Background
The Tigger and Raja Tiger Balm Company (TRTBC) recruited and selected almost 1,000 employees at the new Los Gatos, California warehouse. The Human Resources (HR) team has analyzed whether adverse impact by sex or race/ethnicity occurs during the TRTBC hiring process. Adverse impact happens when members of a protected class (sex, race/ethnicity, age, etc.) are selected at a lower rate than the number applied. The 4/5thrule states that less than 80 percent selection ratio is deemed substantial enough to quantify that adverse impact is occurring. While this is not illegal per se, it can be when hiring processes include assessment tools that are not related to Bona Fide Occupational Qualifications (BFOQ).
Methodology
Using the Equal Employment Opportunity demographic data collected during the application process, the selection rate of each protected class (sex, race/ethnicity) was identified for both job roles – the Warehouse Associate and Front-Line Supervisor. The selection rate was calculated by dividing the hired applicants by the total applicants. After selection rates were determined by sex, the selection ratio was calculated by dividing the lower rate by the higher rate. Since analyzing race and ethnicity involves multiple groups, a cutoff rate must be identified by multiplying the highest selection rate by 0.80. Adverse impact occurs on any selection rates below the cutoff rate.
Findings
Adverse Impact by Sex
Before examining hired employees against applicants, we reviewed demographic data by job role to identify trends. Below is a chart showing the number of employees by sex and job role.
We see a significantly higher number of male warehouse workers compared to females. On the other hand, front-line supervisors are relatively even between male and female workers. Why is that? How does that correspond to the number of applicants between the two job roles?
Below are two tables, one for each job role, comparing the numbers of male and female applicants to the number hired.
| 1 Warehouse Associate | ||||
| Applied | Hired | Selection Rate | Selection Ratio | |
| Male | 800 | 621 | 0.776 | |
| Female | 195 | 117 | 0.600 | |
| 77.29% | ||||
| 2 Front Line Supervisor | ||||
| Applied | Hired | Selection Rate | Selection Ratio | |
| Male | 543 | 132 | 0.243 | |
| Female | 326 | 96 | 0.294 | |
| 82.55% |
We are seeing a higher number of male applicants than women for both positions. However, when comparing selection rates by sex for warehouse associates, there is evidence of adverse impact against female applicants. A female applicant has about a 77 % chance of being hired compared to a male applicant. In contrast, female applicants are being selected for front-line supervisors at a higher rate than men (29% vs. 24%). These rates equal a selection ratio that favors women, but since 82 percent passes the 4/5th rule, adverse impact is not occurring.
Adverse Impact by Race/Ethnicity
Below is a chart showing the number of employees by race/ethnicity and job role
For warehouse workers, the groups with significant populations include, in order of highest to lowest, Asian, Hispanic/Latino, White, and Black/African American. For front-line supervisors, the groups with significant populations include, in order of highest to lowest, White, Hispanic/Latino, Asian, and Black/African American. How does that correspond to the number of applicants between the two job roles?
Below are two tables, one for each job role, comparing the numbers of applicants by race/ethnicity to the number hired.
| 1 Warehouse Associate | |||||
| Applied | Hired | Selection Rate | Adverse Impact | Selection Ratio | |
| American Indian/Alaskan Native | 11 | 8 | 0.727 | NO | |
| Asian | 338 | 219 | 0.648 | YES | 74.49% |
| Black or African American | 210 | 140 | 0.667 | YES | 76.65% |
| Hispanic or Latino | 215 | 187 | 0.870 | 0.696 | |
| Native Hawaiian or Other Pacific Islander | 14 | 10 | 0.714 | NO | |
| White | 195 | 164 | 0.841 | NO | |
| Two or More Races | 12 | 10 | 0.833 | NO | |
| 2 Front Line Supervisor | |||||
| Applied | Hired | Selection Rate | Adverse Impact | Selection Ratio | |
| American Indian/Alaskan Native | 6 | 2 | 0.333 | 0.267 | |
| Asian | 197 | 54 | 0.274 | NO | |
| Black or African American | 103 | 30 | 0.291 | NO | |
| Hispanic or Latino | 307 | 58 | 0.189 | YES | 56.68% |
| Native Hawaiian or Other Pacific Islander | 13 | 4 | 0.308 | NO | |
| White | 237 | 78 | 0.329 | NO | |
| Two or More Races | 6 | 2 | 0.333 | 0.267 |
We first identified the cut-off rate for adverse impact analysis by race/ethnicity by calculating the highest selection rate for each job role. For warehouse associates, the benchmarked group was Hispanics/Latinos at 87%, and for front-line supervisors, it was American Indian/Alaskan Native and Two or More Races at 33%. Using the cut-off rates, we can state that groups that are selected at a ratio of less than 80 percent compared to the benchmarked groups are being disparately impacted.
At the warehouse associate level, there is an adverse impact against Asian and Black applicants, such that Asian and Black applicants had approximately a 75 and 77 % chance, respectively, of being hired compared to a Hispanic/Latino applicant. The other races and ethnicities are not being adversely impacted.
At the front-line supervisor levels, there is an adverse impact against Hispanic/Latino applicants, such that Hispanic/Latino applicants had around a 57% chance of being hired compared to a Native American or an applicant of two or more races. The other races and ethnicities are not being adversely impacted.
Recommendations
Adverse Impact by Sex
Firstly, there are opportunities for improvement in the recruitment of applicants at TRBC. The warehouse associate and front-line supervisor job announcements yielded close to 1,000 applicants. However, male applicants comprised 80% of the warehouse applicants, while only 60% were front-line supervisor applicants. We recommend evaluating the job announcement of the warehouse associate to ensure that male-gendered language is avoided and that the job requirements are bona fide occupational qualifications.
Secondly, we should evaluate the screening and interview process to ensure that assessment methodologies are linked to bona fide occupational qualifications. Because adverse impact is occurring against women at the warehouse associate level, there may be selection criteria that are disproportionately favoring men.
Adverse Impact by Race/Ethnicity
Interestingly, Hispanics/Latinos are the demographic group that is selected at the highest rate for warehouse associates. Conversely, Hispanic/Latinos are the only demographic group disparately impacted at the front-line supervisor level. It appears that Hispanics/Latinos are being typecast into the warehouse role.
While American Indians and people of two or more races are being hired at the highest rate for front-line supervisors, the sample size of these groups is small. For large sample-sized groups, White applicants are being hired at the highest rate. This trend follows robust research on White candidates and incumbents being hired and promoted for manager-level positions at a much higher rate than people of color.
We recommend implementing hiring practices that anonymize candidates by hiding names and addresses. This practice should eliminate bias at the resume-screening level. Additionally, reviewing interview rubrics and training managers on scoring rubrics fairly would reduce bias at the interview level.
