Slicing and dicing the gender/racial earnings differentials

Abstract

Purpose -This paper aims to explore an extensive set of determinants of earnings and to offer recent empirical evidence of their effects on gender and racial earnings gaps. Design/methodology/approach -Most previous studies looked at gender and racial comparisons independently of each other. This study extends previous studies by considering the interaction between gender and race. Using administrative data from a large Canadian firm, this paper explores the determinants of earnings based on a standard human capital model, comparing the earnings of white females, minority males and minority females with their white male counterparts. Both the dummy variable approach and a decomposition analysis are employed. Findings -The results show that ranking in the organizational hierarchy accounts for most of the differences in gender and racial earnings, and ranking, together with human capital and job characteristics variables, explains over 90 percent of the earnings gap.Research limitations/implications -The analyses in the paper are based on data from a Canadian organization with nation-wide operations. The findings may not apply to small or medium sized enterprises in Canada and in other non-Western economies. Practical implications -To eliminate the earnings gap, equal pay programs need to be supplemented by effective employers' programs and policies targeted at equal advancement opportunity. Originality/value -The paper uses firm-level data, which provides natural controls for variations across firms and allows for more in-depth analysis of the impact of various factors on earnings differentials.Previous research on gender and racial differences in earnings in both the USA and Canada have generally relied on data from national household or labor force surveys. Although these surveys have large representative samples and are well designed, by their nature they focus on data collection of supply side variables (i.e. information on the employees) and rarely gather relevant information from the demand side (i.e. information on the employers). Researchers and policy makers, therefore, are limited to formulating recommendations based on information only about employees without consideration for the employers. Studies focusing on demand-side variables are needed to complement our extensive knowledge of the labor supply side. In Canada, the most The current issue and full text archive of this journal is available at

Pharmacists' earnings determinants: differences by ethnic group

“…Separate earnings functions were formulated and tested for the three ethnic groups using regression analysis. An alternative model, in which the ethnic effect would have been identified by one or more dummy variables, was discarded because of its probably incorrect assumption that the responses to covariates were equal for all pharmacists regardless of ethnic identification . In fact, differences in these responses, if they existed, constituted the crux of the empirical evidence pursued by this article.…”

“…An alternative model, in which the ethnic effect would have been identified by one or more dummy variables, was discarded because of its probably incorrect assumption that the responses to covariates were equal for all pharmacists regardless of ethnic identification. [47,48] In fact, differences in these responses, if they existed, constituted the crux of the empirical evidence pursued by this article. The three equations contained the same explanatory variables to compare the direction, magnitude and statistical significance of the influence of each set of coefficients on earnings in the presence of all other independent variables, thus avoiding differences that might be attributed to interaction effects among covariates.…”

Ethnicity as a mediator of pharmacists' labour supply

“…According to the empirical evidence, the percentage of male pharmacists was lower within blacks and Hispanics than within non‐Hispanic whites. The literature shows that gender differentiation, including earnings and division of labour, is heavily influenced by ethnic group . How society in general, and ethnic identification in particular, define acceptable patterns of female labour force participation, male commitment to housework activities and sharing responsibilities within the household/family unit vis‐à‐vis remuneration for work affect the role of gender within each group's labour supply function.…”

“…Most labour supply studies focus on workers across occupations . This practice introduces unaccounted stochastic disturbances caused by inter‐occupational differences in education, occupational prestige, relative scarcity, etc.…”

“…The three equations contained the same covariates for purposes of comparing the direction, magnitude and statistical significance of the influence of each set of coefficients on labour supplied in the presence of all other covariates, thus avoiding differences that might be attributed to interaction effects among covariates. An alternative model, in which the ethnic effect would have been identified by one or more dichotomous variables, was discarded because of its probably incorrect assumption that the responses to covariates were equal for all pharmacists regardless of ethnic identification . In fact, disparities in response, if they existed, constituted the crux of the empirical evidence pursued by this article.…”