*The following was written by Dana Britton, faculty co-director of the Rutgers AAUP-AFT Salary Equity Program and Professor and Chair of Labor Studies and Employment Relations.*

There are several steps in the pay equity process as it has been implemented by the administration. **This is an explanation; it is in no way an endorsement**. But knowledge is power, so here we go:

**1. **Applicants to the salary equity program choose salary comparators. In our example, an Assistant Professor in History at Camden chooses one comparator: an Assistant Professor in History at New Brunswick.

**2.** The Dean must approve the comparators or submit a revised set. In our example, let’s assume the Dean accepts the one comparator.

**3.** University Compensation Services (UCS) takes the applicant’s salary and SUBTRACTS the average salary of all approved comparators. This is the *initial gap*. For now, let’s assume our hypothetical Assistant Professor in Camden makes a base salary of $90,000, and their one comparator makes a base salary of $105,000. This gives us an initial gap of $15,000.

**4.** At this point, the regression analysis comes into play. The regression is the result of an analysis of the University’s faculty salary database as of fall 2019, conducted by a consultant from Jackson Lewis, an external management law firm hired by the University. The regression uses four factors to produce a predicted calendar year (CY) salary for every faculty member in the University. These are:

**Rank**: There are nine clusters, from Assistant Professor NTT to Professor II TT..**Position Title**: Also nine clusters, but a much broader range that includes specific titles such as “associate extension specialist,” “teaching instructor,” and Assistant Professor. The predicted CY salary for an Assistant Professor is $122,917.66. This is the “constant” in the equation. More on this later.**Discipline**: This is the factor that affects most people. There are ten clusters in all. These are not disciplines, per se; they are a combination of campus and department. For example, if you are in English at Camden, your “discipline” is FASC-English, and you are in discipline cluster 4. If you are in English at New Brunswick, your “discipline” is SAS-English, and you are in discipline cluster 1. The predicted CY salary of someone in cluster 1 is $122,917.66 (again, the constant), but if you are in cluster 4, your predicted salary is $107,944.83. Essentially, this means the predicted effect of teaching English in Camden versus teaching English in New Brunswick is a CY salary that is lower by $14,972.84.**Pay plan:**Four clusters in all: Cluster 1 has titles from Teaching Instructor to Distinguished Professor (we have no idea why), Cluster 2 is all Assistant Professors, Cluster 3 is Assistant Research Professors, and Cluster 4 is Associate Professors of Law.

**5.** If you fall into the same clusters as your comparators for all four factors, the regression analysis will have no effect. The letter you receive from UCS will say: Gap = ($15,000), Explained = $0, Unexplained = ($15,000), and your recommended salary award will be $15,000. This situation is most likely for applicants who choose comparators from their own department and campus.

**6.** If you don’t fall into the same clusters as your comparators, things are more complicated. What happens then is that your cluster values are used to predict your salary and your comparators’ clusters are used to predict theirs. UCS takes your predicted salary minus the average of comparators’ predicted salaries and that becomes the “explained gap.” Basically, this is an estimate of how much of a salary gap “should” exist, based on the four factors in the regression. This can come out in your favor. But in some cases, the analysis finds that though you are earning less than your comparators, you are actually overpaid relative to your predicted salary (or they are underpaid). In this case, UCS will give you no salary adjustment.

**SIMPLIFIED EXAMPLE**

**7.** Let’s return to the example of an applicant who is an assistant professor in History in Camden. They have one comparator, who is an assistant professor in History at New Brunswick. They vary on only one cluster: discipline. The applicant cluster numbers are 2,1,3,2, and the comparator cluster numbers are 2,1,2,2. Working out the values in the equation (see the technical example below) gives a predicted CY salary for the applicant of $105,306 and a predicted CY salary for the comparator of $119,968. To make this easier, convert back to AY salaries by dividing by 1.15. This gives a predicted AY salary for the applicant of $91,570.44 and a predicted AY salary for the comparator of $104,320.

**8.** UCS takes the difference—$91,570.44 – $104,320 = ($12,749.56)—and calls this the “explained” gap, which you can see on UCS letters. Basically, this is saying that the regression equation predicts your comparator should be earning $12,749.56 more than you.

**9.** Now go back to step 3 above. The initial gap between our applicant and their comparator was $15,000. Since the regression suggests the comparator should only be making $12,749.56 more, the applicant gets the difference. A UCS letter to this applicant will say: Gap = ($15,000), Explained = ($12,749.56), Unexplained = ($2,250.44), and the recommended salary award will be $2,250.44. Add this to the $90,000 base that our applicant started with, and the final salary is $92,250.44.

**TECHNICAL EXPLANATION**

**10.** Predicted salaries are mathematically derived using coefficients generated from a regression on the University salary database. The equation produces a constant, and each cluster has its own set of coefficients, all expressed in the form of logarithms.

- To calculate the CY salary of our applicant, the equation is: 11.71927 + 0 + 0 + –0.0673583 + –0.0872813 = 11.5646304.
- 11.71927 is the constant. To find its value in dollars, take the exponent, which gives $122,917.66. Every other coefficient modifies the constant up or down.
- Assistant professor and assistant professor TT are the “excluded” categories, meaning their coefficients are 0.
- Discipline cluster 3 has a value of -0.0673583, meaning it lowers the value of the constant.
- Pay plan 2 has a coefficient value of -0.0872813, meaning it also lowers the value of the constant.
- To get predicted CY pay, take the exponent of the equation result, 11.5646304, which gives $105,306. Convert to AY by dividing by 1.15 = $91,570.44.
- Follow the same process for all comparators. Add their predicted salaries together and take the average.
- Subtract the result from the predicted salary of the applicant to get the “explained” value.
- Subtract this from the initial gap. Any balance is awarded to the applicant.
- Because of an oddity in the way UCS is using signs, a positive award to the applicant will be noted as a negative value. So in our case: Gap = ($15,000), Explained = ($12,749.56), Unexplained = ($2,250.34), award = $2,250.34. The formula to make this work is: Gap + (Explained * -1) = Unexplained. A negative unexplained amount is an award to the applicant. A positive amount means no award.

And that’s it. As we are in the process of making salary adjustments, the University will have to run this regression periodically, and the coefficients will change. Over time, this should have the effect of deepening inequalities, as the current salary structure is used again to calculate predicted salaries.