Ethnicity, immigration and housing wealth in Toronto.

Gyimah, Stephen Obeng ; Walters, David ; Phythian, Kelli L. 等

Abstract

Past research on ethnic inequality in Canada has focused largely on earnings and occupational stratification, paying little attention to differences in housing wealth. As recent research in the United States demonstrates, however, ethnic differences in homeownership patterns are significant in perpetuating and, to some extent, exacerbating social inequality among groups. Using data from the 1996 Canadian Census Public Use Micro File on individuals, this paper builds on previous research by investigating the impact of ethnicity on two different measures of housing wealth (tenure and value of housing) in Toronto. The study is guided by two theoretical perspectives on ethnic differences in housing wealth: spatial assimilation theory and place stratification theory. Our results reveal considerable ethnic differences in housing wealth after accounting for life-cycle and sociodemographic factors, lending support to both theories.

Keywords: ethnicity, immigration, housing, Canada

Resume

La recherche anterieure sur l'inegalite ethnique au Canada s'est concentree surtout sur les gains et la stratification professionnelle, portant peu ou pas d'attention, sur les differences de la richesse immobiliere. Comme le demontre la recherche recente aux Etats-Unis, cependant, les modeles de differences ethniques sur l'acces a la propriete sont significatives a perpetuer et, dans une certaine mesure, exacerber l'inegalite sociale parmi les groupes. En utilisant les donnees du recensement public canadien de 1996 sur l'usage de micro-fichiers des individus, cet article repose sur la recherche anterieure, par l'investigation de l'impact de l'appartenance ethnique sur deux differentes mesures en matiere de la richesse de logements (occupation et valeur de la propriete) a Toronto. Cette etude est guidee par deux perspectives theoriques portant sur les differences ethniques de la richesse du logement: la theorie spatiale d'assimilation et la theorie sur la stratification de lieu. Nos resultats revelent des differences ethniques considerables sur la propriete du logement apres avoir tenu compte du cycle de vie et les facteurs socio-demographiques en pretant soutien aux deux theories.

Mots cles: l'appartenance ethnique, l'Immigration, le logement, Canada

Research Background

The past few decades have witnessed substantial research interests in ethnic inequality and socio-economic attainment of immigrants in Canada (see, Warner 2003; Pendakur and Pendakur 1998; Hou and Balakrishnan 1996; Maxim 1992, 1994; Shamai 1992; Herberg 1990; Li 1988; Balakrishnan 1988; Beaujot et al. 1988; Kalbach and Richard 1988; Balakrishnan and Wu 1992; Ray and Moore 1991; Darden and Kamel 2000). While much of this literature has focused on labour market outcomes, such as occupational attainment and earning differentials, only a few (see, Balakrishnan and Wu 1992; Ray and Moore 1991; Darden and Kamel 2000) have examined ethnic inequality by addressing differences in homeownership patterns. As recent research in the United States demonstrates, however, ethnic differences in homeownership patterns are significant in perpetuating and, to some extent, exacerbating social inequality among groups (e.g., Alba and Logan 1992; Yinger 1995; Conley 1999; Flippen 2001; Masnick 2001).

This paper contributes to the understanding of how immigrants and visible minorities (1) are incorporated into the Canadian stratification system by focusing on differences in housing wealth assessed through tenure status and value of housing. We investigate homeownership because of its cultural significance in Canadian society, the opportunities it offers for accumulating wealth, and its significance as an indicator of immigrant acculturation and integration into mainstream society (see Murdie and Teixiera 2003). As Alba and Logan (1992) have pointed out, homeownership among immigrants reflects a commitment of access to property within the host country and entails knowledge of local housing markets. Purchasing a home, then, symbolizes commitment to and investment in the host country and suggests that immigrants have internalized norms and values about ownership. As such, variations in patterns of homeownership across ethnic groups may reflect, in part, the difficulties particular immigrant groups and visible minorities encounter when trying to adapt to a new culture.

Aside from physical shelter, homeownership offers many social psychological benefits, including privacy, stability, and a sense of belonging. The latter two may be particularly important for immigrants, many of whom have experienced considerable disruption in their move to a new country. Further, housing brings with it numerous short and long term economic benefits. The home is often the principal component of wealth for most North American households (Adams 1987; Parcel 1982), bringing with it forced savings, investment appreciation, and protection against inflation. Ethnic stratification in homeownership, then, may serve to reproduce socioeconomic inequality between ethnic groups across the life course and at older ages in particular, when income from assets comprise a larger portion of one's wealth portfolio. Given the importance of homeownership for inheritance prospects, differences in ownership rates from one group to another may also have significant implications for inter-generational persistence in inequality (Saunders 1990).

While owning a home is a cherished dream of many Canadians, rates of ownership have been found to vary considerably by ethnicity and immigration status (Balakrishnan and Wu 1992; Murdie 1999; Ray and Moore 1991; Darden and Kamel 2000; Skaburskis 1996). Studies consistently have shown that Italians and Chinese have the highest ownership rates while Blacks and Aboriginals have the lowest. The literature also indicates that immigrants who arrived in earlier periods have higher rates of homeownership than recent arrivals and the Canadian born population. However, the data for much of this literature are now dated and may not capture important recent changes in ethnic composition of immigrants. For example, the studies by Ray and Moore (1991) and Balakrishnan and Wu (1992) were based on data from the 1986 census. According to Canadian census figures, the proportion of visible minorities in Toronto increased from 21 percent in 1986 to 32 percent 1996. Hence, there is a need to re-examine the issue with more recent data given the remarkable growth in the population of visible minorities and recent immigrants in Toronto, Canada's largest Census Metropolitan Area (CMA).

Although previous studies on ethnicity and homeownership have controlled for period of immigration, they have not explored whether differences in ownership across ethnic groups vary according to the period of arrival. Without discounting their merits, these studies may not fully explain ethnic inequality in housing if the interaction between ethnicity and period of immigration is ignored. In the Canadian context, examining such interactions may be relevant considering the changing ethnic composition of immigrants. While the annual influx of immigrants to Canada prior to 1980 were comprised predominantly of Europeans, more than half of immigrants that have arrived after 1980 are from countries outside of Europe. Indeed, roughly 73 percent of immigrants who arrived between 1991 and 2001 were from Asia, Africa, and Latin America (Statistics Canada 2003). Given that different ethnic groups have arrived in Canada at different periods, it is essential that the interaction between period of immigration and ethnicity be explored.

Generally, while extant research has focused on group differences in ownership rates, less attention has been devoted to the quality of owned homes, as measured through housing value. This is an important topic because many neighbourhoods are largely comprised of owner-occupied dwellings. As such, homeownership is often a requirement for living in particular locales. Persons wanting to live in suburban areas and communities with better quality housing, schools, roads, and lower crime rates, for example, are often forced to buy rather than rent. They must also invest more money into the purchase of their homes than if they were to live in lower-quality neighbourhoods. Yet research in Canada has not explored ethnic differences in housing value, which can be linked to quality of homes and neighbourhoods and all of the disadvantages associated with living in such conditions. Nor has research identified whether ethnic differences in homeownership are reproduced when addressing differences in the value of housing. This provides us with an opportunity to make a unique contribution to the research literature. We take a comprehensive approach to this topic by assessing ethnic differences in housing wealth, not only in terms of ownership patterns, but also in the value of owned homes, which allows the incorporation of an additional element of stratification that is not identified by the decision to own.

Theoretical Perspectives

Two theoretical perspectives guide this study: spatial assimilation theory and place stratification theory. The spatial assimilation model identifies residential assimilation as one outcome of the social attainment process. Much of the early work operating within this framework, particularly in the United States, viewed homeownership within the rubric of a micro-economic model of consumer choice, whereby home purchases are made according to individual needs and preferences vis a vis financial resources (Flippen 2001b). Thus, homeownership is closely tied to financial and life-cycle characteristics such as age, family structure, educational attainment, and income, each of which influence the decision to buy. Flippen (2001 b), for instance, found that the never married were substantially less likely to own their homes than married, divorced, and widowed persons (see also Krivo 1995; Flippen 2001 a). Balakrishnan and Wu (1992) also reported that household composition is strongly associated with ownership. Examining various household compositions, the authors found that husband/wife households with children were most likely to own their homes, while non-family households were least likely to own. The results of these and other studies therefore suggest that such life-cycle characteristics as marriage and childbearing act as motivators for entrance into homeownership.

With respect to ethnicity and immigration status, spatial assimilation follows a model of incorporation that continues the Chicago school's ecological tradition. The model suggests that as ethnic and minority groups acquire greater socioeconomic resources, including increased income, education, and occupational prestige, they experience improved residential outcomes (Massey 1985; Massey and Denton 1993; Charles 2003; Fong and Wilkes 1999). Discrepancies in rates of ownership and housing value across groups, then, can be attributed to group differences in the composition of life-cycle and financial characteristics. Once human capital and life-cycle variables are taken into consideration, then, any observed ethnic/racial differences should disappear. The spatial assimilation model has been found useful in explaining residential patterns and tenure choice in some contexts (Alba and Logan 1992, 1993; Massey and Fong 1990); however, it has also been found to be at odds with a body of literature that reveals substantial ethnic differences in homeownership patterns, even after taking into consideration various socio-demographic characteristics (Long and Caudill 1992; Myers and Chung 1996; Darden and Kamel 2000; Flippen 2001 a).

Place stratification theory offers an alternative explanation for observed differences in rates of homeownership, attributing ethnic housing inequality to discrimination in the housing market (Logan and Molotch 1987). According to this model, residential segmentation artificially constrains the housing options facing ethnic minorities, lowering their ownership rates and negatively affecting the value of their homes (Yinger 1995). As Alba and Logan (1992) point out, it in fact "costs" more for some ethnic groups to obtain ownership than others, particularly whites. That is, whites are more able than non-whites to translate their socio-economic resources into residential outcomes. This disparity may be attributed in part to discriminatory mortgage lending practices that favour whites and to the "redlining" of neighbourhoods with certain characteristics. Fox Gotham (1998), for example, found that mortgage lenders in Kansas City rejected minority applicants more frequently than whites of comparable incomes and, more importantly, that high-income African American applicants were rejected at a higher rate than low-income whites (see also Ross and Yinger 2002; Munnell et al., 1996; Massey and Denton 1993; Oliver and Shapiro 1995). Such practices have the effect of lowering the rates of homeownership among ethnic minorities, as well as reducing the value of homes they do own, as minorities are forced to reduce the size of their loan and purchase homes in lower-income areas.

Market constraints not only hinder the ability of minorities to purchase houses, but also the rates of appreciation of their homes and hence the accumulation of equity after ownership is attained (Krivo and Kauffman 2004). Though ethnic differences in housing values have not been subject to the same scrutiny as homeownership, extant research has shown that homes belonging to visible minorities tend to appreciate at a slower pace relative to homes owned by their white counterparts (Horton and Thomas 1998; Oliver and Shapiro 1995; Massey and Denton 1993). According to Massey and Denton (1993), the value and appreciation of homes is affected by residential segregation. Generally, comparable homes cost more in white neighbourhoods than in black and ethnically integrated neighbourhoods, and the value of homes in white areas tend also to increase more rapidly. Thus, while the cost of purchasing a home in a white community may be greater, the long-term financial benefits more than compensate for the initial cost.

Krivo and Kauffman (2004) suggest that white prejudice against living in communities comprised predominantly of visible minorities weakens the demand for housing in nonwhite neighbourhoods. Such prejudice not only has the effect of lowering the value and appreciation of minority-owned homes, but also makes it more difficult for visible minorities to purchase better quality homes. Shapiro (2004) estimated a 16 percent loss in the value of homes located in areas that are more than 10 percent black. In addition, the housing located in neighbourhoods populated primarily by visible minorities tends to be markedly inferior than in white areas. Research from the United States, for instance, indicates that the housing stock of blacks and Hispanics tends to be more crowded (Krivo 1995), of poorer structural quality (Rosenbaum 1996), and in neighbourhoods with fewer amenities and lower-quality schools and parks (Rosenbaum and Friedman 2001; Flippen 2001b; Krivo and Kauffman 2004). The difficulties minority groups face when trying to obtain high-quality housing as a result of residential segregation is further exacerbated by discrimination by brokers. In Canada, discriminatory behaviours in the housing market have been documented through surveys and audits (Chandra 1973; Quann 1979). Steering to ethnic neighbourhoods, withholding information about available homes, and showing lower-quality and lower-value homes all contribute to ethnic stratification in the housing market.

Both spatial assimilation and place stratification theories offer reasonable explanations for ethnic differences in homeownership and, by extension, housing value. The former predicts that as visible minorities achieve socio-economic parity with the majority group, their rate of homeownership and housing value will coincide. The latter theory, in contrast, suggests that differences in ownership and value will persist regardless of socio-economic attainment. This is because majority group members constrain the housing choices available to minorities so that social distance between themselves and visible minorities can be maintained (Myles and Hou 2004). However, this explanation is difficult to test directly because of a lack of existing data that adequately measure discrimination in the housing market. Nonetheless, we are able to test spatial assimilation theory by comparing the ownership rates and house values of various ethnic groups before and after controlling for theoretically relevant sociodemographic and life-cycle characteristics. Support for the spatial assimilation hypothesis would be obtained if we find that ethnic differences in both homeownership and housing value disappear after introducing life-cycle and sociodemographic variables. If, on the other hand, the effects of ethnicity persist even after considering such variables, place stratification theory must be considered as a possible explanation for ethnic differences in housing wealth. That is, if the effect of ethnicity on homeownership and housing value cannot be explained by group differences in age, immigration status, education, income, and marital status, then the possibility that discrimination leads to ethnic stratification in the housing market must be considered. Thus, the results of our analysis, in conjunction with previous research, will help to provide some insights regarding the validity of both the place stratification and spatial assimilation perspectives.

Data, Measures, and Methods

Data for the study are from the 1996 Canadian Census Public Use Micro File on individuals. (2) The file contains data based on a 3 percent sample of the population enumerated in the census and provides information on demographic, social, and economic characteristics of the Canadian population. In keeping with the bulk of the previous literature, we focus on principal household maintainers (3) in the 25-64 age group who live in the Toronto CMA. Limiting our analysis to Toronto minimizes the problem of regional variations in housing market characteristics, such as cost and affordability, which can affect the housing outcome for different ethnic groups. Also, it is well established that the majority of visible minorities and recent immigrants live in Toronto, making it an ideal location to study racial and ethnic differences in housing consumption. We focus on the pre-retirement population because housing purchases and wealth are theorized to peak during these years.

The study compares ethnic differences in two dimensions of housing wealth--homeownership and housing value. In the first set of statistical models, tenure status (rent or own) of the principal dwelling is used as the dependent variable. Given the dichotomous nature of this variable, a binary logistic model is used. In the second set of models, we examine housing quality defined as the self-reported value of the dwelling. Since the value of owned homes is top-coded (i.e., censored), it necessitates the use of a Tobit regression model (Tobit 1958; Breen 1996).

Our key explanatory variables are ethnicity and immigration status. To circumvent the 'single-multiple responses' problems inherent in the ethnicity variable (4), we used the census-derived visible minority indicator instead. This indicator identifies five ethnic groups: blacks, South Asians, Chinese, other visible minorities, and non-visible minorities who are mainly of Caucasian origins (5). Immigration status is derived from the question on period of immigration which includes five categories: 1) Canadian born; 2) immigrated before 1971; 3) immigrated between 1971 and 1980; 4) immigrated between 1981 and 1990; and 5) immigrated after 1990. Consistent with the spatial assimilation hypothesis, previous research has shown a strong association between homeownership and period of immigration, with earlier immigrants having higher ownership rates than those who have arrived more recently (Balakrishnan and Wu 1992; Ray and Moore 1991). Based on these findings, we expect to find that earlier immigrants will have greater housing wealth than later immigrants and the Canadian-born.

In addition to ethnicity and immigration status, several theoretically relevant variables found to be significant correlates of homeownership in Canada, including age, sex, marital status, household size, educational attainment, and income, are included in the analyses (Balakrishnan and Wu 1992; Ray and Moore 1991; Darden and Kamel 2000; Skaburskis 1996; Brereton 1989; Darko 1993). The age and marital status of the principal household maintainer are included in the models because they reflect the lifecycle stage of the respondent. In general, housing wealth has been found to be positively associated with age. This is explained by the fact that older people have likely obtained higher levels of capital and have a greater need of owning a home. Also, married couples have been found to have higher levels of housing wealth than those of other family arrangements.

Household size is included because it may indicate a willingness to commit to a major housing purchase. For example, larger households have greater need for space and therefore may be more likely to own than rent. Also, families with children may seek housing in suburban areas with better schools and amenities which, because of a lack of rental units, would require them to buy. Larger households might also have more wage earners, which would not only enable ownership, but would also permit the purchase of more expensive homes.

Finally, education and income are included in the analysis as life-cycle indicators relevant to the spatial assimilation hypothesis. Previous research has demonstrated that a significant relationship between education and housing wealth exists; however, the association tends to be weak, particularly at higher educational levels (Balakrishnan and Wu 1992). While we expect those with a high school education to have lower levels of housing wealth, as previous research suggests, we do not expect there to be a significant difference between those with non-university credentials and those with a university degree. Lastly, we expect income to be the most important predictor for both of our dependent variables. All things being equal, the likelihood of owning a home and the value of housing are expected to be higher for principal household maintainers with higher incomes.

Findings and Discussion

Descriptive Overview

Table 1 shows the distribution of the selected variables by ethnicity. It is clear that the percentage of Blacks under age 35 is higher than the other groups. Also, the percent of married couples is lowest among Blacks (6), which could have tremendous impact on housing tenure. With respect to immigration status, about 10 percent of Blacks are Canadian born compared with 4 percent of South Asian and 7 percent of the Chinese. Blacks also tend to be under represented at the highest level of education group compared with the other ethnic groups. For instance, only 21 percent of blacks in Toronto have some university education compared with about 40 percent for both the South Asians and Chinese, and 37 percent for non-visible minorities. Given the strong relationship between education and income, it also not surprising that Blacks are under represented at upper end of the income scale, relative to the other groups. Just about 3.8 percent of black earn 60,000 or more compared with 11 percent for South Asians and Chinese, and 21 percent for non-visible minorities. The fact that these differences are so pronounced illustrates why it is all the more important to control for these variables in our regression models.

Homeownership Patterns

In Table 2, we present bivariate findings on homeownership patterns by ethnicity, immigration status and selected socio-economic and demographic characteristics. Overall, about 60 percent of principal maintainers in Toronto are home owners; however, there are striking differences across groups. About 75 percent of Chinese and 52 percent of South Asians own their homes compared with only 28 of Blacks. Consistent with previous research, homeownership rates are highest among immigrants who arrived before the 1970s, and lower for those arriving after 1980 compared with the native born population. This is probably because earlier immigrants are more likely to have had enough time to accumulate capital to be able to purchase a home. As well, the amount of time spent in Canada may be an important determinant of one's ability to establish a good credit rating to procure a mortgage.

Homeownership also increases with the age of the principal household maintainer, which is consistent with our expectations. Among the youngest cohort (25-34), only 38 percent own their homes compared with more than two-thirds for the oldest cohorts. While age can be seen as representing the effect of life time accumulation of wealth, it may also reflect a stage in the life-cycle where a home might or might not be needed. There is also a substantial difference in homeownership by familial structure. Ownership rates are highest among married respondents and lowest among single respondents. Further, while those with higher levels of education have higher rates of ownership, the relationship is weak. For instance, 58 percent of those with non university education are home owners, while 60 percent of those with some university education own their own homes. In contrast, income has one of strongest effects on homeownership. The ownership rate is 85 percent for those with income greater than $60,000 compared with only 38 percent of those with incomes below $20,000.

As presented in Table 1, the ethnic groups vary considerably by demographic and socio-economic factors. Since no control variables are introduced into the bivariate analysis discussed above, we are unable to assess the net effects of ethnicity and immigration status on homeownership. The statistical analysis that follows in explores if and how ethnic differences in homeownership can be explained by the life-cycle and sociodemographic characteristics. Four binary logistic models are estimated and tested against the results obtained in Model 1. Model 2 assesses the net impact of ethnicity in the presence of the immigration status variable, and vice-versa, while Model 3 examines the effects of all the variables. In Model 4, we test for the interaction between ethnicity and immigration status. The results are presented Table 3.

Model 1 indicates that ethnicity has a strong effect on homeownership. The odds of owning a home is highest among the Chinese and lowest among Blacks. Compared with non-visible minorities, principal household maintainers who are Chinese are 80 percent more likely to own their home. Once the effect of immigration status is controlled in Model 2, the risk of owning a home increases to 3.25 for the Chinese. With the exception of South Asians who have become just as likely as the non-visible population to own a home, the ethnic hierarchy remains the same as Model 1. The effects of immigration status are also significant and consistent with theoretical expectations. Compared with Canadian born, the odds of owning a home are considerably higher for immigrants who arrived before 1981, but significantly lower among those who arrived after that date.

In Model 3, significant ethnic differences remain after controlling for the theoretically relevant variables. This is consistent with previous Canadian studies (Balakrishnan and Wu 1992; Darden and Kamel 2000; Skaburski 1996). With the exception of sex, the effects of the sociodemographic variables are significant and consistent with our theoretical expectations. As Model 4 indicates, the interaction between ethnicity and immigration status is statistically significant, suggesting that the effect of ethnicity on homeownership depends on immigration status, and vice versa. In order to make these results more easily interpretable, the estimates for the interaction coefficients are converted into predicted probabilities, shown in Figure 1.

[FIGURE 1 OMITTED]

Some noteworthy patterns in Figure 1 are worth addressing. Perhaps the most obvious is that the Chinese have the highest probability of owning a home when compared with the other groups. The gap in predicted probabilities becomes larger when Chinese immigrants are compared with other immigrants of the more recent cohorts. The greatest gap is among immigrants who entered Canada between the years 1991-1996. Chinese immigrants who arrived within this period have a predicted probability of 0.67, whereas all other immigrants during this period have a predicted probability that is below 0.30. The non-visible group generally have the second highest probability of owning a home. (7) A somewhat surprising finding is that, among Canadian born ethnic groups, South Asian's have the lowest predicted probability of owning a home (0.20). Their predicted probability is much lower than that of Blacks (0.38) and other visible minorities (0.41).

When comparing only immigrants, Blacks have the lowest predicted probability of owning a home, which holds for each immigrant cohort. However, in comparative terms, and with the exception of Chinese immigrants, the difference between Black immigrants and those of other ethnic groups generally declines for immigrants of more recent cohorts. In fact, the gap in predicted probabilities between Blacks and the other ethnic groups is smallest for immigrants of the most recent cohort (1991-1996). With regard to trends by the immigration status variable, all Canadian born ethnic groups are less likely to own a home than are their counterparts who immigrated to Canada prior to 1981. In fact, with the exception of the Chinese, all other ethnic groups are most likely to own a home if they immigrated to Canada before 1981 than if they immigrated anytime later. When looking only at immigrants, there appears to be a general pattern, whereby those who immigrated during an earlier period have a higher probability of owning a home than those who immigrated later)

A Tobit Model Of The Value Of Owned Home

As discussed earlier, ownership of a home is only one indicator of housing wealth. In the following analysis, we advance our understanding of ethnic stratification by investigating differences in the value of owned homes. As was the case with the homeownership models, four models are presented: Model 1 estimates the total effect of ethnicity; Model 2 assesses the net effect of ethnicity after controlling for immigration status, and Model 3 examines the effects of all the theoretically relevant variables. In Model 4, we test for the interaction between ethnicity and immigration status. The results are presented Table 4.

The estimates for the ethnicity variable in Model 1 show a clear hierarchical pattern, whereby the Chinese report owing the highest valued homes. They are followed by non-visible minorities, South Asians, and other visible minorities. Blacks are at the bottom of the distribution. The difference between each of the ethnicity categories and the reference category, non-visible minorities, is statistically significant. The immigration status variable, added in Model 2, is also statistically significant. The estimates for this variable suggest that, when controlling for ethnicity, those who immigrated to Canada prior to 1971 report owning the most expensive homes. Next are those who immigrated between 1971 and 1980, followed by those who are Canadian born. Somewhat surprisingly, those who immigrated between 1981 and 1990 report owning a home that is of lower value than those who immigrated later (1991-1996). The difference between immigrants of the 1971-1980 cohort and the reference category, Canadian born respondents, is statistically significant. Interestingly, when controlling for immigration status, the estimate for Chinese respondents increased, relative to the estimates of all of the other ethnic groups. Across models, the statistical significance level of ethnicity remained unchanged.

As Model 4 indicates, there is a significant interaction effect between ethnicity and immigration status, suggesting that their effects on housing value are dependent on one another. As with the logistic regression results, the estimates for the variables involved in the interaction term are plotted in Figure 2. Similar to the models for homeownership, Chinese respondents are generally at, or near the top, of the housing value hierarchy. The only exception is among those who are Canadian born. The Canadian born Chinese respondents have an average housing cost of $201,188. That is slightly below non-visible minorities ($205,000), and other visible minorities ($207,434). Canadian born South Asians report an average house value of $215,294, which is the highest of all ethnic groups in this category. Conversely, Blacks report an average housing value of $179,431, which is the lowest of those who are Canadian born.

When comparing the immigration status trends, South Asians who immigrated before 1971 report owning a home that is less than their Canadian born counterparts. All other ethnic groups report a higher house values if they immigrated to Canada before 1971, than if they are Canadian born. As expected, all ethnic groups report that they own a less expensive house if they immigrated between 1971 and 1980, than if they immigrated before 1971. As well, all ethnic groups who immigrated to Canada between 1981 and 1990 report owning a more expensive home than their ethnic counterparts who immigrated between 1971 and 1980. The ethnicity trends in house values are mixed when comparing immigrants of the 1981-1990 cohort with those of the 1991-1996 cohort. South Asians and non-visible minorities of the latter cohort reported owning a less expensive home than their ethnic immigrant counterparts of the previous cohort. Conversely, other visible minorities of the 1991-1996 cohort experienced a moderate increase of just over $6,000 when compared to the previous cohort. Blacks reported owning a home of approximately the same value across the two most recent cohorts.

The most surprising and obvious finding in the figure is that Chinese immigrants of the 1991-1996 cohort show the most clear and dramatic increase in the values of their homes when compared with Chinese immigrants of the previous (1981-1990) cohort. The most recent Chinese immigrants report an average house value of $226,833, while Chinese immigrants of the 1981-1990 cohort report owning a home with an average value of $197,983, a difference of nearly $29,000.

Summary and Conclusion

The primary objective of this study was to examine ethnic differences in housing wealth, assessed through tenure choice and the value of housing. We found considerable ethnic differences in both indicators of housing wealth, and further identified that the effects of ethnicity and immigration status on housing wealth are not mutually exclusive. A key purpose of this study was to investigate whether ethnic inequality in housing wealth in Canada is similar in terms of both homeownership and the value of housing. For the most part, we found that homeownership and value of housing patterns are quite consistent, such that the ethnic hierarchical ranking among various immigration status groups is generally the same for both indicators of housing wealth. For example, the Chinese and non-visible minorities are generally at the top of the hierarchy for every immigration status category, while, with few exceptions, blacks are at the bottom. The results pertaining to blacks mirror findings in the United States (e.g., Long and Caudil 1992; Horton and Thomas 1998); Horton 1992) and may serve to reinforce perpetuating stereotypes. However, the results pertaining to the Chinese call into question the often made generalization that homeownership levels and housing values of visible minorities are low.

Interestingly, we also found that there is an increasing gap between Chinese immigrants and other immigrants in homeownership rates over the last two cohorts, and in the value of housing for the last cohort (1991-1996). We expect that this finding might be explained by the fact that most of the Chinese immigrants who arrived during this period were successful Hong Kong professionals who were uncertain about their future as the British prepared to hand over the political reigns of the island to the Chinese (Fong and Ooka 2002). Most of this class of immigrants had substantial financial capital and could therefore easily afford very good housing upon arrival. Nonetheless, we think it is an issue needs to be further explored in the future.

Another interesting result warrants further attention. Quite surprisingly, we found that Canadian-born South Asians are the least likely of all ethnic groups to own a home, but, at the same time, own the most expensive homes. Perhaps this discrepancy might be explained by the fact that cultural factors in origin countries coupled with a greater need for social recognition in the host country may compel many South Asian immigrants to become home owners. However, their children born in Canada may not share similar aspirations and values since they may regard themselves as being part and parcel of the Canadian society. Future investigators might also want to examine this finding more thoroughly.

Perhaps most importantly, our results clearly show that ethnicity remains a strong determinant of housing wealth, even after controlling for immigration status and the sociodemographic and life-cycle variables. The findings from this study provide substantial support for the spatial assimilation model; however, the support is not unqualified. Recall that spatial assimilation theory implies that ethnic differences in housing wealth should dissipate or disappear when differences in life-cycle and sociodemographic attributes are taken into account. However, despite significant effects of immigration status, age, education, marital status, household size, and income, ethnicity remained a significant predictor of ownership and house value. The results of this study are consistent with existing literature which suggests that spatial assimilation theory cannot, on its own, explain ethnic variations in homeownership, as ethnic groups with similar life-cycle and sociodemographic characteristics have significantly different levels of housing wealth.

The place stratification hypothesis may be useful for explaining the continued effect of ethnicity, which asserts that ethnic stratification is largely attributable to discriminatory practices in the housing market. This conclusion has been made by many researchers who have investigated ethnic differences in housing wealth in both Canada and the United States (Krivo and Kauffman 2004; Alba and Logan 1992; Darden and Kamel 2000; Horton and Thomas 1998; Flippen 2001 a, 2001b; Balakrishnan and Wu 1992). These researchers have argued that both institutional and individual discrimination contribute to ethnic/racial disparities in ownership and the value of housing. Surveys and audits have documented extensively the ways in which the practices of real estate brokers and mortgage lenders result in residential segregation among visible minorities in Canada and the U.S. (Parcel 1982; Yinger 1995; Munnell et al., 1996; Chandra 1973; Quann 1979). Shapiro (2004:121) explains the importance of the association between residential segregation and housing wealth:</p> <pre> [R]esidential segregation costs African American [and other visible minoroty] homeowners enormous amounts of money by suppressing their home equity in comparison to that of white homeowners. The inescapable corollary is that residential

segregation benefits white homeowners with greater home equity wealth accumulation. </pre> <p>Labour market discrimination has also been identified as a key factor in wage differentials between ethnic groups, which might also account for much of the ethnic group differences in housing wealth that persist after controlling for human capital and demographic characteristics. Unfortunately, the data did not allow us to tap into discrimination in our statistical analysis and we therefore could not assess the extent to which the remaining differences are attributable to discriminatory practices. However, on the basis of qualitative work done by others, there is evidence that numerous barriers to equal housing opportunities exist in Toronto, particularly for women and visible minorities (see, for example, Hulchanski 1997; Murdie 1999; Murdie et al., 1996). In qualitative studies, Murdie and colleagues (1996) observed a strong feeling among Jamaicans that racial discrimination is rampant in all sectors of the housing market. The views expressed by focus group participants and reproduced in Murdie at al. (1996) are worth repeating:

"my sister [black person] dealt with an agent and when they were looking for housing in a particular area, the agent said to her--'this is a nice area, not many black people live here' ..it was shocking!"

An alternative explanation for the persistent effects of ethnicity is offered by proponents of the cultural model, who claim that ethnic differences in homeownership can be attributed to cultural values that various ethnic groups place on homeownership (Ray and Moore 1991; Balakrishnan and Wu 1992; Mormino 1986; Sarre 1986; Zucchi 1988). However, limitations of the Canadian census did not permit testing of this theory. (9) Nonetheless, cultural norms and values may be important determinants of homeownership and hence must be taken into consideration when looking into ethnic differences in housing wealth.

Despite the prevailing effects of ethnicity, the significance of human capital and life-cycle factors must not be underestimated. In addition to income and age, immigration status stands out as a particularly important predictor of homeownership and value. Overall, the longer one has been in Canada, the greater is one's housing wealth. This finding is strongly supportive of the assimilation perspective, which hypothesizes that ownership comes along with cultural and economic integration; over time, as immigrants gain well-paying employment, English or French language skills, and knowledge of the local housing market, ownership will soon follow. The findings of this study suggest that, in general, immigrants to Canada can indeed expect to be homeowners, though not immediately upon arrival. Rather, aquisition of housing wealth, in terms of both ownership and value, occurs over time. The reasons for the lag between time at arrival and entrance into the housing market are worthy of further invesitigation.

The relatively low levels of housing wealth among blacks are of particular concern, and we believe that there is a need to explore alternative explanations for these results. While the literature on this issue is far from complete, there is evidence in the United States that down payment constraints have a substantial impact on homeownership among blacks, whereas whites are more likely to receive greater financial support from their families when purchasing a home (Duca and Rosenthal 1994; Linneman and Wachter 1989). As well, given that about 90 percent of the black population in Toronto is made up of immigrants, there is the need to explore migration motives (permanent or transient) and whether such motives contribute to their lower homeownership rates. The need to examine migration motives for black immigrants is further highlighted by recent studies on residential and homeownership patterns of some African immigrants in Toronto. Research on Ghanaian immigrants in Toronto, for instance, has shown that the majority regard their stay in Canada as temporary and prefer owning a home in their native Ghana (see, Gyimah 2001 ; Owusu 1998). Whether this is the case among other black immigrant groups is not yet understood. With respect to the quality of housing, we also cannot rule out the possibility that certain ethnic groups own less expensive homes because they self-select themselves into poorer and lower quality neighbourhoods in order to gain access to affordable owner-occupied housing. Explanations for ethnic stratification in the housing market is an important direction for future research in Canada, given the significance of housing wealth for inheritence and the intergenerational persistence of inequality.

Acknowledgments

* Kelli L. Phythian's contribution was partially funded by a SSHRC standard research grant on the economic integration of immigrants.

References

Adams, John S. 1987. Housing Americans in the 1980's. New York, NY: Russell Sage Foundation.

Alba, Richard and John Logan. 1992. Assimilation And Stratification In The Homeownership Patterns Of Racial And Ethnic Groups. International Migration Review 26:1314-1341.

Alba, Richard and John Logan. 1993. Minority Proximity To Whites In The Suburbs: An Individual Level Analysis Segregation. American Journal of Sociology 98:1388-1427.

Balakrishnan, T. R. 1988. Immigration And The Changing Ethnic Mosaic Of Canadian Cities. Report for the Review of Demography and Its Implications for Economic and Social Policy. Ottawa, ON: Health and Welfare Canada.

Balakrishnan, T.R. and Zhen Wu. 1992. Homeownership Patterns And Ethnicity In Selected Canadian Cities. Canadian Journal of Sociology 17(4):389403.

Beaujot, R., K. G. Basavarajappa, and Ravi B.P. Verma. 1988. Income of Immigrants in Canada. Ottawa, ON: Minister of Supply and Services Canada.

Breen, Richard. 1996. Regression Models: Censored, Sample Selected, or Truncated Data. Thousand Oaks, CA: Sage.

Brereton, B. 1989. Society And Culture In The Caribbean: The British And French West Indies, 1987-1980. In The Modern Caribbean, ed. F. Knight and C.A. Palmer. London, UK: The University of North Carolina Press.

Chandra, K. 1973. Racial Discrimination in Housing: Asian Minorities. San Francisco, CA: R and E Research Associates

Charles, Camille Zubrinsky. 2003. The Dynamics Of Racial Residential Segregation. Annual Review of Sociology 29:167-207.

Conley, Dalton. 1999. Being Black, Living In The Red: Race, Wealth, And Social Policy In America. Berkeley, CA: University of California Press.

Darden, Joe J. and Sameh M Kamel. 2000. Black And White Differences In Homeownership Rates In The Toronto Census Metropolitan Area: Does Race Matter? Review of Black Political Economy 28(2):53-76. Darko, K. 1993. Race and Homeownership in Toronto. M.A. Thesis, Queen's University, Kingston, Ontario, Canada.

Duka, J and S. Rosenthal. 1994. Borrowing Constraints And Access To Owner-occupied Housing. Regional Science and Urban Economics 24:310-322.

Flippen, Chenoa A. 2001 a. Racial And Ethnic Inequality In Homeownership And Housing Equity. The Sociological Quarterly 42(2): 121-149.

Flippen, Chenoa A. 2001b. Residential Segregation and Minority Homeownership. Social Science Research 30:337-362.

Fong, Eric and Emi Ooka. 2002. The Social Consequences Of Participating In The Ethnic Economy. International Migration Review 36 (1): 125-146.

Fong, Eric and Rima Wilkes. 1999. The Social Consequences Of Participating In The Ethnic Economy. International Migration Review 36(1): 125-146.

Fox Gotham, Kevin. 1998. Race, Mortgage Lending And Loan Rejections In A U.s. City. Sociological Focus 31(4):391-405.

Gyimah, Monica. 2001. The Residential Patterns Of Ghanaian Immigrants In Toronto. M.A. Thesis, Faculty of Graduate Studies, University of Western Ontario, London, Ontario, Canada.

Herberg, E.N. 1990. The Ethno-racial Socio-economic Hierarchy in Canada: Theory and Analysis of the New Vertical Mosaic. International Journal of Comparative Sociology 31:206-221.

Horton, Hayward Derrick and Melvin E. Thomas. 1998. Race, Class, and Family Structure: Differences in Housing Values for Black and White Homeowners. Sociological Inquiry 68(1): 114-136.

Horton, Hayward Derrick. 1992. Race and Wealth: A Demographic Analysis of Black Homeownership. Sociological Inquiry 62(4):480-489.

Hulckanski, J. 1997. Immigrants and Access to Housing: How Welcome Are Newcomers To Canada?" Keynote Presentation, Housing and Neighbourhoods Workshop, Metropolis Year II conference, Montreal, November 1997. http://www.library.utoronto.ca/hnc/publish/keynote.pdf

Hou, F. and T. R. Balakrishnan. 1996. The Integration of Visible Minorities in Contemporary Canadian Society. Canadian Journal of Sociology 21 (3):307-326.

Kalbach, Warren, E. and Madeline A. Richard. 1988. Ethnic-religious Identity, Acculturation, and Social and Economic Achievement of Canada's Post-war Minority Populations. Report for the Review of Demography and Its Implications for Economic and Social Policy. Ottawa, ON: Health and Welfare Canada.

Krivo, Lauren J. 1995. Immigrant Characteristics and Hispanic-anglo Housing Inequality. Demography 32(4):599-615.

Krivo, Lauren J. and Robert L. Kaufman. 2004. Housing and Wealth Inequality. Racial-ethnic Differences in Home Equity in the United States. Demography 41(3):585-605.

Li, Peter S. 1988. Ethnic Inequality in a Class Society. Toronto, ON: Thompson.

Linneman, P. and S. Wachter. 1989. The Impact of Borrowing Constraints on Homeownership. AREUFA Journal 17:L389-402.

Logan, John R. and Harvey L. Molotch. 1987. Urban Fortunes: The Political Economy of Place. Berkeley, CA: University of California Press.

Long, James E. and Steven B. Caudill. 1992. Racial Differences in Homeownership and Housing Wealth, 1970-1986. Economic Inquiry 30:83-100.

Mormino, Gary. 1986. Immigration on the Hill. Urbana, IL: University of Illinois Press.

Masnick, George S. 2001. Homeownership Trends and Racial Inequality in the United States in the 20th Century. Joint Center for Housing Studies W01-4, Harvard University.

Massey, Douglas. 1985. Ethnic Residential Segregation: A Theoretical Synthesis and empirical review. Sociology and Social Research 69:315-350.

Massey, Douglas and Nancy Denton. 1993. American Apartheid: Segregation and the Making of the Underclass. Cambridge, MA: Harvard University Press.

Massey, Douglas and Eric Fong. 1990. Segregation And Neighbourhood Quality Of Blacks, Hispanics, And Asians In The San Francisco Metropolitan Area. Social Forces 69:15-32.

Maxim, Paul. 1994. Self-employment Vs. Wage Labour and the Impact of Ascribed/Achieved Characteristics on Income Among Immigrants To Canada. In Studies in Applied Demography, ed. K.V. Rao and J.W. Wicks, 333-343. Proceedings of the International Applied Demography Conference.

Maxim, Paul. 1992. Immigrants, Visible Minorities, and Self-employment. Demography 29:181-198.

Munnell, Alicia H., Lynn E. Browne, James McEneaney, and Geoffrey M.B. Tootell. 1996. Mortgage Lending In Boston: Interpreting Hmda Data. American Economic Review 86(1):25-53.

Murdie, Robert A. and Carlos Teixeira. 2003. Towards A Comfortable Neighbourhood And Appropriate Housing: Immigrant Experiences In Toronto. In The Worm in a City, ed. Paul Anisef and Michael Lanphier, 132-191. Toronto, ON: University of Toronto Press.

Murdie, R. 1999. Housing Careers Of Polish And Somali Newcomers In Toronto Rental Market. A paper presented at the Fourth International Metropolis Conference, Washington DC. http://www.library.utoronto.ca/hnc/ publish/careers.pdf

Murdie, R.A., A. Chambon, J.D. Hulchanski, and J.C. Teixeira. 1996. Housing Issues Facing Immigrants And Refugees In Greater Toronto: Initial Findings From The Jamaican, Polish And Somali Communities. In Housing Question of the "Others," ed. Ermine M. Komut, 179-190. Ankara, Turkey: Chamber of Architects of Turkey.

Myers, Samuel L. Jr. and Chanjin Chung. 1996. Racial Differences In Homeownership And Home Equity Among Preretirement-aged Households. The Gerontologist 36(3):350-361.

Myles, John and Feng Hou. 2004. Changing Colours: Spatial Assimilation And New Racial Minority Immigrants. Canadian Journal of Sociology 29(1):29-58.

Oliver, Melvin L. and Thomas M. Shapiro. 1995. Black Wealth/White Wealth. A New Perspective on Racial Inequality. New York, NY: Routledge.

Owusu, Thomas Y. 1998. To Buy Or Not To Buy: Determinants Of Homeownership Among Ghanaian Immigrants In Toronto. The Canadian Geographer 42(1):40-52.

Parcel, Toby L. 1982. Wealth Accumulation of Black and White Men: The Case Of Housing Equity. Social Problems 30(2): 199-211.

Pendakur, Krishna and Ravi Pendakur. 1998. The Colour Of Money: Earnings Differentials Across Ethnic Groups In Canada. Canadian Journal of Economics 31 (3):518-54

Quann, D. 1979. Racial Discrimination in Housing. Ottawa, ON: Canadian Council on Social Development.

Ray, Brian K. and Eric Moore. 1991. Access To Homeownership Among Immigrant Groups In Canada. Canadian Review of Sociology and Anthropology 28:1-29.

Rosenbaum, Emily. 1992. Race And Ethnicity In Housing: Turnover In New York City, 1978-1987. Demography 29(3):467-486.

Rosenbaum, Emily. 1996. Racial/Ethnic Differences In Homeownership And Housing Quality, 1991. Social Problems 43(4):403-426.

Rosenbaum, Emily and Samantha Freidman. 2001. Differences In The Locational Attainment Of Immigrants And Native-born Housholds With Children In New York City. Demography 38(3):337-348.

Ross, Stephen, and John Yinger. 2002. The Color of Credit: Mortgage Discrimination, Research Methodology, and Fair-Lending Enforcement. Cambridge, MA: The MIT Press.

Sarre, Philip. 1986. Choice And Constraint In Ethnic Minority Housing: A Structuralist View. Housing Studies 1:71-86.

Saunders, Peter, R. 1990. A Nation of Home Owners. London, UK: Unwin Hyman.

Shamai, S. 1992. Ethnicity And Educational Achievement In Canada, 1914-1981. Canadian Ethnic Studies 24:43-51.

Shapiro, Thomas M. 2004. The Hidden Cost of Being African American: How Wealth Perpetuates Inquality. New York, NY: Oxford University Press.

Skaburskis, Andrejs. 1996. Race and Tenure in Toronto. Urban Studies 33(2): 223-252.

Statistics Canada. 2003. Canada's Ethnocultural Portrait: The Changing Mosaic. 2001 Census Analysis series. Catalogue 96f0030xie2001008.

Tobit, J. 1958. Estimation of relationships for limited dependent variables. Econometrica 26:24-36.

Warner, Richard A. 2003. Entry Class and the Earnings Attainment of Immigrants to Canada, 1980-1995. Canadian Public Policy XXIX(1):53-72.

Yinger, John. 1995. Closed Doors, Opportunities Lost: The Continuing Costs of Housing Discrimination. New York, NY: Russell Sage Foundation.

Zucchi, John. 1988. Italians in Toronto: Development of a National Identity. Montreal, QC: McGill-Queen's Press.

Notes

(1) The main source of information used by Statistics Canada to identify visible minorities is the ethnic origin question available in the Canadian Census together with supplementary information on cultural variables such as place of birth, mother tongue and religion.

(2) According to Statistics Canada, the 2001 public use micro files will be available until the later part of 2004.

(3) This normally refers to the person who contributes the greatest amount toward the payments for shelter expenses.

(4) Mostly due to inter-marriage, multiple reporting on ethnicity has become especially common among early European immigrants to Canada such as British, French, German, Dutch, etc. Also, changes to the ethnic origin question in the 1996 census resulted in a number of people reporting "Canadian" as ethnic origin. However, both the incidence of multiple reporting and 'Canadian" are much less among recent immigrants who are mostly visible minorities.

(5) While we recognize the heterogeneity within each of these groups, the categories are maintained to be consistent with previous work (e.g., Ray and Moore 1991; Darden and Kamel 2000; Skaburskis 1996).

(6) This should be interpreted with caution since among Blacks of Caribbean origin, consensual unions are very common.

(7) The only exception is South Asian immigrants who entered Canada between 1971 and 1980. Their predicted probability of owning home (0.70) is slightly higher than that of non-visible minority immigrants of the period (0.68). However, this difference is not statistically significant

(8) Two exceptions are Chinese and South Asian immigrants of the 1971-1980 period. Their predicted probability improved slightly when compared with immigrants of the previous cohort (before 1971).

(9) The ethnicity variable used in the current study does not lend itself to making inferences about cultural norms and values surrounding homeownership, given that each category consists of diverse cultural groups. The "black" category, for instance, includes persons from various African countries and the Carribean and it therefore cannot be presumed that those who fall into this category share similar norms and values.

Stephen Obeng Gyimah

Department of Sociology

Queen's University

David Walters

Department of Sociology and Anthropology

University of Guelph

* Kelli L. Phythian, PhD candidate

Department of Sociology

The University of Western Ontario

Table 1: Percent Distribution of variables by ethnicity, Toronto

 South
 Black Asian Chinese

Age group
25-34 33.5 24.5 18.7
35-44 29.9 36.2 37.2
45-54 24.2 26.3 28.1
55-64 12.4 13.0 16.0
Sex
Female 52.7 19.7 24.1
Male 47.3 80.3 75.9
Marital status
divorced/separated 27.3 8.7 7.5
single 32.6 8.5 14.6
widowed 2.0 2.4 2.4
married 38.0 80.4 75.5
Household size
1-2 persons 39.4 45.9 48.6
3-4 persons 43.0 20.7 25.0
5+ 17.6 33.4 26.4
Education
High school of lower 34.3 32.7 35.2
other non univ/trade 44.8 27.6 22.3
some univ+ 20.9 39.6 42.5
Total income in 1995 ($)
under 20,000 42.8 35.5 40.2
20,000-29,000 22.0 19.9 15.6
30,000-39,000 17.8 18.2 14.4
40,000-49,000 9.1 9.3 10.8
50,000-59,000 4.5 6.5 8.1
60,000 and above 3.8 10.6 10.9
Immigration status
Canadian bom 10.5 3.4 6.6
Immmigrant, before 1971 16.0 8.5 10.6
Immmigrant, 1971-1980 31.4 28.9 24.0
Immmigrant, 1981-1990 25.2 31.9 32.4
Immmigrant, 1991-1996 16.83 27.41 26.43
Sample size 2264 2043 2224

 Non Total
 visible population
Age group
25-34 24.6 25.2
35-44 30.3 31.7
45-54 26.2 26.0
55-64 19.0 17.2
Sex
Female 33.5 33.2
Male 66.5 66.8
Marital status
divorced/separated 17.1 16.4
single 20.4 20.0
widowed 3.1 2.9
married 59.4 60.7
Household size
1-2 persons 46.3 42.3
3-4 persons 41.5 41.6
5+ 12.2 16.0
Education
High school of lower 32.8 32.9
other non univ/trade 30.2 30.1
some univ+ 36.9 37.0
Total income in 1995 ($)
under 20,000 22.8 27.8
20,000-29,000 13.4 15.1
30,000-39,000 18.0 17.7
40,000-49,000 14.1 12.8
50,000-59,000 10.7 9.4
60,000 and above 21.0 17.2
Immigration status
Canadian bom 65.8 49.2
Immmigrant, before 1971 19.4 16.8
Immmigrant, 1971-1980 6.7 12.6
Immmigrant, 1981-1990 5.1 12.4
Immmigrant, 1991-1996 3.16 9.00
Sample size 23350 32848

Table 2: Home ownership in Toronto by selected characteristics, 1996

 Sample Percent
 size owned Chi-square Tests

Ethnicity
Black 2264 28.09
South Asian 2043 51.69
Chinese 2224 74.73 chi-square= 1740;
Other visible minorities 2828 38.22 df=4; p<0.000
Non visible 23350 62.17
Immigration status
Canadian born 16164 58.25
Immmigrant, before 1971 5517 78.00
Immmigrant, 1971-1980 4132 63.75 chi-square=2429;
Immmigrant, 1981-1990 4080 45.22 df=4; p<0.000
Immmigrant, 1991-1996 2955 26.87
Age group
25-34 8274 37.66
35-44 10399 56.94 chi-square=2305;
45-54 8535 68.38 df=3; p<0.000
55-64 5640 73.01
Sex
Female 10913 44.73 chi-square= 1147;
Male 21935 64.33 df=1; p<0.000
Marital status
divorced/separated 5397 36.93
single 6567 27.91 chi-square= 5279;
widowed 949 63.86 df=3; p<0.000
married 19935 73.03
Household size
1-2 13673 41.95
3-4 13910 67.57 chi-square=2467;
5+ 5265 73.24 df=2; p<0.000
Education
High school of lower 10816 54.37
other non univ.trade 9893 58.13 chi-square=92;
some univ+ 12139 60.62 df=2; p<0.000
Total income in 1995 ($)
under 20,000 9138 37.62
20,000-29,000 4949 46.25
30,000-39,000 5801 56.28 chi-square=4096;
40,000-49,000 4212 67.24 df=5; p<0.000
50,000-59,000 3089 76.72
60,000 and above 5659 84.77

Table 3: A Logistic Regression Model of Home Ownership in Toronto,
1996.

 Model 1 Model 2 Model 3 Model 4
 Exp (b) Exp (b) Exp (b) Exp (b)

Constant 1.64 *** 1.44 *** 1.86 **# 1.90 ***
Ethncity *** *** *** ***
 Black 0.24 *** 0.25 *** 0.37 *** 0.33 ***
 South Asian 0.65 *** 1.01 0.77 *** 0.13 ***
 Chinese 1.80 *** 3.22 *** 3.36 *** 1.23
 Other Visible 0.38 *** 0.53 0.51 *** 0.37 ***
 Non Visible (reference) 1.00 1.00 1.00 1.00
Immigration status *** *** ***
 Before 1971 2.76 *** 1.80 *** 1.78 ***
 1971-1980 1.60 *** 1.23 *** 1.13
 1981-1990 0.64 *** 0.53 *** 0.48 ***
 1991-1996 0.25 *** 0.23 *** 0.17 ***
Canadian Born (reference) 1.00 1.00 1.00
Sex *
 Female 0.92 0.93 *
 Male (reference) 1.00 1.00
Age Group *** ***
 34-44 1.51 *** 1.51 ***
 45-54 2.15 *** 2.11 ***
 55-64 3.35 *** 3.27 ***
25-34 (reference) 1.00 1.00
Education *** ***
 Non univ/trade 1.37 *** 1.37 ***
 Some univ/more 128 *** 1.28 ***
 High school/less
 (reference) 1.00 1.00
Marital status *** ***
 Divorced/sep 0.25 *** 0.25 ***
 Single 0.24 *** 0.24 ***
 Widowed 0.69 *** 0.70 ***
 Married (reference) 1.00 1.00
Household Size *** ***
 1-2 persons 0.48 *** 0.48 ***
 > 5 persons 1.49 1.48 ***
 3-4 persons (reference) 1.00 1.00
Total Income 1995 ($) *** ***
 Below $20,000 0.50 *** 0.50 ***
 $20,000-$29,000 0.68 *** 0.68 ***
 $40,000-$49,000 1.51 *** 1.51 ***
 $50,000-$59,000 2.27 *** 2.27 ***
 >$60,000 3.25 *** 3.24 ***
 $30,000-$39,000 (reference) 1.00 1.00
Ethnicity*Immigration
 status ***
 Black & < 71 1.37
 Black & 71-80 1.02
 Black & 81-90 1.34
 Black & 91-96 1.64
 South & < 71 4.45 ***
 South & 71-80 8.53 ***
 South & 81-90 6.82 ***
 South & 91-96 6.73 ***
 Chinese & < 71 1.19
 Chinese & 71-80 2.08 **
 Chinese & 81-90 3.43 ***
 Chinese & 91-96 5.16 ***
 Other & < 71 1.66
 Other & 71-80 1.78 **
 Other & 81-90 1.35
 Other & 91-96 1.62 *
Log likelihood 1748 4109 12967 13104
Prob>chi-square 0.000 0.000 0.000 0.000

Statistical significance: *** p<0.000; ** p<0.01; * p<0.05.

Table 4: A Tobit model of house value, Toronto, 1996.

 Model 1 Model 2

Constant 217435 214061
Ethncity *** ***
 Black -34439 *** -33472 ***
 South Asian -17586 *** -10165 ***
 Chinese 6349 ** 16077 ***
 Other Visible -21656 *** -15753 ***
 Non Visible (reference) -- --
Immigration status ***
 Before 1971 15772 ***
 1971-1980 4256 *
 1981-1990 -17818 ***
 1991-1996 -14270 ***
 Canadian Born (reference) --
Sex
 Female
 Male (reference)
Age Group
 34-44
 45-54
 55-64
 25-34 (reference)
Education
 Non univ/trade
 Some univ/more
 High school/less (reference)
Marital status
 Divorced/sep
 Single
 Widowed
 Married (reference)
Household Size
 1-2 persons
 > 5 persons
 3-4 persons (reference)
Total Income 1995 ($)
 Below $20,000
 $20,000-$29,000
 $40,000-$49,000
 $50,000-$59,000
 $60,000
 $30,000-$39,000 (reference)
Ethnicity*Immigration status
 Black & < 71
 Black & 71-80
 Black & 81-90
 Black & 91-96
 South & < 71
 South & 71-80
 South & 81-90
 South & 91-96
 Chinese & < 71
 Chinese & 71-80
 Chinese & 81-90
 Chinese & 91-96
 Other & < 71
 Other & 71-80
 Other & 81-90
 Other & 91-96

 Model 3 Model 4

Constant 189263 189261
Ethncity *** **
 Black -24400 *** -25569 **
 South Asian -17575 *** 10294
 Chinese 12559 *** -3812
 Other Visible -19381 *** 2434
 Non Visible (reference) -- --
Immigration status *** ***
 Before 1971 10705 *** 10446 ***
 1971-1980 1832 5330
 1981-1990 -17028 *** -12457 ***
 1991-1996 -10757 *** -28916 ***
 Canadian Born (reference) -- --
Sex
 Female -1245 -979
 Male (reference) -- --
Age Group *** ***
 34-44 10624 ** 10429 ***
 45-54 22483 *** 21809 ***
 55-64 21726 *** 21038 ***
 25-34 (reference) -- --
Education *** ***
 Non univ/trade 4444 ** 4443 **
 Some univ/more 24153 *** 24440 ***
 High school/less (reference) -- --
Marital status *** ***
 Divorced/sep -26990 *** -26919 ***
 Single -28360 *** -28442 ***
 Widowed -9701 ** -9444 **
 Married (reference) -- --
Household Size *** ***
 1-2 persons -15282 *** -15044 ***
 > 5 persons 10492 *** 10835 ***
 3-4 persons (reference) -- --
Total Income 1995 ($) *** ***
 Below $20,000 2589 2062
 $20,000-$29,000 3264 3294
 $40,000-$49,000 3679 3398
 $50,000-$59,000 10818 *** 10486 ***
 $60,000 33154 *** 32717 ***
 $30,000-$39,000 (reference) -- --
Ethnicity*Immigration status ***
 Black & < 71 -2040
 Black & 71-80 1050
 Black & 81-90 -1361
 Black & 91-96 15300
 South & < 71 -25815
 South & 71-80 -26987
 South & 81-90 -32434
 South & 91-96 -26262
 Chinese & < 71 14934
 Chinese & 71-80 4245
 Chinese & 81-90 9251
 Chinese & 91-96 54561 ***
 Other & < 71 -1060
 Other & 71-80 -30642 ***
 Other & 81-90 -30918 **
 Other & 91-96 -7766

Statistical significance: *** p<0.000; ** p<0.01; * p<0.05.