UCI Pay Equity Study Update, 1997-98

Methodology

The 1997-98 pay equity study at UCI was conducted for generalcampus ladder rank faculty based on October 1997 Payroll/Personneldata combined with data from the Office of Academic Personnel. Thisstudy applied the pay equity model recommended by theAmerican Association of UniversityProfessors (AAUP) to identify women and minority faculty memberswho appear to be underpaid. The Payroll/Personnel System snapshotused for this analysis included only ladder rank faculty members whowere on active pay status as of the end of October 1997. The onlyindividuals excluded from the analysis were faculty administratorswho were compensated on a fiscal-year basis instead of anacademic-year basis.

An ad hoc committee on pay equity methodology was assembled in theSpring of 1997 to assess the reliability of the model. Withmembership from across the campus and including representatives fromthe Office of Academic Affairs, the Committee on Academic Personnel,and the Faculty Women's Association, this committee determined thatthe model being used could be recommended as a screening device --although not as the only screening device -- for an initialassessment of equity in compensation at UCI.

The model is based on a least squares analysis (multivariatelinear regression) that estimates the adjusted academic-year(nine-month) salary for regular rank faculty members. The estimationequation describes the linear relationship between the predictors(independent variables) and the salaries (the dependent variable) forwhite male faculty members.

The estimation equation is included underneath each school'sgraph. Because the regression analysiswas conducted by academic unit, each unit's equation uses differentvalues to calculate estimated salaries. At UCI, we used the followingpredictors or independent variables:

• degree indicator: an indicator of the highest degree (0 = Baccalaureate or lower; 1 = Masters; 2 = Doctorate)
• degree year: the last two digits of the year the highest degree was awarded (e.g., 87)
• year of hire: the last two digits of the year the faculty member was hired at UCI (e.g., 90)
• year of birth: the last two digits of the year of birth (e.g., 60)

The next step in applying the AAUP model was to calculate theresidual for women and minority faculty members,with residual defined as the difference between actual salary andsalary predicted by the regression equation. A negative residualindicates that the salary is lower than the salaries of white malesin the same academic unit who have the sameattributes.

Graphs 1-11 display the negative andpositive residuals for women and minorities from each of the academicunits, as well as those for females and minorities for the generalcampus. Graph 12 displays thedistribution of residuals for white males, calculated separately foreach academic unit by comparing actual salaries with predictedsalaries for all ladder rank faculty -- white males, females, andminorities.

On each graph, the X-axis indicates the size of the residual. TheY-axis indicates the number of individuals. The "-" area indicatesnegative residuals and the "+" area positive ones. The standard errorfor each academic unit, in the footnote at the bottom of each page,may be used as a measure of the significance of the residual.

Please direct questions or comments about thecontent of this page to Associate Executive ViceChancellor Killackey.