ARCHIVED – Health Status and Social Capital of Recent Immigrants in Canada: Evidence from the Longitudinal Survey of Immigrants to Canada
The main data source used in this paper is the Longitudinal Survey of Immigrants to Canada (LSIC), which was designed to study how newly arrived immigrants adapt to living in Canada during their first four years after arrival. The survey’s target population was immigrants who arrived in Canada between October 2000 and September 2001, were 15 years of age or over at the time of landing, and landed from abroad. [ Note 1 ]
The survey addresses a number of issues including demographic and household characteristics of the longitudinal respondent, health, citizenship, social interactions, groups and organizations, language skills, housing, education, employment, values and attitudes, income, and perceptions of settlement.
The LSIC is longitudinal – that is, the same respondents were interviewed at six months (wave 1), two years (wave 2), and four years (wave 3) after landing in Canada, providing a dynamic picture of the integration experiences of these recent immigrants. Approximately 12,000 immigrants participated in the wave 1 interview, representing about 164,200 of the target population. The final survey (wave 3) sample of 7,700 immigrants represents 157,600 immigrants of the target population who still resided in Canada at the time of the last interview (Statistics Canada, 2007). Our study focuses on these 7,700 immigrants who participated in all three waves. The longitudinal weights designed by Statistics Canada are used to account for sample attrition.
In this paper, self-rated health is used as an indicator of immigrants’ health status. [ Note 2 ] The self-rated health indicator measures individuals’ perception of their overall health. It can reflect aspects of health not captured in other measures, such as incipient disease, disease severity, aspects of positive health status, physiological and psychological reserves, and social and mental function. Epidemiologists have demonstrated that self-rated health is an accurate reflection of a person’s health and a valid predictor of incident mortality and chronic morbidity (see Bond 2006; Idler and Benyamini 1997; Huisman et al. 2007).
In all three waves of the LSIC, the respondents were asked: “In general, would you say your health is excellent, very good, good, fair or poor?” Health status was then grouped into two categories according to the answer: healthy (excellent, very good, or good) and unhealthy (fair or poor). The health status variable is the dependent variable in our logit panel regression models.
To determine the extent to which social capital influences the health status of recent immigrants, we used information unique to the LSIC data on social interactions and group organization participation. We employed the social capital indicators developed by Xue (2008), which use a network-based approach to measure social capital. Unlike many social network measures in the literature that use ethnic, linguistic, or neighbourhood characteristics as a proxy for social capital (e.g., Deri 2005; Bertrand et al. 2000; Chiswick and Miller 1996), this network based approach emphasizes both the structure and content of individuals’ networks, using direct measures of social networks.
The structure of networks includes different levels of social networks. Similar to Xue (2008), in this paper social networks are categorized into three types: kinship, friendship, and organizational networks. The kinship network includes relationships with family members and relatives living in Canada. The friendship network consists of ties with friends. The organizational network is defined as the participation of immigrants in groups and organizations, such as community organizations, religious groups, ethnic or immigrant associations, etc.
Within each type of network, the content of networks is defined by the amount of social involvement and social support such as size, diversity, frequency of contact, and network reciprocity. Social network size is defined as the number of people or units with whom immigrants maintain different types of relationships (family, friends, organizations). While the LSIC does not provide information on the absolute numbers of people in all networks, there are some good substitutes for network size. For example, based on information available from the LSIC, we can obtain an approximation of network size for family ties by counting the number of types of relatives in Canada, such as spouse, children, parents, grandparents, brothers and sisters, uncles and aunts, and cousins. For friends network, sources where immigrants met new friends, such as ethnic association or club, religious activity, through relatives or friends, sports, hobby or other club, spouse’s work, ESL or FSL classes, other classes, etc., are counted to proximate the absolute size of the network. For organizational network, LSIC provides a direct measure of absolute number of groups or organizations that immigrants participated in.
Social network diversity represents the social and ethnic heterogeneity of network members, which is measured by the relative numbers of non co-ethnic members and co-ethnic ones in a person’s networks.
Social network density is defined as the frequency of contact between network members. Using the information on the frequency of contact with people in the networks and information on the relative number of co-ethnic members among friend networks and organizational networks, we create both diversity and density indexes for each type of network, which range from 0 to 1. The higher the diversity index, the more diversified the social network is. The higher the density index, the more frequently individuals contact family members, relatives, or friends, and/or the more frequently they take part in group and organizational activities.
Social network reciprocity can be measured as help received from networks as well as contribution made to networks. We create several indicators to measure the different types of help that an immigrant received from a particular type of network. [ Note 3 ] We also use a variable to indicate the number of organizations or groups for which an immigrant volunteered time.
The main social capital variables are shown in Table A1 in Appendix.
Family income is an important factor that significantly affects individuals’ health status (Zhao 2007a). In this paper we group immigrants into four groups by economic family income quartiles from the lowest to the highest: 0-25 percent, 25-50 percent, 50-75 percent and 75-100 percent. “Economic family” refers to a group of two or more persons who live in the same dwelling and are related to each other by blood, marriage, common law, or adoption (Statistics Canada 2007).
Employment status is also significantly related to the health status of immigrants (Zhao 2007a). The employment status of a respondent is grouped into two categories: employed, and not employed. The employment status of a respondent’s spouse is grouped into three categories: no spouse, spouse currently employed, and spouse currently not employed, which also captures marital status.
Other socio-demographic variables that have potential impacts on the health status of recent immigrants and are controlled for in our regression analysis include age, gender, immigrant category, source area, education level at landing, official language ability, and incidence of problems accessing Canadian health care system.
We group immigrants into five categories: (1) family class immigrants, (2) skilled workers – principal applicants, (3) skilled workers – spouses and dependants, (4) refugees, and (5) other immigrants. [ Note 4 ] The source countries of immigrants are grouped into five broad areas: North America, United Kingdom and Western Europe, Europe except UK and Western Europe, Asia and Pacific, Africa and Middle-East, Caribbean and Guyana, and South and Central America. Education at landing is grouped into four categories: high school or less, trade certificate or college/some university, bachelor’s degree, and master’s degree or above.
To investigate age-specific health status, we separate the population into five age groups: 15-19, 20-34, 35-44, 45-64, and 65+ years. Place of residence is grouped into six categories by census metropolitan areas (CMAs): living in one of the top five CMAs and living in an area other than the five major CMAs. [ Note 5 ] Official language ability (self-assessed) [ Note 6 ] is captured by two dummy variables for English and French: speaking English (or French) well (i.e., speaking fairly well, well, and very well with English or French as the native language) or not (i.e., speaking poorly or not able to speak in English or French).
Accessibility to the Canadian health care system is grouped into two categories: having and not having problems accessing health care services. Problems identified include long waiting times, discrimination, problems finding a family doctor, transportation, and/or insurance covering prescription medication, etc. [ Note 7 ] Health care service access is important because it influences immigrants’ health status and quality of life. Health status may deteriorate as individuals become more prone to chronic conditions due to barriers to health care access (Rivers and Patino 2006).
1 Individuals who applied and landed from within Canada are excluded from the survey. Refugees claiming asylum from within Canada are also excluded from the scope of the survey. For detailed information on sample selection of the LSIC and the survey design and frame, please consult Longitudinal Survey of Immigrants to Canada, Wave 3 – Microdata User Guide, Statistics Canada, 2007.
2 Self-reported health is a commonly used measure of health but has limitations. The data from survey that are self-reported and the degree to which they may be inaccurate because of reporting error is unknown (Perez 2002). One issue in particular is that the notion of what constitutes good health may well change not only with age but also with time in the new country. One of the weaknesses of the LSIC is its limited information on health, so we have few alternatives but to use self-reported health.
3 In the following regression analyses, not all indicators for network reciprocity are included. Because of low variability, the indicators for number of types of help received from a certain kind of network are not included.
4 Other immigrants include mostly business immigrants and a very small number of immigrants who landed under the categories not specified in aforementioned categories.
5 The top five CMAs are Toronto, Vancouver, Montreal, Ottawa, and Calgary.
6 It is important to note that since we are dealing with survey data, as with most other variables in the LSIC, knowledge of official language is self-assessed.
7 Incidence of having problems accessing Canadian health care system might be endogenously determined by social network variables, etc. To address this issue, we compare the GEE population averaged logit estimates with IV estimates, where incidence of barriers to the health care system is treated as endogenous, and education at landing is used as an instrument. We argue that education variables are correlated with incidence of reporting problems accessing health care services, but not with the health status of immigrants. The Wald test statistic of exogeneity from the IV results is not significant, suggesting that there is no sufficient information to reject the null that there is no endogeneity. With the thought that the instruments may not be adequate, we also ran a simultaneous bivariate probit regression of whether an immigrant reported as healthy and whether the immigrant reported having problems accessing health care services, based on the other covariates. The results of social capital effects on health are quite similar to what we report in Table 5. Both the IV results and the bivariate probit regression results are available on request from the authors.
Page details
- Date modified: