Human Resources and Skills Development Canada
Symbol of the Government of Canada

Human Resources and Skills Development Canada

www.hrsdc.gc.ca

A Study of Poverty and Working Poverty among Recent Immigrants to Canada


 
Previous Contents Next

6. Factors associated with low income among recent immigrants

All characteristics being equal, recent immigrants are at greater risk of poverty than are other Canadians (including earlier immigrants and native-born Canadians).

However, having paid employment, accumulating a reasonable number of hours of work and being part of a family with more than one potential breadwinner are characteristics that help recent immigrants avoid poverty, as they do for other Canadians.

Nonetheless, recent immigrants do not benefit as much from personal characteristics favourable to labour market participation such as having a higher level of education; having more labour market experience, and not having work-limiting disabilities.

It is not surprising to see that, in 2004, the incidence of low income was greater among immigrants who had arrived in the 1990s and 2000s than among other Canadians, since studies have already shown that recent immigrants are especially vulnerable to short- and longer-term low income.13 However, as with other Canadians, not all new immigrants live in poverty. Even within this group, certain characteristics can be associated with greater or lower risk of ending up in a low income situation.

In this section, the characteristics that make recent immigrants more vulnerable to low income are examined. We also verify if those characteristics are the same or not as those observed for the rest of the Canadian population. In order to answer these questions, two specifications of logistic regressions were estimated for each of the groups. The first (Model A) includes explanatory variables relative to labour market status, while the second (Model B) does not. Conducting logistic regressions to identify the determinants of poverty among recent immigrants has the advantage of controlling for possible interactions between explanatory variables and allows for the identification of the characteristics that have the greatest influence on the probability under study.14 However, because some demographic and socioeconomic characteristics (such as age, sex and level of education) have an incidence on the labour market status of individuals, it was decided that this category of variables be excluded in Model B in order to capture the direct impact of these variables on the probability of low income.

According to Model A, it is clear that family circumstances as well as labour market status are the most important determinants of poverty among working-age recent immigrants. The findings that appear in Table 8a indicate that recent immigrants who are heads of lone-parent families, or who are not living with relatives, are particularly at risk of experiencing poverty. For recent immigrants, being a full-time student or, for those who are not in school, not working a sufficient number of hours or being self-employed, are also characteristics that significantly increase the likelihood of experiencing poverty. Recent immigrants of Asian and Arab origin are more vulnerable to poverty than others in Canada.

Another interesting question that regression results allow us to answer is whether the determinants of poverty among recent immigrants are the same as those identified for the rest of the population.

Tables 8a and 8b clearly indicates that, regardless of their characteristics, recent immigrants are always at greater risk of poverty than other Canadians with similar characteristics. These Tables nonetheless shows that many of the characteristics that significantly increase the risk of poverty among recent immigrants also affect the rest of the population.

In particular, regardless of whether or not adults are recent immigrants, Table 8a demonstrates that being part of a lone-parent family, being unattached, not having accumulated at least 910 hours of paid work, and being self-employed during the year are characteristics that greatly increase the risk of living in poverty.

However, in comparison with the rest of the working-age population, few socioeconomic characteristics enable us to predict whether or not recent immigrants will experience poverty, when labour market status is controlled for. While education level, number of years of experience, and having work limitations have a significant impact on the likelihood that an adult who is not a recent immigrant will end up in poverty, among recent immigrants, these same characteristics are not significantly linked to the risk of experiencing poverty. All other things being equal, recent immigrants with a high level of education are not significantly less likely to live in poverty than those who are less educated. The same holds true for recent immigrants with more experience in the labour market and a better physical or mental condition.

Those results confirm an observation made in other studies which is that human capital factors are not as rewarding for immigrants as they are for non-immigrants. As Picot, You and Coulombe (2007, p.34) underlined "The movement to having more highly educated immigrants and the shift to accepting more skilled class immigrants had only a small effect." since "the differences in the probability of the outcomes between the less and more highly educated is not as great as one might expect."

Nonetheless, it is important to underline that when the labour market status category of variables is not included in the regression (Model B), the influence of sex, education level and presence of work-limiting disability on the risk of low income for working-age individuals is more important. This indicates that these characteristics indeed have an influence on labour market status of individuals which is partially captured through the effect of labour market status in Model A. It should be noted however, that among recent immigrants, only the impact of experience in the labour market becomes statistically significant when labour market status is omitted. The influence of sex, education level and work-limitation remains non-significant.

Table 8a
Results of model A which estimates the impact of some explanatory variables —including labour market status- on the risk of low income (LI), by immigrant status, 2004
  Recent immigrants Other Canadians
  Estimated coefficient* Predicted prob of LI15 Estimated coefficient Predicted prob of LI
All working-age adults   21.5%   9.5%
1. Demographic Characteristics
Time since immigration  
5 years of less (omitted) - 18.7% - -
Between 5 and 10 years 0.3362 22.9% - -
Between 10 and 15 years 0.2743 22.1% - -
Is an earlier immigrant  
Yes (omitted) - - - 11.7%
No - - -0.3680* 9.2%
Sex  
Female (omitted) - 21.6% - 9.5%
Male -0.0176 21.4% -0.0069 9.5%
Age  
18-29 (omitted) - 16.8% - 10.8%
30-44 0.7628* 26.3% 0.1116 11.5%
45-64 0.2243 19.3% -0.6612* 7.5%
Area of residence16  
Montréal 0.4330 23.5% 0.2577 10.9%
Toronto 0.3005 21.3% 0.0816 9.8%
Vancouver 0.6308 25.8% 0.0280 9.4%
Other area (omitted) - 17.7% - 9.2%
Visible Minority  
Black 0.5602 19.9% 0.2707 10.9%
Asian 1.2164* 29.4% 1.2961* 20.0%
Arab 1.4837* 34.0% 0.1225 9.9%
Other minority 0.2234 16.0% 0.3294 11.3%
Not a member of a visible minority (omitted) - 13.8% - 9.1%
2. Socioeconomic Characteristics
Highest level of education  
Less than a high school diploma 0.4581 26.9% 1.2946* 13.4%
High school diploma -0.1297 19.1% 0.8852* 10.3%
Post-secondary studies 0.1790 23.0% 0.6906* 9.1%
University degree (omitted) - 20.7% - 5.6%
Labour market experience  
Less than 3 years 0.3958 23.6% 0.3957* 10.9%
3 or more years (omitted) - 18.6% - 8.4%
Work limitations  
Yes 0.3317 25.7% 0.5485* 12.7%
No (omitted) - 21.2% - 8.9%
Labour market status  
Full-time students (FTSs) 0.7637* 23.5% 1.5964* 12.5%
Non-FTSs, 0 hours of work 1.4649* 35.0% 2.4324* 21.1%
Non-FTSs, self-employed 0.7275* 23.0% 1.5713* 12.3%
Non-FTSs, salaried, 1-909 hrs. 0.7047* 22.6% 1.5174* 11.9%
Non-FTSs, salaried, 910+ hrs. (omitted) - 14.2% - 3.6%
3. Family Characteristics
Family type  
Unattached, no relatives 2.3549* 56.2% 2.9789* 31.5%
Childless couple (omitted) - 15.8% - 3.7%
Two-parent family 0.0482 16.3% 0.4372* 5.4%
Lone-parent family 2.7728* 64.5% 2.2830* 21.1%
Other family 0.3445 19.3% 0.2429 4.5%
Pseudo R2 22.2 30.6
* These are estimated coefficients for which P<0.5 is below 0.05. Thus, they are statistically significant (i.e. different from 0 at the 95% confidence level).

 

Table 8b
Results of model B which estimates the impact of some explanatory variables —excluding labour market status- on the risk of low income (LI), by immigrant status, 2004
  Recent immigrants Other Canadians
  Estimated coefficient* Predicted prob of LI Estimated coefficient Predicted prob of LI
All working-age adults   21.5%   9.5%
1. Demographic Characteristics
Time since immigration  
5 years of less (omitted) - 18.0% - -
Between 5 and 10 years 0.4212 23.5% - -
Between 10 and 15 years 0.3255 22.1% - -
Is an earlier immigrant  
Yes (omitted) - - - 12.2%
No - - -0.3954* 9.2%
Sex  
Female (omitted) - 22.7% - 10.2%
Male -0.1975 20.1% -0.1994* 8.8%
Age  
18-29 (omitted) - 15.7% - 10.1%
30-44 0.8156* 26.0% -0.0485 9.7%
45-64 0.4595 21.1% -0.1632* 8.9%
Area of residence17  
Montréal 0.4816 23.6% 0.1818 10.7%
Toronto 0.2568 20.5% -0.0031 9.3%
Vancouver 0.7638 27.9% 0.0282 9.5%
Other area (omitted) - 17.4% - 9.3%
Visible Minority  
Black 0.6513 20.9% 0.1704 10.3%
Asian 1.2982* 31.1% 1.3051* 21.7%
Arab 1.6159* 37.0% 0.3700 11.9%
Other minority 0.0969 14.4% 0.2331 10.8%
Not a member of a visible minority (omitted) - 13.4% - 9.1%
2. Socioeconomic Characteristics
Highest level of education  
Less than a high school diploma 0.5702 28.0% 1.3949* 14.9%
High school diploma 0.0192 20.1% 0.8972* 10.5%
Post-secondary studies 0.1999 22.5% 0.6509* 8.8%
University degree (omitted) - 19.9% - 5.2%
Labour market experience  
Less than 3 years 0.7500* 27.2% 0.9685* 15.3%
3 or more years (omitted) - 17.0% - 7.6%
Work limitations        
Yes 0.4743 27.9% 0.9563* 16.6%
No (omitted) - 21.0% - 8.3%
3. Family Characteristics
Family type  
Unattached, no relatives 2.3462* 57.6% 2.7079* 31.1%
Childless couple (omitted) - 15.5% - 3.5%
Two-parent family 0.0922 16.6% 0.4543* 5.3%
Lone-parent family 2.6857* 64.5% 2.0940* 20.6%
Other family 0.3003 19.3% 0.3811* 5.0%
Pseudo R2 19.0 21.8
* These are estimated coefficients for which P<0.5 is below 0.05. Thus, they are statistically significant (i.e. different from 0 at the 95% confidence level).


13 Hatfield, M. (2004).

14 For further details on the methodology used, see Annex C.

15 See Annex C for further details concerning the methodology used and the manner in which the predicted probabilities were derived and interpreted.

16 It should be noted that the low income measure used in this model specification does not take into consideration the differences existing in the cost of living across different Canadian regions. This may explain why the area of residence does not appear as a factor explaining low income among immigrants and other Canadians

17 It should be noted that the low income measure used in this model specification does not take into consideration the differences existing in the cost of living across different Canadian regions. This may explain why the area of residence does not appear as a factor explaining low income among immigrants and other Canadians.

Previous Contents Next

Footer

Date Modified:
2007-08-02