Intersectional Analysis in Quantitative Research: Analyzing Sex/Gender/Sexuality Based Discrimination Before and During Covid-19


Jesse Henstridge-Goudie, Memorial University

Research has shown the covid-19 pandemic had a disproportionate impact on LGBTQ2 individuals, females, and visible minorities in terms of health outcomes, job insecurity, and increased discrimination. Intersectionality, an approach that involves considering the unique experiences of those who exist on multiple axes of power structures and oppression, is a framework that is often considered in qualitative work, but less so in quantitative research. Considering this, i ask the following research questions: first, did discrimination on the basis of sex, sexual orientation, or gender identity increase during the covid-19 pandemic? And second, how did having multiple marginalized identities affect reported rates of discrimination? In answering these questions, i used the dataset “experiences of discrimination from statistics Canada’s crowdsourced survey “impacts of the covid-19 on Canadians” (icc), with data from 36674 individuals collected from august 4th to 24th, 2020, and focused on reported discrimination on the basis of sex, sexuality, or gender identity/expression. To operationalize intersectionality in this analysis, i created a variable called “intersect” with eight possible categories for each combination of factors of interest. Intersect was then used as a focal independent variable, with discrimination before and after covid-19 as dependent variables. I also controlled for respondents’ socio-economic and demographic characteristics. Descriptive (frequency tables, crosstabs) and multivariate analyses (binary logistic regression) were used. Results from the initial univariate analysis of the dependent variables showed a 13.7% decrease in the number of people who reported discrimination on the basis of sex, sexual orientation, or gender identity/expression during covid-19 compared to before covid-19. Bivariate analyses showed that before covid, 54.8% of people who identified as LGBTQ2 reported discrimination on the basis of sex/sexual orientation/gender identity compared to only 19.9% of straight people. During covid 27.6% of LGBTQ2 respondents reported discrimination and 7.9% of non LGBTQ2 reported experiencing discrimination. Before covid, a higher percentage of females reported discrimination; 29.7% compared to 9.7% of males. During covid this dropped to 12.3% of females and 5% of males. Further, visible minorities were less likely to report discrimination on this basis than those who were not visible minorities. However, this analysis did not address if the chance of experiencing discrimination was higher for individuals who fit into multiple of these categories at once, or whether being LGBTQ2 or female was a greater indicator of if someone was likely to experience discrimination. Therefore, i used the intersect variable which allowed me to do a more in-depth analysis of each of these categories by considering their intersections. Results showed that before and during covid-19, LGBTQ2 females who were not visible minorities reported the highest percentages and odds of experiencing discrimination. Compared to LGBTQ2 males who are visible minorities, non-LGBTQ2 females who are not visible minorities were 48% less likely to experience discrimination. Meanwhile LGBTQ females who are not visible minorities were 2.4 times more likely to experience discrimination. This clearly shows LGBTQ2 identity is the largest indicator of if someone was likely to report discrimination, and this likelihood was increased for LGBTQ2 females; with LGBTQ2, female, non-visible minorities as the most likely to report. These results go against what dominant literature and empirical studies tell us about how discrimination was experienced during the covid-19 pandemic. Possible reasons for this could be the timing of the survey (2020), as people had not yet felt the full ramifications of covid at this point. Focus on increased racial discrimination during this time and the rationalization of harassment may explain why visible minorities are less likely to report sex, sexuality, or gender-based discrimination. Further, this study demonstrates the methodological value of creating variables that capture multiple intersecting identities instead of those that attempt a siloed analysis of these marginalized identities. These results also show the need for increased research into this particular time period in Canada, and its effect on individuals with marginalized identities.

This paper will be presented at the following session: