Compression / Inversion Report

Compression / Inversion Report

UFF Summary/Interpretation

In CBA 2007-2010“C & I student completed by Feb. 1, 2009”Started by Joint Committee. Completed by UFF-FGCU to make an offer for bargaining. No agreement. In CBA 2009-11 Supplement23.8 Compression and Inversion Study. Recognizing that the University and the UFF-FGCU Chapter did not jointly complete a Compression and Inversion (C&I) Study as outlined in the 2007-2010 Collective Bargaining Agreement, the parties acknowledge that the parties will commission an external consulting firm having a regional and/or national reputation and experience in performing compensation studies to conduct a C&I Study to inform bargaining.

C & I History

Salary Compression occurs when newly hired, or junior faculty members receive a rate of pay that approaches, or is approximately equal, to the rate paid to faculty of higher, or senior, professional rank.Salary Inversion arises when junior faculty members earn higher salaries than senior faculty.Within rank and between ranks.

What are Salary Compression and Inversion?

Study because individual faculty members saw newer faculty getting salaries the same as, or more than, experienced faculty members. Faculty members felt this was unfair -- Climate Study, turnover, lower morale, less willingness to serve the college, less willing to mentor new faculty, had completed an external study (Market Study) several years before and had discussed doing the internal study in the future at that time.Compression and inversion exists across the university. For example:

Why?

Examples of C&I: Actual Salaries

CIP

Coll

Subj

Rank

Yr

Hire

Salary

03

CAS

Ecology

Asst

09

49,980

Prof

02

88,900

09

CAS

Communication

Assoc

97

64,460

Prof

05

64,666

Prof

06

71,735

14

Eng

Engineering

Asst

06

70,754

Asst

10

75,000

Assoc

06

79,400

Assoc

08

96,900

Examples of C&I: Actual Salaries

CIP

Coll

Subject

Rank

Yr

Hire

Salary

22

CPS

Justice Studies

Asst

04

50,330

Asst

07

50,480

51

Hlth

Nursing

Asst

06

66,813

Physical

Therapy

Asst

08

70,086

Health Sciences

Assoc

96

68,290

52

COB

COB

Asst

09

105,060

Assoc

99

88,400

Prof

96

92,527

27

CAS

Science

Inst

1

03

35,800

Inst

I

09

36,720

Inst

II

06

38,990

Even though the C&I report states that there is no C&I in the university, except for Health and Business (H&B). This conclusion is WRONG—it needs to be understood in light of the statistics used.C&I is particularly large in H&B, but there is also C&I in the rest of the University.Let’s understand what the report does and does not say, especially in light of the statistics we needed to use.

Report Findings

The study provided an extra analysis requested by administration that compares our average JUMPS between ranks with other universities’ average jumpsThis is NOT a market survey of salary levels.Rank Ratio — compares average salaries between ranks and compares to other universities. Mean salary/rank by mean salary/ discipline. Time in rank not included.“Do our salary jumps between ranks look like those of other universities?” Yes, in all but 2 disciplines. Based on salary averages by rank

Methods

Average salaries by rank Instructor; Jr. Asst. Professor/less than 3 years; Asst. Professor; Assoc. Professor, Professor Progressively increase as faculty rise through ranksAverage jumps, not including seniorityThe jumps are similar to peer schools. However, FGCU has larger jumps between levels .Not true in:Heath Professions (Instructors higher than Asst.)Business (Assistants almost = Associates)

Rank Ratio Analysis

Predicts what salary “should be” What would salaries be if we paid everyone consistently, on the same basis.Taking into consideration: discipline, rank, years-at-rank, and market conditions at time of hire to predict “should be”There are variation within rank Develops different regression equations to predict salaries for different disciplines and different “ranks”There are 23 disciplines using CIP at 2-digit level“Ranks:” Instructor, Jr. Asst. Professor, Asst. Professor, Assoc. Professor, Professor. Also, Library, Advisors.Regression compared to rank-ratioConsidered in the C&I literature (and by the consulting firm) to a better method than rank-ratio.Looks at salaries internal to FGCU only, not compared to other universities.Not a market study, we have done two of those previously

Regression Analysis

Considered: 9 month salaries, CIP code for discipline, rank, years of service at rankMost of university paid in a similar mannerUniversity-wide rank and years at rankIdentified the university’s “policy” for paying peopleCreated a “line of best fit” for predicted salariesSome paid differentlyCIP 51 (mostly Health Professions) and CIP 52 (most of Business faculty) not paid on the same policy as rest of UniversityMultipliers differTwo groups in business identified as having high pay premium in equation (finance and accounting)Developed equations for calculation of predicted salaries for all individual faculty members

Regression

Dispersion around “Best Fit” for Most Disciplines

Salary

Time in

Rank

Assistant

Associate

NOTE: Observations “fit” the regression model, because they are close to the line. But some people are underpaid,

others overpaid according to the line

Dispersion around “Best Fit” Problems

Salary

Time in Rank

Assistant

Associate

NOTE: Observations do NOT “fit” the regression model, and so the model is “significant”. But some people are underpaid, others overpaid according to the line

Regression Equation Predictions using average salariesNew assistants make $12,739 more than Inst.New associates make $18,934 more than Inst.$6195 more than assistantsNew professors make $30,355 more than Inst.$17,616 more than assistants$11,421 more than associatesNot compressed when experience is not considered.Additional years of experience are rewarded at an increasing rate as rank increases.

University-Wide Faculty Salaries Compared to New Instructors

Regression Equation Predictions using average salariesAdditional pay to new assistants, associates, and professors in Health is negative, means increased salary for higher ranks are less than for average faculty.Average salary for instructors $64,027New assistants $ -3255 (less) than instructorsNew associates $3463 more than instructors$6720 more than new assistants.New professors $7236 more than Inst.$3772 more than new associates

Health Professions

Regression Equation Predictions using average salariesAvg salaries for instructors $57,588New assistants earn $32,268 more than instructorsNew associates earn $27,505 more than instructors$4763 less than new assistantsNew professors earn $46,724 more than instructors$19,219 more than new associatesFaculty in finance and accounting earn $20,375 more than other business sub-disciplines.

Business

Estimate predicted salaries at the individual faculty member level. Identify discipline (CIP). $ CIP InterceptIdentify current rank. + $ Rank modifierIf accounting or finance add + $ hi-businessMultiply years at ranktimes years at rank modifier + years*rank modPredicted Salary Predicted SalaryActual current salary - current salaryDesired adjustment $

Predicted

vs.

Actual Salary

Comparisons for an Individual Person

Example of How to Compute, NOT

Actual Salaries

Predicted salary

Cip

intercept

Rank Modifier

Hi-Bus

Yrs

at Rank modifier

Your years at rank

Total Predicted

Actual

Difference

Assistant Professor in CIP 1 (Agriculture), 5 years of experience at Assistant level

Y

=

39,606

14,177

(484)

5

51,363

50,000

1,363

Professor, CIP

1 (Agriculture,

10 years of experience as Professor

Y

=

39,606

30,355

1,266

10

82,621

80,000

2,621

Sample Predictions and Differences between Predicted and Current Salary

CIP

Coll

Subj

Rank

Yr

Salary

Predict

Diff

03

CAS

Asst

09

49,980

55,374

5,394

Prof

02

88,900

91,675

2,773

09

CAS

Asso

97

64,460

64,850

390

Prof

05

64,666

70,366

5,700

14

Eng

Asst

06

70,754

75,325

4,570

Asst

10

75,000

75,822

822

Asso

07

88,442

84,900

---

Is the study perfect? NO, but it can be informative as long as we understand what it shows and doesn’t. We must also understand the statistics and limitations of the study.Should we have done it a different wayWe reviewed the literature on C&I to find the way that other universities have handled this issue. Regression is what they used.Previous internal C&I committees have tried to develop “better” approaches, but these were not agreed to by both sides.We want something done now—finally--rather than wait for a perfect study.By all means, continue to identify and report to UFF and Administration your concerns with your experience with C&I. I hope that these reports will influence FGCU NOT to sweep the issue under the unexamined findings of the study.

Final Note