The maps (shown below) and the
least squares models developed unearth significant characteristics of
recent immigrants. Firstly, all three cities show similar pattern
of
recent immigrant characteristics as regards education, low income
economic families, government transfer payments and persons without
income. While all three cities have their highest coefficient of
determination among the regression models as that for low income
economic families, both Montreal and Toronto have their lowest
coefficient of determination among the least squares models as that for
education level. Population of 15 years and more without income is the
second highest correlate with recent immigrant population for the three
cities. This is a tremendous trend that the social influences that
shape recent immigrant attributes do not change and recent immigrant
characteristics might be common in the three cities.
In each city, the suites of socio-economic variables are more
significant in their covariation with the incidence of low-income
economic families than the ethno cultural set. The incidence of
female-led families, male unemployment, and government transfer
payments supply the strongest correlations; interestingly, failure to
complete high school does not figure as prominently. It is not
difficult to establish causal arguments linking poverty and these
factors -- though presumably in most accounts the level of transfer
payments would feature as an effect rather than a cause of poverty
(Ley, D and Smith, H 1998).
Secondly, though all three cities exhibit similar patterns, the impact
of recent immigrant population would be more significant in different
cities. Impact of low income economic families for instance, rose from
1996 to 2001 in Montreal and Vancouver but decline in Toronto with
Vancouver experiencing the highest percentage change, almost double in
2001. Toronto revealed quite a consistent and a distinct trend with the
socio-economic variables chosen while Montreal and Vancouver are more
radical towards the variables. These relationships necessitate a
further statistical analysis to disclose how selected variables combine
in defining poverty in a census tract.