomic prognoses. In this article we develop the concept of displaced care and apply it to Sri Lanka in order to demonstrate how factors of uneven development – i.e. spatially imbalanced patterns of investment, production and services – have forced the economic migration of domestic workers (internally and internationally). Our concept of displaced care is rooted in the commodification of care work and speaks not only to the creation of complex care deficits for vulnerable migrant households, as per global care chains literature, but the constrained choice and involuntariness that frequently accompany domestic worker migration. Structural economic disparities and socio-cultural inequalities, relevant to both international and internal migration, are seen as integral drivers of such displacement. A fundamental outcome of this process is a transformation of care roles: women who migrate as domestic workers cease ‘caring for’ and ‘caring about’ their families in the immediate sense – i.e. physical and emotional care work (Yeates, 2012) – and instead provide material care through remitted income. Though material care encompasses aspects of caring for and caring about, the spatio-temporal liminality of temporary labour migration ensures that these forms of care are necessarily abstracted in remittances. This transformation of care intersects with women’s productive realities and reproductive expectations, and in turn reorients familial relations. It leads to unconventional and constrained expressions of agency.

Much of the positive spin placed on migration-development in recent years is based on the presumption of skill gains or ‘brain circulation’. This ignores the work that a majority of temporary labour migrants actually engage in. Throughout South Asia, large numbers of ‘low-skilled’ workers enter foreign employment under highly constrained circumstances. Often this is under short-term contracts, as part of the kafala system – the employer-tied visa sponsorship scheme used extensively in the Gulf – or it is part of a web of private recruitment brokers or agents who manipulate both ends of the migration process. This commodification of care work is unlikely to lead to any acquisition of ‘human capital’. Recent research among Sri Lankan returnees has indicated that time spent as a migrant domestic worker ‘had no impact on standards, wages, bargaining power or the assertiveness of the worker’ (Verité Research, 2015: 20). Prevailing gender norms in Sri Lanka and West Asia see the provision of care as a natural and inalienable facet of South Asian womanhood (de Alwis, 1996). Domestic work treats care as an innate cultural characteristic; ‘skills’ are not recognised, let alone developed. This devaluation of care has far reaching implications for the sending country, in this case Sri Lanka. The benefits of male migration are set against the opportunity cost of forgone domestic employment, while care is treated as a ‘free’ resource, to be commodified in exchange for remittances. The invisibility of reproductive work obfuscates the developmental repercussions of these resultant care deficits. This neglect is integral to the dynamic of displaced care in Sri Lanka. The exodus of migrant domestic workers following Sri Lanka’s 1977 neoliberal turn ¹ has been met with enthusiasm for the large remittance inflows it has generated, but without sufficient consideration for the socio-economic or developmental implications of the widespread care deficits it creates (Kottegoda, 2004).

In outlining how uneven development and temporary labour migration intersect to produce displaced care, this article puts the extensive global care chains literature to one side, in order to engage and extend bodies of scholarship that emphasise the internal dynamics and developmental intricacies of migrant-sending countries. In mapping the links between internal and international migration, while differentiating these patterns by gender, ethnicity and class, we build upon a theoretical schema developed by Skeldon (2006) to illustrate how different forms of care work relate to forced economic migration in Sri Lanka, and hence constitute a form of displacement.

Using our concept of displaced care, we emphasise the structural overlap between existing development hurdles and new ones set by the systematic generation of widespread care deficits, and against this backdrop identify unconventional expressions of domestic worker agency that arise within the constraints of commodified care work. The course of this discussion frames an interrelationship between structure and constrained agency (see Goss and Lindquist, 1995) that situates the tensions of displaced care across multiple scales of analysis. We use fieldwork findings from two separate studies – PhD research conducted among returned international migrants (Withers, 2016) and a study into local domestic workers emerging from Piper’s work with Verité Research in Sri Lanka (2015) – that both employ a mixed-methods approach to interviewing migrant workers. Limitations of space prevent us from addressing these findings in detail, but together they provide an empirical backdrop to explore internal and international domestic worker dynamics. We start with an analysis of gendered subordination within Sri Lanka’s uneven development, before deconstructing the patterns of internally and internationally displaced care, and, finally, exploring agency and attitudinal change among domestic workers.

Situating gender within Sri Lanka’s uneven development

Sri Lanka’s economic development has been markedly uneven. Successive waves of colonial occupation under the Portuguese, Dutch and British led to an incomplete spread of capitalist social relations and the reorientation of economic life, centred on urban port cities and chief among them Colombo (De Silva, 2013; Jayawardena, 2000). Despite attempts to decentralise the economy under a socialist model following independence in 1948, the turbulence of competing articulations of political and economic nationalism eventually led to a comprehensive embrace of neoliberalism in 1977, and this has further concentrated development within the urban domain (Biyanwila, 2011). The progressive concentration of production and services in and around Colombo has produced a lopsided pattern of economic growth, one that undermines and disrupts the social and economic sustainability of traditional rural livelihoods, without producing remove decentralised alternatives. Over time, Sri Lanka’s economic mainstay has shifted: from inland agricultural production in the pre-colonial era, through economies of extraction and plantation during the colonial and immediate post-colonial periods, then garments manufacturing in urban export processing zones (EPZs) under neoliberalism, to the current dependence on worker remittances from international labour migration (Kelegama, 2006). However, there has been an embedded common logic of ‘accumulation by subordination’ across colonial and post-colonial periods, where pervasive discrimination along gendered, ethnic and caste lines has kept vulnerable and marginalised populations in degraded forms of employment (Withers, 2016).

This dynamic has played out under neoliberalism by way of a sexual division of labour that locates women’s employment in labour-intensive work. It is a division of labour sustained by countervailing social expectations and material realities. Sri Lankan women are at once portrayed as chaste standard-bearers of tradition and morality reified through their domesticity and care work, while having also become de facto breadwinners amidst the feminisation of labour-intensive occupations that devalue women’s work (de Alwis, 1996). The tensions engendered by these competing pressures are stark. As garment workers, plantation labourers and migrant domestic workers, women make up a majority of the workforce in Sri Lanka’s key export-earning sectors (Samarasinghe, 1998), where gendered tropes of ‘nimble fingers’ are deployed to socialise women’s work as naturally subordinate (Elson and Pearson, 1981). Yet participation in paid work entails a compromised fulfilment of socially ascribed care roles, resulting in pronounced double burdens of productive and reproductive labour or cultural chastisement for neglecting traditional gender roles, as has been well documented in the case of garment workers (Lynch, 2007). Women are thus caught between material pressures to enter paid employment at the lowest tiers of the economy and social pressures to fulfil care work and preserve cultural identities at odds with work in ‘corrupting’ urban milieus.

Sri Lanka’s uneven development thus complicates gender-based identities: as traditional livelihoods in the rural hinterland wane, and gainful employment becomes increasingly associated with the city, so subordinated populations have to migrate in search of work. The marginalised are displaced and become more vulnerable still, both in their susceptibility to exploitation and – in the case of rural women – the stigma of their physical separation from the site of performative domesticity. The contradiction between women’s productive and reproductive pressures could hardly be more evident than in the case of internal and international migrant domestic workers. Their provision of care has itself been commodified, either as hired help for wealthy urban households, or as an export of labour to the oil-rich economies of the Persian Gulf. ² Whether domestic workers find employment in Colombo or migrate abroad is tempered by a multitude of factors, affected by considerations such as geographical location, ethnic and religious difference, the extent of existing wealth and the institutional characteristics of local migration practices. However, both forms of migration share a common root cause in the experience of forced economic migration (Delgado Wise, 2009). The displacement of care that migrant domestic workers embody is less a product of ‘rational choice’ – as conceived by the conservative mainstream of labour migration theory (Borjas, 1989) – but more a result of ‘survival migration’, as it has been termed in the Sri Lankan context (Kottegoda, 2004). A sheer lack of alternative employment or income sources in rural Sri Lanka means that migration for work is a necessity.

Internally and internationally displaced care

Despite sharing a common root in forced economic migration arising from declining rural livelihoods, the demographics of domestic worker migration vary substantially in the contexts of internal or international employment. The following section relates this observation to the contextual specificity of internal and international domestic worker migration in Sri Lanka as evidence of differing dynamics of displaced care.

Temporary labour migration to West Asia has mushroomed since restrictions on foreign employment were lifted following neoliberal reforms in 1977 (Gamburd, 2000). This gradually became cemented as a livelihood strategy for poorer households with limited access to decent work locally, and was facilitated by an expanding brokerage system. Migrant domestic work was at the heart of this trend. Sri Lanka began encouraging temporary labour migration at a time when other large migrant-sending countries in South Asia had placed restrictions on domestic worker employment due to reports of widespread abuse, the resultant demand skewing the gender balance of Sri Lankan departures (Eelens and Speckmann, 1992). Overlapping with the Sri Lankan civil war, the foundational decades of temporary labour migration to West Asia did not involve many Sri Lankan Tamils, as formal recruitment channels did not extend into contested territories in the north and irregular migration from these conflict-stricken provinces was far more common. Rather, it was not until the end of the war in 2009 that Sri Lankan Tamils were drawn into temporary labour migration, though under pronouncedly more costly circumstances. Given a lack of existing assets to leverage for placement fees and an unfamiliarity with the wages and conditions of foreign employment, Sri Lankan Tamil communities were treated as a ‘greenfield’ recruiting ground by agents and brokers inclined to levy higher fees for poorly remunerated positions (Withers, 2016). The Indian Tamil population – still concentrated within the plantations of Sri Lanka’s Central Province – is under-represented in temporary labour migration (SLBFE, 2014), reflecting an enduring confluence of socio-cultural discrimination and productive subordination that results in insufficient means to pursue foreign employment as an alternative.

Indian Tamils are conspicuously over-represented in internal domestic work, both as live-in assistance for wealthy Colombo residencies and also as live-out domestic workers servicing one or more households (Verité Research, 2015). For Indian Tamils the contrast between their opportunities for internal and international domestic migration is a product of the stigma attached to local domestic work and the low cost of taking up these positions. Local domestic work is seen to infer subordinate social standing, quite distinct from the sense of heightened social status often attributed with domestic work abroad (Withers, 2016), as class and ethnic differences stand out in clear relief within private homes and the communities surrounding them. Sinhalese women make up roughly half of local domestic workers (Verité Research, 2015), but avoid such work where possible, generally preferring to work abroad where circumstances permit. Muslim women are typically bound by religious and cultural expectations not to leave the home for work, although international migration is a common exception (Withers, 2016). Indian Tamil women, already positioned as a social and economic underclass, are more responsive to the demand for domestic workers in Colombo.

On the other hand, while the costs of domestic worker migration have gradually been reduced by legal reforms, it is typically easier and less costly to find domestic work locally. There is an important care dimension to this form of employment too, as local domestic work predominantly involves ‘live-out’ positions that allow workers to fulfil care obligations for their own families, although this dynamic tends to apply less for Indian Tamil women, whose families often live alongside the plantations and as such are geographically removed from the site of employment (Verité Research, 2015). The ‘flexibility’ of such an arrangement, in allowing women to balance work with the expectations of familial unpaid care work, was the single largest reason why women had opted for local domestic work (Verité Research, 2015).

Both types of domestic worker migration, internal and international, have implications for the creation of care deficits within the context of uneven development, but whereas live-out domestic work in Colombo facilitates a double burden of paid and unpaid care work, international migration necessarily causes a displacement of care where rigid patriarchal norms prevail. For poor households, the forced economic migration of domestic workers has meant a disruption to familial life and access to care among those who need it most, as men whose spouses have migrated rarely pick up the reproductive slack (Gamburd, 2000). Kottegoda (2004) describes how displaced care work due to widespread migration can leave entire communities wanting for care, in turn fostering a reliance on kinship networks and extended families that turn typical care recipients – such as elderly women and young girls – into care givers. Indeed, another recent study found that such care deficits were frequently filled by grandmothers, aunts and children – that is, almost entirely by other women rather than fathers who stayed behind (Cooray, 2015: 66). Other studies have turned attention to the wellbeing of children left behind, who more frequently drop out of school to assume care roles, experience emotional and psychological harm, suffer from ill-health, and experience abuse at the hands of relatives (Perera and Rathnayaka, 2013). Popular concern for the wellbeing of the children left behind, alongside the treatment of women themselves, has come to the fore in tabloid journalism in recent years, framed in a patriarchal rhetoric that accords moral responsibility to domestic workers themselves. This has resulted in the government introducing selective bans on domestic worker migration, where women are under 25 years of age, or have a child under five years of age in their care. This has not served to ameliorate the situation, so much as to disallow participation in ‘survival migration’ – a path that is a direct result of previous and continuing government policy to promote foreign employment (IPS, 2014).

Lastly, as Skeldon (2006) also notes, there can be significant interlinkages between internal and international migration and this holds for the Sri Lankan case too. Skeldon proposes two general forms of interlinkage: internal migration that leads to international migration, and international migration that leads to internal migration. In Sri Lanka, the former describes domestic workers frequently travelling to Colombo as part of the recruitment process for foreign employment (given the overwhelming concentration of registered recruitment agencies in urban Colombo). Women may migrate to Colombo to work in the garments industry (mainly younger Sinhalese women) or as a domestic worker, but later leave those avenues of employment to undertake international migration. Recruitment options are more regulated and freely accessible in urban areas due to the concentration of registered agents. Conversely, for the latter form of interlinkage, where international migration leads to internal migration, the evidence in Sri Lanka is more tenuous. Findings from studies into domestic worker employment indicate that a domestic worker with international experience has no discernible advantage when pursuing work as a local domestic worker; as they have a greater sense of status or ‘success’ through foreign employment, they may be more reluctant to do so (Verité Research, 2015). The mobility of highly-skilled women workers may have a catalysing effect on internal migration, stimulating demand for domestic workers, but as it is a marginal category of international migration from Sri Lanka we cannot presume this has any meaningful influence on the dynamics of displaced care. Nonetheless, it would be disingenuous to treat either form of domestic work as entirely distinct from the other – it is often where internal and international migration overlap that agency and attitudinal change arises, as we argue in the final section of this article.

Structure–agency interplay: Attitudes towards domestic work

Internationally, the dynamics of migrant domestic work have become increasingly complex, in structural as well as in regulatory terms. New trends indicate a general decline in the supply of workers set against an increase in demand, arising in part from restrictions on domestic worker migration, imposed by origin countries – like Sri Lanka, but also Indonesia and the Philippines – that present travel bans as a form of worker protection (Jayasundere et al., 2015). Such restrictions fundamentally limit migrant agency, by curtailing freedom of movement, while heightened demand has not translated into any real improved bargaining power for extant or prospective migrant domestic workers. Destination countries have not offered improved working conditions: instead they have reoriented recruitment networks to draw from other segments of the global reserve army of care. The agency of international domestic workers remains largely unchanged beyond demographic shifts (Piper, 2008). The same increase in demand for domestic work has also been seen within Sri Lanka itself. Changes in demand emanate largely from a new generation of middle-income employers. Such families tend to comprise double-income-earning couples (itself a reflection of increasing female labour force participation rates, and the increasing cost of living in Colombo). This results in an apparent need to hire a domestic worker, rather than it being a cultural expectation or a sign of social class (Verité Research, 2015). It depends on the specific stage of their life course and personal circumstances. Families may seek full-time live-in workers when they have small children or frail parents. Others prefer live-out workers, whom they hire on an hourly basis. For domestic workers, increasing demand for live-out workers can be beneficial: such arrangements often allow for higher incomes and provide greater control over time and privacy (Verité Research, 2015).

The expectation that returned foreign domestic workers will have their experience and skills recognised, and that this will translate into greater bargaining power as local domestic workers, does not hold true in the Sri Lankan case. There are two principal considerations in examining why international migration does not, in this instance, lead to improved circumstances for internal migrants.

First, international migration connotes a semblance of prestige. It compares favourably with the stigma associated with local domestic work. Working abroad affords domestic workers a degree of ‘invisibility’ from peer scrutiny, facilitating a form of constrained agency in which migrants can re-author their experiences as personal narratives of ‘success’ upon return. Through conspicuous attitudinal change, and the embrace of ‘modern’ values, they can establish an improved social standing relative to non-migrants (Withers, 2016). Participation in local domestic work would undermine this construction of status; repeat international migration is a far more common outcome for foreign domestic workers.

Second, in terms of political agency, neither local domestic workers nor foreign returnees have collectively organised themselves in Sri Lanka. Improving their lot therefore largely depends on other civil society organisations, but these efforts have been constrained by an insufficient understanding of domestic work within a ‘rights’ or ‘development’ framework (Withers and Biyanwila, 2014). Many NGOs have specific mandates determined by their particular manner of engagement, and are reluctant to expand their scope, while others have been preoccupied with issues arising from Sri Lanka’s three-decades long civil war. Unless domestic workers take issues into their own hands, through self-organisation, increased local demand for domestic workers is unlikely to result in any increased bargaining power (Verité Research, 2015).

We should also acknowledge the more unconventional expressions of agency that are rooted in the transformation of care: from caring for and caring about to an abstraction of both by caring materially via remittances. In her work exploring the in-between identities of garment workers in Colombo, Hewamanne explored how the countervailing social and material expectations of Sinhalese women were navigated through ‘suicide narratives’ – futility expressed through jokes, conversation and threats that nevertheless served as a coping mechanism for acute structural constraints on workers’ basic humanity (Hewamanne, 2010). While not ‘agency’ in any traditional sense of empowerment, these narratives represent personal ways through which individuals reconcile, and find new meaning within, trying circumstances. Some returned domestic workers expressed satisfaction in being able to support their families by way of remittances. While unable to provide the physical care demanded by patriarchal framings of motherhood, domestic workers could fulfil their broader obligation to care for the family, through attitudinal changes creating a social acceptance of material care via remittances, that then substituted for traditional care roles. In this way, foreign domestic workers are able to negotiate the double burden of productive realities and reproductive expectations, but not without detriment to personal wellbeing and familial relationships that become estranged between borders.

Indeed, in making sense of international migration, individual wellbeing is often articulated as others’ wellbeing, and the sacrifice necessary to materially benefit the family remains the dominant theme. As one returned migrant from Withers’s research remarked, ‘the daughter studied well, she was staying in a convent. We paid the bills, the son was with the father, it was a good thing, that I went abroad … [but] I feel that they got distant from me. I regret so much, leaving my children’ (Withers, 2016: 156).

Concluding remarks

In this article, we have conceptualised migrant and non-migrant domestic work as displaced care – as such work is explicitly bound up in the forced economic migration that arises from uneven development. For Sri Lanka, this dynamic is moored in a continuum of gendered and ethnic subordination in labour markets, resulting in the creation of ‘invisible’ care deficits, that accrue most harmfully for those at the base of the care chain – i.e. for the households of domestic workers from the poorest backgrounds. Displaced care results from international and internal migration alike, but they have distinct characteristics: international and internal migrants have markedly different demographics, socio-economic circumstances and scope for exercising agency. While foreign domestic workers seem to fare materially better than local domestic workers, and are better able to navigate their in-between identities, neither has been able to negotiate real improvements in their working conditions. A persistent lack of organisation among returned and active domestic workers, both local and international, is the key hurdle to be overcome in order to transform increased demand for domestic work into better outcomes for workers.