Regardless of education or training, racialized immigrant women earn less
May 21, 2021
Conventional wisdom suggests that first-generation immigrants often struggle in their careers for reasons related to being new to the domestic workforce.
For example, offshore degrees, poor language fluency or an absence of networking contacts are cited as common reasons for career challenges. This means most efforts to improve immigrants’ career outcomes often focus on things first-generation immigrants can do for themselves, like networking and training.
We wanted to know whether this is actually the best approach. Our evidence about which groups get paid the most versus which earn the least points to an alternative that may be even more effective: reducing workplace discrimination based on gender and race.
Who earns the most and the least?
To find out why immigrants receive lower pay, we looked at who earns more or less as a result of their combination of immigrant generation, gender, race and native language. We also wanted to find out whether being a first-generation immigrant is really the most important factor in terms of salary. (Spoiler alert: it’s not, according to our findings.)
We compared the annual pay for a sample of 20,000 employees within 6,000 firms that represent Canada’s workforce. Canada is a particularly good test case for examining immigrants’ outcomes because of its high proportion of first- or second-generation immigrants (37.5 per cent of the population).
We took an “intersectional” approach to this research. That means we looked at all 24 possible combinations of these four characteristics, rather than measuring how much each characteristic impacts pay on its own:
· Immigrant generation: First-generation, referring to those born abroad; descendants, referring to their Canadian-born children and grandchildren; and non-immigrants, referring to those who are not recently descended from immigrants
· Gender: Men and women
· Race: Self-identified as a person of colour, or not
· Language: Whether the person works in their native language or in an additional language.
We recognize that these are all rough categories that ignore important complexities. For example, because of data limitations, we couldn’t test for different outcomes between racial groups or across the full gender spectrum.
Despite this lack of nuance, rough categories can still be useful as a starting point for making big-picture comparisons between groups.
Finally, we compared “apples-to-apples” by controlling for five individual characteristics (age, experience, education, occupation and unionization) plus four characteristics of their workplaces (size, industry, performance and international competition). That means we aren’t comparing entrepreneurial housekeepers in one group against corporate data analysts in another.