* We are grateful to Derby Voon for research assistance.
THE IMMIGRANT HOUSING MARKET: ANALYSES FOR AUSTRALIA
This paper examines the immigrant adjustment process in Australia from the perspective of the housing market. It shows that immigrant “catch-up” to the native born in the housing market is much more rapid than in the labour market. A decomposition of the estimated coefficients of a logit model of tenure choice is developed that gives formal recognition to the immigrant adjustment process. The results from this decomposition demonstrate the importance of taking account of immigrant adjustment when seeking to understand variations in rates of home ownership across birthplace groups.
THE IMMIGRANT HOUSING MARKET: ANALYSES FOR AUSTRALIA
The immigrant adjustment process has been the focus for a large number of studies of labour market outcomes in Australia (see Miller and Neo (2003)). Other studies have widened the scope of enquiry into immigrant adjustment by considering participation in political and social matters (Kelley and McAllister (1984)), proficiency in English (Chiswick and Miller (1995)) and immigrants’ investments in the housing market (Kee (1992), Bourassa (1993)(1995)). The latter line of research compares the home ownership rates of immigrants and the native born, and examines how the native born advantage in this regard varies by the length of residence of immigrants in Australia. Convergence of the rate of home ownership of the foreign born to that of the native born has been proposed as an index of economic and social adjustment (Chiswick and Miller (2003)).
In recent years there has been concern over the impact of rising housing prices on the ability of the young to enter the housing market (Mudd, Tesfaghioris and Bray (2001)). As recently arrived immigrants share many of the experiences of young adults—they are entrants into the Australian labour market and generally have little wealth—it is likely that these housing pressures on the young will carry across to the immigrant experience, and impact on the immigrant adjustment process. This matter has been unexplored to date, and is the focus of the current study. Hence the paper uses 2001 Census data to examine home ownership patterns among immigrants, and compares these with the home ownership patterns among the native born. Chiswick’s (1978)(1979) theory of immigrant adjustment is the underlying model applied to explore the factors that affect the home ownership decision and the immigrant adjustment process in Australia.
Data analysis is carried out on the three groups distinguished in much of the recent literature on the labour market outcomes of immigrants, namely the Australian born, those born abroad in English-speaking countries and those born abroad in non-English-speaking countries. Analyses are also conducted for a number of separate birthplace groups (Germany, Italy, Greece, Philippines, Vietnam, China). Three main issues are addressed. The first is whether home ownership rates differ across birthplace groups. The second issue concerns the determinants of immigrants’ home tenure choice: are these the same as for the native born? The third issue explored is whether findings for Australia are consistent with those reported for the United States.
The paper is organised as follows. Section II reviews literature that is pertinent to the immigrant adjustment process, presents Chiswick’s (1978)(1979) theory of immigrant adjustment, and shows how this can be applied to the housing market. The third and fourth sections present the data and methodologies used, and discuss results from a logit model of the determinants of house ownership. The fifth section develops a method for categorising the differences between birthplace groups as being due to the different endowments of the groups and due to behavioural or other differences between the groups. The final section summarises the major empirical findings and implications of the results.
II. LITERATURE REVIEW
The immigrant adjustment literature seeks to assess whether new immigrants are able to assimilate themselves in the destination country. It has utilised proxies such as earnings, the employment rate and language skills to capture immigrant adjustment effects. The studies have found that immediately after arrival immigrants experience lower income and higher unemployment rates than the Australian born, though this disadvantage is smaller for those from English-speaking countries. As the length of residence increases, the immigrant experience becomes similar to that of the Australian born (Chapman and Miller (1986)).
The use of home tenure choice adds an additional dimension to the previous studies, and allows for a more comprehensive understanding of the immigrant experience. The studies that have explored issues associated with the home tenure choices of immigrants in Australia have examined the relative importance of various economic and demographic variables, including household income, occupation, age, marital status, language proficiency, household size and the number of dependent children (Kee (1992), Bourassa (1993)(1995) and Junankar et al. (1993)). The typical model utilised has the following form:
Probability of owning a home = f(Birthplace, Length of Residence, Educational Attainment, Family Income, Marital Status, English Proficiency, Occupation, Age, Family Status). (1)
Junankar et al. (1993) report that housing tenure choice is related to immigrant’s country of birth, with immigrants from New Zealand having a higher likelihood of renting, presumably owing to high rates of return migration for this group. Similarly, Kee (1992) reports that immigrants from Italy, Malta, Greece and Yugoslavia have above average rates of home ownership. Bourassa (1995), however, found that after controlling for differences such as duration of residence in Australia and marital status, seven out of the ten immigrant groups1 studied in Sydney and Melbourne had similar home tenure choices as the Australian born. The other three immigrant groups (from Italy, Lebanon and Malta) actually had higher home ownership rates than the Australian born. This could be due to behavioural differences of the three immigrant groups, or other factors that affected the home ownership decision that were not captured in the model specifications. Despite this apparent anomaly, Bourassa (1995) concluded that, in general, immigrants’ home purchasing behaviour is not too dissimilar to that of the Australian born.
The length of time immigrants have lived in Australia has been shown in a number of studies to have an important bearing on their housing tenure choice. Junanker et al. (1993) argue that immigrants appear to go through a tenure cycle, which begins with renting or shared accommodation, and then owning their own home. The probability of home ownership for immigrants is relatively low in the immediate post-arrival period as, being new to the housing market, they may not possess the collateral needed to secure a housing loan.2 There may also be instances of discrimination. Junankar et al. (1993) cite evidence of immigrants being particularly subject to such exploitation because of language barriers and lack of knowledge about industry practices.
With time in Australia, however, immigrants acquire the resources necessary to switch from renting to owning. They will also accumulate the knowledge that will enable them to make better-informed decisions. Consequently, rates of home ownership among longer-term residents even often surpass those of the Australian born. At the same time, it would be expected that rates of home ownership will vary across immigrant groups owing to different levels of resources and different expectations. For example, some foreign born may feel that buying a house is not a viable long term investment.3 The patterns of change in immigrants’ housing tenure choice with duration of residence in Australia appear to be similar to those reported for the US by Coulson (1999) and Chiswick and Miller (2003). Coulson (1999, p.214) remarks that as the immigrant status dissipates over a period of time, the probability of home ownership in the US increases. Similarly, using home ownership as an indicator of immigration adjustment, Chiswick and Miller (2003) found that it took just several decades of living in the United States before immigrants have the same rates of home ownership as the native born.
These observations have similarities with the patterns of immigrant adjustment reported in the labour economics literature concerning earnings, employment and the language skills of immigrants. As such, it is likely that the focal model of immigrant adjustment proposed by Chiswick (1978), that has been used to good effect in the labour economics literature, may be able to guide the study of the immigrant housing market, and a brief outline is provided below.
Chiswick (1978)(1979) proposed a simple but powerful model to explain the economic progress of immigrants relative to the native born. Specifically, he distinguished between “economic” and “non-economic” immigrants. All immigrants are assumed to make decisions based on maximisation of their economic well-being. The strict distinction between the two is that economic immigrants are motivated by monetary returns, rather than by social and political factors. He further categorised them into three sub-groups; economic immigrants from English-speaking countries, economic immigrants from other countries, and refugees. Consistent patterns of earnings are expected across the three sub-groups, as illustrated in Figure 1.
Figure 1 captures the general patterns of earnings determination that have been reported in Australia and for many other countries. In particular, immigrants earn less than the native born in the first year of residence in Australia, though the extent of this disadvantage varies across immigrant groups, being greatest for refugees and least for economic immigrants from English-speaking countries. This disadvantage is argued to reflect the less-than-perfect international transferability of the skills immigrants acquired in their country of origin.
The post-arrival improvement in the earnings of immigrants reflects their investments in skills relevant to the Australian labour market. These investments may not only be valuable in their own right, but they could also increase the international transferability of immigrants’ stock of pre-immigration human capital. For example, learning English in the post-arrival period may enhance the value of both the immigrants’ pre-arrival formal schooling and the skills they acquire through labour market experience.
The initial earnings disadvantage experienced by the immigrants is expected to vanish and their earnings may even exceed those of the native born as their duration of residence increases. This effect could be associated with the self-selection process in migration. Hence, Chiswick (1979) argued that immigrants often have a higher level of innate ability and work motivation than their fellow countrymen with similar characteristics who remain at home. In the United States, it was found that, other things being equal, foreign born white men had earnings comparable to those of the native born within 13 years of immigration (Chiswick (1978)).
Chiswick’s model of the economic progress of immigrants appears to be well suited to account for the typical observations of immigrant outcomes in the housing market. It suggests that attention should be placed on the way rates of home ownership vary across groups of immigrants that differ in the international transferability of their pre-immigration experiences. Attention should also be placed on how rates of home ownership change with duration of residence. The analyses of immigrant housing tenure choices in Australia that follow focus on these issues.