Analytical Studies Branch Research Paper Series
Economic Immigrants in Gateway Cities: Factors Involved in Their Initial Location and Onward Migration Decisions

Acknowledgement

This research was funded by the Research and Evaluation Branch at Immigration, Refugees and Citizenship Canada (IRCC). Excellent comments were received from Ying Gai, Sébastien Vachon and Xiaoyi Yan at IRCC.

Abstract

Immigrants tend to reside disproportionately in larger Canadian cities, which may challenge their absorptive capacity. This study uses the linked Longitudinal Immigration Database and T1 Family File to examine the initial location and onward migration decisions of immigrants who are economic principal applicants (EPAs) and who have landed since the Immigration and Refugee Protection Act was passed. The main objective of the study is to identify the factors associated with initially residing and remaining in Canada’s three largest gateway cities: Montréal, Toronto and Vancouver (referred to as MTV). The findings suggest that many factors are strongly associated with the probability of initially settling in MTV. Prominent among these is immigration category, followed closely by country of birth, intended occupation, knowledge of official languages and pre-landing work experience in MTV. After the initial decision is made, the geographic location of economic immigrants does not change much. The initial location decision of EPAs is a very strong predictor of their location years later. Indeed, only about 11% of EPAs have moved to or out of MTV by 10 years after landing. That being said, some important factors are associated with the onward migration that does take place. Key among these is country of birth, especially for EPAs who initially settled outside MTV. The number of residents in the city of initial location who come from the same country is positively associated with remaining in MTV and negatively associated with moving to MTV for those residing elsewhere. Notably, immigration category plays less of a role in explaining the onward migration decision, compared with the initial location decision.  

Keywords: immigrants, geographic mobility

Executive summary

Immigrants tend to reside disproportionately in larger Canadian cities, such as Montréal, Toronto and Vancouver (MTV). Although the share of immigrants settling in MTV has been declining, the influx remains large. This may challenge the absorptive capacity of these gateway cities in at least two ways: by increasing the competition for jobs with similarly skilled Canadian-born workers and by increasing the demand for social services (particularly immigration settlement services, including housing).

The main objective of the study is to identify the factors associated with initially residing and remaining in MTV. Specifically, the study uses the linked Longitudinal Immigration Database (IMDB) and T1 Family File (T1FF) to examine the initial location and onward migration decisions of immigrants who are economic principal applicants (EPAs) and who have landed since the Immigration and Refugee Protection Act (IRPA) was passed. The study results are divided into two types: the factors associated with the probability of initially residing in MTV; and, conditional on the initial location decision, the factors associated with remaining in or moving to MTV several years later. The first part of the analysis focuses on immigrants who landed in 2004, 2009 or 2014, whose initial location decisions are captured in the first full tax year following landing. In the second part, the 2004 and 2009 landing cohorts are separated by their initial location decision (in MTV or not, one year after landing), and their subsequent location decisions five years after landing are examined. For the 2004 cohort, location decisions 10 years after landing are also examined. In all cases, linear probability models are estimated, where the dependent variable is a binary variable indicating residence in MTV.

The findings suggest that many factors are strongly associated with the probability of initially settling in MTV. Prominent among these is immigration category, followed closely by country of birth, intended occupation, knowledge of official languages and pre-landing work experience in MTV.

After the initial decision is made, the geographic location of economic immigrants does not change much. The initial location decision of EPAs is a very strong predictor of their location years later. Indeed, only about 11% of EPAs have moved to or out of MTV by 10 years after landing. That being said, some important factors are associated with the onward migration that does take place. Key among these is country of birth, especially for EPAs who initially settled outside MTV. The number of residents in the city of initial location who come from the same country is positively associated with remaining in MTV and negatively associated with moving to MTV for those residing elsewhere. Notably, immigration category plays less of a role in explaining the onward migration decision, compared with the initial location decision.

1 Introduction

Canadian society is aging, partly because of increased life expectancy, as well as a relatively low fertility rate. One of the natural consequences of an aging society is increased economic pressures on the existing labour force to provide for the elderly through social programs. As a countermeasure, Canada has adopted an active immigration policy through the IRPA. The country has accepted more than 250,000 new entrants per year over the last decade and is expected to accept 310,000 immigrants in 2018, and 340,000 in 2020.^{Note 1} Immigration under the IRPA has three main objectives: benefitting the economy, reuniting families and meeting humanitarian or compassionate ideals. Overall, the approach is still largely geared towards the economic prosperity of Canada, as 60.6% of immigrants who landed in 2015 were classified as economic.^{Note 2}

However, many studies over the last two decades or so have highlighted the relatively poor and declining labour market outcomes of immigrants (Picot and Sweetman 2012).

Some of the key reasons cited include changing source countries, declining returns to foreign experience, the information technology sector crash in the early 2000s, poor language skills and a lower rate of return associated with foreign credentials.   

Another potential factor is the tendency of immigrants to settle in larger Canadian cities, such as MTV and others. Although the share of immigrants settling in larger cities has declined in recent years (Bonikowska, Hou and Picot 2015), the influx remains large. This may challenge the absorptive capacity of these larger centres in at least two ways: by increasing the competition for jobs with similarly skilled Canadian-born workers and by increasing the demand for social services (particularly immigration settlement services).

As a result, immigration policy makers have already begun to address the issue through the Provincial Nominee Program (PNP), which is designed to encourage a more even dispersion of economic immigrants throughout Canada. The Evaluation of the Provincial Nominee Program by Immigration, Refugees and Citizenship Canada concluded that it is generally effective at retaining program participants, although retention rates vary considerably across the country (lowest in Atlantic Canada, highest in Alberta and British Columbia).^{Note 3} Pandey and Townsend (2011) also found that PNPs are generally effective at attracting and retaining economic immigrants.

More broadly, the issue of immigrant location (both initial and onward migration) remains understudied. Ostrovsky, Hou and Picot (2011) found that recent immigrants responded to the economic boom in Alberta to a greater extent than Canadian-born individuals. Bonikowska, Hou and Picot (2015) found that changes in immigration policy (particularly the PNP) and shifts in immigrant source regions were both associated with changes in the initial destination of immigrants. Hou (2007) looked at the size of the pre-existing immigrant community as a factor in the initial destination and found no relationship.

This study contributes to the literature by comprehensively examining the initial location and onward migration decisions of EPA immigrants who have landed since the IRPA was enacted in 2003, using the linked IMDB and T1FF. The study focuses on identifying the key factors associated with initially settling in the gateway cities (namely, MTV) and with remaining in, or moving to, these cities. Several potential factors are available in the IMDB, while the T1FF allows for immigrants to be followed up geographically for several years after entry.

The next section describes the methodology used in the study. This is followed by a description of the detailed results. The study concludes with a summary of the main results.

2 Data and methods

The study results are divided into two types: the factors associated with the probability of initially residing in a gateway city (MTV);^{Note 4} and, conditional on the initial location decision, the factors associated with remaining in or moving to MTV several years later. The first part of the analysis focuses on immigrants who landed in 2004, 2009 or 2014, whose initial location decisions are captured in the first full tax year following landing. In the second part, the 2004 and 2009 landing cohorts are separated by their initial location decision (in MTV or not, one year after landing), and their subsequent location decisions five years after landing are examined. For the 2004 cohort, location decisions 10 years after landing are also examined. In all cases, linear probability models are estimated, where the dependent variable is a binary variable indicating residence in MTV.^{Note 5}

The analytical sample includes all EPAs who landed in the target years. Economic class immigrants include those who fall under the following categories: skilled worker, business, provincial nominee, live-in caregiver and Canadian experience (introduced in 2009). The principal applicant is simply the main applicant in the family; this excludes accompanying spouses and dependent children.^{Note 6}

The data come from the IMDB and the T1FF, which have been linked from 1982 to 2015. The IMDB collects information from all immigrants who landed in Canada (i.e., became permanent residents or citizens). The T1FF contains tax-relevant information on individuals and their families for all taxfilers and their non-filing dependants (including spouses, children and adult children who are disabled).

This study focuses exclusively on the period since the IRPA was enacted (in 2003), as this is the existing legislation governing immigration policy, and earlier results may no longer apply.^{Note 7}

The IMDB–T1FF data contain several important variables for the analysis. The dependent variable (residing in MTV or not) is captured in the T1FF through the family postal code, which is recoded to a census metropolitan area (CMA).^{Note 8} The CMA is then used to create the MTV indicator.^{Note 9} With only a few exceptions, the explanatory variables come from the IMDB and are measured at landing. They include the following:

• country of birth
• immigration category (the five economic classes described above)
• intended occupation (derived from the first digit of the National Occupational Classification)
• highest level of education of the principal applicant (no postsecondary qualification, trades certificate, other non-university postsecondary diploma, bachelor’s degree, master’s degree or doctorate)
• self-declared (untested) knowledge of the official languages (English only, French only, English and French, or neither English nor French)
• sex
• age
• marital status of the principal applicant (married or common law, or single—separated, divorced, widowed or never married)
• number of dependants
• number of years of employment^{Note 10} in MTV during the five years before landing
• number of years of postsecondary enrolment^{Note 11} in MTV during the five years before landing.^{Note 12}

For the onward migration models, two additional covariates are included. These are linear measures of family income and the number of residents in the same city who were born in the same country (both created from tax data and measured in the first full tax year after landing).^{Note 13} Both of these variables are divided by 10,000, since the coefficient estimates per unit were very small.^{Note 14} Family income is measured after taxes and adjusted by dividing by the square root of the number of family members. This approach expresses income in “adult equivalents,” which produces a per capita measure while still accounting for economies of scale in household production associated with larger families. Income is expressed in 2015 constant dollars in the models.

3 Results

3.1 Initial location

Table 1 shows the frequency distribution of the variables used in the initial location analysis. More than half of EPAs initially resided in MTV, regardless of the landing cohort; however, the proportion declined consistently over the years (from 72.1% of the 2004 cohort to 51.2% of the 2014 cohort).^{Note 15} The top 20 countries of birth are shown in the table (in descending order of frequency among the 2014 landing cohort). They collectively account for 70% to 80% of landed EPAs throughout the study period. Among EPAs in the 2014 cohort, more than one-fifth (21.3%) came from India (up from 12.0% of the 2004 cohort), followed by 15.4% from the Philippines (up from 9.4% of the 2004 cohort), 8.9% from China (down from 15.8% of the 2004 cohort), and 7.9% from Iran. There is a big decline after Iran, with EPAs from France (the next country in terms of size) comprising 3.1% of the total.

In the 2004 landing cohort, most EPAs were skilled workers (89.4%). This was still the top category among members of the 2014 cohort, but to a far lesser extent (only 39.2%). An increasing proportion of EPAs have entered as provincial nominees or in the Canadian Experience Class (a new category introduced in 2009) in recent years. In the 2014 cohort, very few EPAs entered in the business or live-in caregiver classes.

For the most part, the distribution of intended occupations of EPAs has remained fairly consistent over time. However, there are two exceptions. The proportion intending to work in natural and applied sciences and related occupations declined considerably (from 40.5% to 24.1% over the period), while the proportion intending to work in sales and service occupations tripled (from 5.9% to 17.4% over the period). Nevertheless, the proportions remained fairly stable, for the most part. Overall, most EPAs intended to work in occupations typically associated with university qualifications.

Indeed, at least two-thirds of EPAs held a university degree at landing (regardless of the cohort). However, the proportion declined from 79.9% of the 2004 cohort to 71.2% of the 2014 cohort. This decline was counterbalanced by increases in the proportion who held a trades certificate (from 3.7% to 7.0%) or who had no postsecondary qualification (from 5.6% to 11.9%).      

Although most EPAs knew English or French (92.7% of the 2014 landing cohort), the depth of knowledge is unknown. Almost two-thirds of EPAs were male. EPAs were also young—60.5% of the 2014 cohort were younger than 35 at landing (up from 51.8% of the 2004 cohort). In addition, members of more recent cohorts were less likely to be married or in a common-law relationship, and they had fewer dependants.

While the majority of EPAs did not work in MTV in the five years before landing, this has become less evident over time (73.0% of the 2014 cohort, down from 89.5% of the 2004 cohort). Similarly, the majority of EPAs did not enrol in a Canadian postsecondary institution in MTV in the five years before landing, but this has also become less evident among more recent cohorts (88.5% of the 2014 cohort, down from 95.8% of the 2004 cohort).

Table 2 shows the linear probability model results of the initial location decision. Since the number of observations is high, most of the coefficients are statistically significant. Consequently, the size of the coefficients is more important than their statistical significance, in most cases. The coefficients are interpreted as the percentage point difference in the probability of residing in a gateway city one full tax year after landing, compared with the reference category.

The results suggest that the probability of initially settling in MTV varies considerably across several immigrant characteristics. With regard to source country, EPAs from France, Nigeria, the Philippines, the United States and the United Kingdom were the least likely to initially reside in MTV. However, while this has increasingly been the case for EPAs from France, the Philippines and Nigeria over the study period, it has become less so for those from the United States and the United Kingdom. In contrast, EPAs from countries such as Algeria, Iran, Haiti, China and Morocco were the most likely to settle in MTV—a pattern observed across the study period.   

The immigration category is even more strongly correlated with the probability of settling in MTV. Provincial nominees were far less likely to do so than skilled workers (the reference category). The difference is very large, at 33.4 percentage points for the 2014 landing cohort. EPAs in the Canadian Experience Class and live-in caregivers were also less likely to settle in MTV than skilled workers (a difference of about 25 percentage points for each). On the opposite end of the spectrum, business class immigrants were more likely than skilled workers to reside in MTV (a difference of 18 percentage points).

Not surprisingly, EPAs intending to work in natural resources, agriculture and related production occupations were the least likely group to settle in MTV (13.7 percentage points less than those intending to work in management occupations, the reference category). Most other groups were either slightly less likely than the reference category, or about as likely, to initially reside in MTV.

The highest level of education is not strongly correlated with location decisions. EPAs with a trades certificate or a doctorate were the least likely to settle in MTV. Tradespeople may be required in all parts of the country (particularly Alberta, Saskatchewan, and Newfoundland and Labrador, all of which experienced resource-based economic booms during the period). Doctoral degree holders may be more likely to work in higher education, which employs workers in many different parts of the country.

The knowledge of official languages is more strongly correlated with location decisions. EPAs with knowledge of English only were the least likely to settle in MTV, while those with knowledge of French (either alone or with English) were the most likely.

Age, sex, marital status and the number of dependants are not particularly strong correlates of location decisions. Women were more likely to reside in MTV than men, but the difference is only 1 to 3 percentage points, depending on the landing cohort. The differences by age group, marital status and number of dependants are even smaller and generally not statistically significant.

Having worked in MTV for one additional tax year before landing is associated with a significant increase in the probability of landing in MTV (14 to 18 percentage points, depending on the cohort). Having worked for two additional years in MTV is as strongly associated with initially residing in MTV as being a provincial nominee (compared with the skilled worker reference category). In contrast, enrolment in a postsecondary institution in MTV before landing is weakly correlated with settling in MTV (and negatively so). This is perhaps not surprising, as postsecondary institutions present only a short-term geographic commitment to students, as opposed to work, which may offer a long-term opportunity.

3.2 Onward migration

The section presenting results for onward migration is subdivided according to the EPAs’ initial location decision (MTV or not). In both instances, the outcome of interest is an indicator of living in MTV years later.

3.2.1 Initially landed in Montréal, Toronto or Vancouver

Among EPAs from the 2004 and 2009 landing cohorts who initially settled in MTV, 93% remained in MTV five years later. For the 2004 cohort, 89% remained in MTV 10 years later. Thus, the initial location decision is a strong predictor of later geographic decisions.

Despite the low transition rate out of MTV, some factors are more strongly associated with migration than others. This is most evident across the country of birth dimension (Table 3). EPAs from Nigeria, the United Kingdom, Pakistan, India and Bangladesh were generally the least likely to stay in MTV 5 and 10 years after landing. (Results from Table 2 had indicated that EPAs from Nigeria and the United Kingdom were among the least likely to initially reside in MTV.) In contrast, EPAs from Algeria, Romania, Ukraine and Morocco were the most likely to remain in MTV.

Very few of the other factors examined are strongly correlated with onward migration. These include the immigration category, which was closely associated with the initial location decision, and the intended occupation (also associated with initial location). Women and married or common-law EPAs were slightly more likely to stay in MTV than men and single individuals, respectively. The same can be said for EPAs with more dependants, with a higher family income, who were surrounded by more immigrants from the same country in the initial location (CMA) of choice, or who worked more in MTV before landing. However, in all of these cases, the associations are generally weak, at least compared with the country of birth. For example, a $10,000 increase in adjusted family after-tax income is associated with a 0.1 to 0.3 percentage point increase in the probability of remaining in MTV 5 or 10 years after landing. The number of residents from the same country is somewhat more strongly correlated with the probability of staying in MTV. Specifically, an increase of 10,000 in the number of immigrants from the same country is associated with a 1.7 to 3.1 percentage point increase in the probability of remaining in MTV. The number of years of postsecondary enrolment in MTV before landing is negatively associated with remaining in MTV, although the relationship is very weak and statistically significant in only one instance.

3.2.2 Initially landed outside Montréal, Toronto or Vancouver

There is also very little onward migration among EPAs who initially resided outside MTV. About 9% of EPAs from the 2004 landing cohort who initially settled outside MTV had moved to MTV by five years after landing. That number rose to 11% by 10 years after landing. Among their counterparts from the 2009 landing cohort, 6% had moved to MTV by five years after landing.

Many of the factors examined in this study are more strongly associated with the onward migration decision of EPAs who settled outside MTV (Table 4) than with the onward migration decision of those who settled in MTV (Table 3). This is especially true with the country of birth. EPAs who settled outside MTV and came from Iran, China, South Korea or India were often significantly more likely than their counterparts from the “rest of world” category (the reference category) to have moved to MTV, by both 5 and 10 years after landing. In contrast, EPAs from Romania were less likely to move to MTV years later.

A second important factor is the number of residents in the CMA who come from the same country. For every increase of 10,000 in this figure, the estimated probability of moving to MTV declines by 3.5 to 5.6 percentage points (5 years after landing) and by 5.0 percentage points (10 years after landing).

Another important factor is knowing English and French: EPAs in this group were more likely to have moved to MTV by 5 and 10 years after landing.

Women, older individuals and individuals who are married or in a common-law relationship were less likely to move to MTV after initially residing outside MTV. The same can be said for individuals with more years of employment outside MTV before landing, as well as for those who initially settled in a city with more residents from the same country of birth.

4 Conclusion

A large portion of immigrants who land in Canada initially settle in larger urban areas, including Montréal, Toronto and Vancouver (MTV). This may challenge the absorptive capacity of the host cities since demand for social services increases, and the new arrivals do not always successfully integrate into the Canadian labour market. Immigration policy makers have already begun to address the issue through the Provincial Nominee Program, which is designed to encourage the dispersion of new economic immigrants throughout the country.

To better inform immigration policy on the issue, this study comprehensively examines the factors associated with the initial location and onward migration decisions of economic principal applicants (EPAs) in the period since the Immigration and Refugee Protection Act was enacted. It does so using the linked Longitudinal Immigration Database and T1 Family File.

The findings suggest that many factors are strongly associated with the probability of initially residing in MTV. Prominent among these is immigration category, followed closely by country of birth, intended occupation, knowledge of official languages and pre-landing work experience in MTV.

After the initial location decision is made, the geographic location of economic immigrants does not change much. The initial location decision of EPAs is a very strong predictor of their location years later. Indeed, only about 11% of EPAs have moved to or out of MTV by 10 years after landing. That being said, some important factors are associated with the onward migration that does take place. Key among these is country of birth, especially for EPAs who settled outside MTV. The number of residents in the city of initial location who come from the same country is positively associated with remaining in MTV and negatively associated with moving to MTV. Notably, immigration category plays less of a role in explaining the onward migration decision, compared with the initial location decision.  

References

Bonikowska, A., F. Hou, and G. Picot. 2015. Changes in the Regional Distribution of New Immigrants to Canada. Analytical Studies Branch Research Paper Series, no. 366. Statistics Canada Catalogue no. 11F0019M. Ottawa: Statistics Canada.

Frenette, M. 2017. Postsecondary Enrolment by Parental Income: Recent National and Provincial Trends. Economic Insights, no. 70. Statistics Canada Catalogue no. 11-626-X. Ottawa: Statistics Canada.

Hou, F. 2007. “Changes in the initial destinations and redistribution of Canada’s major immigrant groups: Reexamining the role of group affinity.” International Migration Review 41 (3): 680–705.

Ostrovsky, Y., F. Hou, and G. Picot. 2011. “Do immigrants respond to regional labor demand shocks?” Growth and Change 42 (1): 23–47.

Pandey, M., and J. Townsend. 2011. “Quantifying the effects of the Provincial Nominee Programs.” Canadian Public Policy 37 (4): 495–512.

Picot, G., and A. Sweetman. 2012. Making It in Canada: Immigration Outcomes and Policies. IRPP Study no. 29. Montréal: Institute for Research on Public Policy.