Immigration in the Long Run
The Education and Earnings Mobillity of Second-Generation CanadiansOctober 9th, 2008
The real test of a society’s inclusiveness toward newcomers is not only the socio-economic well-being of immigrants in the years following their arrival, but also – and possibly more important – the socio-economic progress of their children in the decades afterwards. In this study, Miles Corak examines how Canada has performed in that latter respect. More specifically, he focuses on the education and earnings outcomes of second-generation Canadians and on the relationship between those outcomes and the educational attainment and earnings of their immigrant parents.
Schooling is often seen as an important avenue to greater engagement and participation in society, and as a gateway to successful labour market outcomes. But these positive effects cannot occur, Corak explains, if higher education is the preserve of those born to highly educated parents and, similarly, if having parents with little education predestines their children to low education levels. It is also a problem, he adds, if the most important factor explaining the schooling outcomes of second-generation Canadians is parental earnings.
At the same time, promoting higher educational attainment will not be sufficient if the labour market does not also reward aspiration, energy and talent, as opposed to privilege and pedigree. A scenario in which disadvantage is transmitted across generations – where low income in combination with low education in one generation begets low income and low education in the next – is the most obvious indication of a lack of equal opportunity.
In examining these cross-generational equity concerns, Corak answers five relevant public policy questions: What is the degree of intergenerational education mobility in Canada, and is it different among immigrants and their children? What are the main determinants of educational outcomes for second-generation Canadians? Has the link between parent and child education outcomes strengthened or weakened over time? What is the degree of intergenerational earnings mobility among the children of immigrants? And, finally, how does that earnings mobility compare with the mobility of the population as a whole and with the experience of other countries in this regard?
Overall, the author finds that Canada is a highly mobile society in which the education and earnings outcomes of children are not strongly related to those of their parents. His analysis indicates that:
- Second generation Canadians are generally younger, more educated, do no worse and sometimes do better than their counterparts in the labour market.
- the tie between the education outcomes of parents and those of their children is looser among immigrants than among the Canadian-born population, and the education outcomes of second-generation children, particularly those with less-educated parents, are not predetermined by family background;
- the educational attainment of Canadian-born children of immigrants are only weakly related to the earnings of their parents;
- adult earnings are not strongly tied to parental earnings, and this holds true for both immigrants and the population as a whole; moreover, this tie is much weaker in Canada than it is in countries such as the United States, the United Kingdom, France and Germany; and
- the degree of educational mobility for children who grew up during the past two decades is no different than it was for children raised during previous decades.
Corak concludes that, in terms of the education and earnings outcomes of Canadian-born children of immigrants, Canada’s performance is encouraging. There are, however, two caveats. First, some immigrant communities, particularly those from Caribbean, West African and Latin American countries, are experiencing low earnings across the generations, even when both parents and children have above-average education – an intergenerational dynamic that should raise a red flag for policy-makers. Corak argues that we need to understand how this situation has come about and whether it is changing among more recent cohorts of immigrants from these countries.
Second, Corak cautions that the intergenerational analyses he undertakes in this paper are inherently backward looking, illustrating the experience of Canadian-born children whose immigrant parents came to this country before 1980, who attended primary and secondary school in the late 1970s and early 1980s, went to university in the 1980s and worked in the labour market of the 1990s. However, while there is no guarantee that more recent cohorts are experiencing similar outcomes, preliminary data does suggest the persistence of earlier trends.
In his conclusion, Corak makes the case for adopting a longer-term perspective on the meaning of social inclusion, and argues that, when devising immigration policy, governments should balance shorter-term labour market goals and longer-term societal objectives such as family and education.
Words like “inclusion,” “exclusion,” “accommodation,” “integration” and “social cohesion” have become important touchstones for the discussion of public policy in Europe, North America and other countries of the Organisation for Economic Co-operation and Development (OECD). Much of this discussion, of course, focuses on the place of immigrants in the economies and societies of these mature democracies, and is particularly controversial in the wake of events during the first years of the new century. Race-related riots in the streets of Paris and on the beaches of Australia, controversial publications in newspapers — most notably in Denmark but also in other European countries — and heated debate and protests over immigration legislation in countries as different as the Netherlands and the United States all represent particular flashpoints of much deeper and persistent challenges associated with adapting to and accepting increasingly diverse immigrant and visible minority populations into host countries that are themselves characterized by more inequality and less solidarity than in past decades.
At the same time, many commentators have also come to appreciate that an important test of a society’s ability to both adapt and integrate concerns the children of immigrants — the so-called second generation. The individuals throwing Molotov cocktails in the suburbs of Paris during the autumn of 2005 were not, after all, immigrants to France, but the French-born children of immigrants. Discussions concerning the place of Muslim immigrants in German society, to cite another example, became much more salient with the growing realization that their adult offspring, particularly the women, retain traditional values and are not integrating into the mainstream of society. In other words, the clearest marker of whether a society is inclusive is the extent to which the native-born children of immigrants grow up to be fully engaged and self-reliant adults, contributing to and influencing the mainstream, as well as benefiting from it.
In this context, the schooling of immigrant children is often cited as an important outcome related to their capacity to succeed in the labour market and to adapt to the values of the mainstream. In a lead editorial in the aftermath of the French riots, an influential Canadian newspaper asked the question that was on the minds of policy-makers throughout the OECD: “Could it happen here?” The editorial concluded by suggesting that even the citizens of Canada, which is often held up as a model for successful integration of immigrants, “can only hope that their public education system and their public institutions can somehow impart a sense of shared values…That dream must be matched with the promise of equal opportunity.”1
The Canadian experience is an interesting case to examine because it is indeed often held up as an international success story. Accordingly, the integration of immigrants and second-generation immigrants has been the subject of a good deal of discussion and study, but with differing interpretations. Soroka, Johnston and Banting, for example, conclude their analysis of data on national identity and belonging, social values and trust and social and political participation of both firstand second-generation Canadians by stating that “the largest challenges to social cohesion in Canada remain rooted not in the attitudes, beliefs and attachments of relative newcomers but in the historic fault lines between the oldest nations that make up the country” (2007, 4-5). Reitz and Banerjee also look at similar indicators, but are less sanguine in their conclusions, suggesting that “experiences of discrimination and vulnerability remain, slowing the social integration of minorities. Furthermore, these effects may be intensified for the children of immigrants, whose expectation of equality may be greater than was the case for their parents” (2007, 34).
In this study, I hope to inform this sort of discussion by focusing on the education and labour market outcomes of the children of immigrants. These dimensions relate directly to the concerns expressed in the newspaper editorial referred to above — namely, the values that lead to social integration, and the education system as the main instrument in promoting them, but also the opportunities that are based on a labour market that, in a sense, is a level playing field on which newcomers and their children can succeed on the basis of their aspirations and talents.
Questions to be examined
My frame of reference is a pair of studies by Aydemir, Chen and Corak (forthcoming, 2008) and a growing literature on the generational mobility of earnings and education that has come to complement the large amount of research on the social and economic position of immigrants. The perspective on “integration” that I adopt is based on the strength of the tie between the situation of immigrants — in particular, their level of education and their labour market earnings — and the adult outcomes of their Canadian-born children, the second generation.
In this context, I ask a series of questions related, first, to the education outcomes of the children of immigrants and, second, to their earnings as adults. One hopes that these questions are framed in a way that is relevant to an appreciation of the accomplishments of the past and the challenges of the future.
A focus on education is important because schooling is often seen as an important avenue to engagement and participation in society and as a gateway to successful labour market outcomes. Accordingly, policy-makers concerned with issues of “inclusion” need to know the answers to at least three questions dealing with the schooling outcomes of the children of immigrants.
First, what is the degree of generational education mobility in Canada as a whole, and is it different for immigrants and their children? In other words, is higher education the preserve of those born to highly educated parents and, similarly, does having parents with lower levels of education predestine children to low education? The answer to this question highlights how the education system functions and whether it does so differently for immigrants. If the education outcomes of the Canadianborn children of immigrants are closely tied to parental education levels — and, indeed, more closely tied than for the children of Canadian-born parents — then there is a greater presumption that values and opportunities are based in and transmitted from the home rather than the broader community. An extreme example might be the occasional claim that less-educated families in some immigrant communities actually discourage their children, particularly daughters, from pursuing higher education. In this context, relying on the education system to promote rather lofty integrative goals might be an overly optimistic strategy — that success requires institutional reform or behavioural change.
Second, what factor — either parental earnings or parental education — is most closely related to the schooling outcomes of second-generation Canadians? The answer would help to shed light on the worry that the current economic situation of immigrants has strong implications for the next generation. In the following pages, I briefly review our understanding of the relative decline in the economic status of immigrants, particularly recent immigrants, which has been well documented by a series of recent research reports.2 But the salience of all of this from a longer-term perspective is that, if money (i.e., parental income) matters a good deal in determining the ultimate education attainment of the children of immigrants, there might be long-term challenges for social and economic integration in the sense that low income now leads to lower education in the next generation. On the other hand, if things other than money — such as the expectations, aspirations and values that immigrants have for their children and pass on to them — might be related to their own education level, then current economic difficulties might not echo into the next generation; if that is so, it might be appropriate to adopt a more sanguine perspective on the long-term integration of immigrants and their children.
The third and final question relates to matters of schooling: has the strength of the tie between parent and child education outcomes changed over time? The answer would help to put current challenges into context. If the patterns in the degree of intergenerational transmission of education are no different now than they were a generation or two ago, this would suggest a continuity in the capacity of Canadian society to deal with the challenges it currently faces, rather than an indication that the current situation is something different and untested. It might be that, although Canadians face challenges with respect to the integration of new immigrants, particularly those from visible minority groups or from vastly different origin countries, these challenges might be no different than those faced generations ago, when the Irish, Italians, Ukrainians and others came to this country and were seen at the time as outsiders. In other words, to understand the nature and extent of current challenges, it is necessary to determine whether there is continuity with the past or whether things are in some way fundamentally different now.
As important as education may be, however, it is not enough to promote social and economic integration if the labour market is not structured in a way that rewards aspiration, energy and talent, as opposed to privilege and pedigree. A scenario in which disadvantage is transmitted across generations, one in which low income in combination with low education in one generation begets low income and low education in the next, is the most obvious indicator of something less than equality of opportunity.
But another scenario could also be as socially corrosive, or even more so. It would be particularly frustrating for both immigrant parents and their children if parents were not able to provide financially for their children despite having reasonable or even high levels of education, and then to witness the same scenario playing out in the next generation, with their grandchildren suffering material want because the Canadian labour market does not offer opportunities and just economic returns to their children even though they were born in Canada and might have reasonable or even high levels of education. Such an intergenerational economic dynamic — high education and its attendant expectations but diminished economic rewards — could also be the basis for social exclusion.
So I pose two additional, and related, questions on the economic side of the ledger that are also relevant to understanding the extent of integration among the children of immigrants, and that shed light on the possibility of these scenarios playing out. First, what is the degree of generational earnings mobility among the children of immigrants? Second, how does their mobility compare to that of the population as a whole or, for that matter, to that in other countries? In other words, to what extent do the children of immigrants occupy the same relative position in the earnings distribution as their parents — lower-income parents raising children who grow up to be lower-income adults, while higher-income parents raise higherincome children — and does this tendency differ from the experience of Canadian-born parents and their children or from the situation in other countries?
The remainder of the study is structured as follows. First, I review the basic facts of immigration to Canada and the integration of recent immigrants into the Canadian labour market. It is well known in public policy circles that immigrants form a significant fraction of the Canadian population, that the inflow of immigrants is also significant but increasingly from more diverse parts of the world and that the earnings of recent immigrants have fallen relative to those of the Canadian population, so that their chances of living in low income are much higher. I suggest, however, that the common interpretation of these facts likely overstates the extent to which they represent challenges for social inclusion. On this basis, I imply that it is the expected and actual outcomes of the children of immigrants that need to be taken into account to appreciate fully the challenges to Canada’s social and economic fabric.
Accordingly, in the second section, I offer a statistical overview of the second-generation population, and I point out that this is, in fact, an advantaged group with respect to education and labour market outcomes. Second-generation Canadians are generally younger and better educated and do at least no worse and sometimes better in the labour market than the offspring of Canadian-born parents. These are the children of immigrants who came to Canada before 1980, who were educated in the 1970s and 1980s and who entered the labour force in the 1990s.
The third section is the heart of the matter. Here, I present the major results of the analysis of generational mobility and offer answers to the questions I posed above in a way that casts a positive light on the immigrant and second-generation experience in Canada. (An appendix outlines the framework used in the academic research from which this section borrows, and offers a brief overview of the major findings in this literature.) At the same time, the analysis suggests that this sanguine perspective might not hold for certain immigrant communities. The major message is that, even though the concerns and experiences of particular groups need to be addressed, overall the experience of second-generation Canadians is something to be celebrated.
In the final substantive section of the study, I attempt to address some issues and cautions for the conduct of public policy, arguing that the relatively positive conclusions reached here might still be appropriate for more recent cohorts.
A Portrait of Recent Immigrants in Canada
Alook at some broad facts about immigration to Canada and the labour market experience of recent immigrants might reasonably suggest that immigrants find themselves socially and economically excluded, which could lead to the kind of ferment and social unrest experienced by other countries with which Canada is often favourably compared. A critical eye needs to be used, however, in interpreting some of these data, as well as an appreciation of the institutional context in which these trends are playing out.
First, a relatively large proportion of Canada’s population is foreign born: census data for 2005 show that 21 percent — more than one in five — of the population were born outside the country, the highest proportion since 1931, when it stood at a historical high of 22 percent (Diversity Watch 2008). Canada is hardly unique, however, in having a large proportion of immigrants in its population: as figure 1 shows, the foreign-born population of a number of OECD countries exceeded 10 percent in 2001. Although these data should be accepted with a certain amount of caution, as they might not account fully for illegal immigrants — particularly in the cases of the United States and, probably, Italy — Canada clearly is among the more important OECD host countries for immigrants, although it is not in any sense an extreme outlier. At the 2001 level of 18 percent, the proportion of immigrants in Canada is significantly higher than that in France, Germany or the United States, countries that have experienced a good deal of controversy on issues related to immigration, but the proportion is higher still in Australia and Switzerland, and about the same in New Zealand. Some countries with significantly lower fractions of immigrants in their population than Canada are experiencing challenges associated with social inclusion, so that, on their own, these proportions do not tell us much about the degree and nature of such challenges.
Second, not only is the stock of foreign-born among Canada’s population significant; the flow has become increasingly diverse. Before 1970, the very large majority of newcomers originated in Europe and the United States, and were often thought to have arrived with values and outlooks relatively similar to those of the Canadian mainstream. The change in the selection process in the mid-1960s, however — from one based on national origin to one eventually based on a points system that reflected language skills, work experience and other criteria associated with labour market success — resulted in a shift in the countries from which immigrants came to Canada. As figure 2 illustrates, the proportion of immigrants coming from Asia, the Caribbean, Latin America, the Middle East and Africa rose in each subsequent decade, reaching three-quarters in the 1990s. As a result, it is often suggested that, since the proportion of Europeans among newcomers has greatly declined, the challenge of integrating the flow of immigrants into Canadian society has increased (see, for example, Sykes 2008). This view might be an overstatement, however, for at least two reasons. First, it presupposes that, in earlier times, there was as much commonality of values among Europeans and with those of the Canadian mainstream as there might be now. In fact, there were sharp distinctions between the values of immigrants from southern and eastern Europe and those of immigrants from northern and western Europe; there were also sharper distinctions in terms of religious background (Protestants versus Catholics), as well as language. Second, the suggestion that the challenge of integration has increased also presupposes that the “mainstream” is static, that it is not influenced or changed by immigrant flows, which is certainly not the case in Canada. Thus, if the meaning of “mainstream” values adapts and is influenced by a more diverse flow of immigrants, then the challenges to integration might be lessened.
None of this is to deny that, as figure 2 shows, there have been important changes in the nature of the flow of immigrants and, consequently, in the characteristics and values they bring with them. But these changes are playing out in the context of a country that has a long history of interacting with newcomers, and one in which identity has always been something to be negotiated and renegotiated. Biles, Burstein and Frideres have called this a “twoway street” in which “both immigrants and current citizens are expected to adapt to each other, to ensure positive outcomes for everyone” (2008, 4).
Immigration and the notion of citizenship have interacted and evolved throughout the last century, possibly to a degree not matched in many other OECD countries: from a starting point in the Immigration Act of 1910, which conferred on cabinet the authority to exclude “immigrants belonging to any race deemed unsuited to the climate or requirements of Canada”; through the Canadian Citizenship Act of 1947, which distinguished Canada as the first Commonwealth country to create its own class of citizenship separate from that of Great Britain; and ultimately to the 1977 Citizenship Act and the interplay between multiculturalism and official bilingualism since that time (Citizenship and Immigration Canada 2000). This interaction continues today with the increasing participation of immigrant groups in the political process (Anderson and Black 2008). As such, changes in the origin countries and diversity of recent immigrants may pose different challenges than they might in other countries with a different historical backdrop and a more rigid sense of national identity.
That said, it is also often pointed out that the labour market situation of new arrivals has deteriorated significantly relative to that of earlier immigrants. Indeed, as figure 3 shows, over the past few decades, there has been a quite notable deterioration in the relative earnings of immigrant males. Immigrants who came to Canada during the 1960s earned essentially the same as Canadian-born males, but the relative earnings of each subsequent cohort of newcomers have declined; by the 1990s, recent male immigrants were earning half as much as Canadianborn males of the same age. Aydemir and Skuterud (2005) suggest that about one-third of this decline reflects the changing mix of language skills and country-of-origin effects, another third is due to a reduction in the value of foreign work experience and the remainder reflects adverse labour market conditions at the time immigrants arrived and the lingering effects of this unfortunate timing on future earnings — in part, associated with a policy change in the mid-1980s to stop reducing total immigrant admissions during economic recessions (see Institute for Competitiveness and Prosperity 2007, 34-6).
The upshot of these labour market developments has been, in each subsequent decade since 1980, an ever-larger fraction of immigrants living on low incomes. As illustrated in figure 4, about 15 percent of non-immigrants live on low incomes, a fraction that has declined slightly since 1980; for recent immigrants, however, the fraction has risen significantly so that, by 2000, more than one in three were living on low incomes. Indeed, Hou and Picot (2003) point out that all of the increase in low-income rates in Montreal, Toronto and Vancouver is due to the increase in low-income rates among immigrants.
These developments are quite rightly viewed with concern. But it should also be recognized that “low income” and “poverty” are relative concepts in mature economies like that of Canada, always entailing a comparison with the prevailing patterns of consumption that characterize the mainstream. That is, to understand what it means to live in low income or poverty, it is important to appreciate to what living standard individuals are comparing themselves. Public policy discussions assume that trends in the low incomes of more recent immigrants have the same salience as they would for the Canadian-born population, but no evidence is offered to support this assumption or to make clear what the appropriate reference point should be. We really do not know how immigrants gauge their situation in Canada or whether low income has the same implication for recent immigrants who might tend to compare their current circumstances with those they left behind, rather than with those of other Canadians. Many immigrants have left behind challenging and even desperate situations — one need only think of the experiences of some political refugees — to build a better life in Canada, and it is reasonable to suppose that the past weighs heavily in their evaluation of their current living standards. To some extent, one can generalize this perspective to cover most immigrants, who recognize that their decision to emigrate will involve a period of adjustment.
For all immigrants, however, regardless of their starting point, the realization eventually takes hold that they have to live their daily lives in the Canada of the here and now, not in their county of yesterday. How quickly this happens will vary from community to community and individual to individual. We simply do not have hard evidence on how the process unfolds. The point to keep in mind is that the reference points embedded in figure 4 are neither totally accurate nor unchanging, and assuming that all recent immigrants instantly compare themselves to the typical Canadian overstates the degree to which relative economic hardship translates into failed expectations.
Moreover, for those who come to Canada with high skills and high expectations that are not immediately met, there is often the option of leaving. Aydemir and Robinson (2008) document that up to one-third of immigrants leave the country, most within a year of arrival. For some, this “exit” option offers an escape valve that prevents the buildup of discontent. And for many of those who decide to stay, they do so in the hope of improving not just their own living standards but also the opportunities and lives of their children — indeed, immigrants often shoulder the sometimes very large costs of settling in a new land precisely because of the benefits they perceive for their children, a longterm consideration that is often forgotten in public policy discussions based on information of the sort offered in figures 3 and 4.
So could the social disruptions that have occurred in other countries happen in Canada? While it is true that the proportion of the population that is foreign born is higher in Canada than in most OECD countries and that the inflow of new immigrants has become more diverse in its origins, it is also the case that identity is something fluid in this country, the mainstream having a capacity to accommodate diversity in a way that is unparalleled elsewhere. Being a significant fraction of the population of Canada’s open and democratic society affords immigrants the opportunity to have a political influence that they would not have if they represented only a small proportion of the total (Anderson and Black 2008).
And while it is also true that economic outcomes have deteriorated significantly for recent cohorts of male immigrants relative to the average Canadian, it is also the case that the labour market is characterized by a good deal of flexibility and that living standards are high in an absolute sense compared with those in many of the countries immigrants have left. This is not to say that the relative earnings decline and the increase in relative poverty are not concerns, but they are unlikely to lead to sharp ruptures in the sense of identity and participation of the immigrant population.
To some degree, this is also the result of a sense of hope for the future — in particular, a sense that Canadian society and labour markets offer opportunity for the next generation. Therefore, it is important to appreciate and understand some of the facts and patterns about the cross-generational transmission of social and economic status. How the children fare plays an important part in developing an accurate understanding of whether it could happen here.
Second-Generation Canadians: A Significant and Advantaged Part of the Population
The 2001 census reveals that second-generation Canadians are not only a significant fraction of the population, but also a relatively advantaged group that performs better than third-generation Canadians and those with even deeper roots in the country across a host of education and labour market outcomes, as summarized in tables 1 and 2. Since, in the long run, integration is related to acquisition of the host country’s languages and the age of the immigrant on arrival, the tables divide the immigrant population into two groups: those who arrived before the age of 12 and those who were 12 or older when they arrived. The former group are likely to have had some schooling in the Canadian elementary system and are more likely to have developed better language skills, which could mean that they do not differ in their adult outcomes from those who were born in Canada to immigrant parents (see Schaafsma and Sweetman 2001).
As shown in tables 1 and 2, according to the 2001 census, almost 65 percent of the population ages 16 to 65 were at least third-generation Canadian or of Aboriginal origin, around 20 percent were immigrants and the remaining 15 percent were second-generation Canadians with at least one parent born outside the country (those with both parents born outside the country represented 7-8 percent of the population). In short, immigrants and second-generation Canadians form a sizable proportion of the country’s population.
According to the 2001 census, second-generation Canadians (again, among those between the ages of 16 and 65) with both parents born abroad were, on average, 35 years of age, compared with an average age of 39 for those with both parents born in Canada; just over 50 percent of the former group were less than 35 years of age, compared with less than 40 percent of the latter.
Second-generation Canadians also tended to have more education: those with both parents born abroad had, on average, about 14 years of schooling, one year more than those whose parents were born in Canada; around a third of the former group had at least 16 years of education, with more than 20 percent of men and almost one-quarter of women having at least an undergraduate university degree. In contrast, about 22 percent of men whose parents were born in Canada had that many years of education, and fewer than 15 percent had at least an undergraduate university degree. In fact, while 16 percent of second-generation Canadian men and 14 percent of women had less than 12 years of schooling, almost 30 percent of men and about a quarter of women whose parents were born in Canada had that amount of education.
These comparisons continue to favour second-generation Canadians when one looks at birth cohorts. According to the 2001 census, every age cohort of second-generation Canadians had a higher proportion with 16 or more years of education than did those whose parents were born in Canada (not shown here). For the younger cohorts, the differences were even more striking: more than 44 percent of men ages 25 to 34 whose parents were born abroad had at least 16 years of education, compared with 30 percent of men in that age cohort whose parents were born in Canada. Slightly more than half of second-generation Canadian women in this age group had at least 16 years of schooling, compared with 35 percent of women whose parents were born in Canada. Indeed, that age group of second-generation Canadian women were the most educated group in the country.
In terms of labour market outcomes, second-generation Canadians tend to fare no worse, and sometimes better, than those whose parents were born in Canada (see tables 3 and 4). In fact, according to the 2001 census, about 72 percent of Canadian-born women with both parents born abroad were employed, the highest employment rate among all the groups listed in table 4. Furthermore, average annual earnings tended to be higher for foreign-born and second-generation men, and noticeably higher for women, than for Canadian whose parents were born in this country: second-generation women with both parents born abroad earned on average just over $27,000 in 2000, or about $630 per week, compared with less than $25,000, or about $575 per week, for women with Canadian-born parents. As tables 3 and 4 show, second-generation Canadians, particularly women, whose parents were born abroad were also more likely to be in the top earnings quartiles.
To add more detail to this broad portrait of the children of immigrants, second-generation Canadians tend to be more urban than the population as a whole, with fully 30 percent of them living in Toronto and another 10 percent in Montreal and Vancouver at the time of the 2001 census; to be fluent in a language other than English or French; to be no more likely to rely on income assistance from the state than those with Canadian-born parents; and to be more likely to work in professional and administrative occupations as opposed to production jobs (Aydemir, Chen, and Corak forthcoming).
While a good deal of attention has been focused on the economic challenges that recent immigrants face, the story is not the same for the children of a previous generation of immigrants. As the 2001 census data make clear, second-generation Canadians are an advantaged group, integrated and contributing to the Canadian mainstream. It is the relationship between the education and earnings of this group and the circumstances of their parents to which I now turn.
Education and Earnings Mobility across the Generations
A well-developed body of Canadian research confirms that the educational attainment of children is tightly correlated with the education levels of their parents (see the appendix, where I also note that this is characteristic, albeit to differing degrees, of all OECD countries). Here, however, I focus not on the underlying reasons for this correlation but on the strength of this relationship and on whether it is different for immigrants and their children.
As in previous work on the transmission of education across generations, Aydemir, Chen and Corak (2008) find that better-educated parents have bettereducated children. When they compare the patterns of educational attainment among the immigrant and nonimmigrant populations, however, they develop two insights. First, highly educated immigrant parents are as capable of passing on their educational attainment to their children as are highly educated Canadianborn parents; further, children of immigrant parents with above-average levels of education also go on to obtain above-average levels of education. Second, the children of immigrant parents with lower levels of education are much more likely to attain many more years of schooling than their parents than is the case for children of Canadian-born parents, particularly parents with little education.
As figure 5 shows, the pattern that highly educated parents are most likely to have highly educated children is clearly present. The majority of Canadian-born children of fathers with a university degree go on to obtain a university degree, regardless of gender and whether or not the father was born in Canada, although highly educated Canadian-born parents have a slightly higher tendency to pass on their education advantage than do immigrant parents: almost 70 percent of daughters of university-educated Canadian-born fathers go on to obtain a university degree, compared with 66 percent of daughters of university-educated immigrant fathers. Sons of university-educated immigrant fathers, however, are rather less likely to obtain a degree than those of Canadian-born fathers (52 percent versus 62 percent).3
However, figure 5 shows very different patterns for children whose parents have less than a university education: in this case, sons of immigrant parents are more likely to obtain a university degree than are sons of Canadian-born parents (almost 25 percent versus less than 10 percent). This difference is also apparent for all other levels of paternal education below the university level, and the pattern also holds for daughters. Immigrant parents with lower education are, in general, more likely to have highly educated children than is the case for Canadian-born parents, which contributes to the higher overall levels of education among second-generation Canadians. That the children of immigrant parents are more educated than those with Canadian-born parents derives not simply from the fact that their parents are more educated than Canadians, but also from the fact that they are more “education inclined,” regardless of their level of education.
Aydemir, Chen and Corak (2008) give more precision to these findings. For Canadians in the 25-to-37year-old cohort whose parents were born in this country, each additional year their parents went to school is associated with four-tenths of a year of additional schooling. This implies that a child of university-educated, Canadian-born parents is likely to obtain, on average, about two years more schooling than a child of Canadian-born parents with high school or less, which significantly lowers the former group’s risk of dropping out of high school — indeed, it is rare for the children of university-educated, Canadian-born parents to lack high school certification. On the other hand, 10 percent of children of Canadian-born parents with less than high school also do not obtain their high school diploma.
Figure 6 shows the association between the average years of schooling of immigrant parents from 70 countries and those of their Canadian-born sons in the same 25-to-37 age cohort and compares them with the educational attainment of Canadian-born fathers and their sons.4 That many of the observations are bunched in the top right-hand quadrant of the figure makes it clear that most immigrant communities and their Canadian-born sons have more education, on average, than do Canadian-born parents and their sons in this age cohort; the pattern is broadly similar for daughters. Notable, however, is the general tendency of the relationship: broadly speaking, more parental education is associated with more education for the child, but, at 0.13 years for each additional parental year of schooling, this relationship is actually quite weak. That is, every additional year of schooling is associated with around a tenth of an additional year for Canadianborn sons (for daughters, the association is even a bit weaker). In other words, these relationships are less than half as strong as they are for comparable Canadians whose parents were born in this country.
Furthermore, the children of highly educated immigrants are no more and no less likely also to have higher levels of education than are the children of Canadianborn parents: highly educated parents in both communities have a strong tendency to pass this advantage on to their children. What makes the immigrant experience different, however, is that, for this group, educational disadvantage is not transmitted across the generations. Figure 6 echoes the results in figure 5, illustrating that, in almost all immigrant communities with relatively lower education levels, their Canadian-born children attain more years of schooling than the Canadian average — in only four communities is this not the case (Portugal, Paraguay, Ecuador and Other Central America). Thus, relatively less-educated parents seem able to pass on to their children values and behaviours that, in the context of the Canadian education system, encourage them to obtain higher levels of schooling; moreover, this tendency is stronger among immigrant families than among Canadian-born families, and is the reason for the relatively weaker overall correlation between the educational attainment of immigrant parents and their Canadian-born children.
While there is an association between education levels across the generations, in contrast there is virtually no relationship between parental earnings and the educational attainment of children. Aydemir, Chen and Corak (2008) examine in a rather broad way some of the other factors associated with generational mobility and find that income on its own plays no role. Rather, it is parental education that dominates income, and it turns out that, when a father’s education is accounted for, his earnings are actually negatively associated with his children’s years of schooling. Thus, the second-generation children making the gains in educational attainment are those of low-income but highly educated immigrants. There is something in the family life of these children that, in the context of an education system that does not appear to limit access according to income, permits them to succeed.
The final observation that Aydemir, Chen and Corak (2008) make is that these patterns have not changed over the course of the postwar period. In particular, their analysis shows that, for both men and women, the intergenerational association in educational attainment, including overall average attainment, has been stable across all birth cohorts. Canadian-born children of immigrants obtain about 0.10 to 0.15 years more schooling for every additional year of education their fathers have; this is significantly lower than the tie between Canadian-born parents and their children, who, by contrast, obtain an additional 0.3 to 0.4 years of schooling for each additional year of education their fathers have. This degree of intergenerational mobility on the part of the most recent second-generation Canadians is no stronger or weaker than it has always been, and has not changed during most of the postwar era relative to that of other Canadians.
In summary, second-generation Canadians are more educated than those whose parents were born in Canada. They have made these gains not just because their immigrant parents are more highly educated and are able to pass on this advantage — something that highly educated Canadian parents are also able to do — but also because, even when their parents are less educated, they are more likely to move up the schooling ladder. These patterns do not seem to be associated with how much money immigrant parents earn, and they characterize the Canadian experience for those who went through the education system during the 1980s and 1990s, as well as for those who went through during the 1950s, 1960s and 1970s.
Is educational attainment rewarded equally in the labour market? That is, do the Canadian-born children of immigrants, despite their educational attainment, experience discrimination, blocked access to good jobs and, ultimately, limits to their earnings? Is the Canadian labour market a level playing field or do connections and networks matter in a way that gives those from privileged backgrounds a head start, and that limits the returns on the education obtained by others who might be relative newcomers? These questions are not easy to answer, particularly when they deal with issues such as discrimination. But they are important because they ultimately determine whether the promise of immigration to Canada is a reality or a fiction for those who make the move “because of the children.”
One should first point out that, even if educational advantage is passed on strongly from parent to child among the Canadian population at large, earnings advantage is not. In a series of research reports, I and a series of co-authors have documented the degree of cross-generational earnings mobility in Canada, and have placed this in an international context (see Corak 2001, 2006). In Canada, every 10-percentage-point increase in fathers’ earnings is associated with only a 2percentage-point increase in the adult earnings of children. That is, on average, parents pass on only a fifth of their earnings advantage — or, for that matter, disadvantage — to their children. There might be an association in earnings across generations, so that the children of fathers with above-average earnings tend also to have above-average earnings, but the association is so weak as to imply that there will be none between a grandparent’s earnings and those of grandchildren. In fact, this pattern is so weak as to place earnings mobility in Canada among the highest in the OECD, similar to that in Finland, Norway and Denmark in having relatively little inequality transmitted across generations. The OECD countries with the greatest stickiness in earnings across generations are the United States, the United Kingdom and France, where between four-tenths and one-half of parental earnings advantage is passed on to children.
Aydemir, Chen and Corak (forthcoming) show that this pattern of a fifth of parents’ earnings advantage being inherited by children also holds for immigrants and their Canadian-born children — or, more accurately, their sons. The transmission of earnings advantage across generations of immigrants is no different on average than it is for the population as a whole, and, in fact, is a good deal less for daughters: there is virtually no association between paternal earnings of immigrants and the earnings outcomes of their daughters.
In an important sense, this is a good-news story, suggesting that Canadian labour markets are flexible and merit based. But the story pertains to overall average performance. That is not to say that particular groups and individuals do not experience intergenerational cycles of poverty or, for that matter, high income. The average hides a good deal of variation, and, in fact, particular groups among the immigrant population do face challenges that are not characteristic of the whole.
Figure 7 illustrates this point in a very broad way, showing that generational progress in earnings varies by region of origin, with the largest gains made by daughters whose parents were born in Asia and virtually no gains made by sons whose parents originated in the Caribbean or South America. For example, the figure shows that, during the early 1980s, when they were raising their children, immigrant fathers from North America or Europe actually made 12 percent more per week than a comparable group of Canadians. This advantage was perfectly preserved in the next generation, with sons and daughters also earning 12 percent more than their counterparts when they were adults in 2000. All other groups of immigrant fathers, with the exception of those from Africa, were at a relative earnings disadvantage when they were raising their children, making 10 to 15 percent less than the Canadian average for men of a similar demographic. Yet their children made substantial gains. The most notable were those of the daughters of Asian immigrants: their fathers earned 90 percent of the average, yet the daughters made almost 30 percent more than the average.
A similar pattern is evident among all other groups, but it is more muted for sons. In fact, sons of African immigrants made no progress over the relative position of their fathers, nor did the sons of Caribbean and South American immigrants. This latter case is particularly notable because this is the only one in which there is a strong risk that poverty will be transmitted across generations. These fathers earned about 15 percent below the Canadian average, and the next generation of sons did no better — the only case in which the second generation did not earn more than the average weekly amount.
Some high-risk communities
It is important to appreciate this variation in outcomes, as it offers a more precise picture of where discontent and disengagement might arise in Canadian society. The risks can be illustrated more clearly by using information on both education and earnings in combination, and by focusing in detail on the country of origin. Table 5 catalogues by 70 countries of origin individuals who were raising children during the 1980s according to whether fathers had aboveor below-average education and aboveor below-average earnings, and according to the position of sons and daughters in the distribution of education and earnings.
With respect to sons, note, first, that there are no cases of downward education mobility: when immigrant fathers have above-average education, so do their Canadian-born children. In other words, when immigrants have an educational advantage, they are able to pass it on to the next generation.
Second, the intergenerational transmission of disadvantage is, in fact, rather rare. There are no examples of communities where fathers with below-average education and earnings raise children who also have below-average education and earnings. In fact, in only 12 of 70 communities do fathers have belowaverage years of schooling and earnings, and in all but two of these cases sons go on to have above-average earnings, breaking out of this potentially challenging starting point. In six cases sons do so by having above-average education; others have belowaverage education despite having higher earnings. Only in two communities of immigrants — those from Cyprus and Greece — do below-average education and below-average earnings in the parental generation lead to the below average earnings in the sons’ generation, and neither country has been highlighted by the literature or by anecdote as being a particular flashpoint for discontent. There are no communities in which this outcome occurs for daughters.
Finally, one could imagine that integration into the broader community could be particularly challenging under certain circumstances. Consider the case of a father who comes to Canada with higher-than-average schooling, yet who earns less than the average, perhaps because of the difficulty in having his foreign education credentials recognized in the Canadian labour market. If this father then witnesses a similar pattern among his children, one might reasonably imagine that a sense of frustration or lack of belonging to the host country could develop in both the father and his adult offspring.
In 57 of 70 immigrant communities, fathers have above-average education; in 38 cases, they also have below-average earnings. But in only 11 of these cases, representing a small total population, do sons go on to earn less than the Canadian average despite having above-average education. That said, these cases are dominated by the Caribbean countries and, with the addition of West Africa, likely represent a visible minority group that Reitz and Bannerjee (2007) highlight and that form the basis for their rather pessimistic perspective on the integration of second-generation immigrants. There are also two Latin American countries in this group.
The situation for daughters is somewhat better. In only one case (Norway) is there downward education mobility; and in one other case (a group labelled Other South America), daughters of fathers with above-average education and below-average earnings also have the same outcomes. In this sense, the situation is better for daughters than for sons. In 37 of the 38 immigrant communities with fathers in this situation, daughters go on to have both higher education and higher earnings than their counterparts of Canadian-born parents. Furthermore, there are no examples of the intergenerational transmission of relative disadvantage in education and earnings. In 13 cases, fathers have less education than the Canadian average, and in 12 of these they also have below-average earnings. But there is only one case in which daughters find have both below-average education and lower earnings.
In summary, there is only a loose tie between the educational attainment of second-generation Canadians and that of their parents, and the relationship is looser still between the education levels of Canadian-born parents and their children. Degree of affluence has little to do with this intergenerational tie — if anything, lower-earning immigrants have better-educated children. Moreover, the strength of this relationship has not changed across the birth cohorts of the postwar era.
There is also only a loose relationship between the earnings of second-generation Canadians and those of their foreign-born parents — indeed, the tie is significantly looser in Canada than it is for counterparts in the United States, the United Kingdom or France. For second-generation men, the tie is no different than that for the population as a whole; for women, there is no intergenerational link at all, unlike the case for the population at large. While this overall average pattern is quite loose, there is a good deal of variation in outcomes; in some immigrant communities, low earnings are transmitted across the generations despite above-average levels of education among both parents and children.
A Caution: Does the Past Imply the Future?
It might be appropriate to adopt a sanguine perspective on the overall average performance of Canadian immigrants and social institutions, but this view should be tempered by the fact that specific, yet important, challenges remain, particularly with respect to the sons of immigrants from the Caribbean, West Africa and some Latin American countries. Such a view should also be tempered by recognizing that intergenerational analyses of this sort are inherently backward looking, illustrating the experiences of children whose parents came to Canada before 1980 and who attended primary and secondary school in the late 1970s and early 1980s, went to university in the 1980s and worked in the labour markets of the 1990s. One should stress, therefore, that it is the extent to which these intergenerational patterns hold in the future that is relevant for the children of the more recent cohorts of immigrants and for current public policy.
In fact, it is the significant numbers of immigrants since 1980 who have suffered the relatively significant deterioration in immigrants’ economic circumstances highlighted in figures 3 and 4. Indeed, more recent data from the 2006 census show that, of the 6.2 million foreign-born in the country, 2.8 million, or 45 percent, have arrived since 1991, when the decline in their economic prospects began to be particularly notable (Chui, Tran, and Maheux 2007). Moreover, the economic situation of these most recent immigrants has not improved since 2001 (Picot and Hou 2008). In large measure, this seems to be the result of a decline in the returns on education, which, in turn, is related to lower literacy levels.
Nonetheless, the 2006 census data also show that the immigrant population continues to grow strongly: fully 1.1 million immigrants entered Canada between 2001 and 2006, raising the proportion of immigrants in the total population to 20 percent. The country-of-origin pattern highlighted in figure 2 has also continued, with almost 60 percent of the most recent arrivals coming from Asia and the Middle East. And, most notably, given the findings presented in the previous section, the proportion of recent immigrants from the Caribbean, Central and South America and Africa increased to just over 10 percent between 2001 and 2006, up from 8 to 9 percent in 2001 (Chui, Tran, and Maheux 2007). These magnitudes are new, and the sheer numbers involved might raise issues for public policy that differ from those of the past.
That said, the educational attainment of immigrants is now much higher than it was for the parents of the second-generation children who were the focus of the analysis in the previous section. The proportion of immigrants with a university degree reached a historic high of 42 percent among those arriving between 1995 and 2000; in contrast, just 19 percent of those who arrived in the five-year period before the 1981 census had a university degree. Further, close to 55 percent of immigrants are now admitted under the economic class, up from 37 percent in the late 1970s and early 1980s (Picot and Sweetman 2005).
Does an intergenerational analysis of children whose parents immigrated before 1980 still have relevance in the current environment? There are a number of reasons to suggest that, in fact, the children of more recent immigrants are faring at least as well as previous cohorts. There is no doubt, however, that these children face considerable challenges. For example, a study of the literacy and numeracy skills of immigrant children in the midto late 1990s whose parents’ mother tongue is neither English nor French suggests that this group is particularly at risk (Worswick 2004). As figure 8 illustrates, when these children enter the Canadian education system at four to five years of age, they score 20 to 25 percent lower than children of Canadian-born parents on a battery of tests designed to assess age-appropriate reading, writing and math skills. But when they are ready to leave primary school at age 11, their test results are only slightly different from those of their Canadian-born counterparts — indeed, they might be slightly better. Considering that these children likely face greater challenges than do second-generation Canadian children of immigrant parents, their progress is all the more remarkable, and suggests that second-generation Canadians are likely to do just as well. Overall, during the 1990s, Canada’s primary public education system seems to have done an admirable job of giving children of more recent immigrants the opportunity to develop the skills needed for the next step in their education.
There are also reasons to believe that these opportunities continue to exist and are seized on during the teen years. Marshall (2007) looks at how teenagers spend their time, and begins by noting that “skills and knowledge acquisition from schooling is a teen’s most important asset for ensuring a positive socio-economic outcome later in life” (9). She finds that, on any given day, 50 percent of boys and 58 percent of girls 15 to 19 years of age whose parents were Canadian born did some homework; for the children of immigrants, the corresponding figures were 71 percent of boys and 74 percent of girls.
Likewise, a recent OECD study confirms that the patterns Worswick observes for primary-school-aged children continue to hold at age 15, with no significant difference at least in mathematics scores between those of children who immigrated with their parents and those of children born in the country to Canadian-born parents (2006b, 30). Immigrant children might start off with numeracy and literacy disadvantages when they enter the Canadian school system, but they catch up by the end of primary school and keep up in secondary school.
The OECD study also notes that, in many countries, there are considerable differences in the outcomes of secondgeneration teens and those of their native-born counterparts. Canada is one of only three countries (out of 17 studied) in which second-generation children did better than native-born children in mathematics, and one of only two in which second-generation children outperformed native-born children in reading. Of the countries in which second-generation children scored lower in mathematics tests than native-born children of nativeborn parents, the disparity in scores exceeded 20 percent in the United States, 40 percent in France and 90 percent in Germany; similar patterns were observed in reading tests. Further, the OECD suggests that the superior performance of second-generation Canadian children to that of immigrant children “give[s] some insight into the effectiveness of countries’ school systems in developing immigrant students’ mathematical literacy skills” (2006b, 32). While this might be true, it is also important to recognize that such a comparison might not control sufficiently for the fact that immigrants to Canada might have been selected differently than those in other countries. In other words, the home environment could also play a role in determining these superior outcomes.
There is also evidence to suggest that, among more recent cohorts, the lowest-performing second-generation children do better than the lowest-performing children of Canadian-born parents. Schnepf (2007) compares fourthand eighth-grade mathematics and reading test scores from 2001 and 2003 of firstand second-generation children with those of children of native-born parents in eight countries. She finds that, with a couple of exceptions, only in Canada do second-generation children outperform children of native-born parents. In fact, their scores are above those of children with native-born parents over the entire test score distribution. The difference is particularly notable, however, at the bottom: the marks of second-generation children with test scores in the bottom fifth of their group are 5 to 10 percent higher than those of children of Canadian-born parents in the bottom fifth of their group. This echoes the earlier findings that the educational attainment of secondgeneration children with less-educated parents is higher than that of children with less-educated native-born parents.
Collectively, these findings seem to suggest that, despite the possibly increased difficulties of their immigrant parents in the labour market, more recent cohorts of second-generation children continue to do well in school and to achieve a positive head start in life that likely will augur well as they move from their primary and secondary years to their young adult years. At the same time, one should note again that these are overall results pertaining to average performance; they do not illustrate how particular groups are faring, nor do they give policy-makers a hint as to whether the high-risk groups of the previous generation — those highlighted in table 5 — will remain highrisk groups in the coming years or whether other groups will join or replace them. This is an important issue for public policy that needs to be addressed with case studies of particular communities and the knowledge of analysts and advocates closer to them.
The real test of a society’s inclusiveness toward newcomers comes not simply through the socioeconomic well-being of immigrants in the years following their arrival, but also — and possibly more important — from the socio-economic progress their children make in the decades afterwards. To many newcomers, the often huge costs of emigrating are compensated for by the hope and expectation that life will be better not just for themselves but also for their children. Immigration policy thus needs to be assessed and judged in a cross-generational perspective.
In this study, I have focused on the education and earnings outcomes of second-generation Canadians and on the relationship of those outcomes to the education and earnings of their immigrant parents. The evidence suggests that there is a good deal to celebrate in the Canadian model.
- There is a good deal of mobility across the generations in Canada in general, and the education outcomes of immigrant children, particularly those of less-educated parents, are not predetermined by family background. The intergenerational tie that does exist is less than half as strong as it is for children of Canadian-born parents. Both the family values and aspirations of immigrant parents and the functioning of the education system play a role in this fluidity.
- Parental earnings are very weakly related to the educational attainment of second-generation children, suggesting that children who are raised under economic disadvantage do not experience educational disadvantage in adulthood.
- The degree of generational education mobility for children who grew up during the 1980s and reached young adulthood early in the new millennium is no different than it was for children raised during the 1950s, 1960s or 1970s. The continuity of the strength of this tie suggests that, whatever challenges Canadian public policy currently faces, they might not be any more or less manageable than in the past.
- Adult earnings are not strongly tied to parental earnings, whether among immigrants or in the population as a whole. Moreover, this tie is much weaker in Canada than it is in other countries — such as the United States, the United Kingdom, France and Germany — with which Canada isoften compared and which traditionally have had more difficulty successfully integrating immigrants socially and economically into the fabric of the marketplace.
All of this is not to suggest that there were not challenges and shortcomings in the past. While these findings refer to the overall average socio-economic progress of immigrants, there is a tremendous variation in outcomes among the many immigrant communities. In some, an intergenerational dynamic raises a red flag for public policy: above-average education combined with below-average earnings in one generation is replicated in the next generation for members of certain visible minority groups from particular countries and regions. In this sense, the data I present echo other research and anecdotal evidence. Public policy needs to understand how this has come about and how it is changing among more recent cohorts of immigrants. More research is needed for recent cohorts of immigrants, but the preliminary evidence suggests that the trend is continuing. The bulk of evidence presented here relates to immigrants who came before 1980.
But immigration policy is more than just labour market policy; it is also social policy — particularly education policy — and family policy. Since the introduction of the points system, Canada’s immigration policy has been increasingly concerned with labour market goals. The selection of immigrants has longer-term consequences, however, and it is reasonable to suggest that choosing immigrants on the basis of their education, language and occupational intentions might be echoing across the generations. In fact, there is an advantage to choosing more highly educated immigrants that extends beyond immediate labour market goals: this is the immigrant baby bonus, perhaps unintended and unforeseen, that Canadian society has reaped.
The child development literature points out that among the most important influences on longer-term outcomes of children is the education level of parents. If the “integration” of immigrants needs to be considered from a longer-term, cross-generational perspective, rather than with regard just to labour market outcomes, then it is also reasonable to ask if selection rules can be designed with an eye to the ability of immigrant families to help their children to become full participants in Canadian society.
Even if a lack of literacy skills in some way deters immigrants’ successful integration into the labour market or if their formal credentials are not fully recognized as applicable in Canada, immigrants still value formal education and pass that on to their children. If Canada has inadvertently reaped social gains by slanting the selection process toward better-educated immigrants, this process is likely to be at work to an even greater degree for more recent immigrants than for those who came before 1980, because even larger fractions of the immigrant inflow since then have been selected under the points system. In fact, the advantages to choosing better-educated immigrants extend beyond immediate labour market goals.
The characteristics, education and attitudes of immigrants are not the only factors, however, that lead to this favourable intergenerational dynamic. The immigrant community is distinguished from the population as a whole in that the children of lesseducated parents do much better than one might have expected. This is certainly a tribute to the aspirations and values such parents pass on to their children, but it is also a reflection of an education system that is inclusive, avoids the early streaming and tracking of children characteristic of other countries, focuses resources on the relatively disadvantaged and continues to be accessible even at the highest levels. In this sense, immigration policy is also education policy, with implications for provincial governments in ways that have long lags and that might be unforeseen when the lens focuses entirely on near-term labour market outcomes at the national level.
This said, while it is difficult to disentangle the independent role families play from the role domestic public policy plays, there is no doubt that different immigrant communities have different values concerning the appropriate way to promote the socioeconomic attainment of their children through the education system and the labour market. But it is also fair to say that, whatever the characteristics and values of their country of origin, immigrants, through their decision to leave and through the administrative rules used to accept them, are a select group. Only certain people come, with aspirations that condition their ability to raise their children successfully in the new country. Family is important, which also explains why immigrants with relatively little education often raise highly educated children. They themselves might not have had the opportunity to pursue their education fully in their country of origin, but they value this opportunity for their children once in Canada.
As such, immigration policy is also family policy, and the selection process also needs to recognize the independent role that families play in promoting socioeconomic progress. This line of thought might provide a different perspective on, for example, the use of temporary workers. There is anecdotal evidence that, in the past, the use of temporary workers to fill the demand for low-skilled employment might have led to more challenging family circumstances and riskier intergenerational dynamics. For instance, many female immigrants from some Caribbean countries were admitted in the past as temporary domestic workers, but initially without their children, husbands or families, and then permitted to become permanent residents.5 Low-skilled immigrants who lack both family support and the ability to raise their children themselves elevate the challenges their children ultimately face.
This perspective offers a different lens through which to view current initiatives to meet labour market shortages associated with, for example, the resources boom in Alberta, British Columbia and Saskatchewan. Policy changes to increase the flow of less-educated workers — admitting only single men and women, or using temporary worker permits so that their families have even fewer rights and less access to benefits — do not resonate with the longerterm cross-generational goal of integration, although much would depend on how such programs are implemented and the extent to which temporary status leads to permanent status.
The point is, however, that we need to reflect on the longer-term implications of these policies. Immigrants with poor education traditionally have tilted the playing field in favour of their children’s education because they themselves aspire to higher education and because the family is in a position to pass on that expectation. Immigration policy in a cross-generational context is inherently linked with broader social policies having to do with the education system, access to higher education and the creation of opportunity in the labour market. Shifting the selection process to immediate labour market goals that require less-skilled labour and weaker family attachments likely would make these goals more of a challenge to reach if the family and social policy environment has less to do with the future of children.
If Canada has managed to avoid the kind of social disruptions and challenges that many other OECD countries are facing with respect to the integration of immigrants, that is something certainly to be celebrated, but it is also something that needs to be better understood. This requires that there be a longer-term focus on social inclusion and that public policy balance shorter-term labour market goals with longer-term social consequences and the needs of families and society as a whole — a balance that is all the more challenging because the fulcrum lies within the realm of both federal and provincial responsibilities.
- “How Immigrants Fit In.” Editorial, Globe and Mail, November 12, 2005.
- In this paper, recent immigrants are those who have come to Canada within the past 10 years.
- It is certainly the case that the child development literature often cites the strong correlation between the educational attainment of the mother and that of her children, but this finding has been the subject of some recent reconsideration given that the education of the father is often not simultaneously controlled (see, in particular, Behrman and Rosenzweig . In their analysis, Aydemir, Chen and Corak  control for the education of both parents, and find that the relationship between the father’s education and that of his children is more consistent, so that is the variable I use in this exposition. This is not a central point, however, and should not be used to attribute causality to what is a descriptive statistic associated with population-wide tendencies.
- To be strictly correct, these data were calculated net of a number of other factors that might influence education levels. In particular, differences in the age and province of residence of the 70 immigrant groups were removed through regression analysis, and Ontario was used as the reference case. These adjustments, however, do not influence the results.
- “Today’s Young Black Women.” The Agenda, TV Ontario, November 23, 2007. www.tvo.org/cfmx/tvoorg/theagenda.
Anderson, Christopher G., and Jerome H. Black. 2008. “The Political Integration of Newcomers, Minorities, and the Canadian-Born: Perspectives on Naturalization, Participation, and Representation.” In Immigration and Integration in Canada in the Twenty-first Century, edited by John Biles, Meyer Burstein and James Frideres. Montreal and Kingston, ON: McGill-Queen’s University Press.
Aydemir, Abdurrahman. 2003. “Effects of Selection Criteria and Economic Opportunities on the Characteristics of Immigrants.” Analytical Studies Branch Research Paper 182. Ottawa: Statistics Canada.
Aydemir, Abdurrahman, Wen-Hao Chen, and Miles Corak. (Forthcoming).“The Intergenerational Earnings Mobility among the Children of Canadian Immigrants.” Review of Economics and Statistics. (Available as Statistics Canada Research Paper No. 267.)
————. 2008. “Intergenerational Education Mobility among the Children of Canadian Immigrants.” Statistics Canada Research Paper No. 316.
Aydemir, Abdurrahman, and Chris Robinson. 2008. “Global Labour Markets, Return and Onward Migration.” Canadian Journal of Economics 41 (4).
Aydemir, Abdurrahman, and Mikal Skuterud. 2005. “Explaining the Deteriorating Entry Earnings of Canada’s Immigrant Cohorts: 1966-2000.” Canadian Journal of Economics 38 (2).
Behrman, Jere R., and Mark R. Rosenzweig. 2002. “Does Increasing Women’s Schooling Raise the Schooling of the Next Generation?” American Economic Review 92 (1).
Biles, John, Meyer Burstein, and James Frideres. 2008. “Introduction.” In Immigration and Integration in Canada in the Twenty-first Century, edited by John Biles, Meyer Burstein and James Frideres. Montreal and Kingston, ON: McGill-Queen’s University Press.
Borjas, George J. 1992. “Ethnic Capital and Intergenerational Mobility.” Quarterly Journal of Economics 107 (1).
————. 1993. “The Intergenerational Mobility of Immigrants.” Journal of Labor Economics 11 (1).
Card, David, John DiNardo, and Eugena Estes. 2000. “The More Things Change: Immigrants and the Children of Immigrants in the 1940s, the 1970s, and the 1990s.” In Issues in the Economics of Immigration, edited by George J. Borjas. Chicago: University of Chicago Press for the National Bureau of Economic Research.
Chui, Tina, Kelly Tran, and HeÌlène Maheux. 2007. Immigration in Canada: A Portrait of the Foreign-born Population, 2006 Census. Cat. 97-557-XIE. Ottawa: Statistics Canada.
Citizenship and Immigration Canada. 2000. “Forging Our Legacy: Canadian Citizenship and Immigration, 1900–1977.” Accessed October 1, 2008. http://www.cic.gc.ca/English/resources/publications/ legacy/index.asp
Corak, Miles. 2001. “Are the Kids All Right? Intergenerational Mobility and Child Well-Being in Canada.” In The Review of Economic Performance and Social Progress: The Longest Decade: Canada in the 1990s, edited by Keith Banting, Andrew Sharpe, and France St-Hilaire. Montreal: Institute for Research on Public Policy and Centre for the Study of Living Standards.
————. 2004. “Generational Income Mobility in North America and Europe: Introduction.” In Generational Income Mobility in North America and Europe, edited by Miles Corak. Cambridge: Cambridge University Press.
————. 2006. “Do Poor Children Become Poor Adults? Lessons from a Cross Country Comparison of Generational Earnings Mobility.” Research on Economic Inequality 13 (1).
Corak, Miles, and Andrew Heisz. 1999. “The Intergenerational Earnings and Income Mobility of Canadian Men: Evidence from Longitudinal Income Tax Data.” Journal of Human Resources 34.
de Broucker, Patrice, and Laval LavalleÌe. 1998. “Intergenerational Aspects of Education and Literacy Acquisition.” In Labour Markets, Social Institutions, and the Future of Canada’s Children, edited by Miles Corak. Cat. 89-553-XIB. Ottawa: Statistics Canada
Diversity Watch. 2008. “Group Backgrounds: Immigrants and Refugees.” Toronto. www.diversitywatch.ryerson.ca/ backgrounds/immigrants.htm.
Fortin, Nicole M., and Sophie Lefebvre. 1998. “Intergenerational Income Mobility in Canada.” In Labour Markets, Social Institutions, and the Future of Canada’s Children, edited by Miles Corak. Cat. 89-553- XIB. Ottawa: Statistics Canada.
Fournier, Elaine, George Butlin, and Philip Giles. 1995. “Intergenerational Change in the Education of Canadians.” Education Quarterly Review 2. Cat. 81-003. Ottawa: Statistics Canada.
Grawe, Nathan D. 2004a. “Intergenerational Mobility for Whom? The Experience of High- and Low-Earning Sons in International Perspective.” In Generational Income Mobility in North America and Europe, edited by Miles Corak. Cambridge: Cambridge University Press.
————. 2004b. “Reconsidering the Use of Nonlinearities in Intergenerational Earnings Mobility as a Test for Credit Constraints.” Journal of Human Resources 39 (3).
Hertz, Tom, Tamara Jayasundera, Patrizio Piraino, Sibel Selcuk, Nicole Smith, and Alina Veraschagina. 2007. “The Inheritance of Educational Inequality: International Comparisons and Fifty-Year Trends.” B.E. Journal of Economic Analysis and Policy 7 (2).
Hou, Feng, and Garnett Picot. 2003. “The Rise in Low Income Rates Among Immigrants in Canada.” Analytical Studies Branch Research Paper no. 198. Ottawa: Statistics Canada.
Hum, Derek, and Wayne Simpson. 2004. “Reinterpreting the Performance of Immigrant Wages from Panel Data.” Empirical Economics 29 (1).
Institute for Competitiveness and Prosperity. 2007. “Prosperity, Inequality, and Poverty.” Working Paper 10. Toronto.
Marshall, Katherine. 2007. “The Busy Lives of Teens.” Perspectives on Labour and Income. Cat. 75-001-XIE. Ottawa: Statistics Canada.
Morgan, Stephen L., and Mark W. McKerrow. 2007. “Human Capital Development and the Frontiers of Research in the Sociology of Education.” In Perspectives on Human Capital Development, edited by Garnett Picot, Ron Saunders and Arthur Sweetman. Montreal; Kingston, ON: McGill-Queen’s University Press.
Mulligan, Casey B. 1997. Parental Priorities and Economic Inequality. Chicago: University of Chicago Press.
OECD (Organisation for Economic Co-operation and Development). 2006a. International Migration Outlook 2006. Paris: OECD.
————. 2006b. Where Immigrant Students Succeed: A Comparative Review of Performance and Engagement in PISA 2003. Paris: OECD.
OECD and UNESCO. 2003. Literacy Skills for the World of Tomorrow — Further Results from PISA 2000. Paris; Montreal: OECD and UNESCO, Institute for Statistics.
Pendakur, Krishna, and Ravi Pendakur. 2002. “Colour My World: Have Earnings Gaps for Canadian-Born Ethnic Minorities Changed over Time?” Canadian Public Policy 28 (4).
————. 2007. “Minority Earnings Disparity across the Distribution.” Canadian Public Policy 33 (1).
Picot, Garnett, and Feng Hou. 2008. “Immigrant Selection, the IT Bust, and Earnings of Entering Immigrants.” Paper presented at the meetings of the Canadian Economics Association, Vancouver, June 6-8.
Picot, Garnett, and Arthur Sweetman. 2005. “The Deteriorating Economic Welfare of Immigrants and Some Possible Causes: Update 2005.” Analytical Studies Branch Research Paper 262. Ottawa: Statistics Canada.
Reitz, Jeffrey G., and Rupa Banerjee. 2007. “Racial Inequality, Social Cohesion and Policy Issues in Canada.” In Belonging? Diversity, Recognition and Shared Citizenship in Canada, edited by Keith Banting, Thomas J. Courchene and F. Leslie Seidle. Montreal: Institute for Research on Public Policy.
Schaafsma, Joseph, and Arthur Sweetman. 2001. “Immigrant Earnings: Age at Immigration Matters.” Canadian Journal of Economics 34 (4).
Schnepf, Sylke Viola. 2007. “Inequality of Learning amongst Immigrant Children in Industrialized Countries.” Paper presented to the 29th Annual Research Conference of the Association for Public Policy and Management, Washington, DC, November 8-10.
Sen, Anindya, and Anthony Clemente. 2006. “Intergenerational Correlations in Educational Attainment and Birth Order Effects: Empirical Evidence from the 1986, 1994, and 2001 Canadian General Social Surveys.” Unpublished paper, University of Waterloo.
Soroka, Stuart, Richard Johnston, and Keith Banting. 2007. “Ties That Bind? Social Cohesion and Diversity in Canada.” In Belonging? Diversity, Recognition and Shared Citizenship in Canada, edited by Keith Banting, Thomas J. Courchene and F. Leslie Seidle. The Art of the State, vol. 3. Montreal: Institute for Research on Public Policy.
Statistics Canada. 2003. Ethnic Diversity Study: Portrait of a Multicultural Society. Cat. 89-593-XIE. Ottawa.
Sweetman, Arthur, and Gordon Dicks. 1999. “Education and Ethnicity in Canada: An Intergenerational Perspective.” Journal of Human Resources 34 (4).
Sykes, Stuart. 2008. “A Story of Reefs and Oceans: A Framework for the Analysis of the ‘New’ Second Generation in Canada.” Policy Research Initiative Discussion Paper. Ottawa: Government of Canada.
Worswick, Christopher J. 2004. “Adaptation and Inequality: Children of Immigrants in Canadian Schools.” Canadian Journal of Economics 37 (1).
Appendix – Intergenerational Analysis — Framework and Methods
This study is concerned mainly with how and to what extent the education and earnings outcomes of children are rooted in family background. The research literature on which it relies addresses such concerns with a “regression to the mean” model. This approach is used in economic analysis to measure mobility in earnings, income and other indicators of socio-economic status across generations (see, for example, Mulligan 1997; Corak 2004). Equation (1) and equivalently figure A-1 summarize this model:
Yi,t = α + βYi,t-1 + εi,t (1)
To use the example of education, in this equation the educational attainment of family i’s child would be Yi,t, which is equal to the average years of education of the cohort of children born in generation t, as represented by α, plus two factors that determine the deviation from this average: a fraction of parental education (βYi,t-1) and other influences not associated with parental income (εi,t). A similar framework can be used to analyze earnings and incomes.
Average educational attainment evolves through time, and it might be that many or all members of a generation will obtain more education than their parents. This is captured in equation (1) by the value of α. However, and just as important, the equation reflects the idea that an individual’s education is nonetheless related to his or her parents’ education. This is captured by the value of β, which represents the fraction of education advantage that, on average, is transmitted across the generations.
In other words, β summarizes in a single number the degree of generational education mobility in a society. It could conceivably be positive or negative. A positive value indicates the generational persistence of educational advantage, where higher parental education is associated with higher child education; a negative number indicates generational reversal, where higher parental education is associated with lower child education. In fact, the published research shows that this relationship is always positive, though it varies significantly across countries and with the level of development — as, for example, in the analysis of 30 countries by Hertz et al. (2007). Research also shows that, for the most part, this relationship is less than 1 in value.
This is the situation represented in figure A-1, which implies that, while there is an association between the outcomes of parents and children, parents pass on only a fraction of their relative advantage to their children. In other words, if one parent is twice as educated as another — that is, if the parent has a 100 percent advantage — this advantage will shrink to βx100 percent in the next generation. Parents with above-average education will have children who obtain above-average education, but not to the same extent. The lower the value of β, the flatter the relationship in figure A-1 and the more likely the child will be at the average for his or her cohort. In the extreme, when β is zero, parental background offers no information about the child’s adult outcomes, and children will deviate from the average of their cohort only because of luck or other unobserved influences not related to the parents’ outcome.
Intergenerational mobility in education, of course, has been a long-standing concern in both economics and sociology. Morgan and McKerrow (2007) offer a partial survey of both the Canadian and US sociological literature. Some of the most related recent Canadian work in this area includes Fournier, Butlin and Giles (1995); de Broucker and LavalleÌe (1998); and Sen and Clemente (2006). All these studies find a strong positive association between the education of parents and that of children. More recently, attention has shifted to the relationship between family background and actual literacy and numeracy outcomes for children, as opposed to formal schooling (see, for example, OECD and UNESCO 2003). Most of this research does not use the exact framework depicted in figure A-1, preferring simply to calculate raw correlations of the outcomes of parents and children. Further, none of these studies focuses explicitly on immigrants, though Sen and Clemente (2006) is closest in spirit to the methodology used in Aydemir, Chen and Corak (forthcoming, 2008).
It should also be noted that there is an extensive international literature on the degree of generational income and earnings mobility using this framework. In Corak (2006), I offer an overview of this literature and develop a set of internationally comparable estimates of the relationship between the earnings of fathers and sons. I find there is a good deal of variation across the rich countries in the degree to which paternal earnings advantage is passed on to sons, by at least a factor of two, from 20 percent or less to 40 percent or more. Further, in no country is the inherited parental advantage much lower than one-fifth. The United States, the United Kingdom and, to a slightly lesser extent, France stand out as being the least mobile societies, with 40 to 50 percent of fathers’ earnings advantage being passed on to sons. At the other extreme are Denmark, Norway, Finland and Canada, where about 15 to 20 percent of earnings advantage is passed across generations; in the intermediate position are Germany and Sweden, where about 30 percent is passed on. Again, there is little information of this sort directly related to the experience of immigrants.
It might be reasonable to expect that the degree of generational mobility of immigrants and their children differs from that of Canadian-born parents and their offspring. The framework offered by equation (1) and illustrated in figure A-1 provides an understanding of these differences in several ways.
First, there might be differences in the characteristics of immigrants and the native-born that are not related to parental education or income. The literature on immigration shows that, as a select group reflecting individual decisions to migrate and the administrative rules used by host countries, immigrants differ from the population as a whole (Aydemir 2003). Selection according to traits such as ability or motivation might make immigrant parents more efficient in fostering their children’s health, education, behaviour and outlook in a way that positively influences their adult labour market outcomes. Or it might just be that immigrant parents are more concerned about their children’s outcomes, and transfer larger amounts of time and resources to them than does the typical native-born family.
To the extent that these characteristics are not correlated with education or income, they influence the estimate of α. If immigrants are more effective or more concerned parents, then a will be higher for their children than for those of native-born Canadians. In this case, educational advantage will be passed on to children in a way that is independent of parents’ own education levels, and all immigrant children will do better than the population as a whole, regardless of parental education. Obviously, all these tendencies work in the opposite direction if immigrants are less effective or less concerned parents.
It might also be, however, that mobility across generations differs among immigrants because the relationship between adult and child outcomes differs — in other words, because β is higher or lower in value. If, for example, one correlates the characteristics of immigrants that make them different from the native-born with education or income, immigrants will have a greater tendency to pass on existing educational advantages (or disadvantages) than does the population as a whole.
It is sometimes argued that this relationship occurs because of what has been referred to as “ethnic” or “social” capital (Borjas 1992, 1993). This implies that β in equation (1) might be higher for immigrants if the average value of education in their community plays a role in determining longer-term outcomes. The nature and degree of this influence certainly might vary across immigrant communities, but the presumption in the literature appears to be that, on the whole, it is more important than for the population at large. The approach Aydemir, Chen and Corak (forthcoming, 2008) use is most closely related to that of Borjas (1992, 1993) and particularly of Card, DiNardo and Estes (2000).
Fortin and Lefebvre (1998); Corak and Heisz (1999); Corak (2001); and Grawe (2004a, 2004b) present evidence for the generational earnings mobility of the general Canadian population, while Sweetman and Dicks (1999) and Pendakur and Pendakur (2002, 2007) offer analyses by ethnicity. Hum and Simpson (2004) note that the wages of second-generation children are slightly higher than those of their native-born counterparts, but the available data do not permit them to relate this to generational mobility. Aydemir, Chen and Corak (forthcoming, 2008) are the first to study the degree of generational mobility among immigrants with Canadian data.
This publication was produced under the direction of Geneviève Bouchard and Sarah Fortin, Research Directors, IRPP. The manuscript was copy-edited by Barry Norris, proofreading was by Barbara Czarnecki, production was by Chantal Létourneau, art direction was by Schumacher Design, and printing was by AGL Graphiques.
Copyright belongs to IRPP. To order or request permission to reprint, contact:
1470 Peel Street, Suite 200
Montreal, Quebec H3A 1T1
All IRPP Choices and IRPP Policy Matters are available for download at www.irpp.org
To cite this document:
Corak, Miles. 2008. “Immigration in the Long Run: The Education and Earnings Mobility of Second- Generation Canadians.” IRPP Choices 14 (13).
About the author
Miles Corak is a professor in the Graduate School of Public and International Affairs at the University of Ottawa. He has published numerous articles on topics dealing with child poverty, access to university educa- tion, intergenerational earnings and education mobili- ty, and unemployment. He is a co-chair of the Canadian Employment Research Forum and a research fellow of the Institute for the Study of Labor (Germany), and until 2007 was the director of the Family and Labour Studies Division at Statistics Canada.
I acknowledge with gratitude very helpful com- ments from Genevieve Bouchard, Sarah Fortin and two anonymous referees. In particular, I also acknowledge the help and support of my colleagues Abdurrahman Aydemir and Wen-Hao Chen as co- authors on previous research that underpins many of the facts presented here, and that was originally developed as part of the research program of the Family and Labour Studies Division at Statistics Canada. This essay has also benefited from presenta- tions and discussions as keynote addresses to the Population Association of New Zealand at its July 2007 meeting in Wellington; to the November 2007 conference “Applications in Measuring Poverty” organized by the Maastricht Graduate School of Governance, in Maastricht, the Netherlands; and to the September 2008 conference “Inequality: Mechanisms, Effects and Policies,” organized by the Centro Ricerche Interuniversitario sullo Stato Sociale (CRISS) and the University of Rome “La Sapienza” in Rome. The responsibility for the content of this essay and any interpretations and errors it may contain remains, however, entirely mine.