National and Urban Contexts. for the Integration of the Immigrant Second Generation. in the United States and Canada

Transcription

1 National and Urban Contexts for the Integration of the Immigrant Second Generation in the United States and Canada Jeffrey G. Reitz and Ye Zhang University of Toronto March 2005 (Final draft for conference publication) An earlier draft of this paper was presented at a conference on The Next Generation: Immigrant Youth and Families in Comparative Perspective, Harvard University, Radcliffe Institute for Advanced Study, th th October 29 30, Helpful comments from Kara Somerville, and editorial assistance of Elizabeth Thompson are acknowledged with thanks. Jeffrey G. Reitz is Professor of Sociology, and R.F. Harney Professor of Ethnic, Immigration and Pluralism Studies, at the University of Toronto. Ye Zhang is Ph.D. candidate in the Department of Sociology, University of Toronto. Address correspondence to Jeffrey G. Reitz at Munk Centre for International Studies, University of Toronto, 1 Devonshire Place, Toronto ON, Canada M5S 3K7, Tel , Fax ; jeffrey.reitz@utoronto.ca;

2 National and Urban Contexts for the Integration of the Immigrant Second Generation in the United States and Canada Abstract The integration of the black, Chinese and other Asian immigrant second generation in the United States and Canada is affected by institutional and cultural differences between the two nations, as manifested at the urban level. Urban contexts matter because second generation minorities tend to remain in the high-immigration cities settled by their parents, and because of distinctive characteristics of these cities in each country. Immigrant parents in U.S. immigration cities faced relatively high native-born educational levels; their own educational levels were relatively lower, and this educational polarization reduced their earnings. Albeit parallel, the Canadian polarization pattern was much less pronounced. In the analysis presented here, variations across nations and cities in the local economic standing of immigrant parents have little relation to educational attainments of second-generation children, which are affected more by local standards for the native-born white urban middle-class. In effect, assimilation is segmented by nation and urban area, and not by social class. Ethnic differences from the local native-born educational standard are similar in the two countries lower for second-generation blacks, higher for second-generation Chinese and other Asians. Data are from a merged file of five U.S. Current Population Surveys, , and the 2001 Canadian census public use microdata file.

3 National and Urban Contexts for the Integration of the Immigrant Second Generation in the United States and Canada Given that characteristics of host societies shape the reception and integration of immigrants (Reitz 2003), the question arises whether these effects, and possibly others, carry over to affect the second generation, and if so, how and why. Comparative research can be useful in answering such questions, and this study focuses on the comparative experiences of the second generation in the United States and Canada. Previous research shows substantial differences in the economic standing of immigrants in the two countries, as a result of several significant institutional and cultural factors (Reitz 1998). These institutional and cultural factors could also affect the second generation. The differing circumstances of the immigrants may be important as well. This study explores the extent to which differing American and Canadian circumstances may or may not affect the second immigrant generation, and why. The paper begins with an outline of potential cross-national differences for the immigrant second generation that might be inferred from previous theory and comparative research. This considers crossnational differences in the aggregate, and also differences as manifested at the urban level where immigrants are concentrated. Then, previous research on the integration of the immigrant second generation within each country is considered in light of these possibilities. None of this previous research is comparative, and the data analysis which follows represents the first quantitative comparison of the integration of the immigrant second generation in the two countries. It is based on pooled data

4 from several Current Population Surveys over the period 1995 to 2003, and the Canadian census public use microdata sample file for Potential Implications of National Differences for the Second Generation The most pervasive theme in contrasting American and Canadian societies, as identified by S.M. Lipset (1968, 1989), is the greater American emphasis on the value of individualism and individual liberty, and the correspondingly greater Canadian concern with collective welfare. There is considerable evidence of this difference, drawn from attitude surveys and institutional analyses (Grabb and Curtis 2004). Although it may not be as extreme as the broader contrast between the two North American countries and Europe, as Esping-Anderson (1990) has pointed out, it is significant nonetheless. The American-Canadian contrast, which has been related to the experiences of recent immigrants (Reitz 1998), may also affect their children, and there are two dimensions to this potential effect on the second generation. One is related to equality of opportunity, the other to equality of outcomes. Individualism s concern with fairness in the competitive process, and equality of opportunity in the pursuit of happiness, has a direct bearing on the possibility of discrimination based on racial or ethnic origins, including both the original immigrant minorities and the second generation. In fact, the issue of discrimination has been seen a critical issue for the second generation, in contrast with the immigrant generation, which might face greater barriers to equal opportunity arising from foreign credentials, language problems, or cultural differences. Arguably, if Americans place a greater emphasis on the value of individualism, then the potential for such discriminatory treatment of the minority immigrant second generation may be less in that country. Such an equal-opportunity implication is far 2

5 from automatic, as individualism may also help justify freedom from interference in personal decisions, potentially shielding discriminatory practices from regulation. And Canadian pride in tolerance as a national style essentially claims that its more laid-back, live-and-let-live approach to life helps restrain the impulse to discriminate in private affairs. Speculation about segmented assimilation for the second generation in the United States (Portes and Zhou 1993) is a related concern. For racial minorities, there are alternative paths of assimilation toward different segments of the host society. That is to say, a society segmented along racial lines offers different possibilities for the assimilation of newcomers. Immigrants and their descendants may assimilate into the mainstream society, into their own minority community, or into a racial minority community within the host society. This raises the possibility that existing patterns of inequality experienced by racial minorities will be repeated for racial minority immigrants. In the United States, inequalities might be reinforced by patterns of racial segregation and polarization historically experienced by native-born African Americans. New native-born minorities in countries such as Canada which do not share the legacy of mass slavery might escape the burdens entailed by this social framework (Boyd 2003). Against this speculation stands the argument that the American racial struggle has yielded many institutional protections and resources for racial minorities, offsetting any negative effects on immigrants or their children (Foner 2003, Nee 2003). The evidence suggests that any differences between countries in patterns of racial or ethnic discrimination is small, and this would imply small differences for the minority second generation. Reitz and Breton (1994) focused on cross-national differences for racial minority black and Asian immigrants, and showed that differences in earnings net of measured qualifications such as education and work experience are small. They further found that results of discrimination field-trials show that it occurs to 3

6 a similar degree in the two countries. Moreover, evidence shows that with sufficient controls for human capital, there appears to be no widespread racial discrimination against all second generation minorities in either country (though without direct cross-national comparison; see Zeng and Xie 2004; Hum and Simpson 1999). Field research (e.g. Kasinitz et al. 2003) also suggests no negative impact of American race relations involving the African-American community on the black second generation. To date, there has been no direct comparison of the experiences of the second generation minorities in the United States and Canada with regard to the question of racial discrimination, however, and this will be one objective of the present analysis. As noted above, the second dimension of the impact of individualism is related to equality of outcomes. The cultural contrast of American individualism and Canadian collectivism projects greater inequality of overall institutional outcomes in the U.S. This has affected immigrants and could carry over to affect the second generation. Borjas (1988, 1990, 1999) noted that immigrant earnings for all major immigrant groups were lower in the United States than in Canada, and linked this to lower relative educational levels for immigrants in that country. While Borjas pointed to immigration policies as a possible source of this difference, Reitz (1998) showed that a range of institutional sectors is involved, including labor markets, which affect earnings distributions and welfare policies. Labor markets are interdependent with other sectors, including immigration policy. However, the most important difference concerned educational institutions in the two countries. As Lipset (1989) had argued, if individualism produces greater inequality of outcomes, then participants have an incentive to maximize their personal resources for competition, one of the most important of which is educational attainment. High educational levels of the native-born represent obstacles for immigrants, lowering their relative earnings whatever other effects they may have, such as boosting productivity. 4

7 Lower relative education and earnings for immigrants in the United States might suggest more negative outcomes for the second generation, specifically with respect to educational opportunity. First, the greater educational gap for immigrants in the U.S. would mean that the second generation faces a greater challenge in terms of the degree of inter-generational educational mobility required to achieve equality with the native-born. Second, the relatively lower earnings of immigrants in the United States creates a greater educational funding problem for their children. Immigrants who struggle in economic terms, and who more frequently live in conditions of poverty, will have greater difficulty providing the resources for their children to do well in school, and to finance a post-secondary education. And finally, compounding the funding problem are both the higher level of tuition for many post-secondary institutions, particularly those of the highest quality, and the lower level of social welfare protection in the United States. There is nothing automatic about these effects, however. Social class differences in educational attainments which exist to a degree for the native-born population, may or may not apply to the children of immigrants. Differences in educational attainments related to family origins for the native-born are related to economic or financial matters, but only partly so. To some extent, they reflect social aspirations, and the existing social hierarchy of society. The children of immigrants may or may not follow any particular socially-defined pattern in the mainstream society. They may not necessarily define their educational goals by following an existing pattern based on parental status or resources. This possibility, that the integration of the second generation may follow patterns set by a variety of different components of the native-born population, and not simply by the native-born population as a whole, represents an extension of the processes recognized in segmented assimilation. The process of segmented assimilation suggests that the children of immigrants may be variously oriented toward 5

8 particular ethno-racial groups within the host society. Here, the question is whether the second generation may adopt behaviour distinctive to particular social class groupings in the host society, whether or not a minority group assimilates into its native-born social class counterpart. In sum, there are two possible lines of inquiry regarding U.S.-Canada differences in the experiences of the second generation: those related to discrimination and equality of opportunity, and those related to equality of outcomes. Regarding the former, previous research suggests that differences for the second generation may be small. Regarding the latter, an important question is how the greater inequalities affecting the immigrant generation in the United States may carry over to affect the second generation. Implications of Urban Systems and Immigrant Concentrations The specification of American-Canadian differences at the urban level must take into account the differing urban systems of the two countries. In fact, cross-national differences in the experiences of immigrants in the United States and Canada cannot be understood properly without considering differences in inter-urban variation. This may be seen as a question of segmented assimilation along urban lines. The second generation may be affected both by inter-urban differences in the situation of the immigrant generation itself, and by other aspects of inter-urban differences impinging directly on the second generation. The concept of global cities put forward by Sassen (1988, 1991) suggests that those cities which play a major role in management of the global economy especially New York experience 6

9 considerable labor market polarization which particularly affects immigrants. Sassen s theory was that the super-affluent elites in global cities create a demand for personal services which attracts immigrants, who then are compelled to work in these low-paying and demeaning jobs. The supposition of greater inequality for immigrants in New York turns out to be correct, but the reasons for this affect are not quite what Sassen suggested, and this affects the comparison of New York with other American cities, and with cities in other countries, such as Canada. Evidence from the census (Reitz 1998) suggests that the inequalities affecting immigrants in New York arise from a more pervasive skills polarization, rather than specific labor demand. This is important because such skills polarization affects not only New York but other high-immigration cities in the U.S., regardless of their occupational structure or position in the global economy. There are two aspects to this skills polarization. One is the high level of education of the native-born workers, which reflects the advancing postindustrial development in technology, finance, and other high-end services characteristic of all urban areas experiencing economic expansion, not just the global cities. The other is the attraction of large numbers of immigrants. Again this occurs not just in global cities, but in all cities experiencing economic expansion. In the United States, high immigration is associated with lower levels of immigrant education, as Bartel (1989) noted. In the U.S., this pattern of immigrant settlement contributes to educational polarization. The result is more extreme inequalities for minorities in many highimmigration cities in the United States, not just New York as a global city. Such educational polarization is less characteristic of Canadian cities, for reasons which relate to the overall pattern of Canadian development and the continuing importance of immigration within it. Many major Canadian cities are regional capitals, and the occupational and educational profiles of such cities to some degree reflect this similarity. Moreover, Canada s immigrants have been recruited as part 7

10 of a national program of nation-building (Reitz 2004). Hence, the proportions of immigrants in Canadian cities are not only higher than in the United States, they tend to be distributed more uniformly across urban areas. Educational levels also tend to be more uniform, possibly because settlement policies matter more than family networks in determining the location of newly arriving immigrants. The result has been that immigrant inequalities are not only less than in the U.S., but more uniform across Canadian cities. (Australia is an even more extreme contrast with the U.S. in this regard; see Reitz 1998, 2001). For the second generation, the hypothesis is that these local urban contexts matter in mediating cross-national differences. It may be suggested that assimilation is segmented by urban area, just as it obviously is segmented by national setting. Segmentation by national context is taken for granted, predicated on the assumption that the children of immigrants most often remain in the host society, neither returning to the home country, nor moving on to another society. Similarly, the urban area may be the context for second-generation assimilation, to the extent that patterns of urban concentration established by the immigrant generation are maintained in the second generation. The second generation clearly remains within the city of their immigrant parents at least while growing up, and during the period in which they receive the early part of their education. But there may be a number of reasons for the second generation to remain resident in immigration cities, beyond the general tendency for children to remain near their parents. For one, the ethnic community is an attraction for the second generation, as it is for the parents. The second generation may retain a degree of involvement in or commitment to the ethnic community, and these involvements and commitments are quite often local and not necessarily available in another urban area. For another, cities lacking an immigrant community may be seen as less hospitable to minorities with origins outside the country, and therefore as less attractive places in which to live or work. 8

11 If the urban setting is a context for assimilation, then the questions raised about the impact of variations in immigrant economic success at the national level can be raised again at the urban level. It was assumed at the national level that national educational standards were the framework for the second generation in setting their own aspirations. This might also be the case at the urban level. Because the native-born educational levels are higher in the key high-immigration cities, such as New York, the second generation may attempt to meet that standard. At the same time, the question of the impact of social and economic inequality can be put forward. If in the high-immigration cities in the United States, immigrants have lower relative earnings and higher rates of poverty, what is the impact on the educational attainments of the second generation? Previous Research on Second Generation Integration of Minorities within Each Country One might begin by comparing the education and labor market integration of the second generation, using research conducted within each country separately. Studies of the immigrant second generation in the United States (Portes 1996; Portes and Rumbaut 2001; Rumbaut and Portes 2001; Kasinitz, Mollenkopf, and Waters 2003; Farley and Alba 2002) show very positive results for Asians, also positive results for blacks though somewhat less so, and significant inequality and poverty for Mexicans and a number of other Latin American groups. In Canada, native-born visible minorities have attained levels of education comparable to the native-born majority group or higher, in both secondary school and university completion (Davies and Guppy 1998: 136; see also Worswick 2001). While analyses of educational opportunity by Boyd (Boyd and Grieco 1998; Boyd 2002) show positive results in survey data for visible minorities overall, however, Simmons and Plaza (1998), using 1991 census data, examined university attendance among the second generation in Toronto, and found that 9

12 young black men in Canada show some disadvantage, though a modest one. Although for specific groups, such as blacks and Asians, these results seem broadly similar, it is difficult to draw precise comparative conclusions. Boyd (2003: 111) states that contrary to the second-generation decline and segmented underclass assimilation models found in the United States, adult visible minority immigrant offspring in Canada do not have lower educational attainments than their parents or their not-visible-minority counterparts. Yet in the U.S., for the black and Asian groups comparable to those studied by Boyd, researchers have found a lack of support for the segmented assimilation model. For example, Farley and Alba (2002: 697) say, We find no evidence supporting the segmented assimilation hypothesis, a finding which mirrors what Kasinitz, Mollenkopf, and Waters (2003) found in New York. These results mainly concern education, and relative educational levels are important, particularly because the labor market experiences of native-born racial minorities have been a bit less positive. Most analyses in both countries have shown that the earnings of native-born black children of immigrants are less than expected based on educational attainment (Reitz and Breton 1994). For Asian groups, there is more controversy, with some studies showing no net disadvantages in the United States (Zeng and Xie 2004) or in Canada (Hum and Simpson 1999), while other studies show that there are differences, in both the United States (Hirschman and Wong 1981, 1984) and Canada (Pendakur and Pendakur 1998, 2002). The possibility of lower earnings than expected based on equivalent education means that the second generation groups may have to attain higher levels of education than the nativeborn in order to achieve a comparable level of economic well-being. 10

13 Data Analysis The data analysis here will compare a merged file from five Current Population Surveys in the United States (1995, 1997, 1999, 2001 and 2003) with the 2001 Canadian census public use microdata file. The value of the CPS data for studying the second generation has been noted (Model 2002, Farley and Alba 2002; Rumbaut 2003), and in the analysis presented below, the five independent surveys are used yielding an overall N = 759,065. The Canadian census for 2001 is the first since 1971 to include a question on parental birthplace, and the microdata file for 2001, N =801,055, is large enough to contain significant second generation samples for many groups of interest. Within the second generation, the analysis identifies a single young age cohort, 25-39, most of whom have completed their education and entered their working careers. In both countries, most of the children of immigrants of non-european origin are relatively young, so the age cohort is the appropriate focus (see Reitz and Somerville 2004). These persons, including racial categories within this group, will be compared with three other groups. To provide an indication of progress toward equality, an age-specific comparison with the mainstream population will be made, as represented by native-born whites of third and higher generation, also aged To provide an indication of intergenerational mobility, the young second generation group also will be compared with an older immigrant group, aged 50 and over, taken to represent the parental generation. Because the native-born norm for inter-generational educational mobility is a variable across contexts, the comparison provides a nativeborn benchmark. Hence, the third comparison group is the older native-born population, aged 50 and over. The purpose is to provide a comparison of inter-generational mobility for immigrant groups and the native-born. 11

14 Variables in the analysis and their distributions, for the U.S. and Canadian samples selected as indicated, are shown in Table 1. The CPS analysis captures white, black and Asian racial categories, within which hispanic origins (identified in a separate question) are excluded. The 2003 file includes mixed-race categories (White-Asian, Black-Asian, and so on), which are re-coded here as other and hence excluded. (Alternative codings for 2003 do not affect the results.) In the Canadian data, the same ethno-racial origins categories can be identified, using a combination of a question similar to the race question used in the U.S., and a question on ethnic origins. Persons of hispanic origins far less prominent than in the U.S. are also excluded in the Canadian analysis. The composition of the Asian second generation is diverse in both countries but differs somewhat: Filipinos constitute a larger part of the U.S. Asian group, while Chinese are relatively more numerous in Canada. In the U.S. data, specific origins within the Asian category can be identified on the basis of parental birthplace, and the analysis focuses on the Chinese as a category which can be examined on a cross-national basis. In the Canadian data, Chinese origins are distinguished in the race question. Two educational attainment criteria are examined: attainment of some post-secondary education but no bachelor s degree, and attainment of a bachelor s degree or higher including graduate and professional degrees. These categories appear to be broadly similar in the two countries, but some potential differences should be kept in mind. On the one hand, in these and other data, the proportion of Americans with university degrees is much higher than Canadians. For the young native-born white populations in these data (see bottom panel of Table 1), the proportion with bachelor s degree or higher is 26.1 percent in the U.S., compared to 15.9 percent in Canada. On the other hand, the proportion of Canadians with other types of post-secondary education is higher than in the U.S., 36.8 percent compared to 26.1 percent in the U.S. Therefore the proportion of Canadians with any type of postsecondary education is as high as, or slightly higher than that of their American counterparts (

15 percent in Canada, compared to 52.2 percent in the United States). Two cross-national differences may be reflected in these data: a greater Canadian emphasis on a more vocationally-oriented post-secondary education system; and a difference in the labeling of postsecondary institutions. What Americans call colleges are more likely to include academic institutions with programs in many ways comparable to universities, whereas in Canada, colleges are generally more vocationally oriented. American colleges which offer bachelor s degrees are quite diverse, however, and the mix of academic and vocational content may vary. Bachelor s degrees in Canada come from universities, and the larger proportion of Canadians who identify their degree as not in the bachelor s category may or may not reflect a different level of education as compared to the Americans who identify their college degree as a bachelor s degree. The possibility that the difference is partly if not largely a question of labeling is reflected in the fact that immigrants in the United States are far more likely to describe their post-secondary degree as a bachelor s degree while immigrants in Canada are more likely to describe their post-secondary degree as not at the bachelor s level. This difference does not appear to relate to the specific wording of the question, and may instead reflect a different conventional meaning of the term in each country. In this analysis, then, two measures of educational attainment are reported: proportions with bachelor s degrees or higher, and proportions with any postsecondary education. Findings Overall Cross-national Differences in Second Generation Education. Overall the positive educational performance of the minority second generation compared to the native-born whites which 13

16 had been found in each country is reflected in the present cross-national comparison, with the specific difference emerging that the most positive performance for the second generation is most often in the United States. Table 2 presents comparisons for bachelor s degree attainment, and for all post-secondary rd education combined. Data on educational attainments for the native-born white 3 generation is presented in the first row for each country, followed by data for immigrant origins groups. As explained above, inter-generational differences can be assessed based on age-group comparisons. All data are presented in the four left-hand columns of Table 2: younger persons aged (columns 3 and 4) compared to older persons aged 50 and over (columns 1 and 2). For the immigrant origins groups, the young group aged is the second generation (columns 3 and 4), and the older group aged 50 and over is composed of immigrants (columns 1 and 2), who are taken to represent the parental immigrant generation. To assist in these comparisons, various computed differences are reported in the six righthand columns of Table 2. These include the relative education of older immigrants compared to older native-born (columns 5 and 6), relative education of the younger second generation compared to sameage native born (columns 7 and 8), inter-generational differences reflecting mobility (in columns 9 and 10), and comparisons of inter-generational mobility for immigrant origins groups and the mainstream population (the final columns 11 and 12). In both the United States and Canada, educational levels of the younger second generation minorities exceed those of same-age native-born whites, but there are cross-national differences which vary depending on the group and the educational criterion. The relevant comparisons of the second generation groups with same-age members of the native-born white population are in columns 7 and 8. For the young black second generation in the United States, the percentage attaining bachelor s degrees is 8.6 percent higher than for the mainstream whites, compared to 5.9 percent higher in Canada. On the other hand, the percentage of the black second generation attaining any type of post-secondary education 14

17 in the United States is 10.2 percent higher than for the mainstream white population, but 15.0 percent in Canada. The second generation of Asian origins has even higher overall levels of education. Although the cross-national comparison across the entire heterogeneous Asian category is ambiguous favoring Canada somewhat in the bachelor s category, but not in overall post-secondary attainment when the comparison focuses on the Chinese as a specific origin group, the comparison clearly favors the United States. The impact of cross-national differences in the composition of the Asian category is likely at work here; for example Filipinos who are relatively more numerous in the United States have lower levels of education. Most of the subsequent analysis focuses on the Chinese rather than the broader Asian category. The percentage of the young Chinese second generation with bachelor s degrees is a rather spectacular 45.1 percent higher than the same-age white mainstream population in the United States, compared to a more modest but still very impressive 37.3 percent higher in Canada. For all postsecondary education, the figures for the Chinese second generation still favor the United States though less so: 28.4 percent higher in the United States, and 23.6 percent higher in Canada. The comparison shows the significance of types of post-secondary education. For both Chinese and blacks, the comparisons show higher relative attainment at the bachelor s level for the second generation in the United States, but when the comparison includes all post-secondary education as opposed to only bachelor s degrees, the attainment levels in Canada improve in relative terms (though in the case of the Chinese the comparison still favors the United States). Strong educational attainments for second generation blacks in the United States in the Canadian comparison are particularly impressive given the lower relative educational levels of the immigrant generation in that country. In effect, rates of inter-generational educational mobility for the black second generation in the United States have been so high that they have more than overcome the educational deficits of the immigrant generation. For older black immigrants in the United States, 15

18 educational levels were lower than for the older mainstream population, whereas for black immigrants in Canada the opposite was the case. For the black second generation to outperform the mainstream population in the United States required particularly high rates of inter-generational mobility. As the figures in columns 11 and 12 show, in the United States inter-generational differences in bachelor s degree attainment were percent for the black second generation compared to the mainstream population, compared to +5.5 percent difference in Canada. Inter-generational differences in any postsecondary education were percent in the United States compared to only +2.6 percent in Canada. The very highly-educated Chinese second generation also achieved much higher rates of educational mobility than the mainstream population, but the cross-national comparison shows that the extent of this is about the same in each country. The Chinese second generation in the United States is better educated relative to the native-born than its counterpart in Canada, and better educated in absolute terms at the bachelor s degree level. However, educational levels for older Chinese immigrants are higher both relatively and absolutely in the United States compared to Canada, so the extent of upward educational mobility reflected in these outcomes is roughly the same in each country. Three significant conclusions can be drawn from the evidence of strong educational attainments, and the positive inter-generational change. First, within each country, the economic and other difficulties faced by black and Chinese immigrants have not prevented their children from succeeding educationally. Second, the greater economic difficulties that these two immigrant groups experienced in the United States have not carried over to produce either lower educational attainments for the second generation in that country or lower rates of educational mobility, compared to the native-born, or compared to what is observed in Canada. To the contrary, second generation patterns of educational attainment on average exceed what is typical for the native-born, who are often in more favorable 16

19 economic circumstances. And third, the expectation that the black second generation in the United States might be negatively affected by the racial polarization of the host society the segmented assimilation hypothesis is not supported. Although the black second generation has lower educational attainments than Chinese or other Asians in both countries, a relatively higher proportion have bachelor s degrees in the United States compared to Canada, and overall the black second generation in the United States achieved greater educational mobility both with respect to bachelor s degrees and with respect to all post-secondary education, than did their counterparts in Canada. Urban Contexts for the Second Generation. Understanding the sources and implications of these patterns requires consideration of the urban context for immigration. In the United States, immigrants have been heavily concentrated in key urban areas. These are all large cities, but among large cities they form a distinctive group, the so-called gateway cities. New York and Los Angeles as the prime gateway cities are destinations for about one in four of all immigrants arriving in the U.S. though they contain only about 7 percent of the overall population. Chicago, Miami and Washington are home to another 11 per cent though they also contain only about 5 percent of the population. At the same time some large American cities contain relatively few immigrants. In some cases, such as Detroit or Philadelphia, the reason seems to be a relative lack of labor demand. In other cases, such as Atlanta or Minneapolis, a reason may simply be the lack of existing networks of chain migration. Immigration to some of these cities has been increasing, but remains low relative to the main immigration centers. In considering any particular origins group, however, it is important to note major differences in their specific patterns of urban concentration. Table 3 lists cities of urban concentration of origins groups for immigrants over 50 years of age, who represent the parents of the younger second generation population examined here. While New York is a gateway for many origins groups, most groups have 17

20 somewhat distinctive patterns of settlement in other cities. In particular, for older black immigrants among whom the concentration in New York is extremely high (38.5 percent), there are also substantial concentrations in Miami, Washington, Boston, Newark and Ft. Lauderdale (between 4-7 percent in each). These six cities accounted for two-thirds of all black immigrants over 50 years of age in the United States. For Chinese, although there is also a large concentration of older immigrants in New York (16.1 percent), much of the rest of the pattern of urban concentration is rather different. There is a large group in Los Angeles comparable to the one in New York (17.7 percent), and significant groups also in San Francisco (6.9 percent ), Honolulu (4.9 percent), Washington (3.5 percent) and Philadelphia (3.3 percent). These leading cities for Chinese immigrants contain nearly 52.3 percent of all Chinese immigrants over 50 years of age in the United States. Among these only New York and Washington are also cities of black immigrant concentration. In Canada, there is also a pattern of urban immigrant concentration similar to what is seen in the U.S., but with some differences. Canada has three gateway cities, with Toronto having over one-third of Canada s immigrants, and together with Montreal and Vancouver containing over 60 percent of all immigrants to Canada. This represents a high immigrant concentration though of course it must be remembered that these three cities also contain about one-third of the entire Canadian population. Toronto is most comparable to New York in that it is the key destination for most immigrant groups, and in fact Toronto is a relatively more prominent destination in Canada than New York is in the U.S. For Canada virtually all other groups are also concentrated in either Vancouver or Montreal, or both. Black immigrants in Canada are very heavily concentrated in Toronto, which contains over half (51.1 percent) of all black immigrants in Canada. Most of the others are in Montreal (23.2 percent), which contains both French- and English-speaking Caribbean black immigrants. For Chinese, the largest 18

21 group is also in Toronto (41.6 percent), and although Vancouver is most prominent as a second city for Chinese (34.2 percent), there is also a significant group in Montreal (5.2 percent). The data show that these immigration cities in both countries are also the context for the second generation. Table 3 shows that the second generation minorities continue to be concentrated in the main immigration cities in both countries. To some extent this is simply a home-town effect, whereby all children tend to remain in the city in which they grew up. They want to remain close to friends, relatives, and familiar places. However, the racial difference for immigrants suggests that additional factors are at work for the minority second generation. One may be the attraction of an immigrant community, and larger numbers of co-ethnic persons, who represent social and economic opportunity. Another may be a perception of the key immigration cities as more receptive to members of one s minority group. Higher urban concentrations of the second generation suggest that these urban areas represent a substantial degree of social segmentation, somewhat restricting the process of their integration into society. The second generation does not become distributed across society in the same way as the native-born, but to some extent remains tied to the immigrant settlement area. Educational Attainments within Urban Contexts. Because of the urban concentration of the second generation, educational attainments and inter-generational mobility should be examined within the relevant urban context. This is done in Table 4, Parts A and B, for the black and Chinese second generation populations, respectively. For convenience, the table repeats the basic national analysis in Table 2, and then shows the breakdown according to urban areas of immigrant concentration. 19

22 Table 4 part A for the black second generation shows two major findings. First, relative to urban contexts of concentration, and the mainstream population within those contexts, the educational attainments of black second generation in the two countries are generally lower than appears in the national comparison, and show no consistent cross-national difference. In both countries, educational attainments for the black second generation in the urban areas of major concentration the six cities in the U.S., and Toronto in Canada are as high as in the national comparison, and in the case of the U.S. are significantly higher. However, because educational standards in these urban areas are so much higher than the national standard, the attainments of the second generation relative to that standard appear much less than in the national comparison. In terms of bachelor s degree attainment, they fall short by 6.5 percent in the six U.S. cities, fall short by 9.0 percent in Toronto, and about equal the (lower) native-born standard in Montreal. In terms of overall post-secondary educational attainment, they are about equal with the mainstream standard in the U.S. cities, and 7.1 percent higher in Toronto, and 12.3 percent higher in Montreal. The second finding is that these roughly similar U.S.-Canada results in terms of educational attainment for the second generation evolved from very markedly different starting points in terms of the relative immigrant-generation educational levels. Therefore producing similar (local) results entailed widely varying (local) patterns of inter-generational educational mobility. The different educational starting points essentially reflect the more extreme educational polarization for immigrants in U.S. cities both the higher educational levels for the native-born and lower educational levels for immigrants. In New York and the other five key cities of their concentration in the U.S., the black second generation achieved bachelor s degrees at a rate of 49.7 percent, fully 36.7 percent higher than for their parents, and reflecting a rate of inter-generational mobility over 20 percent higher than for the mainstream population. By contrast, in Toronto the inter-generational educational mobility of the black second 20

23 generation in terms of bachelor s degrees was much lower and only slightly higher than for the nativeborn. They achieved bachelor s degrees at a rate of 22.2 percent, only 13.1 percent higher than for their parents, reflecting a rate of inter-generational educational mobility only 3 percentage points higher than for the mainstream population. A parallel situation exists in Montreal: the black second generation achieved bachelor s degrees at a much lower rate than those in the major areas of black immigrant concentration in the U.S., and the rate of inter-generational mobility is about comparable to that of the native-born population. Hence the combined results for the six U.S. cities and the two Canadian cities is very similar. Data on post-secondary education in general also show a similar pattern. Results for the second generation in the two countries were relatively lower in the major areas of urban concentration, and evolved from very different starting points. Rates of inter-generational mobility relative to the mainstream population were by far the greatest in the main black immigration cities in the U.S. What is quite interesting in these findings is the way in which these second generation minorities seem have responded to local native-born standards in both countries. Inter-urban and cross-national variations in rates of educational attainment among second generation minorities match local native-born white standards more closely than they do those of their parents. Within the high immigration cities, particularly in the U.S. despite the significant skills gap experienced by the older immigrant generation and the resulting lower relative earnings the rates of educational mobility achieved by their children were higher than for the native-born whites, and attainment levels were high enough to completely eliminate the immigrant-generation skills polarization. In the cross-national comparison, the relative position of the second generation within local urban contexts of greatest concentration looks somewhat similar. In the Canadian case, in the cities outside Toronto and Montreal where the black immigrant population had educational rates substantially higher than for the native-born, the second generation showed significantly lower rates of educational mobility compared to the native-born population. This 21

24 produced a black second generation with roughly similar educational levels to what is observed in such cities across both countries. For the Chinese second generation (see Table 4 - Part B), although generally the educational attainments are much higher than for blacks, there are two major findings which at least in certain respects parallel those for blacks. First, the relative educational attainments for the Chinese second generation in the two countries are lower when viewed in urban context then they are in the national comparisons, particularly in the United States. This is similar to what was found for blacks. In the United States, the Chinese second generation bachelor s degree attainment rate was 45.1 percentage points above the mainstream, but only 28.8 percent above in the six cities of greatest concentration. In Canada, the Chinese second generation bachelor s degree attainment was 37.7 percent above the mainstream in the national data, down to 33.6 percent in Toronto, and 25.4 percent in Vancouver. If overall rates of attainment of any post-secondary education are considered within urban areas, it is found that such rates are in fact extremely high both in Toronto and in Vancouver, and actually exceed mainstream rates a bit more than for their counterparts in the U.S. centers of Chinese immigration. The second finding is that the very high rates of inter-generational educational mobility for the Chinese second generation also appear at the urban level in both countries. However, even these high rates do vary by urban area and cross-nationally. In the U.S., the pattern of educational polarization for Chinese immigrants in the six cities of greatest Chinese concentration is reflected both in higher education for the native-born whites, and lower education for the older Chinese immigrants in such cities. For the second-generation this means that educational success in the high immigration cities requires much higher rates of inter-generational educational mobility. As a result, although rates of educational mobility are in fact substantially higher for the Chinese second generation in these cities, 22

25 their levels of educational attainment relative to the mainstream populations are less. In Canadian cities, the increase in post-secondary education below the bachelor s level is very marked, and given that rates for the Chinese immigrant generation were relatively low, unusually high rates of inter-generational change were required to produce results which are fairly similar to what is seen in the U.S. It is interesting also to compare the Chinese and black second generation with regard to the significant variability of inter-generational educational change. At the national level the Chinese have higher levels of educational attainment than blacks relative to the mainstream standard, and higher levels of inter-generational educational mobility. At the urban level, the Chinese also have higher levels of educational attainment than blacks. However, circumstances dictate that wide variations in intergenerational change lie behind the observed results, and these variations may disrupt the usual betweengroup comparisons. In one instance, namely that of inter-generational increase in all post-secondary education in high-immigration U.S. cities, the rates are about the same for blacks and Chinese just short of 50 percent. As well, the rate of inter-generational increase in bachelor s or higher degree attainment for blacks in high immigration U.S. cities (36.7 percent) is higher than the corresponding rate for Chinese outside the high immigration cities (27.3 percent). For both groups the rates of change in the high immigration cities are substantially higher than for either group in U.S. cities outside these prime immigration cities. Conclusions Although instances in which comparisons depend on specific educational criteria suggest the need for further assessment of the labeling of forms of post-secondary education in each country, the 23