![Sample our Global Development journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5929/TOC-Subject-Sample-2021-Global-Development.jpg"})
Abstract
The purpose of this article is to investigate one possible mechanism by which the Canadian labour markets adjust to immigration. Despite the fact that Canada is one of the major immigrant receiving countries in the world, most studies that look across Canadian local markets have found immigration’s effects to be weak. The well-known argument is that rising immigration levels in an area may result in the out-migration of an area’s residents if the immigrants displace the local workers in employment, bid down wages, or cause housing prices to rise through increased demand for shelter. The present study investigates this bias by estimating the mobility responses of local residents to immigrant inflows based on a spatial equilibrium model for 28 census metropolitan areas (CMA) from 2000 to 2009. The results show that immigration has a significant displacement effect between CMAs.
Acknowledgements
I thank Dr Charles Beach (Queens University) and Dr Barry R. Chiswick (George Washington University) for their valuable comments on an initial draft. I also thank my research assistant Maggie Jones (PhD student at Queens University) for her outstanding job, not only in extracting and organizing the data but also for her very handy comments on analysing the data. Responsibility for any shortcomings remains with me.
Notes
1. Two recent studies on housing markets (Akbari & Aydede, Citation2012; Ley & Tutchener, Citation2001) also indicate a weak negative linkage between immigration and local housing prices.
2. The list of included CMAs is given in Appendix 1.
3. An excellent literature review can be found in the Hou and Bourne (Citation2004) paper.
4. Saiz (Citation2007) uses a similar utility function to identify the relationship between immigration and local housing markets in the USA without an explicit disutility that local residents may receive from immigrants.
5. We assume that ρ > 0.
6. In the literature of urban economics, it is assumed that there exists a reservation locale that yields a constant level of utility to all consumers.
7. To incorporate moving costs, we can discount by μ; hence, the equilibrium condition becomes Uir = µ
where 0 ≤ µ ≤ 1 depending on the distance. Although discounting does not change the fundamental results, it implies that spatial differences and immigration have weaker effects on relocation decisions as the distance increases. This is taken care of by using a fixed effect model in the estimations.
8. From (3), it is obvious that an increase in R reduces N. It can also be shown that, at any given level of H (= h × P) and N, an increase in M raises R in the short run.
9. Note that (4) puts a constraint on the coefficient of (Mr−Mj) by assuming that the degree of substitutability (ε) and the disutility from immigration are identical across regions.
10. This is a similar framework as that applied by Hatton and Tani (Citation2005) for the UK.
11. See http://www5.statcan.gc.ca/cansim/a00?lang=eng&mode=aboutCansim for more details.
12. Tables 2820110 and 2820114 are extracted from the Labour Force Survey, Table 1110025 is extracted from administrative files, and Table 3270005 is extracted from the New Housing Price Index collected monthly, which gives a measure of changes in new housing prices relative to the base year 2007.
13. This is similar to Frey’s models that cover all people.
14. CANSIM Table 2820110 provides the distribution of occupations for each CMA based on National Occupational Classification for Statistics (NOC-S). We used the following categories to find the number of ‘high-skilled’ workers: C + (D0 – D1) + (E0, E2) + (E1, E130)/total employment. Please see http://dc2.chass.utoronto.ca/cgi-bin/cansimdim/c2_getArrayDim.pl for their definitions.
15. We specially thank Charles Beach at Queens University for bringing this to our attention.
16. House price indices are not available for six CMAs (Abbotsford, Kingston, Saguency, Sherbrooke, Trois-Riviere, Thunder Bay). Therefore, the number of observations decreases when Housing Price Index is included in the regressions.
17. We also tried different strategies to improve the test results. For example, the coefficient of interest gets bigger (0.0145) and the test results significantly improve when both Wage Earnings and Net Immigration Ratio are instrumented together with lagged values of Employment Growth Rate, LFPR, Net Immigration Ratio and Wage Earnings at time t – 1. Using lagged or contemporaneous values of included instruments does not make a significant difference.
18. See Glaeser (Citation2008) for more details.