- Dataset: In this article, we will be focusing on base salary analysis using IBM’s fictional data sets (HR Employee Attrition)[1]. Though this dataset includes terminated employees, I will only be using current employee data in order to replicate real-world compensation analysis.
2. Limitations: This analysis has some limitations because the dataset used is fictional. These limitations include:
a. Limited number of job levels and job titles: Originally the dataset included 5 job levels and 10 job titles. In order to make this dataset more realistic, I changed the number of job levels to 7 and the number of job titles to 55.
b. No pay differential: I didn’t consider pay differential factors including cost of living even though we are faced with this in the real world.
c. Compensation Model is based on Monthly income: An individual’s compensation is paid hourly or annually. Therefore, I changed each compensation line item to annual salary by applying a random variable to make a more realistic distribution of compensation levels based on the salary range within a given job level.
d. Data aging: Depending on the age of the dataset, we may need to account for inflation to make it comparable to real world figures. However, since I don’t know the age of the original data, inflation was not accounted for.
e. Lack of market data/ salary structure: In reality, we analyze internal and external equity using market data and internal salary structure. However, since the fictional dataset has no salary structure and market data, I simulated market data as well as individual salary data by multiplying random variable factors to the original dataset
3. Analysis to validate pay equity
a. What is pay equity?
Pay equity means fairness of compensation paid by employers to individuals or groups of employees. Pay equity is crucial for employers not only to attract and retain talent, but also to comply with labor regulations concerning pay discrimination.
b. Internal Pay Equity Analysis
i. Segmented Approach: For smaller organizations or individual departments at larger organizations, it may be helpful to distribute all employees along a given salary structure (see Exhibit 1). By taking this measure, we can narrow down the priory group and isolate employees who are paid under the midpoint of their respective job level. These employees are underpaid relative to other employees at the same job level. Before deciding whether or not an underpaid employee warrants a salary adjustment, review of additional factors may be necessary. These factors include the amount of time since employee’s last promotion, their performance appraisal rating, their level of experience, whether or not this employee manages any direct reports and whether this employee occupies an in-demand position.
[1] https://www.kaggle.com/pavansubhasht/ibm-hr-analytics-attrition-dataset
ii. Organizational approach: Larger organizations may benefit by conducting a Compa-ratio analysis by employee. Compensation metrics like the Compa-ratio helps firms analyze pay equity effectively because it allows segmentation of employees into groups relative to the salary midpoint. Compa-ratio is derived by dividing salary by the salary midpoint (P50) within a company’s salary structure. 100% represents the midpoint, or said another way, an employee who is at the 50th percentile of the range. If the compa-ratio falls between 85–110%, that is generally considered a fair compensation level. A compa-ratio falls below 85% indicates that the employee is underpaid relative to the salary midpoint. It is generally prudent to review these cases to understand the reason why. If the compa-ratio is above 130%, the employee in question is overpaid relative to the salary midpoint. Again, it is prudent to examine why in order to justify the higher level of compensation.
Key findings of Exhibit 2 are as follows:
· 27% of employees are underpaid (Compa-ratio: < 85%). There may be many reasons why this is the case including lack of experience or a lower education level. Additionally, employees in this category may not have the requisite skill level vis-à-vis other employees in their job level.
· 47.2% of employees are paid at fair rate (Compa-ratio: 85% — 110%). Be that as it may, 11.9% of these employees are paid at a slightly higher rate compared to the midpoint of their respective salary range.
· 19.3% of employees are paid a premium (Compa-ratio: 110% — 130%). This might be explained by the employee living in an area with a higher cost of living area or in-demand skill that warrants a premium compared to the jobs at the same level. Additionally, a higher level of compensation may be necessary to reward or retain top-performers. Just as was the case with underpaid employees, further review may be necessary in order to validate whether the increased compensation levels are justified.
· 6.5% of employees are paid significant premium (Compa-ratio > 130%). This might be explained by the employee living in an area with a higher cost of living area or in-demand skill that warrants a premium compared to the jobs at the same level.
c. External Pay Equity: External equity refers to the employee’s perception that he/she is compensated similarly to an employee at a comparable job level working for a competing firm. A key metric that employers need to become familiar with is the Market Index. The Market Index is the ratio of an employee’s base salary to the actual market rate for that job. It provides a measure of how employee pay compares to the market rate specific to their position. In this analysis, I used a fictional Market P50 (market median) figure to determine Market Index. I assumed that this data is not location specific (e.g. nationwide) and does not need to reflect a pay differential. Additionally, although we usually complete job matching on a 1:1 basis (e.g. between our employee’s job and jobs in the market), I will not cover this topic.
Key findings from Exhibit 3 are as follows:
· 36.6% of employees are underpaid compared to the market rate (Market Index < 85%). This could be explained by a number of factors including a lack of experience or education vis-à-vis other employees within that job level. Further review is likely required in order to mitigate employee dissatisfaction.
· 53.4 % of the employees are paid fairly (Market Index: 85% — 110%).
· 8.5% of the employees are paid a slight premium (Market Index: 110% — 130%). This could be attributed to the employee having more experience, a desirable skill or is geographically located in a high-cost of living area.
· 1.5% of the employees are paid a significant premium (Market Index: >130%). Although employers may need to pay higher compensation levels in a highly competitive job market or to retain top performers, further review may be required to justify the higher cost.
d. Comparison of Market Index vs Compa Ratio
As show in exhibit 4, the total number of employees with a compa-ratio below 110% falls below the market index range. On the other hand, the total number of employees with a compa-ratio above 130% well exceeds the Market Index. This could mean that the company’s salary midpoint is lower than the market midpoint. Additional analysis should be conducted.
