Gender and Migration: West Indians in Comparative Perspective

Transcription

1 doi: /j x Gender and Migration: West Indians in Comparative Perspective Nancy Foner* Emigration has a long history in the West Indies. Its roots go deep traceable to the legacy of slavery, the distorting effects of colonial rule, the centuries-long domination of the islands economies by plantation agriculture, and, in recent years, continued dependence on world powers, lending institutions, and corporations. Over the years from the end of slavery to the present West Indian women as well as men have been part of various migrant streams as the search for a better life has taken them all over the globe, to Central America, Britain, Canada, and the United States. 1 It is only in recent decades, however, that women have come to dominate major West Indian migrant flows, so that questions pertaining to gender and migration have taken on special relevance in contemporary studies of West Indian migration. This article explores gender issues in West Indian migration by taking a comparative cross-national perspective. The focus is on the three major West Indian migration movements of the mid- and late-twentieth and early twenty-first centuries to the United States, Britain, and Canada. A comparative approach has a number of benefits for the study of West Indian migration (Foner, 2005). It not only points to similarities and contrasts in gender-related patterns among West Indian migrants in the United States, Britain, and Canada but also forces us to try to account for them. It brings out, in an especially dramatic way, the role of the context of reception and the receiving country s immigration policies in shaping male-female differences in West Indian migration flows as well as immigrant adaptation. * Department of Sociology, Hunter College and Graduate Center, City University of New York, USA. Published by Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK, International Migration Vol. 47 (1) 2009 and 350 Main Street, Malden, MA 02148, USA. ISSN

2 4 Foner The comparative analysis of the three migrations in this article explores the reasons for and patterns of West Indian migration as they relate to gender, including the practice of leaving children behind in the Caribbean, as well as aspects of the labour market incorporation of West Indian men and women when they have arrived and settled in the migrant destination. More specifically, the comparisons raise some intriguing questions. Why, for example, did West Indian women comprise a greater proportion of the migrations to the United States and Canada than to Britain? Why were West Indian women more likely to work in caregiving jobs in private homes in the United States and Canada than in Britain? And have the dynamics of transnational motherhood differed in the North American and British contexts? The three migration movements, taken together, have involved hundreds of thousands of people. The mass inflow to Britain began after the end of World War II, lasting from the early 1950s to the 1960s. In 1951, the West Indian immigrant population was tiny, only a little over 17,000; by 1966, it had grown to 269,000, with more than half living in London. The movement had slowed to a trickle by the late 1960s, after passage of restrictive immigration laws in Britain; by 1974, the whole cycle of primary migration from the Caribbean was basically over, and, after that, the Afro-Caribbean population grew through increases in the number of British-born. Just as West Indian migration to Britain was ending, it began to take off in North America, most notably in the United States, where the 1965 amendments to the Immigration and Nationality Act made it possible. After several decades of heavy migration, the West Indian immigrant population of the United States stood at about 1.1 million in 2000, with about half living in New York State, and many thousands have continued to arrive since then (in 2007, 18 per cent of the Jamaican immigrants in the United States 101,000 people had arrived between 2000 and 2007) (Camarota, 2007). Canada, too, opened its doors to West Indian migration in the 1960s, following reform of racially discriminatory immigration policies; between 1961 and 1996, according to 1996 census data, 271,000 West Indians immigrated to Canada, with Toronto the most popular destination (Darden, 2006: 152). What follows draws heavily on the secondary literature on West Indian migration but it is also based on my own research among Jamaican migrants in London (in the 1970s) and New York (in the 1980s) as well as in Jamaica itself (in the 1960s) (Foner, 1973, 1978, 1986, 1987, 2001,

3 Gender and migration ). For this reason, but also, in many cases, because of data availability, I often use material on Jamaicans who are, it should be noted, the largest West Indian nationality group in all three countries. The spotlight in this article is on first-generation immigrants in the periods of mass immigration, which is why I focus on the 1950s and 1960s in the British case and the decades of the late twentieth century in Canada and the United States. GENDERED PATTERNS OF MIGRATION For much of the past, West Indian migration was dominated by men who left their home islands to work, often on agricultural and construction projects abroad. The late nineteenth and early twentieth century movements to Panama and other Central American countries, to other parts of the Caribbean (including Cuba), and to the United States were heavily male as one scholar has put it, remarkably so (Senior, 1991: 108). In Barbados, fully 70 per cent of the total net emigration over the period was of males (Senior, 1991: 108). 2 Overall, external migration from Jamaica between 1888 and 1921 was highly sex-selective with a much greater proportion of males than females involved (Roberts and Sinclair, 1978: 8). 3 In the first attempt to build the Panama Canal in the 1880s, one-fifth of the Jamaican male population of reproductive age went to Panama. After World War I, many thousands of Jamaicans went to work in Cuba s sugar industry, an estimated 60 per cent of them male although some 25,000 women were lured by opportunities in personal service, dressmaking, and laundry work (Eisner, 1961:151). Fifty-five per cent of early twentieth century foreign-black migrants to the United States were men (Reid, 1969 [1939]: 236). This male-dominant pattern continued after World War II, characterizing the huge migration to Britain, but it changed once the destinations shifted to the United States and Canada in the late 1960s. Causes of migration Does the change to a female-dominant pattern in West Indian migration flows in recent years mean that the underlying causes for leaving the Caribbean have also shifted? The answer is no. Scarce resources, overpopulation, high unemployment and underemployment, limited possibilities for advancement these have long spurred West Indian men and women to look abroad for economic security, better job prospects, improved living standards, educational opportunities, and ways to earn money and

4 6 Foner get ahead. In the past few decades, economic crises along with inflation and unemployment have continued to fuel migration fever. West Indian small islands simply cannot deliver the kinds of jobs, lifestyles, and consumption patterns that people at all levels of society want. And increasingly they want more, due to such factors as improved communications, the promises of new political elites, the expansion of educational opportunities, as well as reports and visits from migrants themselves. In this regard, it should be emphasized that limited employment opportunities in the West Indies have long had a deep impact on women as well as men. Women in the Anglophone Caribbean have always worked, going back to the days of slavery when women toiled on the plantations (Massiah, 1986). Since emancipation in the 1830s, West Indian women have been actively involved in occupations outside the home, although they have suffered high rates of unemployment, typically higher than men (Massiah, 1986; Senior 1991). In the late 1970s women constituted about 40 per cent of the Jamaican labour force, although they had more than twice the rate of male unemployment, a situation that had not improved by the late 1990s, when women were 46 per cent of the labour force (Austin, 1979: 513). Whenever they have migrated abroad, West Indian women have intended to work for wages. This has been a basic part of their plan. The British case Given these intentions, one might expect that male-female migratory patterns would be similar in both the British and North American cases. This is, in fact, not the case. In Britain during the years of the mass inflow, West Indian men were generally in the majority. The man in the family often emigrated first, later followed by his wife (or common-law wife or female partner) and children. As Ceri Peach puts it, women following their menfolk represented a continuing characteristic of the mass movement to Britain in the 1950s and early 1960s (Peach, 1968: 45). To be sure, West Indian migration to Britain was characterized from the start by a high percentage of women; they even outnumbered men in a few years of the massive inflow. In her study of migration from the island of Nevis, Margaret Byron notes that migration to Britain was the first labour movement in which women migrated in almost equal proportions to men [emphasis mine] (Byron, 1998: 221; Byron, 1994). Overall, however, for West Indians in the peak years of the migration, between 1952 and 1962, the proportion of men was higher than for women in nearly all years; about 70 per cent of West Indian migrants in

5 Gender and migration were men, and close to 60 per cent in 1955, 1956, and After the imposition of immigration controls in Britain in 1962 and throughout the rest of the 1960s, the migration consisted mainly of those classified as dependents, overwhelmingly children but also a good many women (Rose, et al., 1969: 76, 89). Between 1961 and 1966, there was a substantial swing towards female immigrants, with net arrivals of West Indian women exceeding those for men (Rose et al., 1969: 105). Still, in 1966, males dominated England s West Indian immigrant population 1,066 males per 1,000 females among Jamaican immigrants, 1,026 males per 1,000 females among other West Indian immigrants (Rose et al., 1969: 105). At first glance, this migration pattern might suggest that West Indian men are more adventurous, independent, and ambitious than West Indian women, setting out for foreign parts while their womenfolk trail behind. This is hardly the case. West Indian women have tremendous ambition and desire to work and be financially independent. As a number of scholars have noted, the desire for economic independence is a strong characteristic of West Indian women and a vital component of their self-image (Senior, 1991: 115; Anderson, 1986: 315). Elizabeth Thomas-Hope has written that the lower class West Indian woman just as much as males of the same social status, has a high migration potential because of her equally high level or motivation for securing employment and ultimately some measure of upward mobility (1992: 4). Just as wage work is normal and accepted for women in West Indies, so, too, there is no social barrier preventing women from migrating without and prior to a man. Nor do wage-earning opportunities in Britain seem a significant factor in explaining why men often led the way, since these opportunities were available for both men and women in Britain. (This, it should be noted, contrasts with the migrations a century ago to Panama and Central America, where the employment opportunities were mainly for male workers on plantations, canal-building, and railroad construction.) In most cases, there was simply not enough money for the whole family to emigrate to Britain together, and men, as the main expected family providers, probably received preference in raising the rather considerable funds to pay for the passage. The preference for men to leave for Britain as the solid core of migrants is part of Peach s argument about the impact of labour demand on gender ratios among West Indian migrants. He argues that the economic climate in Britain affected the migration of West Indian men more than women, the women in many cases joining relatives who had already

6 8 Foner migrated; during the 1958 recession the proportion of men arriving fell, but rose with recovery (1968: 43 45). 4 The US case If anything explodes the myth of the dependent female West Indian migrant, it is the movement to the United States and Canada of the past few decades. There, West Indian women, not men, have dominated the migration, and women frequently came on their own, before their spouses or children. In both the United States and Canada, the nature of immigration policy indeed the very fact that legal rules defined who could enter the United States and Canada in contrast to the open immigration from the West Indies to Britain prior to 1962 in combination with job opportunities were largely responsible for this pattern. At the time of the last US census, in 2000, 55 per cent of West Indian immigrants in the United States were women (Holder, 2007: 677). 5 In virtually every year since 1967, West Indian women in the legal stream to the United States have outnumbered men. The proportion of women was particularly high in the first few years of the new (post-1965) immigration, in the case of Jamaicans as high as 76 and 73 per cent for 1967 and 1968 and levelling off in the 1970s as well as in the 1980s and 1990s to between 52 and 54 per cent (Donato, 1992; Zhou, 2002). The 1965 immigration legislation and subsequent revisions favoured persons in particular occupations as well as close relatives of US citizens and permanent residents. In the early years of the new immigration, it was easier for women than men to qualify for labour certification largely due to the demand for domestic labour in American cities. Figures for Jamaicans show that the percentage of total workers immigrating who were classified as private household workers peaked in the very same years that the percentage of women migrants was so high: in 1968, 50 per cent of total Jamaican workers were listed as private household workers; in 1967, 48 per cent. 6 Women could also easily obtain visas as nurses, and in fact about one-third of legal Jamaican immigrants classified as professionals between 1962 and 1972 were nurses (Palmer, 1974: 576). Several decades later, between 1990 and 1992, according to Immigration and Naturalization Service data, 28 per cent of the professional and technical immigrants from Jamaica were nurses (Palmer, 1995: 33). As the migration progressed, and a larger proportion qualified for immigrant status on the basis of family ties rather than occupation, women

7 Gender and migration 9 were probably as likely as men to have relatives in the United States to sponsor them a major reason why the gender ratios became less lopsided in favor of women after the first three years of the new immigration. Between 1983 and 1991, of the over 90,000 Jamaican immigrants legally admitted to the United States, 90 per cent entered the country through family reunification provisions of US immigration law. Of the 10 per cent who came through occupational categories, the large majority were in the nonprofessional category frequently used by live-in domestics, which was cut back by US immigration legislation in 1990 (Kasinitz and Vickerman, 2001: ). Unfortunately, estimates of the gender composition of the undocumented stream are not available, but women no doubt have made up a substantial proportion, partly because of their ability to find jobs in private households as domestics, childcare workers, and companions to the elderly. Undocumented West Indians, men and women alike, have nearly always entered the United States legally on temporary visitors visas and then overstayed the visas. The Immigration and Naturalization Service estimated that in 2000, 44,000 Jamaicans and 34,000 Trinidadians resided in the United States illegally. This was a time when, according to the census, 513,000 foreign-born Jamaicans and 194,000 Trinidadians lived in the United States. It was common in the post-1965 period for women to migrate to the United States first, often followed by their husbands. Many women in the Jamaican village I studied in the late 1960s left their families to take jobs in New York (Foner, 1973). Mrs. R., for example, a middle-aged higgler (market woman) lived on a hill behind the house where I stayed. Soon after I left Jamaica, I learned that she had moved to New York as a live-in domestic worker. Like Mrs. R., once Jamaican women established themselves in New York, they frequently sent money and plane tickets for children and sometimes husbands as well. A few Jamaican women I interviewed in New York in the 1980s told me they had been reluctant to emigrate but were pushed to pave the way for men. They came first because they, rather than their husbands, could get an immigrant visa. Some Jamaican migrants never sent for their husbands at all, and a number of women I interviewed had actually moved to New York to bring about or formalize a separation. The New York migration stream has included many West Indian women without spouses single, divorced, separated, or widowed who, once in the United States (at least at the beginning), often took up live-in employment as private household workers. The easy availability of this type of work may well

8 10 Foner have encouraged or enabled such women to come to the United States in the first place. Family arrangements in West Indian societies are also part of the picture. Although at any given time about half or more of the women in West Indian societies are married or in common-law unions, large numbers have relatively loose ties to men in what have been called visiting unions or are single, usually through separation, divorce, or widowhood (Powell, 1986). The Canadian case In the Canadian case, the role of immigration policy stands out even more starkly. Since the 1960s, women have always outnumbered men in the West Indian migration streams; according to 1996 census data, black women immigrants from the Caribbean exceeded the number of black men by 1.28 times for all periods between 1961 and 1996 (Darden, 2006: ). Or to put it another way, among West Indian immigrants counted in the 1996 Canadian census, women made up 57 per cent arriving in Canada in the period (Darden, 2006: ). 7 The gender imbalance is related to Canadian immigration policies. What is often pointed out is the role of government-run domestic schemes in which West Indian women were allowed entry into the country if they agreed to work as domestics (Henry, 1994: 30). In the 1970s, the West Indies constituted the largest source region for foreign domestics in Canada (Bakan and Stasiulis, 1997). Many West Indian women arrived in Canada from the mid-1950s to 1960s under the Caribbean Domestic Scheme, a programme that lasted until 1966 and permitted single women between eighteen and forty with no dependents and at least an eighth grade education to move to Canada as landed immigrants on condition that they remain in live-in domestic service for at least one year. Between 1955 and 1960, an average of three hundred West Indian domestics a year were admitted to Canada, the number increasing to one thousand a year between 1960 and 1965 (Darden, 2006: 157). West Indian domestics continued to arrive in the 1970s under the Temporary Employment Authorization programme, this time on short-term work permits allowing them to stay in Canada as long as they worked for a designated employer. In 1981, with the implementation of the Foreign Domestic Movement programme, West Indians admitted to Canada on employment authorization could apply for permanent resident status after they had been in the country for two years. Ten per cent of the domestics admitted under this programme between 1982 and 1990 were from the Caribbean.

9 Gender and migration 11 By the 1990s, the number of Caribbean women entering Canada as livein domestics had dropped sharply, and a shift to Filipino domestics had to do with government-mandated education requirements in the live-in domestic programme and the availability of a more malleable labour supply a topic I explore in more detail below. It should be noted, too, that West Indian domestics, as well as others admitted through the Canadian point system on the basis of skills or employment, often have paved the way for relatives (females as well as males) to follow on the basis of family ties. LABOUR MARKET INCORPORATION This leads to the question of jobs. West Indian women left the Caribbean to work and wherever they went, they entered the workforce in exceedingly high proportions. In Britain in 1966, when most West Indian women were first-generation immigrants, two-thirds were economically active in London and the West Midlands; in 1990, 75 per cent of foreign-born Jamaican women 16 and over in the United States were in the labour force; and in 1985, more than 75 per cent of foreign-born West Indian women 15 years and older in metropolitan Toronto were in the workforce (Rose et al., 1969; Zhou, 2002: 31; Henry, 1994: 103). 8 In all three countries and in the three major cities of West Indian settlement (London, New York, Toronto) West Indian immigrant men and women tended to cluster in different occupational spheres, in quite a number of cases the same concentrations. Much of the employment for West Indian immigrant women has been in women s work. When I did my research in London and New York, for example, West Indian immigrant women were often found in the health-care field as nurses and nurse s aides in hospitals or in various kinds of clerical and sales jobs, and in Toronto, West Indian women often work in health care as well. 9 West Indian immigrant men in London often worked in transport on the railroads, buses, and underground just as many in New York City became truck, bus, or jitney van drivers. A combination of skills West Indians brought with them nurses being especially noteworthy in this regard and available opportunities in the destination city mostly explain these similarities. Once established, occupational niches became self-reproducing as new arrivals learned about and got help finding jobs through personal networks in the immigrant community (Waldinger, 1996; on the development of the West Indian niche in health care in New York City see Foner, 2000: 91 92).

10 12 Foner This is one part of the job story. Another is differences in West Indian immigrant women s (and men s) job patterns in the receiving societies. The structure of demand, together with ethnic racial patterns in the labour market, helps account for the contrasts. This comes out in a comparison of New York and London. In London of the 1970s, for example, many West Indian women were office cleaners a field they did not enter in New York City, where West Indians English language ability made other options more attractive and which, instead, drew on the large number of newly arrived Latino and other non-english speaking immigrants. Largely for the same reason, West Indian immigrant women in New York City did not go into the garment industry an ethnic niche there for Chinese and Latina immigrant women whereas quite a number of their counterparts in London worked in clothing factories and other manufacturing jobs when they arrived (Rose et al., 1969; Davison, 1966). That the West Indian migration to New York, and elsewhere in the United States, in the last few decades has included a good many men and women with a college education also helps explain why a substantial minority are found in various professional and managerial positions there as compared to Britain in the period of mass migration when the proportion in the professions was lower and the migration was, overall, less skilled (Foner, 2005: ). 10 The context of reception has operated in yet another way to influence labour market incorporation. This has to do with the system of healthcare in the three countries. In the United States, in the absence of national health insurance, women s as well as men s job choices often have been influenced by the availability of health benefits. In my New York research, I found that West Indian women the unmarried or those whose spouses lacked health benefits often sought out (and stayed) in particular jobs, despite unpleasant conditions and poor pay, in order to provide health coverage for their families (Foner, 1994). National health insurance, and other government benefits available to all British citizens (including West Indians), led to a different job calculus in London and no doubt in Canada as well, where national health insurance is available to permanent residents in Toronto after a threemonth waiting period. WEST INDIAN DOMESTIC WORKERS And then there is the West Indian domestic caregiver, a position held by many West Indian women in New York but never prominent in London

11 Gender and migration 13 and experiencing marked decline in Toronto. The complex factors involved come out in an examination of the major cities for West Indians in the three countries. In New York City, West Indian women make up a sizable proportion more than one out of ten of the city s domestic workers. One might almost say that domestic work is part of the New York West Indian female experience. Whatever job West Indian immigrant women now have, large numbers, at some point in their work careers in New York almost always at the start had a spell as a domestic worker. 11 On sidewalks and in parks and apartments throughout New York City, the nanny caring for a young child or the woman helping an elderly, usually white, person is often West Indian. One New York-based journalist has written that as our parents and grandparents live longer, a caring woman from the Caribbean will very likely be the last human contact many of them will have (Conover, 1997). No such image exists in London, where far fewer West Indians have been engaged in private household work. In Canada, according to various accounts, the heyday of the West Indian domestic is a thing of the past. In New York, a number of factors explain the development of this occupational niche which, incidentally, has a long history. In the early twentieth century, West Indian migrant women were mainly employed as household workers or in other personal service jobs in 1925, this description fit three-quarters of foreign-born West Indian women in Manhattan s labour force as well as two-thirds of African American working women (Model, 2001). Whereas European immigrant women during and after World War I were able to find better-paid work in New York City s factories and as clerks, southern and Caribbean black migrant women were largely excluded from these occupations. In New York s racialized landscape, it was difficult or impossible for most to find alternatives to household work (Watkins-Owens, 2001: 35 36). By the end of the twentieth century, racial exclusion from other jobs was no longer this kind of barrier as employment options for black women widely expanded. It should also be noted that race generally has not been an obstacle to domestic employment; although employment agencies report that some parents refuse to hire black childcare workers and if they do, expect to pay them less than whites New Yorkers have a long familiarity with hiring black domestics (Wrigley, 1995). Demand has been critical for live-in and live-out nannies housekeepers, housecleaners, and home-care attendants to the elderly. In the past

12 14 Foner few decades, domestic caregiving jobs have grown in number as more middle-class women with young children entered the labour force. Alternative child-care arrangements are often unavailable: grandmothers are often working or do not live nearby and high-quality child-care centers are in scarce supply. In any case, many middle-class New Yorkers view day-care centers as offering institutional, second-class child-care and prefer the convenience and flexibility of having someone look after their children at home. At the other end of the life course, demand for homecare attendants has expanded with the growing number of elderly, the increased likelihood that adult daughters will be working outside the home or living far away, and changing state regulations which facilitate or encourage home (rather than nursing home) care. While demand has been growing, African American women have largely withdrawn from the field; in the post-civil rights era they have been unwilling to take jobs imbued with racial subordination and servitude, and many of the better educated have taken advantage of opportunities in public sector employment. The very supply of immigrant workers has fueled demand for domestic workers. The top-ranked US cities in paid domestic work (measured by the proportion of employed women in this field) have large concentrations of Caribbean or Latina women (Hondagneu-Sotelo, 2001: 6). The increasing number of immigrants searching for work in New York has made modestly priced domestic services more widely available by keeping them affordable and relatively inexpensive. English language facility has given West Indian immigrants an advantage over non-english speaking immigrants in competing for jobs caring for children and the elderly. Although poorly paid, many domestic positions (especially live-out jobs) are more lucrative than low-skilled manufacturing jobs that have employed large numbers of Latina and Chinese immigrant women. For undocumented West Indian women, whose employment options are severely constrained by their immigration status, domestic work has provided the chance to work where papers are not checked carefully, or not checked at all. (Many West Indian women arrive on and overstay their tourist visas, thereby becoming undocumented.) In the United States, especially in the late 1960s, 1970s, and 1980s, domestic employment offered a way that many West Indian women could qualify through sponsorship on the job for immigrant visas, thereby legalizing their status. Generally, it has been the newly arrived, typically undocumented, West Indian women who have been willing to take live-in

13 Gender and migration 15 domestic jobs. 12 Often, they moved to the United States on their own no doubt in many, or perhaps most, cases because they knew of the availability of private household jobs. Women on their own have been less averse to live-in domestic service than those arriving with spouses, partly because such jobs assure them a place to live when they first arrive (Marshall, 1987). Once West Indian women got a foothold in domestic service, friends and relatives followed as social networks passed along valuable information about how to find jobs (for instance, advertisements in the Irish Echo) and contact with agencies (Kasinitz and Vickerman, 2001: 202). As Pierrette Hondagneu-Sotelo shows in her study of Los Angeles domestics, employers prefer informal referrals from coworkers, neighbours, friends, and relatives. Not only is it cheaper avoiding an agency fee but network hires inspire automatic trust; employers feel they can find someone reliable and trustworthy to look after their children or an elderly parent through referrals from employers they know (2001). The very development of an ethnic niche can help reinforce it in yet another way: as groups cluster in occupations, this tends to foster prejudice among employers about desirable and undesirable ethnic traits with preference for those who have established themselves in a field (Marshall, 1987). To many New Yorkers, nanny and West Indian ( from the islands ) have become synonymous, increasing their comfort with hiring someone from the West Indies for a caregiving job. Why did this not happen in London? Lack of demand in the years of the mass migration, racial preferences, male-female migration patterns, and the legal context of colonial migration, I would argue, go a long way in providing the answer. Given the male-female migration pattern among West Indians in the 1950s and early 1960s, newly arrived West Indian women were often joining their spouses and therefore not open to live-in jobs, like many of their counterparts in New York who arrived after Nor of course was there a concern about legal status in London: when most West Indians moved to Britain, they came freely, without quotas or visas, as citizens of the British colonial empire. Women leaving for Britain, in short, had no need to think about getting a job that would lead to an immigrant visa an important reason why many came to work in domestic jobs in New York and were the family pioneers in the first place. The fact is, too, that others white others already dominated the domestic work field in London. In the days of colonial rule, British

14 16 Foner settlers in the West Indies may have hired black women there to mind their children a legacy of course from the days of slavery but in Britain itself there is no tradition of black nannies. In the 1950s and 1960s, Londoners seeking caregivers for their children or women to clean their houses preferred to hire domestic servants from groups they had long depended on for such work Irish women, working-class Londoners, young women from northern England or au pairs from the continent rather than unfamiliar Caribbean immigrants, with their black skin and strange accents. The upshot is that London West Indians did not become associated with domestic work. When I did research in London in the 1970s, not one Jamaican woman in the 110 households I surveyed then worked in a private household or told me she had done so since coming to England; not one English person I knew hired a West Indian domestic worker. (Many Jamaican women I interviewed had cleaning jobs, but they worked in offices.) 13 By the late 1990s and early twentyfirst century, when West Indian immigration had long stopped, secondgeneration West Indian women, born and educated in England, had other job opportunities and no motivation to take on the low-paid servant role in the homes of whites. 14 As in New York, in Toronto, demand for domestic workers and requirements for legal entry meant that many Caribbean women in the 1960s, 70s, and 80s were domestics. As Ninette Kelley and Michael Trebilcock write in their history of Canadian immigration policy, the Canadian government agreed to allow Caribbean domestics to enter as landed immigrants in the 1950s when it became clear that European domestics could not meet the demand. This programme was so successful that it was expanded in the following years (1998: 336). By the end of the twentieth century, however, the Philippines had become the major source of domestics in Canada, accounting for 68 per cent of all participants in the Foreign Domestic Movement programme (Darden, 2006: 161). Filipinos were also the vast majority entering under the Live-In Caregiver Program instituted in 1992, requiring the equivalent of a Canadian grade twelve education and six months of full-time formal training in a field or occupation related to the job sought in Canada as a live-in caregiver (in 1993, the government allowed applicants to substitute twelve months of practical experience for the six-month training). It is likely that the increased educational and experience requirements discouraged Caribbean applicants. Moreover, Canadian employer preferences changed, too. In the 1950s and 1960s, Caribbean nannies were seen as obliging and loving mammies, but these stereotypes

15 Gender and migration 17 altered, according to Abigail Bakan and Daiva Stasiulis, after Caribbean domestics began to organize for their rights. Filipino live-in nannies became the ones seen as naturally nurturing, docile, and good with children perhaps partly due to their precarious status as temporary migrants (women admitted under the Live-In Caregiver Program could seek permanent residence only after working as a live-in caregiver for at least two years) (1997). Filipino immigrants, it should be noted, also are English-speaking, English being the official language of schools and universities in the Philippines. TRANSNATIONAL MOTHERS A final topic has to do with transnational motherhood and thus relates to migrants experiences in the receiving society as well as patterns of migration. 15 The similarities in the three countries stand out. Whether they headed for Britain, Canada, or the United States, at least initially migrants often left children behind with relatives in the Caribbean. Because of West Indian women s responsibilities for child-care, and the centrality of their role as mothers, separation from children has had a much greater impact on them than on men. In one sense, this pattern of leaving children with relatives can be viewed as the internationalization of a long-standing West Indian tradition of childminding or child-fostering by female kin (Ho, 1993). 16 Indeed, leaving children with a grandmother (nearly always the mother s mother) in rural areas while the mother goes to work in an urban area or town has long been an accepted practice. Nonetheless, many migrant women left children behind because they felt they had no choice. Davison s early 1960s study of Jamaicans who had been living in England for two years found that 89 per cent of the children of the migrants (male and female) were still in Jamaica (1966: ). 17 In Britain, according to Davison, financial reasons including lack of money to pay the children s fare were major considerations in leaving children in Jamaica; housing difficulties were also cited and, in some cases, inability to find an adult escort for the child on the journey (1966: 117). In the United States and Canada, legal issues caused additional problems, not just leading to separation in many cases but also often lengthening the time that parents and children were apart. In the United States, because many women arrived with tourist visas and overstayed them to become undocumented they were unable to bring their

16 18 Foner children with them. For those who entered with, or eventually obtained, green cards, sponsoring their children under family reunification provisions of US immigration law involved a wait after making the application. In Canada, too, requirements for entry played a role in family separations, as already explained in the case of the various domestic worker programmes. Indeed, in a recent article, Darden accuses Canadian immigration policy of requiring West Indian women to disconnect from their roles as mothers, partners, and caregivers in their own families and to become resources that nurtured white Canadian families (2006: 162). As in Britain, financial considerations played a role in Canada and the United States in that many women preferred to leave children (or some children) in the Caribbean, at least at first, because they worked long and often irregular hours or because they wanted to save on the expense of child care or housing when they were getting settled. In the United States, women without dependent children were free to accept live-in domestic jobs; even in live-out jobs, employers prefer child-care workers who do not have their own children and can maintain constant availability (Kaufman, 2000). In Canada, the Caribbean Domestic Scheme of the mid-1950s to mid-1960s required the women to be single so that some applicants lied about having children (Darden, 2006: 158). 18 Research in New York City reveals that another factor entered into some women s calculations there: fears that their children would be affected by violent crime in New York City and have to attend inferior, and dangerous, schools in the poor neighbourhoods in which they live (Soto, 1987; Waters, 1999). Whatever the motivations and despite the cultural roots of child-fostering transnational mothering has often created strains for the women involved. In Davison s British study, there were many reports of great distress felt by the parents concerning their children and yet they felt unable to surmount the formidable obstacles in the way of a family reunion (1966: 117). In London in the 1960s and 1970s and New York in later decades mothers missed children left behind and often worried about the care they were receiving back home or whether they would get in trouble in adolescence. 19 For women who worked caring for other people s children in the United States or Canada the feelings of loss were complicated by the nature of their jobs. Shellee Colen notes that West Indian nannies attachments to their charges in New York often assumed more importance when they were separated from their own children. I give them more, said one worker, because I just

17 Gender and migration 19 think of them as my own. Just because I was lonely I gave them all I have (Colen, 1990: 106; also see Wrigley, 1995). And while working mothers who leave children behind are free of child-care worries in New York as they were in London and save on expenses abroad, the women are obliged to send money to relatives in the West Indies for their child s (or children s) support. Indeed, studies in the Caribbean note that although grandmothers and aunts are usually emotionally attached to the children in their care, they worry that when the children join relatives in New York or London, the flow of funds will be reduced or dry up altogether. In some cases, this has led relatives to try to delay the children s migration. 20 How many of the children left behind have ended up joining their mothers (or fathers) abroad is an open question. In Davison s British study, only about a quarter of the female migrants with children in Jamaica intended to bring them to England; a quarter were still undecided; and half had no intention of bringing them. Many may well have changed their mind, since the interview took place only two years into their stay abroad and since plans to return to Jamaica a major reason why many intended to leave children there may not have panned out. Still, as Davison observes, some of the children were happily settled in Jamaica with a grandmother or other relative, and in a few cases the parent so disliked living in England that he or she did not want to bring them over (1966: 117). When children did finally join their mothers, the reunions often led to different problems. For the children, the separation from relatives who cared for them in the West Indies many called the grandmother they grew up with mother was often wrenching. Richard, a twenty-one year old Jamaican living in Toronto, was left with his grandmother at the age of three when his mother moved to Canada. When his mother sent for him to join her in Canada, Richard felt torn away from a grandmother he loved and respected, and even today, tears come to his eyes when he talks about her (Henry, 1994: 82). 21 Added to this grief, children experienced the shock of finding themselves in an unfamiliar country with people they hardly knew or remembered (Cheetham, 1972: 87 88). 22 These dynamics have been replayed in New York and London, where, as also happened many times in Toronto, the children often found their mother with a new partner so that they not only had to become reacquainted with her but also had to adjust to a stepparent and sometimes new siblings as well (Waters, 1999: ). While

18 20 Foner they were coping with racial hostility, an unpleasant climate, and new schools, teaching methods, and life styles, their mothers (and fathers) were working long hours and often unable to give the children much attention or supervision. One woman I knew brought her two sons to New York in the early 1980s, and while the oldest one did well, the youngest, who had been robbed at school on his second day, was so terrified that he vomited before leaving for school every morning. Although the mother was able to take him to school every day, the demands of her full-time child-care job and her own schooling she was attending community college to upgrade her skills and get a better job meant that she was unable to spend much time with him. For their part, mothers (and fathers) have often been bitterly disappointed if their children were confused, resentful, or withdrawn instead of grateful for the reunion, which usually entailed great financial sacrifices to bring about (Cheetham, 1972). Trying to establish discipline over children they had spent little time with or often had not seen for over five or six years posed another, often frustrating, challenge. CONCLUSION This comparative analysis of West Indian migration makes clear the important role of the receiving society s immigration policy and context of reception in shaping gender-linked patterns of migration and incorporation. I have focused on the United States, Britain, and Canada, but within each country, West Indians have headed for particular urban destinations, New York, Toronto, and London being the most popular and where the vast bulk of research on West Indian migrants has been done. Important as these three cities are, others, too, have received large numbers of West Indian migrants, and one question is whether and how gender-related patterns including occupational distributions and gender ratios vary among cities in each country. In the United States other cities of settlement for investigation might include Miami, Hartford, and Washington, DC, in Canada, Montreal, Vancouver and Edmonton, and in Britain, Birmingham, Bristol, and Leeds. 23 In emphasizing the role of the national and urban context I do not mean to imply that transnational ties are unimportant, as my discussion of transnational motherhood makes clear. Gender is implicated in a broad range of transnational practices and relations that involve West Indians wherever they have moved. Not only do West Indian migrants often maintain ongoing links to their home society as they reside abroad

19 Gender and migration 21 for example, sending back remittances, building houses in the home community, making home visits, and sometimes circulating back and forth but many also keep up ties with relatives and friends who have moved to other countries. The relevant question here is what difference it makes that women s and men s transnational relationships encompass different nations and cities London and Jamaica, for instance, rather than New York and Jamaica. Systematic research is required on this topic, but one factor that should be considered is basic geography. Modern transportation and technology have made the world a smaller place and facilitate the maintenance of transnational ties for West Indians today wherever they go. Yet greater geographic proximity to the Caribbean from the United States (especially as compared to Britain) is bound to have an impact, if only on the ability to make more frequent visits home which may affect women differently than men given, among other things, West Indian women s centrality to kinship networks. Distance between cities across borders in the diaspora also influences transnational relations and practices. A fascinating practice has recently been documented regarding connections between West Indians in Toronto and New York. The two cities location, a driving distance from each other, has given rise to a business opportunity for some Toronto-based West Indian female entrepreneurs. They charter buses and fill them with other West Indians (mainly women) in the Greater Toronto area who travel to Brooklyn for the weekend to visit relatives and friends and attend community events or who take day-excursions to outlet malls and flea markets in upstate New York and Detroit (Trotz, 2006). 24 This Toronto-New York example offers just a hint of the way that West Indian women s transnational relations and practices may develop in particular ways depending on the national and urban context of settlement. Another area for further research and exploration concerns the children of immigrants. I have looked at the first generation, but a critical topic for comparative investigation is the role of gender among second, and indeed third, generation West Indians born and raised in Britain, the United States, and Canada,. There is, in short, much on our research agenda. What can be said at this point is that a comparative approach focusing on West Indians in different destinations has much to offer. It can help us to better appreciate the complex ways that migration and gender are interwoven and interlinked in the West Indian case and, in this way, also ultimately make a contribution to a general understanding of migration and gender in global perspective.

20 22 Foner NOTES 1. In this essay, I use West Indian to refer to people from the Anglophone Caribbean, including the mainland nations of Guyana and Belize. 2. While Barbadians who went to Panama to work on the canal were mostly men, some women travelled there to look for work or to join husbands; Barbadian women were shopkeepers and worked as laundresses and seamstresses there (Richardson, 1985: 122). 3. On these early movements see Eisner (1961), Newton (1984), and Petras (1988). 4. Peach cites Davison s sample survey of Jamaican emigrants which showed that 72 per cent of the women, in contrast to 29 per cent of the men, had their passages paid for by migrants established in England. 5. Among Jamaican immigrants in the United States, according to 2000 census figures, 56 per cent were female, 44 per cent male (Jones, 2006). In New York City, West Indian immigrants have been disproportionately female: in 2000, the number of males for every 100 females was 71 for Jamaican immigrants, 70 for Trinidadian immigrants, and 87 for Guyanese immigrants (Lobo and Salvo, 2004: 150) See Lobo, Salvo, and Virgin on female dominance of West Indian migrant flows to New York City in the and periods (1996: 20). 6. After 1970, the percentage of total Jamaican workers listed as private household workers was much lower, ranging from 26 per cent in 1972 to 12 per cent in In 2001, women made up 58 percent of Jamaican immigrants in Canada (Jones, 2006). 8. By way of comparison, in 1966, the economic activity rate for all women in London and the Midlands in 1966 was a little under 50 per cent; the labour force participation rate for all women in the United States 16 and over in 1990 was 53 per cent. A national survey conducted by Political and Economic Planning in Britain in the 1970s found 74 per cent of West Indian women working compared with 45 per cent of women generally (Stone, 1983: 35). 9. In 1990, of the Jamaican women aged in New York City who reported an occupation to the census, nearly a third were in the health care field as nurses, orderlies, attendants or nurses. In Greater London, in 1966, 14 per cent of West Indian women in the labour force were nurses (Kasinitz and Vickerman, 2001; Rose et al., 1969: 157). For Toronto, see Jansen and Lam (2003). 10. In the United States as a whole, according to the 2000 census, 28.7 per cent of West Indian immigrants worked in managerial, professional, and related occupations. (Holder, 2007: 682); according to the March 2007 Current Population Survey, 28 per cent of Jamaican immigrants (25 64) in the United States had a college degree or more (Camarota, 2007).