Determinants Of Annual Salary In Canada

Ordered Probit Regression Approach

Abstract

Several studies provided descriptive analyses of 2016 Canadian census data; however, there are rarely advanced econometrics studies showing the relationship between ranges of annual salary and several socioeconomic, institutional, and location (or mobility) variables in Canada. A sample of 12,228 was extracted from 2016 census data. The annual salary data were grouped into seven non-overlapping homogeneous categories as outcome variables to reduce very high heterogeneity in the observations. The Ordered Probit regression underpinned the covariate analysis to understand the influence of various independent variables on annual salary for full-time employees in Canada. The prediction probability for an annual salary ranging from CDN$28,000 to 77,999 is 65.6 percent while for the CDN78,000 to 127,999 range is 32.5 percent; implying that 98.1 percent of full-time employees who completed at least high school and were aged 18 to retirement age received between minimum annual salary CDN$28,000 and maximum CDN$127,999. The remaining 2 percent received the highest salary ranging from CDN$128,000 to 400,000 during the survey. The results indicate that all age groups, ethnicity except white, residential types, education levels, household size, marital status, and all the provinces (or location) variables significantly (p-value < 0.05) influenced annual earnings;  however, the magnitude and the direction (positively or negatively) of influence depends on the annual salary earning brackets.