Declining Initial Earnings of Recent Immigrants in Canada by Admission Class, 1980-1995

Shiva Halli
Department of Sociology
University of Manitoba

An examination of the economic performance of different classes of immigrants has been hindered until very recently mainly because of a lack of appropriate data. For instance, the Census data, which were heavily used in immigration research, did not contain any information on the immigrant classes. With the availability of a new data set – the Longitudinal Immigration Database (IMDB) – this task is now possible. The present study has employed the IMDB to examine the income determination structures for different classes of immigrants.

The project derives information from a number of immigrant "profiles" that the research division of Citizenship and Immigration Canada (CIC) has been developing from information available from the IMDB. These profiles speak to the effects of different immigrant attributes on their subsequent labour market behaviour at the time immigrants land in Canada, which is derived from the information collected from their tax returns. Some of the descriptive information on earnings performance will then be framed out – demonstrating the relative worsening position of those who immigrated between 1980 and 1995 with regard to their earnings in the paid labour force. We will attempt to show that old success stories no longer correspond to the social reality of today for all immigrants.

Generally speaking, immigrants in the 1990s have not performed as well as previous arrivals in terms of their initial employment earnings. Immigrants landing in the early 1990s experience significantly lower initial earnings than that of previous cohorts of immigrants. The starting position for each immigrant cohort seems to be getting progressively worse. These outcomes occurred in spite of the fact that recent immigrants are better educated than were previous cohorts of immigrants.

Results by the class of immigrants show some variations. For instance, since 1980, economic principal applicants reported higher initial employment earnings than all other categories of immigrants regardless of the year of admission. They also show that they and the other categories of immigrants landed since 1989 have experienced lower employment earnings relative to the Canadian average than those who landed earlier.

As obvious as the worsening relative position is for newer arrivals’ cohorts, we also know that the immigrant flows themselves have changed between the early and late 1980s and into the 1990s. The composition of flows by category of immigration has changed, as have many other characteristic factors that have been shown to significantly affect earnings ability – such as education and language ability. In order to control for these characteristics along with others such as age, place of birth, marital status, sex, occupation, etc., we performed multiple regression analysis. Although most of the characteristics show expected relationships with the dependent variable, employment earnings, there are some interesting variations. However, what is more interesting is the changing nature of the income determination process. In other words, the same equation that has explained more income variations for earlier immigrants cohorts does not seem to explain the earnings of later arrivals, with the exception of economic immigrants. More importantly, the variables included only explain less than 15 percent of the variation, even for the economic class of immigrants. There seems to be something happening to the process of earnings determination, especially for more recent immigrant arrivals. It is clear that a framework and methodology need to be more fully developed to investigate the relative worsening lower market position for immigrants landed through the 1990s.