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.
The experiences of recent immigrants in Canada have received considerable attention from the media and researchers since the mid-1990s. Much of this attention has focused on the labour market and financial outcomes of immigrants, such as the extent to which they find employment in their intended field, experience upward earnings mobility and obtain incomes above low-income thresholds. A recurring message is that immigrants who landed in the 1990s and 2000s are not faring as well as those who landed in earlier decades (Picot and Sweetman, 2005; Picot 2004). Deteriorating economic outcomes, coupled with rising levels of educational attainment among more recent landing cohorts, raise the question of whether there is a large and perhaps widening gap between immigrants' expectations of life in Canada and their subsequent experiences of it.
Other evidence offers a more favourable starting point. Immigrants in the 2000-2001 landing cohort express positive views of the social and political environment in Canada, pointing to the importance of safety and security, rights and freedoms and peace and stability as aspects of Canadian life they like most (Schellenberg and Maheux 2007). They also point to these factors as central in their decision to settle permanently in Canada. Furthermore, most immigrants, including those in economic admission categories, say they came to Canada for non-economic reasons, such as joining family members already here, providing a brighter future for their children and, enjoying a high quality of life. In this context, assessments of life in Canada may be more positive than economic outcomes alone might suggest.
In this paper, we use the Longitudinal Survey of Immigrants to Canada (LSIC) to examine how immigrants in the 2000-2001 landing cohort subjectively assess their life in Canada. More specifically, to what extent are they personally satisfied with their life in Canada? How has life in Canada measured up to their expectations of it? If given the opportunity, would they make the same decision again to come Canada? Responses to these questions are examined across a broad range of demographic, social and economic characteristics.
Immigrants' assessments of their life in Canada warrant investigation for several reasons. First, the well-being of all Canadians is a central objective of public policy and is an important goal in its own right. Measures of subjective well-being, such as those used in this paper, offer a useful complement to other approaches that focus on employment, income or health. Second, immigrants' assessments of their life in Canada can shed light on the factors that contribute to (or detract from) a positive settlement experience, with potential implications for settlement programs. Third, the capacity to attract and retain skilled immigrants is increasingly viewed as a key ingredient for sustaining economic growth in western nations. Such capacity may be reduced if immigrant dissatisfaction is associated with higher rates of onward or return migration, or if dissatisfied immigrants tend to dissuade friends and family abroad from joining them in the host country.
2 Literature review
The question of what makes people satisfied has been central to philosophy since antiquity and has been a focus of research in a variety of disciplines in modern times as well. Research in the field of subjective well-being is of particular relevance to studies of immigrants' assessments of life in their host country (see reviews in Veenhoven 1996; Diener, Suh and Oishi 1997). In this approach, an individual's well-being is not defined by external criteria, such as income, wealth or body mass index, but rather in terms of his or her own subjective assessments, thus giving "…priority and respect to people's own views of their lives." (Diener, Suh and Oishi 1997). This internal perspective is typically conceptualized along two dimensions. The first involves the cognitive appraisals people make of their life; that is, conscious evaluative judgments about satisfaction with their life as a whole or with specific aspects of it, such as work or marriage (Diener, Suh and Oishi 1997). The second dimension involves emotional states or moods, including positive affect such as happiness, joy or pride and negative affect such as sadness, anger or anxiety. Cognitive appraisals and emotional states are correlated, although the evaluative judgments people express have been found to be less responsive to short-term situational contexts than emotional states. Our analysis is most closely akin to cognitive appraisals, focusing on the evaluative assessments that immigrants make of their life in Canada.
Subjective well-being research has identified a broad range of factors associated with life satisfaction, some of which have been replicated in a smaller subset of literature on life satisfaction among immigrants. Our literature review draws on both and is deliberately broad in scope. 1
2.1 Demographic characteristics, health status and personality traits
A number of demographic variables are generally found to be correlated with life satisfaction among the general population. There is a well-documented U-shaped correlation between age and satisfaction, with satisfaction levels lower among individuals in their thirties and forties than among individuals in younger and older age groups (Helliwell and Putnam 2004). Individuals who are married or living in common-law are generally found to have higher levels of life satisfaction than those who have never been married, widowed, separated or divorced (Helliwell and Putnam 2004). In the aggregate, women tend to report higher life satisfaction than men, although gender differences are not significant in several studies focused on immigrants (Fugl-Meyer, Melin and Fugl-Meyer 2002; Remennick 2005).
Education is an often-reported correlate of subjective well-being, with levels of life satisfaction rising in tandem with educational attainment. However, Helliwell and Putnam (2004) note that this correlation tends to diminish or disappear when other factors such as health and employment status are taken into account. In this respect, education may be an instrumental variable that improves health, and in turn satisfaction, but which does not influence subjective well-being directly.
There is a well-documented positive correlation between health and subjective well-being. Helliwell and Putnam (2004, p. 1440) report in their work that "…as in many other studies, self-assessed health status is the single most important correlate of subjective well-being..."
Another central theme in subjective well-being research is the role played by personality characteristics. Traits such as social assertiveness, empathy, extraversion and internal locus of control are among those consistently found to be correlated with subjective well-being among the general population. (For a recent meta-analysis see Steel, Schmidt and Shultz 2008). Correlations between personality traits and life satisfaction have been documented in a few studies of immigrants. For example, Young (2001) reports that refugees who have high self-esteem and internal locus of control maintain higher levels of life satisfaction under conditions of migration stress. Uksul and Greenglass (2005) find that proactive coping and optimism are negatively correlated with depression and, to a lesser extent, positively correlated with life satisfaction among Turkish migrants in Toronto, while Vohra and Adair (2000) report that guilt over leaving the country of birth is negatively correlated with life satisfaction among Indian immigrants in Canada.
2.2 Settlement experiences
The immigrants' experiences in the host country have also been found to be correlated with levels of satisfaction. Perceptions of acceptance and welcome, particularly perceived discrimination, have received some research attention. Sam (2001) and Chow (2007) find a negative correlation between perceived discrimination and life satisfaction among immigrant and international students enrolled in colleges and universities, with students who said they had experienced discrimination expressing lower levels of satisfaction than students who said they had not. Vohra and Adair (2000) and Ying (1998) document the same correlation among samples of adults. The negative correlation between perceived discrimination and life satisfaction is also reported in a broad study of immigrants in 13 countries (Vedder, van de Vijver and Liebkind 2006). In an Australian study, Fozdar and Torezani (2008) note the apparent paradox of high levels of perceived discrimination in combination with positive reports of subjective well-being among refugees. Fozdar and Torezani (2008, p. 30) argue that their negative experiences and perceptions were expressed as 'contained disappointment,' rather than as "…serious dissatisfaction with life generally, orientation to Australia or negative subjective well-being."
In the field of cross-cultural psychology, a central theme is how individuals adapt to the changing cultural contexts in which they are located. New immigrants may find themselves in a social and cultural milieu that is far different from that of their country of origin, facing different sets of norms, attitudes and behaviours. A variety of responses may be undertaken in this context (see Berry and Sam 1997; Berry 1997). An individual's capacity to adjust 'behavioural repertoires' to the new milieu vary, and for some the strains associated with this process may have implications for well-being, including satisfaction with life (Roccas, Horenczyk and Schwartz 2000; Berry and Kim 1988).
In addition to the social and psychological aspects of settlement, immigrants also face a variety of logistical challenges. Some of these, such as navigating an unfamiliar city or finding housing, may be overcome fairly soon after arrival. Others, such as accessing health services, may be an ongoing challenge (Schellenberg and Maheux 2007). Obstacles or frustrations of this sort may influence immigrants' assessments of life in their host country. Tran and Nguyen (1994) find that unmet health-care needs are associated with lower levels of life satisfaction among male Indochinese refugees in the United States. More broadly, the 'hassles' and 'uplifts' individuals encounter on a daily basis have been found to be correlated with satisfaction (Hart 1999).
2.3 Social capital
Studies of life satisfaction among the general population show a positive correlation between social ties and well-being. For example, Helliwell and Putnam (2004, p. 1435) report that "…social capital is strongly linked to subjective well-being through many independent channels and in several different forms." The relationship between social ties and life satisfaction has also been documented among immigrant populations. For example, Ying (1992) finds that among Chinese-American immigrants, life satisfaction is negatively correlated with social isolation and positively correlated with friendship ties and participation. Likewise, a positive correlation between satisfaction and contact with friends is documented among immigrants by Sam (2001) and Chow (2007).
2.4 Material well-being
The relationship between material well-being and subjective well-being has been central in life satisfaction research for over 40 years and has been advanced using both nations and individuals as the unit of analysis (Veenhoven 1996). At the individual level, the relationship between income and subjective well-being is complex. As noted, "differences in life-satisfaction across individuals are not proportional to differences in their income." (Boarini, Johansson and Mira d'Ercole 2006, p. 35); the correlation between decreases in income and declines in life satisfaction tend to be stronger than the correlation between increases in income and improvements in life satisfaction; and changes in non-financial characteristics, such as health, education, and social ties, are associated with larger changes in life satisfaction than gains in income.
The relationship between income and satisfaction may have particular relevance for new immigrants given variations in the strength of the correlation between income and satisfaction across the income distribution. Specifically, Diener et al. (1993) report that the correlation between income and satisfaction is strongest at the bottom of the income distribution (less than US$15,000) and the correlation decreases steadily above this threshold. Given the over-representation of recently-landed immigrants at the bottom of the income distribution (Picot 2008), it may be that income is a more important correlate of life satisfaction for them than for other groups. The strength of this correlation may also diminish over time, as earnings tend to rise with years of residence in the host country.
The employment and income characteristics of new immigrants may be correlated with life satisfaction in additional ways. As Diener, Suh and Oishi (1997) note, some theories of subjective well-being posit that individuals are satisfied when they are able to achieve their goals. While some goals are universal, such as the desire to meet one's basic needs, others are more idiosyncratic. Hence, the goals people seek to achieve are shaped by their location in the life course, their circumstances and histories, their preferences and tastes and so on. In this context, the goals sought by immigrants in the years immediately after landing, and in turn, the factors correlated with their subjective well-being, may be different from those of more well-established immigrants or native-born persons. Subjective well-being is also conceptualized in comparative terms. One variant of this is Michalos' (1985) Multiple Discrepancy Theory, which posits that life satisfaction is inferred from the discrepancy between 'how life is' and 'how life ought to be.' Individuals may use a variety of benchmarks for such comparisons including (i) what they want, (ii) what they had earlier in life, (iii) what they expected to have, (iv) what they think other people have and, (v) what they feel they deserve (Veenhoven 1996).
3 Data and methods
Data for this study were drawn from the Longitudinal Survey of Immigrants to Canada (LSIC). The target population of the survey, which was conducted jointly by Statistics Canada and Citizenship and Immigration Canada, includes all immigrants who (1) landed in Canada between October 1, 2000, and September 30, 2001, (2) were age 15 or older at the time of landing, and (3) landed from abroad having applied through a Canadian Mission Abroad. 2 The sampling frame for the LSIC was an administrative database of all landed immigrants to Canada maintained by Citizenship and Immigration Canada. 3
Three LSIC questionnaires were fielded during the course of the survey. Approximately 12,000 immigrants were interviewed between April 2001 and May 2002, about six months after landing in Canada. Approximately 9,300 of the same immigrants were located and interviewed in 2003 (about two years after landing), and about 7,700 were located and interviewed a third time (about four years after landing). These three interviews/questionnaires are referred to as 'Wave 1,''Wave 2' and 'Wave 3.' The approximately 7,700 respondents included in all three waves are nationally representative of approximately 157,600 new immigrants, of whom 104,400 are economic immigrants, 42,600 are family-class immigrants and 9,700 are refugees.
LSIC respondents were asked several evaluative questions about their life in Canada. They were asked about their satisfaction with their life in Canada during the Wave 1 and Wave 3 interviews. The Wave 1 question read:
Now I would like to finish with a few general questions about life in Canada. Generally speaking, how satisfied are you personally with your experience in Canada so far?
Respondents answered using a five-point scale ranging from "completely dissatisfied" to "completely satisfied." In Wave 3, respondents were asked:
Using a scale of 1 to 5, where 1 means "Very dissatisfied" and 5 means "Very satisfied," how would you rate your level of satisfaction with your life in Canada?
In all three waves, LSIC respondents were asked to compare their experience in Canada with their expectations of it and also to reflect on their immigration decision:
Generally speaking, would you say that your experience in Canada has been (i) much better than you expected; (ii) somewhat better; (iii) about what you expected; (iv) somewhat worse; (v) much worse than you expected?
If you had to make the decision again, would you come to Canada?
We use the term 'assessments of life in Canada' to refer to these three outcome measures. Given the design of the LSIC, these questions refer to the period from six months to four years after landing, which is a more specific reference period than is typically used in studies of general life satisfaction. Furthermore, subjective well-being research generally distinguishes between satisfaction with life overall and satisfaction in specific domains, such as job or marital satisfaction. Our analysis falls somewhere between, since 'life in Canada' is more limited in scope than 'life in general,' but may reflect evaluative judgments across multiple domains, such as family, employment and community ties.
Our analysis is deliberately broad in scope, drawing on the breadth of information available in the LSIC and key findings from the subjective well-being research. The following independent variables are included our analysis. The distribution of respondents across each is shown in Table 1 .
3.1.1 Demographic characteristics, health status and coping
LSIC respondent's sex, age and education at landing are included in our analytical models. 4 Self-assessed health status is included, using respondent's evaluations of their health on a five-point scale ranging from poor to excellent. 5 Consistent with the literature, we expect self-assessed health status to be positively correlated with favourable assessments of life in Canada.
The LSIC does not include the detailed battery of questions that psychologists use to construct psychometric measures of traits, such as assertiveness or locus of control. However, LSIC respondents were asked:
In general, how would you rate your ability to deal with the everyday tasks in your life, such as family and work responsibilities?
This measure of self-assessed coping capacity is included to, at least partially, account for individual personality differences since these likely affect self-assessments of coping skills. 6 The variable is expected to be positively correlated with favourable assessments of life in Canada.
3.1.2 The immigration process and settlement experiences
Canada's immigration policy has been guided by three broad objectives: to reunite families, to fulfill the country's international obligations and humanitarian tradition with respect to refugees, and to foster a strong viable economy in all regions of Canada. These objectives are reflected in the admission categories of immigration through which people are admitted to Canada as permanent residents. Five admission categories are identified in our analysis: principal applicants in the skilled worker category; spouses and dependents of skilled workers; family-class immigrants; refugees; and others. 7
About 1 out of 10 LSIC respondents had lived in Canada prior to immigrating, often on a student or work visa. Prior residents who subsequently immigrate to Canada may be a self-selected group, comprised of individuals whose past experiences in the country were particularly positive. They may also have more realistic expectations of life in Canada than those with no prior residence. Sam (2001) reports that life satisfaction among international students is positively correlated with information received prior to arrival. Furthermore, individuals who had previously resided in Canada may have social networks, work experience or other assets that facilitate settlement. For these reasons, we expect prior residents to express more favourable assessments than other immigrants.
Other aspects of the immigration process were considered in earlier versions of our analysis. The presence of family members in Canada at the time of landing was examined, with the expectation that this would be positively correlated with favourable assessments. However, this was not the case and the variable was dropped. Reasons for immigrating to Canada were also considered. In his study of immigrant high-school students in Toronto, Chow (2007) finds that non-economic motivations for immigrating were positively correlated with life satisfaction. Reasons for immigrating were included in earlier versions of our models but did not yield significant results, and separate multivariate models were run for immigrants who had (or had not) immigrated for economic reasons. Again, this did not yield noteworthy results, and reasons for immigrating were dropped from the analysis.
In terms of experiences after landing, LSIC respondents were asked if they had experienced discrimination or had been treated unfairly by others because of their ethnicity, culture, race or skin colour, language or accent, or religion. Those who said "yes" were asked about the frequency of such experiences. This variable is included with the expectation that perceived discrimination will be negatively correlated with favourable assessments of Canada. In addition, respondents were asked if they felt they had to change their values or way of thinking or behaving in order to adapt to living in Canada, and if so, how difficult they found it to make such changes. This variable is also included with the expectation that individuals experiencing difficulties will have less favourable assessments.
Turning to some of the logistical aspects of settlement, LSIC respondents were asked if they had encountered problems accessing health care, enrolling in training or educational programs or finding housing. A set of 'yes/no' dummy variables is included with the expectation that logistical problems will be negatively correlated with assessments of life in Canada.
3.1.3 Social capital
Four measures of social capital are included in our analysis: frequency of contact with friends; perceived friendliness of neighbours; frequency of attendance at religious services; and frequency of involvement in other groups or organizations. All four measures are expected to be correlated with positive assessments of Canada.
3.1.4 Material well-being
Three variables pertaining to material well-being are included in our analysis model. The first is employment status, as satisfaction levels are generally found to be lower among unemployed than employed individuals. The second is personal income 8 and the third is housing tenure. All three are expected to be associated with positive assessments of Canada.
3.1.5 Comparisons with the country of origin
In addition to the variables listed above, a set of supplementary models is presented including three additional variables: (i) respondents' assessments of their material well-being after two years in Canada ("things like a car, home and disposable income"), compared to their situation before coming to Canada, (ii) respondents assessments of their quality of life in Canada ("things like safety, freedom and pollution"), compared to their situation before coming to Canada, 9 and (iii) gross domestic product (GDP) per capita at purchasing power parity in their country of birth. 10 These variables are included on the grounds that immigrants' assessments of Canada may be informed by comparisons against other benchmarks, such as prior circumstances. Positive assessments of Canada are expected to be more prevalent among immigrants who say their material well-being and quality of life are better now than they used to be. However, these variables are subjective assessments of life now and in the past, and may capture the same underlying issues as our dependent variables. We present the results given the conceptual relevance of these variables, but do so in a set of supplementary models given the potential endogeneity.
3.1.6 Changes in assessments of Canada over time
Our analysis largely focuses on respondents' assessments of their life in Canada at Wave 3—four years after landing—rather than on changes in individuals' assessments over time. This is because the satisfaction question was not asked at Wave 2 and because wording and placement of the satisfaction question changed from Wave 1 to Wave 3, which raises the possibility that differences over time reflect survey design rather than respondents' views. 11 Moreover, several key independent variables, including capacity to cope with daily tasks, perceived discrimination, adaptation to life in Canada and perceptions of neighbours were not asked at Wave 1. That being said, multivariate models, including a reduced set of independent variables, are run using pooled information from all three LSIC waves. These models include a variable identifying whether an individual's assessment of his or her life in Canada was provided six months, two years or four years after landing, allowing us to determine if his or her assessment became either more positive or more negative over time.
In terms of statistical techniques, ordered logistic regression models are run on the satisfaction and expectation questions to take full advantage of the five-point response categories. A logistic regression model was run on the question regarding the decision to come to Canada. Predicted probabilities were calculated from the models, setting other covariates to their mean values. We focus on these predicted probabilities in our discussion for ease of presentation. All models were calculated using bootstrap weights to correct variance estimates for survey design. Before turning to our results, it is critical to first address the issue of sample attrition, as this has fundamental implications for this study.
3.2 Sample attrition
Of the 12,040 immigrants who completed the LSIC six-month questionnaire, 7,716 were subsequently located and completed questionnaires two and then four years after landing. The other 4,324 were not retained in the sample either because they were not relocated, or were unwilling to complete subsequent questionnaires. 12 This attrition rate of 37% 13 raises the question of whether it was the most dissatisfied immigrants who were lost. If dissatisfied immigrants were more likely than others to move, either within Canada or abroad in search of better opportunities, the chances of retaining them in the sample would have been disproportionately low. This would also be the case if dissatisfied immigrants were more likely than others to refuse to be re-interviewed. In either case, the remaining LSIC sample would disproportionately comprise satisfied respondents, thus casting doubts on any conclusions drawn regarding immigrants' assessments of life in Canada. Our analysis indicates that this is not the case.
To address this issue, respondents who completed the Wave 1 questionnaire were divided into two groups: those who completed both the Wave 2 and Wave 3 questionnaires (and hence were retained in the sample), and those who did not. The percentage of respondents with specific characteristics who were retained in the sample; the odds of respondents with specific characteristics being retained in the sample, after controlling for other factors; and the compositional characteristics of respondents who were and were not retained in the sample are shown in Table 2 .
Overall, 63% of the respondents who completed the first LSIC questionnaire (Wave 1) went on to complete the second and third questionnaires as well (Waves 2 and 3). The shares that were retained in the sample did not vary across responses to the questions on satisfaction, expectations or the decision to immigrate. For example, 63.8% of immigrants who were completely satisfied with life in Canada after six months were retained, compared to 64.7% of those who were completely dissatisfied. Likewise, 63.5% of those who said life in Canada was much better than expected were retained, compared to 65.3% of those who said life was much worse than expected. The same patterns (or lack thereof) are evident when satisfaction at Wave 2, and the likelihood of being retained in the Wave 3 sample, are considered (data not shown).
When assessed in terms of sex, age group and immigrant admission categories, differences in retention are modest. The largest differences are found among respondents aged 35 to 44, who were slightly more likely than average to be retained (67.6%) and among the small number of respondents aged 55 or older, who were somewhat less likely (54.2%). Across immigrant admissions categories, 69.5% of refugees were retained in the sample compared with 58.4% of family-class immigrants. 14
Turning to other possible sources of selection bias, one might expect the likelihood of remaining in the LSIC sample to be correlated with the occupations in which immigrants intended to work. Immigrants in the 2000-2001 landing cohort, who had planned on working as electrical engineers, computer programmers and software developers, were not in Canada long before the downturn of the high-tech sector. Canadian-born and immigrant workers alike had to find new jobs (Frenette 2007), and under these circumstances, immigrants may have opted to return to their country of origin or migrate onward in search of employment opportunities. If this were the case, one would expect a smaller-than-average share of high-tech workers to be retained in the sample. 15 More broadly, given the well-publicized challenges that many highly-educated immigrants face in finding work in their field, one might expect satisfaction levels and, in turn, retention rates to be lowest among university graduates. Evidence does not support either hypothesis. LSIC respondents intending to work in high-tech occupations were just as likely to be retained in the sample as respondents intending to work in other occupations, and university graduates were just as likely as respondents with less than high school education to be retained. Overall, it is the lack of variability across occupation and education categories that is most striking.
All in all, recently landed immigrants are a highly mobile population (Hou 2007), and locating each LSIC respondent two and again four years after landing was a challenge. Yet even though the LSIC has an attrition rate of 37%, there is no evidence to indicate that this introduced a systematic bias in reported assessments of life in Canada.
4.1 Descriptive results
Most LSIC respondents have positive assessments of their life in Canada. Six months and four years after landing, just over 54.2% said they were satisfied with life in Canada and almost 19% said they were very, or completely, satisfied (Table 3 ). Combined, 73% of respondents provided a favourable response to this question, 17.7% of respondents were neutral in their assessment, saying they were neither satisfied nor dissatisfied with life in Canada, while 9.4% said they were either dissatisfied or very dissatisfied.
Respondents were somewhat less positive about the extent to which their expectations about life in Canada had been met. In all three waves, almost 25% said that life in Canada was somewhat, or much worse, than they had expected. Whether or not these individuals had overly optimistic expectations at the outset cannot be determined. The largest share of respondents— 39% in Wave 1 and about 33% in Waves 2 and 3—said that life in Canada was about what they expected it would be. Finally, 36% of respondents said life in Canada was somewhat better or much better than they had expected at Wave 1, with this share increasing to 41% and 43% at Waves 2 and 3.
Considering our third measure, the vast majority of LSIC respondents (87% to 91%) said that, if they had to make the decision again, they would still come to Canada.
4.2 Multivariate results
Our multivariate analysis begins with a set of models run on assessments of life in Canada provided four years after landing in Canada.
4.2.1 Demographic characteristics, health status and coping
LSIC results are consistent with many of the findings reported in the research literature. For example, the often-found U-shape relationship between age and satisfaction is evident, with immigrants aged 35 to 44 less likely to be satisfied with life in Canada than their younger and older counterparts (Table 4). Holding other characteristics at their mean values, the predicted probability of being satisfied or very satisfied is 2 percentage points higher among immigrants aged 25 to 34, and 7 percentage points higher among those aged 55 or older, compared with the 35-to-44 year age group. 16 Similarly, 'middle aged' immigrants are less likely than their younger counterparts to report that their expectations have been met or that they would make the same decision again to come to Canada. 17
There is a negative correlation between educational attainment and satisfaction, as university graduates are least likely to be satisfied with life in Canada (a predicted probability of 72%) and immigrants with less than high school education are most likely (a predicted probability of 79%). Immigrants with high school diplomas or non-university credentials fall within this range. The negative correlation between education and met expectations is stronger, with a 12-percentage-point difference between university graduates and individuals with less than high school education (at 39% and 51% respectively). This correlation remains significant with health status, employment status and personal income in the model, suggesting that the correlation between education and satisfaction is more direct among recent immigrants than it may be among the general population of immigrants. Turning to our third outcome measure, immigrants with less than high school education are most likely to say they would make the same decision to come to Canada, but responses do not vary across other groups. Considering gender, the responses of men and women do not differ significantly. 18
Consistent with the literature, there is a strong correlation between health and assessments of life in Canada. Immigrants who have less favourable assessments of their health (i.e., they rate it as fair or poor) are significantly less likely to say they are satisfied with life in Canada than those rating their health as excellent (a 13-percentage-point difference in predicted probabilities). Differences in the other two outcome measures range from 8 to 10 percentage points between these groups. A similar correlation is evident between self-assessed capacity to deal with everyday tasks in life and assessments of life in Canada, with those individuals who rate their capacity as fair or poor, being 7 to 13 percentage points less likely to provide positive assessments than those rating their capacity as excellent.
4.2.2 The immigration process and settlement experiences
Immigrants' assessments of Canada are consistently and strongly correlated with the admission category by which they entered the country. Principal applicants in the skilled worker category express less favourable assessments than immigrants in the family class and refugee categories. This is most evident in terms of the extent to which life in Canada has been somewhat better or much better than expected (differences of 11 and 16 percentage points). Turning to prior residence in Canada, immigrants who have lived here before are more likely to express positive assessments than those who have not, with differences of 4 to 5 percentage points on the three outcome measures.
During their third interview (which occurred 4 years after landing), LSIC respondents were asked if they had experienced discrimination or unfair treatment over the previous 24 months and if so, how often this had occurred. Most respondents (72%) said they had not experienced discrimination or unfair treatment, while 28% said they had. Looking more closely at the response categories, 9% of all respondents said they had rarely experienced discrimination or unfair treatment over the previous 24 months, 15% said they had sometimes experienced such treatment, and 4% said they had experienced such treatment most, or all, of the time. There is a strong monotonic correlation between perceived discrimination and assessments of life in Canada. Respondents who said they had not experienced discrimination had a 77% predicted probability of being satisfied or very satisfied with life in Canada, with that probability declining to 66% among those who said they had experienced discrimination sometimes, and declining further to 55% among the small group that said they experienced discrimination most of the time or all of the time. The same large, monotonic decline is evident on the 'met expectations' and 'come again' outcome measures.
Considering the adjustment of 'behavioural repertoires' to a new social and cultural context, 41% of LSIC respondents said they did not have to change their values or behaviours to adapt to life in Canada, 38% said they had to make changes, but did not find this difficult, while the remaining 20% said they found it difficult making such changes. Individuals experiencing difficulties were less likely than others to express satisfaction with life in Canada, or to say they would make the same decision again to immigrate (differences of 6 percentage points). Difficulties adapting were not significantly correlated with the extent to which expectations had been met.
Turning to the logistical aspects of settlement, problems encountered accessing housing, health care and education/training are all negatively correlated with satisfaction with life in Canada and the extent to which expectations have been met. Problems encountered accessing health care and education/training are also negatively correlated with whether immigrants would make the same decision again to come to Canada. Differences in predicted probabilities across these variables range from 3 to 8 percentage points.
4.2.3 Social capital
Of the four social capital variables (frequency of contact with friends; perceived friendliness of neighbours; frequency of attendance at religious services, and frequency of involvement in other groups or organizations) included in the analysis, three are positively correlated with assessments of life in Canada. Immigrants who perceive their neighbours as very friendly are more likely to be satisfied with life in Canada than immigrants who do not know their neighbours or express neutral perceptions of them (a difference of 13 percentage points). A similar pattern is evident in the extent to which expectations have been met (a difference of 9 percentage points). Likewise, more frequent contact with friends is positively associated with satisfaction and met expectations. Turning to ties with groups and institutions, participation in religious services is positively correlated with met expectations in Canada and the willingness to make the same decision again. The fourth social capital variable—participation in other types of groups or organizations—is not correlated with any of these outcomes. 19
4.2.4 Material well-being
Turning to measures of material well-being, there is a positive correlation between employment status and satisfaction, but this difference is fairly small (at 4 percentage points), relative to other variables. Employment status is not significantly associated with the other two outcome measures. Alternative formulations of the employment status variable yield similar results.
Personal income is correlated with two of the three outcome measures—satisfaction and met expectations. Compared with individuals who have no personal income, those with incomes of $40,000 or more are significantly more likely to be satisfied with life in Canada (differences of 5 to 9 percentage points), and to say life in Canada is somewhat better or much better than expected (differences of 11 to 14 percentage points). Among the 80% of LSIC respondents with personal incomes below $40,000, assessments of Canada do not vary significantly across income groups. The same patterns are evident when household income is substituted in the models.
Finally, housing tenure is positively correlated with assessments of life in Canada, as home owners are more likely than renters to say they are satisfied with life in Canada and that their expectations have been met.
4.2.5 Comparisons with the country of origin
Turning to the variables included in our supplementary models, 43% of LSIC respondents said their material well-being in Canada is better than it was prior to immigrating, 29% said it is about the same, and 27% said it is worse. Immigrants in the skilled worker category expressed the least favourable views in this regard. 20 Considering perceived quality of life, immigrants in all admission categories (83% overall) said their quality of life is better in Canada than it was prior to immigrating, with only modest differences across admissions categories. 21
The inclusion of these additional variables has fairly modest effects on the other covariates in our analysis. For example, the predicted probability that satisfaction with life in Canada exceeded one's expectations varies by 13 percentage points across educational categories in the base model, with this difference declining to 10 percentage points with the inclusion of the additional variables. Similarly, predicted probabilities decline by, at most, 2 or 3 percentage points on most other covariates. One exception is immigrant admission category. The 16-percentage-point difference between skilled workers and refugees in the predicted probability of life in Canada exceeding one's expectations declines to 7 percentage points with the inclusion of the material well-being, quality of life and GDP per capita variables. Differences between immigrant categories on the other two outcome variables also diminish or disappear.
Immigrants' assessments of their relative material well-being and quality of life are significantly associated with their assessments of life in Canada. Compared with individuals who say their material well-being is better in Canada than it was prior to immigrating, those who say their material well-being is worse are significantly less likely to be satisfied with life in Canada (a difference of 10 percentage points), significantly less likely to say they would make the immigration decision again (a difference of 12 percentage points) and significantly less likely to say their expectations of life in Canada have been exceeded (a difference of 20 percentage points).
Perceptions of relative quality of life in Canada are also correlated with these outcomes, with differences in predicted probabilities across the better and worse categories ranging from 11 to 15 percentage points, although readers are reminded that very few LSIC respondents rate their quality of life in Canada as worse than it had been before.
GDP per capita in the respondent's country of birth is not strongly associated with the likelihood of expectations being met or making the same immigration decision again, but is positively correlated with satisfaction with life in Canada.
4.2.6 Changes in assessments of Canada over time
The results thus far are based on information collected from LSIC respondents approximately four years after arriving in Canada. One question that arises is whether their assessments changed over that four-year period. Outlooks may have been particularly positive during the first six months in Canada, when the promise of a new future lay ahead, but may have become less positive over time, if plans and expectations did not subsequently come to fruition.
A pooled sample combining information collected from respondents six months, two years and four years after landing was constructed, and assessments of Canada were compared across these points in time. Information on four variables—self-assessed coping capacity, perceived discrimination, difficulty adapting to Canada and perceptions of neighbours—are not available in either Wave 1 or Wave 2, and are consequently excluded from these models. Over the period from six months to four years after landing, the likelihood of being satisfied or very satisfied with life in Canada declined by 4 percentage points (net of other factors), while the likelihood of saying that coming to Canada was the right decision declined by 6 percentage points. The likelihood of expectations being met did not change over time.
Other results from the pooled model also warrant comment. First, the change in the predicted probability of being satisfied with life in Canada associated with problems concerning access to housing, health care and training/education are somewhat larger in the models based on the pooled sample (at 6, 6 and 10 percentage points respectively), than they are in the model using Wave 3 data (at 4, 3 and 6 percentage points respectively). While part of this difference may be attributable to the smaller set of covariates included in the pooled model, the difference may also be attributable to the fact that problems accessing housing and training are more frequently reported at six months and two years after landing than they are four years after landing (Schellenberg and Maheux 2007). In this context, problems accessing goods and services would appear to be stronger negative correlates of assessments of life in Canada at earlier points in the settlement process. Second, the change in the predicted probabilities of being satisfied with life in Canada and with life in Canada exceeding one's expectations varies more widely across personal income categories in the pooled sample (achieving significance in four categories), than they do in the Wave 3 data models (achieving significance in two response categories). This may be because the income variable is picking up some of the effects captured by perceived discrimination and other variables not included in the pooled models.
5 Discussion and conclusions
The objective of this paper has been to examine how immigrants in the 2000-2001 landing cohort assess their life in Canada. Four years after landing, about three-quarters of those still in the LSIC sample said they were satisfied, or very satisfied, with life in Canada, and a comparable proportion said their expectations had been met or exceeded. Nearly 9 out of 10 said that, if given the chance, they would make the same decision to come to Canada. One shortcoming of the LSIC is the absence of a control group (such as Canadian born) against which these responses can be compared. Without such a benchmark, it is difficult to say whether the 'three-quarters' figures cited above should be interpreted as 'high' or 'low.' One might have expected positive views to be less prevalent, given the labour market challenges immigrants have generally faced through the 1990s, or perhaps more prevalent given immigrants' favourable views of Canada's social and political environment. Interpretation of absolute levels of satisfaction is thus subject to debate.
An advantage of the LSIC is the opportunity it provides to examine how subjective assessments of life in Canada vary across a broad range of characteristics. Our analysis includes a number of variables—such as age, health status, coping capacity and social ties—that are generally found to be associated with life satisfaction in the subjective well-being literature. Our results, while focusing on a somewhat different set of outcome measures, are consistent with those in that literature. Our analysis also includes numerous other variables that are of particular relevance to immigrants and immigration research.
Within the 2000-2001 landing cohort, positive assessments of life in Canada are less prevalent among immigrants admitted through the skilled worker category, among university degree holders, and among persons aged 35 to 54. The disjuncture between expectations and outcomes is particularly large among the latter group, and they also score lower on the satisfaction question. Assessments of material well-being matter, as the likelihood of life in Canada falling short of expectations is especially evident among immigrants who believe their material well-being is worse after immigrating than it was before. This takes on added importance, given that one-third of immigrants in the skilled worker category felt worse off in material terms than they were prior to migration.
While economic factors are strongly correlated with subjective assessments of life in Canada, so too are social factors. For example, positive ties with neighbours are associated with an increase in the predicted probability of positive assessments of satisfaction and expectations in the range of 9 to 13 percentage points, while perceptions of discrimination or unfair treatment experienced sometimes, most of the time, or always, are also associated with large decreases in the predicted probabilities of positive assessments. While the deteriorating economic outcomes facing immigrants may appear to be an appropriate starting point for examining subjective assessments of life in Canada, social contexts and viewpoints should not be overlooked.