Pay Equity Study
June 2019 - Results
The pay equity analysis focused on two aspects:
- Gender and race pay gap within the 绿奴天花板 System, and
- Whether there are any indication of systemic pay disparities between employees of different race or gender.
Statistical analyses identified potential pay differences between protected groups and other employees that are statistically significant. The analyses also determined whether the differences could be explained by a factor other than gender, race or age. Statistical analyses were performed in accordance with methods recognized by the Equal Employment Opportunity Commission (EEOC). To the extent that the statistical analysis identified potential pay differences, a deeper incumbent level analysis is being conducted. Once the review has been completed, if an inequity is identified employees will be contacted directly.
Where we are:
- A high-level review of compensation data is underway
- Working on regression analysis for staff, faculty, and executives
The statistical analysis includes reviewing the effect of the following elements on pay differentials:
- Gender
- Age
- Race/Ethnicity
- Faculty Rank
- Terminal Degree
- Tenure Track
- Years of Service at 绿奴天花板 (years since hire)
- Years in Position/Rank (years in current position or current faculty rank)
- Job Value (represented by Pay Grade Midpoint, or Market Median where available)
What we know:
- Consultants are identifying positions that need additional analysis
Next steps:
- Conduct a detailed incumbent level comparative analysis to validate potential pay disparities
- Identify areas to address first (departments, discipline areas, job titles)
- Develop controls to mitigate risk and ensure consistency moving forward
Regression Analysis Methodology
Gallagher's analysis will include all of the applicable variables to determine which have a significant impact on pay.
- Statistical significance for inclusion in the formula is defined as p < .05. This is the accepted level of statistical impact on the result.
- Variables that do not have a significant impact on pay will be identified and removed from the analysis until the best set of variables that impacts pay is identified.
- This analysis requires multiple 鈥渞uns鈥 of data to obtain the best set of variables that impact pay.
- We will also review the regression model R square to evaluate predictive strength of the regression model.
- R Square definition: the percentage of the response variable variation that is explained by the regression
model.
- R square value of 1.0 (or 100%) indicates that the model explains all variability of the response data.
- R square value of 0 (or 0%) indicates the model explains none of the variability of the response data.
- R Square definition: the percentage of the response variable variation that is explained by the regression
model.
- Exclusions
- 139 employees who chose not to self-disclose their ethnicity will be excluded from the regression analysis. However, employees with 鈥淥ther鈥 under the ethnicity category will be kept.
- Outliers
- While the regression analysis identifies standard deviations, since we are using multiple variables to 鈥榩redict鈥 the dependent variable, we cannot statistically exclude individual employees as 鈥榦utliers鈥.
- In order to identify systemic pay equity issues, we must use the full data set, including 鈥榦utliers鈥.
- Regression Approach
- Gallagher utilizes several regression models system wide, and groupings by campus/staff, both with and without market data as a variable.
- Regressions use hourly pay for comparison in order to account for differences in annual hours.
- Exception: When using market data (9/10 month), we will utilize current 9 month salary equivalent to run regression analysis