Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.
Immigrants in the hinterlands
by André Bernard
Full article in PDF format
- Poor economic outcomes of immigrants
- Immigrants less apt to settle in small urban centres
- Immigrants generally earn less, but gap smaller in less urbanized areas
- Economic integration faster in smaller areas
- The most vulnerable immigrant groups integrate rapidly in small areas
- Only immigrants from the United States and Oceania integrate better economically in larger centres
- Advantages of smaller regions persist after controlling for characteristics of immigrants
- Factors in the better economic integration of immigrants in less urban areas
- Discrepancies stable between urban and rural areas
Recent immigrants have experienced more difficulty integrating into the labour market than previous cohorts in the 1970s and 1980s. Since the 1990s, immigrant cohorts have earned significantly less income during their first years in Canada than other Canadians, and earnings growth in subsequent years has not been sufficient to achieve income parity (Frenette and Morissette 2003).
The immigrant population has changed greatly over the last few decades, one of the most dramatic changes being country of origin. Immigrants are now increasingly coming more from Asia (China, India and the Philippines, in particular) than from European countries such as the United Kingdom and Italy or from the United States. As a result, the proportion of immigrants who speak a language other than English or French at home has increased sharply (Citizenship and Immigration Canada 2005a).
At the same time, immigrants with university degrees are becoming more and more common. Of the immigrants who arrived between 1996 and 2001, more than one-third had a university degree, twice the proportion of native-born Canadians (CIC 2005a). Recent immigrants are also much more likely to be 'economic' immigrants, who qualified on the basis of admissibility criteria resulting from policies specifically intended to promote their entry into Canada. Because this should normally result in improved economic outcomes for immigrants, the deterioration observed over the last few years has caused serious concern (Picot, Hou and Coulombe 2007).
One trend, which has garnered considerable attention, is the increasing concentration of immigrants in Toronto and Vancouver. The proportion settling in those two cities rose from 43% for those immigrants admitted before 1986 to 61% for those admitted between 1996 and 2001 (CIC 2005a). Even though relatively few immigrants are choosing to settle outside the large urban centres, immigration is attracting a great deal of interest from smaller communities. These communities, especially in rural areas, often face declining populations, and immigration can represent a potential means of revitalizing their economies. A more balanced geographic distribution of immigration is generally acknowledged as being desirable (CIC 2001). Some specific policies have already been put in place to attract more immigrants to rural parts of the country.1
This concentration of new immigrants settling in very large urban centres raises the question of the differences between large urban centres and the rest of the country: could economic integration difficulties simply reflect problems encountered in large urban centres?
Of course, every newcomer to the labour market, immigrant or otherwise, must overcome certain challenges, such as a lack of work experience, a mismatch between knowledge gained in school and industry requirements, and a lack of information on employment opportunities. However, immigrants face additional hurdles, including recognition of foreign qualifications, an even greater lack of information on labour market requirements and employment opportunities, and sometimes an incomplete ability to function in one of Canada's official languages. Discrimination may also also be an issue, since immigrants are increasingly likely to be members of visible minorities (Hum and Simpson 2004).
Income trends, in absolute terms, of immigrants over the years provides an incomplete view of economic integration. High incomes do not necessarily mean rapid integration if incomes of other Canadians are even greater. Relative measures are more meaningful.
Economic integration can be measured through two components: the initial income gap between immigrants and Canadians in general, and the rate at which that gap narrows. Given all of the factors, incomes of immigrants can be expected to be lower in the first years after arrival. However, rapid economic integration would result in the rapid closing of the gap and its elimination within a few years. Because economic integration is a complex process that includes components other than income, like labour force participation, this measure of economic integration is not the only one that could be examined.
The distribution of immigrants based on five areas defined for this study (see Data sources and definitions) is very uneven and does not reflect the distribution of the Canadian population. While approximately 34% of Canadians 20 years of age or older live in one of the three largest urban centres (Toronto, Montréal and Vancouver), approximately 75% of immigrants make these cities their homes. Conversely, while slightly more than 1 in 5 Canadians live in a small town or rural area with a population under 15,000, the corresponding proportion of immigrants is less than 1 in 40 (Chart A).
Immigrants in large cities and those in small towns are not all that different (Table 1). Immigrants in the smallest areas are slightly less likely than other Canadians to hold a university degree. However, they are more likely to have pursued postsecondary studies without obtaining a university degree. Immigrants in small areas are also less likely to be refugees, but the proportions of skilled worker economic class immigrants and family class immigrants are similar in all types of areas. Given that knowledge of an official language is more critical in small areas (because of less linguistic diversity), it is interesting that 1 in 4 immigrants living in a small town or rural area did not have official-language knowledge upon settling there (compared with almost 2 in 5 in very large urban areas).
The most striking differences between areas involve country of origin. Immigrants living in small areas come mostly from Europe and the United States, while immigrants in large urban centres come mostly from Asia. Nevertheless, more than 1 in 4 immigrants in the smallest areas come from Asia and the proportions of immigrants from Africa in very large urban areas and small urban areas are similar.
For Canadians in general, living in a large metropolitan area means a higher income. Median incomes of Canadians in very large urban areas and large urban areas were $28,100 and $30,500, respectively, compared with $22,500 in small towns and rural areas (Chart B), a significant difference.
For immigrants, the pattern is reversed. Incomes of immigrants were lowest in very large urban areas (median $16,800) and highest in small urban areas (median $19,500), a difference of 16%. Incomes of immigrants in small towns and rural areas (median $18,800) were also significantly greater (by 12%) than those of immigrants in very large urban areas.
While immigrants have lower incomes in all types of areas, the gap narrows along the gradient from urban to rural. In very large urban areas, the median income gap is very large, at 67%. In small urban areas, the gap falls to 32%, while in small towns and rural areas the gap is only 20%.
Economic integration can be examined by starting with the initial income gap between immigrants and Canadians and then measuring the subsequent rate of convergence or equalization over time.
Integration of immigrants in small, less urbanized areas is more rapid and that advantage increases over time. In very large urban areas, the initial income gap is 37%. It gradually decreases, but rather slowly. After four years, the gap is still 22%, falling below the 10-percent threshold in the twelfth year (Chart C). In contrast, in small urban areas, the initial gap is only 14%, and in the fourth year immigrants are earning 2% more than Canadians. The relative advantage of immigrants continues to increase over time, reaching a peak of 18% following the eleventh year.
In small towns and rural areas, the advantage of immigrants is even more pronounced. In their first year of permanent residence, their average income is 4% higher than that of Canadians. In the thirteenth year, the relative income advantage of immigrants rises to 19%.
Immigrants in the smallest areas, while they have diverse characteristics, are more likely to have prior official-language ability and are less likely to have at most a high school education or to be refugees. Immigrant groups with no more than a high school education and groups with no official-language ability, as well as refugees, are examined in greater detail. Analysis of refugees is especially important since they land in Canada under completely different circumstances from that of qualified economic immigrants.
For each group, economic integration is significantly more rapid in smaller areas than in large urban centres. Immigrants with no more than a high school education earn incomes that are 46% lower in very large urban areas, compared with 23% lower in small towns and rural areas (Chart D). The gap closes very slowly in large cities—after 13 years, the gap is still 20%. However, in small towns and rural areas, the gap closes quite quickly, so that as of the fifth year the gap in most years is significantly less than 10%.2
The pattern is similar for immigrants without prior ability in one of Canada's two official languages. The initial gap is smaller in small towns and rural areas (31%) than in other areas, especially very large urban areas (50%), and the subsequent increase in the relative income of immigrants is also much faster (Chart E).
Refugees, though they represent only 5% of immigrants in small towns and rural areas, integrate very rapidly—so rapidly that, after only one year, their incomes are 10% greater than that of Canadians living in the same type of area (Chart F). By contrast, refugees in very large urban areas earn 43% less and, after 13 years of residence, the gap is barely under 20%. In other areas, refugees generally earn lower incomes. However, in smaller areas, the gap is not as wide.
Only immigrants from the United States (and to a lesser degree from Oceania) integrate more quickly in economic terms in larger centres than in smaller ones (data not shown). All other immigrants, especially those from Asia, show a smaller initial discrepancy and subsequent relatively larger increase in income in smaller cities.
Even after taking into consideration the different characteristics specific to immigrants, as well as other observable characteristics common to Canadians, economic integration is much faster outside the major urban centres (Table 2; see also Linear regression).
Economic class immigrants have difficulty integrating in the major urban centres, regardless of their education, their ability in an official language or their country of origin. For almost every group of immigrant considered, parity had still not been achieved even after 13 years, the maximum observable with the data. In fact, only those with a university degree, ability in an official language and from a region other than Africa and Asia eventually manage to achieve parity—and even then, after seven years.
In contrast, in a small urban or rural area, these same immigrants generally manage to integrate quite rapidly, especially when they have a university degree upon establishment. In fact, every group of immigrants with a degree achieves parity within at most four years, and some achieve it within the first year. Nonetheless, in many cases economic integration is better in smaller regions even for immigrants with at most a high school diploma upon establishment.
For refugees, the contrast between the larger urban centres and the smaller urban and rural areas is even more striking. In the larger urban centres, none achieved parity within 13 years.
Refugees in the small urban areas, smaller cities and rural areas integrate well from the economic standpoint, particularly those arriving with a university degree. For most groups, refugees in the smaller cities and rural areas achieve income parity very quickly. Those with a university degree achieve it within the first year, regardless of their country of origin or their prior ability in an official language. Refugees with at most a high school diploma do better than those living in the larger urban centres.
It is difficult to clearly identify the factors accounting for immigrants' better economic integration in smaller urban areas. Because not many characteristics of individuals are available in the database, it is likely that many of the differences identified are only tied to other unobservable factors specific to immigrants, not to the regions.
Nonetheless, some hypotheses merit consideration. The difficulty associated with the recognition of education obtained abroad is well known, and the lack of information about labour market requirements and job opportunities, and the sometimes imperfect ability in one of the official languages, are examples of factors that can slow the economic integration of immigrants.
With regard to education, the impact of university degrees earned abroad on relative incomes is greater in less urbanized regions. It is difficult to determine the extent to which this is because immigrants living in the smaller areas are better able to translate their education acquired abroad into income and/or because a smaller proportion of people with university degrees live in these areas. Based on the 2001 Census, the proportion of university graduates in the adult population aged 25 to 64 years is 30% in the largest urban centres (Montréal, Toronto and Vancouver) and 16% in the areas with fewer than 100,000 inhabitants. Among new immigrants, the differences according to education upon arrival are much less pronounced (Table 1). Immigrants with university degrees are particularly well represented in the small areas (Chart D), and having pursued postsecondary (not only university) studies abroad greatly improves the advantage of immigrants in the smaller areas. However, even less well-educated immigrants post better results in terms of economic integration in smaller urban areas and the smaller cities and rural areas.
The need for information about labour market requirements and job opportunities suggests that the creation of a network—formal or informal—with non-immigrants would likely be inevitable in smaller areas, precisely because of the smaller proportion of immigrants there. In return, this network may be critical to economic integration, even if the small proportion of immigrants may be a source of other kinds of disadvantages. This does not mean that immigrants living in smaller regions will not face the same difficulties inherent to the local labour market as any of their neighbours. Rather, they will be less likely to be at a disadvantage than immigrants in the major urban centres merely because they are immigrants.
Lack of ability in an official language is not as great a handicap outside the major centres. In the largest urban centres, none of the groups of immigrants without ability in an official language managed to achieve income parity after 13 years. In smaller cities and rural areas, several groups, in particular refugees, managed to achieve it. One could conclude that these immigrants are more likely to learn one of the official languages quickly if they live in an area with a high proportion of French- or English-speakers. This enables them to overcome this barrier more rapidly than in the larger urban centres.
To a large extent, the data also rule out at least one other possible hypothesis. Even though immigrants living in smaller cities and rural areas are more likely to come from Europe and the United States, this does not explain why they do better than immigrants in the major urban centres. The raw regressions suggest that in smaller cities and rural areas, the impact of country of origin on income advantage is very small, and does not necessarily favour immigrants from Europe, the United States or Oceania. Also, immigrants from the United States are the only ones to have integrated more rapidly from an economic standpoint in the larger urban centres. In other words, it is very likely that the discrepancies identified would be even larger if the distribution by country of origin in the smaller cities and rural areas were closer to that in the larger urban centres.
Naturally, several factors could affect immigrants' ability to integrate. These include, in particular, their formal or informal reception by government and community, any discrimination they may face, and their motivation to integrate into the labour market. None of these can me measured from the data.
The three cohorts of immigrants studied show a surprising stability in the differences between urban and rural areas over time. From 1994 to 1996, the relative incomes of immigrants who arrived in 1992 and 1993 were 32% higher in smaller cities and rural areas than in the largest urban centres. For those who arrived in 2001 and 2002, this difference was only slightly smaller, 27% for the years 2003, 2004 and 2005. The comparative advantage of the largest urban centres edged up from 24% to 25% (data not shown).
The economic well-being of immigrants is critical for a country like Canada, which relies heavily on immigration for demographic growth. Where immigrants choose to settle appears to affect their economic integration. It is much faster outside the largest urban centres, which is where most of them settle. In contrast, the incomes of those who choose to settle outside these major centres are similar to those of other Canadians. This initial disadvantage of immigrants, when it exists, generally disappears after a few years.
In contrast, in the largest urban centres, immigrants face a large initial income disadvantage, and subsequent increases are not enough for them to achieve parity. Better economic integration of immigrants outside the largest urban centres is evident even after taking into consideration differences in terms of immigrants' education upon arrival, prior ability in an official language, admission class and country of origin.
These results put the large income differences between recent immigrants and other Canadians, identified in previous studies, into perspective. These differences appear, at least in large part, to result from a dynamic exclusive to the largest urban centres.
Immigrants living outside the largest urban centres can translate their credentials acquired abroad into a relative income advantage more easily. They are more likely to overcome their lack of ability in an official language, quickly learning English or French, enabling them to increase their ability to generate income faster.
The Longitudinal Administrative Databank (LAD) provides a 20% sample of the T1 Family File (T1FF), containing cross-sectional annual data on all Canadian tax filers and their family members. Census family formation is done using information provided to the Canada Revenue Agency each year through individual tax returns and Canada Child Tax Benefit applications. LAD also contains data from the Longitudinal Immigration Database on characteristics of immigrants at the time of landing.
The sample was restricted to individuals 20 years of age or over.
Before-tax income comprises employment income (74% in 2005), other market income, like investment income (14%), and government transfers (12%). All figures are in constant 2005 dollars. The sample includes only individuals whose income exceeds $1,000.
For this study, an immigrant is any person who obtained permanent residence in Canada between 1992 and 2003.
Years since establishment are calculated from the date on which permanent residence was obtained (which may differ from an immigrant's date of arrival in Canada). Only whole years are counted, so that income in the year of establishment, during a portion of which an immigrant was not a permanent resident, is omitted.
Very large urban areas are the census metropolitan areas (CMAs) of Montréal, Toronto and Vancouver.
Large urban areas are other CMAs with populations exceeding 500,000-Québec, Ottawa-Gatineau, Hamilton, Winnipeg, Calgary and Edmonton.
Mid-sized urban areas are the 20 CMAs with 100,000 to 500,000 residents.
Small urban areas are census agglomerations with 15,000 to 100,000 residents.
Small towns and rural areas comprise all other locations.
Income gaps between immigrants and the population as a whole, by year, as of the year of landing, are adjusted for age. Incomes of immigrants are compared with the median income of the general population for the same type of geographical area and for the same age group (13 defined age groups). This adjustment is required because years since establishment are correlated with age, and age is correlated with income.
Regression models are used to measure how certain key factors account for a phenomenon after controlling for other observable characteristics. This study used an ordinary least squares linear regression model:
Ln(yitr/Yr) = α + β1X' + β2IMMIGRANT + β3(YEARS)' + β4(ORIGIN)' + β5(EDUCATION)' + β6(CLASS)' + β7(LANGUAGE)' + εitr
The dependent variable is a measure of the individual's income advantage. This is the ratio of, on the one hand, individual i's income in year t in region r and, on the other hand, the median income (Y) of the entire population in region r (median income in constant dollars, all years combined).
A ratio of one indicates parity between the income of an individual and that of his/her neighbours; a ratio higher (lower) than one indicates a relative advantage (disadvantage) in terms on income. To facilitate the calculations, we used the logarithm of the ratio as the dependent variable for the regression. This way, the explanatory variable coefficients could be added and interpreted as the percentage impact on the ratio, or, in other words, the impact in percentage terms on the income advantage. The construction of this dependent variable is similar to that of Li (2003). Only individuals whose incomes are greater than $1,000 are included in the regression models (as in the descriptive tables), in order to exclude those who are not in the labour market or are dependents.
The regression considers every individual, not only immigrants. Thus, there are two types of explanatory variables. The control variables that are common to immigrants and to all other Canadians are included in vector X. Unfortunately, the administrative data used only contain a limited number of variables on the characteristics of individuals. Nonetheless, the individual's province of residence, type of family, age group and sex are included in the model. Dichotomic variables for every year from 1992 to 2005 are also included in order to account for the impact of the business cycle.
The other explanatory variables included in the model are strictly for immigrants. First, there is a dichotomic variable identifying immigrants as such. This variable identifies an initial impact of immigration in terms of income advantage (or disadvantage). Next, the dichotomic variables are included, representing each of the years following their arrival (starting with the second year). These variables are included in the YEARS vector. Because the data cover years 1992 to 2005, immigrants can be followed for up to 13 years after their arrival (for immigrants who arrived in 1992). The coefficient for the variable identifying immigrants provides an indication of the 'income discrepancy upon establishment' (a coefficient that is assumed to be negative), while the coefficients associated with the different years since establishment provide an indication of the 'catch-up speed' (coefficients assumed to be positive) in the incomes of immigrants with those of all Canadians living in the same type of area.
Other control variables specific to immigrants are added to take the different characteristics among immigrants living in the major urban centres and those living elsewhere in the country into consideration. This variable identifies the admission class of the immigrant (economic class, family class, refugee and other), prior knowledge of an official language, level of education at the time of arrival, and the immigrant's country of origin. All of these variables specific to immigrants (including the number of years since establishment) are multiplied by the indicator (using a value of 0 or 1) identifying immigrants, which is zero for all other Canadians.
Income parity between immigrants in a certain class and all Canadians in the same type of area is considered achieved after a certain number of years when the coefficient associated with the status of immigrant, added to the coefficient associated with the number of years since arrival, is equal to or greater than zero, which means that the catch-up after arrival was enough to make up for the initial unfavourable income discrepancies. To do this calculation for every group of immigrants, the coefficients associated with the different targeted characteristics—which is to say the coefficients associated with the CLASS, LANGUAGE, EDUCATION and ORIGIN variables—have to be added up.
Three more regressions are done to determine whether the differences between urban and rural areas are accentuated or reduced during the period under study. For the most part, the form of these regressions is comparable. However, the data on all types of regions are grouped and the variables identifying the regions are included in the model, while the variables identifying the number of years since arrival are omitted. Thus, the coefficients associated with the different types of regions represent the average of the income advantages associated with the types of regions. The three regressions help compare the change in results for three cohorts of immigrants: those who arrived in 1992 and 1993, those who arrived in 1997 and 1998 and, finally, those who arrived in 2001 and 2002.
- For example, the federal government recently announced new measures to attract French-speaking immigrants to rural parts of Prince Edward Island (CIC 2007). The 2005 Annual Report to Parliament on Immigration (CIC 2005b) has already recognized the potential of the Provincial Nominee Program for "supporting the regionalization of immigrants to centres outside Canada's three largest cities" (page 18).
- The variable used measures education upon landing only. Further education, not observed here, is probably an important contributor to these patterns.
- Citizenship and Immigration Canada. 2001. Towards a more balanced geographic distribution of immigrants. Special Study. May. Strategic Research and Review Unit. 65 p.
- 2005a. Citizenship and Immigration Canada. 2005. Recent Immigrants in Metropolitan Areas: Canada-A Comparative Profile based on the 2001 Census. April. Catalogue no. MP22-20/1-2005E. 105 p.
- 2005b. Citizenship and Immigration Canada. 2005. Annual Report to Parliament on Immigration, 2005. Catalogue no. Ci1-2005. 54 p.
- Citizenship and Immigration Canada. 2007. "Canada's new government to help attract Francophone imimigrants to rural Prince Edward Island." October 12, 2007, News Release. (accessed January 17, 2008).
- Frenette, Marc and René Morissette. 2003. "Will they ever converge? Earnings of immigrants and Canadian-born workers over the last two decades." (PDF) Statistics Canada Catalogue no. 11F0019MIE. Analytical Studies Branch Research Paper Series, no. 215. 20 p. (accessed January 17, 2008).
- Hum, Derek and Wayne Simpson. 2004. "Economic Integration of Immigrants to Canada: A Short Survey." Canadian Journal of Urban Research. June. Vol. 13. Institute of Urban Studies, University of Winnipeg.
- Li, Peter S. 2003. "Initial Earnings and Catch-Up Capacity of Immigrants." Canadian Public Policy. Department of Sociology, University of Winnipeg. Vol. 29, no. 3, p. 319-337.
- Picot, Garnett, Feng Hou and Simon Coulombe. 2007. "Chronic Low Income and Low-income Dynamics Among Recent Immigrants." (PDF) Statistics Canada Catalogue no. 11F0019MIE. Analytical Studies Branch Research Paper Series, no. 294. 48 p. (accessed January 17, 1008).
André Bernard works in the Labour and Household Surveys Analysis Division. He can be reached at 613-951-4660 or email@example.com.
You need to use the free Adobe Reader to view PDF documents. To view (open) these files, simply click on the link. To download (save) them, right-click on the link. Note that if you are using Internet Explorer or AOL, PDF documents sometimes do not open properly. See Troubleshooting PDFs. PDF documents may not be accessible by some devices. For more information, visit the Adobe website or contact us for assistance.
- Date modified: