According to the 2016 Census, 122,510 Canadians aged 25–54 were employed in a policy research occupation, a workforce characterized as predominantly female (61% female) (Table 2). Specifically, one in six (16%) worked in health policy research, a field characterized by more pronounced gender segregation (74% female). Of the eight political research occupations under observation, only the economic research workforce is dominated by men (44% female). The remaining professions under observation are 53-75% women.
All eight occupations are characterized by lower average annual earnings for women than for men, despite similarities in job duties and working conditions. Women in the health policy research workforce earned an average of 88 cents for every dollar earned by men (Table 2). In other occupations, the gender earnings ratio ranged from 72 to 91 cents on the dollar. Occupations in traditionally male-dominated sectors (particularly economics, natural and applied sciences and business development policy areas) tend to offer higher average pay levels than occupations in traditionally female-dominated sectors (including health, education, social and recreation policy areas). Higher-paid occupations are also characterized by wider gender earnings ratios (72-82 cents on the dollar) compared to their counterparts in traditionally female-dominated sectors (88-91 cents on the dollar).
Across occupations, a higher share of women is associated with lower average wages among women (r= − 0.68) (Fig. 1). The negative correlation of occupational feminization is even stronger with respect to the decline in average wages among men (r= − 0.80).
Health policy researchers were mainly employed in health and social care settings (women: 44%; men: 39%) and in public administration (women: 24%; men: 26%), although not exclusively (Fig. 2). A negligible number were engaged in educational services and in other scientific and technical services. In contrast, health and social care institutions engaged high numbers of recreation policy researchers (females: 17%; males: 6%) and social policy researchers (females: 11%; males: 6%). In other words, the boundaries of the health system were not easily delineated by any policy research area.
The age structure of the health policy research workforce differs little by gender (Table 3). Conversely, female policy researchers in the natural and applied sciences and in business development fields tend to be younger than males, i.e. more often in the 25–34 age group – a reflection of the feminization of sectors in which women have traditionally been underrepresented. In terms of other key labor market variables, female health policy researchers were characterized less often than men by having a university degree (58% vs. 61%) and more often by working part-time (11% vs. 8%) . In terms of social identity variables, women health policy researchers were significantly less likely than men to report being the primary breadwinner (50% vs. 72%), but were more likely to live in a household with children (53% vs. 48%). Women were also less likely than men to have been elderly migrants to Canada (15% vs. 26%).
A Bivariate Analysis of the Gender Wage Gap
Based on the simple linear regression model, female health policy researchers were found to have earned 9.0% (95% CI 5.1–12.7%; p< 0.05) less than men. This was the smallest (unadjusted) gender pay gap among the eight occupations observed, which otherwise ranged between 9.2% (among education policy researchers) and 23.9% (among business development policy researchers) (Fig. 3). Bivariate analysis of the political research workforce also confirmed a strong positive correlation between the degree of occupational feminization and the size of the gender pay gap (r= 0.76).
Multivariate and decomposition analyzes of the gender pay gap
Multivariable linear regression analysis confirmed evidence of a significant gender pay gap in the health policy research workforce, with women earning 4.8% (95% CI 1.5–8.0%) less by men, after adjustment for other work, social and housing characteristics (Table 4, model 5). Those in their early careers (ages 25–34) tend to earn less than their more established counterparts, ceteris paribus, and those who immigrated to the country in adulthood compared to their counterparts , who were born or had migrated in childhood or adolescence (ie before exposure to higher education and access to the labor market).
In non-health policy occupations, the gender pay gap is considered significant for five other fields: women’s earnings average from 4.0% less (among social policy researchers) to 12.3% less (among business development policy researchers) than men’s earnings (Table 4). No notable gender pay differences were found for policy researchers in government programs and in the fields of education, among which differences in raw wages were largely due to age, higher education, and senior migrant status.
In a regression model pooling all eight areas of policy research together, each of the seven female-dominated occupations was found to pay significantly less on average than economic policy research (i.e., the only female-dominated occupation men under observation), all else being equal (not shown) ). In particular, the mean annual salary among health policy researchers was 21.1% (95% CI 19.4–22.8%) lower than their counterparts in economic policy research. For economic policy researchers, salaries averaged 15.4% lower (among business development policy researchers) to 36.2% less (among recreation policy researchers). The overall gender pay gap is considered significant, with women’s mean earnings estimated to be 8.1% (95% CI 6.9–9.2%) lower than men’s, regardless of policy field or other professional occupation or personal characteristics.
Decomposition analysis showed that, as might be expected, gender differences in educational attainment and other traditional human capital variables accounted for a large proportion (27%) of the gender pay gap in the political research workforce (Table 5). . However, 15% of the wage gap is due to occupational differences, i.e. from the field of policy and program research, as distinct from other job characteristics. The gender pay gap is less pronounced in health policy research than in the field of (better paid) economic policy. Age differences between women and men account for 6% of the wage gap, and differences in social identity characteristics account for 10% of the difference. After decomposing the gender gap in occupational wages, a substantial 40% of the gap remains unexplained by the measured predictors.