4 Results
Archived Content
Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.
4.1 The level of co-ethnic concentration at the workplace
In Canada's eight largest metropolitan areas, about 10% of non-British, non-French immigrants share a common ethnic origin with the majority of their co-workers (Table 1). Another 6% of immigrants work in settings where about half of their co-workers are from the same ethnic ancestry. The level of co-ethnic concentration at the workplace is much lower among the Canadian born.9 Only 4.5% of Canadian-born workers share a common ethnic origin with the majority of their co-workers.
Among immigrants, the level of co-ethnic concentration at the workplace varies greatly across ethnic groups. Chinese and Portuguese immigrants have the highest levels at 19.9% and 17.6%, respectively. Italian, Filipino and South Asian immigrants also have relatively high levels of co- ethnic concentration. Germans and other European immigrant groups rarely work in an ethnically homogenous environment. Among the Canadian born, the group difference is small, and the level of co-ethnic concentration is generally low. Only the Italians and Portuguese show relatively high levels of co-ethnic concentration.
As a point of reference, Table 2 presents the level of residential concentration measured at the street-block level from the 2001 Census data. Note that the population size at the street-block level—with a median size around 100 or about 30 households—is probably larger than the size of a typical work site—this information is not available in the Ethnic Diversity Survey (EDS). The residential-concentration level could be higher ifit was based on a smaller geographic unit. Keeping this caveat in mind, the comparison of Table 1 with Table 2 reveals two interesting patterns. First, among immigrants, Chinese, South Asians and Italians have high levels of co- ethnic concentration, both at the workplace and in residential neighbourhoods. Filipinos and Portuguese have relatively high levels of workplace concentration, but low levels of residential concentration. Germans and Blacks have relatively low levels of concentration, both at the workplace and in residential neighbourhoods. Second, for each ethnic group, immigrants and the Canadian born have very similar levels of residential concentration. By comparison, for visible minority groups, the relatively high level of workplace concentration among immigrants narrows substantially among their Canadian-born counterparts.
4.2 The characteristics of workers in ethnically concentrated workplaces
Low educational levels characterize both immigrant and Canadian-born workers who work mostly with their co-ethnics. As in the top panel of Table 3, about 49% of immigrant workers who share the same ethnic ancestry with most of their co-workers have education beyond high school, compared with 72% among those with few or no co-ethnic co-workers and 68% among those reporting their ethnic ancestry is not important to them. Even individuals who are not working have higher educational levels than those who mostly work with their co-ethnics. In general, these patterns held for both visible-minority and European-origin immigrants. Among the Canadian born, the disadvantage in educational levels for workers who mostly work with their co-ethnics is concentrated among visible minorities.
Low proficiency in English/French is another characteristic among immigrant workers who mostly work with their co-ethnics. As shown in the bottom half of Table 3, about 31% of immigrant workers who share the same ethnic ancestry with most of their co-workers have English/French as their mother tongue or grew up in an English/French-speaking family environment. This is much lower than the level of 62% among those who had few or no co- ethnic co-workers and 68% among those reporting their ethnic ancestry is not important to them. This is also lower than the level of 45% among those who are not working. The disadvantage in English/French proficiency is particularly severe among visible-minority immigrant workers in ethnically homogeneous settings.
Among Canadian-born visible minorities, workers who mostly work with their co-ethnics show a clear lack of proficiency in English/French. But for Canadian-born European groups, language proficiency does not vary with the level of co-ethnic concentration at the workplace.
Education and proficiency in host-society language may not be the only factors associated with the degree of workplace concentration. To discern the effects of other sociodemographic variables, a multinomial regression model is constructed for immigrants (left panel in Table 4) and the Canadian born (right panel Table 4) separately. The dependent variable for each group is the five categories representing work status and degree of workplace concentration, with the category that most co-workers are co-ethnics as the common reference. Table 4 shows the odds ratios estimated from the multinomial regression models. For example, immigrant women's odds of working with few or no co-ethnic workers compared with working with most of their co- workers being co-ethnics is 1.39 times larger than the corresponding odds for men (the first number in Column 3).
As shown in the first column of the left panel in Table 4, compared with those who are not working, immigrant workers who share the same ethnic ancestry with most of their co-workers tend to have lower proficiencies in English/French, a stronger sense of ethnic belonging and more co-ethnic friends. They also tend to be men, are younger, are less likely to be recent immigrants, are more likely to be married, and are less likely to have young children at home.
Among workers, lower proficiencies in English/French, stronger ethnic belonging and more ethnic friends are also significant factors in distinguishing immigrant workers who share the same ethnic ancestry with most of their co-workers from those with fewer co-ethnic workers or those reporting that their ethnic ancestry is not important (Columns 2 to 4 in the left panel of Table 4). Lower levels of education and being recent immigrants (in Canada for five years or less) are two other important factors for working in ethnically homogenous settings.
Among the Canadian born, using minority languages and having ethnic friends are the two variables with large effects distinguishing workers who mostly work with co-ethnics from those who are not working, as well as from those with fewer co-ethnic co-workers. Unlike immigrants, whose use of a minority language in social functions could at least partially result from lack of proficiency in English/French, not using the host-country language(s) among the Canadian born probably reflects the individuals' preference for using their minority languages. The effect of education is either insignificant or in the opposite direction to that among immigrants. These results suggest that individual preference may play a major role in affecting co-ethnic concentration at the workplace among the Canadian born. By comparison, both the disadvantage in human capital and individual preference are important factors for working in an ethnically homogeneous workplace.
In addition to deficits in human capital, immigrant workers in ethnically homogenous working settings are disproportionately represented in low-paying occupations and industries. As in Table A.1, among immigrants, workers in an ethnically homogenous environment are more likely to work in sales and services occupations (33%) and in occupations unique to primary industries, processing, manufacturing and utilities (21%) than workers with few or no co-ethnic co-workers (20% and 11%, respectively). Compared with workers who have few or no co-ethnic co-workers, immigrant workers in ethnically homogenous settings are over-represented in personal services industries (29% compared with 10%), but less likely to work in business services industries (15% compared with 26%) and public service industries (9% compared with 21%). Immigrant workers in ethnically homogenous settings are more likely to be self-employed than are other workers.
Among the Canadian born, the differences in occupational and industrial distribution by the level of co-ethnic concentration at the workplace are not as large as among immigrants. Still, Canadian-born workers in an ethnically homogenous working environment are over-represented in occupations unique to primary industries, processing, manufacturing and utilities, and in personal services industries.
4.3 Co-ethnic workplace concentration and earnings
The question here is whether the observed disadvantages in human capital and job attributes can explain any earnings gaps associated with working in ethnically homogenous settings. When differences in individual characteristics and job attributes are accounted for, the remaining earnings gaps could be due to lower wage rates specifically associated with ethnically homogenous work settings, other job characteristics that are not measured in the study, such as firm size and union coverage, and/or the tendency among those who are highly motivated and with higher-than-average abilities—rather than those observed in the study—to stay away from an ethnically homogeneous workplace. It is difficult to separate the effects due to unmeasured job attributes and unobserved heterogeneity, but the omission of both factors would lead to an over- estimate of the earnings disadvantage associated with ethnically homogenous work settings (Edin, Fredriksson and Åslund 2003). If the remaining gap is small when observed individual characteristics and job attributes are controlled for, and if unobserved abilities of workers in ethnically homogenous settings are not higher than average, it is reasonable to conclude that working in these settings is not directly associated with large disadvantages in earnings.10
The results in Table 5—based on regression models presented in Table A.2—show that only immigrant male workers in an ethnically homogenous environment show, on average, a large gap in their earnings relative to those with few or no co-ethnic co-workers (as in Model 1's). This observed earnings gap primarily results from the difference in individual human capital and job attributes. Among immigrant men, when controlling for basic demographic variables, including ethnic groups, potential years of work experience, family structure and metropolitan area of residence, those workers in an ethnically homogeneous environment earn 33% less than do workers with few or no co-ethnic co-workers. The gap reduces to 18% when years of residence in Canada, education and proficiency in English/French are controlled for. The gap narrows further to 11% when differences in occupational/industrial distributions, self-employment status and working time are accounted for. Differences in individual human capital and job attributes account for about two thirds of the observed earnings gap. The remaining earnings gap is not statistically significant at p=0.05.
Among Canadian-born women, the earnings gap stands at 17%, sizeable but not statistically significant, between workers in ethnically homogeneous settings and workers with few or no co- ethnic co-workers. This gap narrows to 11%, when controlling for differences in education, language proficiency and job attributes. By comparison, Canadian-born women who report that ethnic ancestry is not important earn significantly more than those who report that ethnic ancestry is important, but work with few or no co-ethnics. Among immigrant women and Canadian-born men, the earnings variation associated with workplace concentration is generally small or statistically insignificant.
The negative association between working with co-ethnics and earnings appears to be in absolute value greater among the Chinese and Italians. This is shown in Table 6, which is based on regression models constructed separately for each large ethnic group (with at least 300 observations when combining immigrants and Canadian born, as well as men and women). When observed individual characteristics and job attributes are controlled for, workers in an ethnically homogeneous environment still earn about 32% less than workers with few or no co- ethnic co-workers, among both Chinese and Italians. Note that the Chinese have the highest rate of working in ethnically homogeneous settings among immigrant groups, while the Italians have the highest rate among the Canadian born.
When analysis is performed separately for the immigrant and Canadian-born Chinese and Italians, the disadvantage associated with working in an ethnically homogeneous environment exists only among Chinese immigrants—mostly among Chinese immigrant women—but it persists for both immigrants and the Canadian born for Italians (tables not included, but available upon request). Note here the comparison is within group. The large earnings gap associated with
workers in ethnically homogenous settings within these two groups does not necessarily mean that these workers earn much less than workers from other groups that also work in ethnically homogenous settings. Indeed, further analysis shows that male Chinese immigrants who work in ethnically homogenous settings have a trivial earnings gap (2%) with their counterparts from all other groups. Female Chinese immigrants who work in an ethnically homogenous environment have a relatively large earnings gap (19%) with their counterparts from all other groups, while female Chinese immigrants who have few or no co-ethnic co-workers enjoy a relatively large earnings advantage (22%) over their counterparts from all other groups. This suggests that the large variation in earnings associated with workplace concentration results, at least partially, from the well-off Chinese immigrants who work in non-ethnic settings relative to their counterparts from other groups.
Similar patters exist for Italian immigrant men and women, and Canadian-born Italian men. Canadian-born Italian women who work in an ethnically homogenous environment have a large earnings gap (21%) with their counterparts from all other groups, while Canadian-born Italian women who have few or no co-ethnic co-workers have similar earnings as their counterparts from all other groups.
4.4 Co-ethnic workplace concentration and life satisfaction
Immigrants working in ethnically homogeneous settings are less likely to report low levels of life satisfaction than workers with few or no co-ethnic co-workers (Table 7, which is based on logistic regression models in Table A.3). This holds true for both men and women. For instance, when the difference in observed individual characteristics and job attributes are controlled for, the odds ratio of reporting low levels of life satisfaction is 0.60 for male immigrant workers in ethnically homogeneous work settings relative to workers with few or no co-ethnic co-workers. This odds ratio is equivalent to an estimated 7-percentage-point difference on an average of 17% reporting low levels of life satisfaction. Further controlling for annual earnings, the odds ratio of reporting low levels of life satisfaction for workers in ethnically homogeneous settings reduces slightly for immigrant men, but it remains the same for immigrant women.
Immigrant men who share the same ethnicity with about half of their co-workers and whose ethnicity is unimportant to them are more likely to report low levels of life satisfaction than workers with few or no ethnic co-workers. This suggests that the degree of co-ethnic concentration at the workplace and self-reported life satisfaction are not linearly associated or that some complicated selection patterns are at play.
Among the Canadian born, workers in ethnically homogeneous settings are not significantly different in reporting low levels of life satisfaction from those with few or no co-ethnic co- workers. Similar to the pattern observed among immigrant men, Canadian-born men who share the same ethnicity with about half of their co-workers and those whose ethnicity is not important to them are more likely to report low levels of life satisfaction than those Canadian-born men with few or no ethnic co-workers. Canadian-born women who share the same ethnicity with about half of their co-workers are also more likely to report low levels of life satisfaction than do women with few or no ethnic co-workers.
For large visible minority/ethnic groups, workers in ethnically homogeneous settings are generally less likely to report low levels of life satisfaction than those workers with few or no co- ethnic co-workers, although the difference is not statistically significant among South Asians, Blacks and Italians. This result is shown in Table 8, where immigrants and the Canadian-born men and women are combined in the models.
9 In this paper, the Canadian born include only those who were born in Canada with at least one immigrant parent.
10 However, if workers in ethnically homogenous settings had lower-than-average unobserved abilities, they would be more likely to be selected into poorly-paid jobs. Thus, it is possible that accounting for differences in occupational and industrial distribution leads to 'over-control.' Therefore we present models with and without controlling for occupational and industrial distribution and working time.
- Date modified: