exhibit d - chicago employment law attorney | the …...1 case no. 15‐cv‐03308‐sem‐tsh, dkt....

Exhibit D

E-FILED Thursday, 10 August, 2017 04:42:32 PM

Clerk, U.S. District Court, ILCD

3:15-cv-03308-SEM-TSH # 48-4 of 107

IN THE UNITED STATES DISTRICT COURT

FOR THE CENTRAL DISTRICT OF ILLINOIS

SPRINGFIELD DIVISION

SAJIDA AHAD, MD, on behalf of herself

and all others similarly situated,

Plaintiff,

vs.

BOARD OF TRUSTEES OF SOUTHERN

ILLINOIS UNIVERSITY, and SIU

PHYSICIANS & SURGEONS, INC.

Defendants

Case No. 3:15‐CV‐03308‐SEM‐TSH

Expert Report of Chen Song, Ph.D.

In Support of Defendants’ Opposition to Class Certification

May 8, 2017

Exhibit D

3:15-cv-03308-SEM-TSH # 48-4 of 107

Table of Contents

I. Assignment ....................................................................................................................................... 2

II. Qualifications ................................................................................................................................... 3

III. Opinions ............................................................................................................................................ 3

IV. Plaintiff’s Allegations and Dr. Sharp’s Findings ......................................................................... 6

V. Background on SIU Physician Faculty Hiring and Compensation Process ............................ 8

A. SIU‐SOM Faculty and SIU‐HC Physician Joint Appointment ............................................. 8

B. Department Chairs as Decision‐Makers for Physician Faculty Compensation ............... 10

C. AAMC Physician Compensation Statistics Support Analysis by Department ............... 13

D. Determinants of SIU Physician Faculty Compensation ...................................................... 16

1. AAMC Salary Benchmark represents one of the factors for physician faculty’s

compensation at hire. .......................................................................................................... 16

2. SIU Compensation Plan Identifies SIU‐SOM Base, Clinical Base, and Clinical

Incentives as Compensation Components....................................................................... 17

3. SIU‐SOM relies on various funding sources to cover physician faculty’s SIU‐SOM

Base, and directorship often helps with securing funding ............................................ 18

4. SIU‐HC Clinical Base and Clinical Incentives are largely driven by clinical revenue,

except during start‐up period when clinical income is guaranteed. ........................... 20

VI. Review of Literature ...................................................................................................................... 27

VII. Statistical Analysis ......................................................................................................................... 31

A. Information Relied Upon ........................................................................................................ 32

B. Construction of Analytical Database ..................................................................................... 33

C. Statistical Analyses and Results ............................................................................................. 38

D. Additional Comments on General Surgery Division at Surgery Department ................ 43

E. Additional Comments on Neurology Department ............................................................. 46

VIII. Evaluation of the Sharp Report ................................................................................................... 48

A. Dr. Sharp asserted what he needed to show ........................................................................ 49

B. Dr. Sharp’s methodology cannot be utilized to calculate class‐wide damages ............... 52

C. Dr. Sharp’s methodology cannot fully address Plaintiff’s claims ..................................... 53

D. Dr. Sharp did not properly model multi‐year compensation correlations ...................... 54

E. Partial correction of Dr. Sharp’s regression models ............................................................ 56

F. Dr. Sharp’s models generated absurd predicted compensation values ........................... 56

G. Other database construction and modeling issues in Dr. Sharp’s regressions ................ 59

1. Erroneous Calculation of SIU Experience (“years_at_siu” Variable) .......................... 59

2. Not Accounting for Division Chiefs ................................................................................. 62

3. Other Data Issues ................................................................................................................ 63

IX. Conclusions .................................................................................................................................... 65

3:15-cv-03308-SEM-TSH # 48-4 of 107

I. Assignment

1. I have been retained by HeplerBroom LLC (“Counsel”), counsel for Defendant

Board of Trustees of Southern Illinois University School of Medicine (“SIU‐SOM”),

and SIU Physicians & Surgeons, Inc., also referred to as SIU Healthcare (“SIU‐HC”),

to conduct economic and statistical analyses related to allegations of gender

discrimination in pay, brought by Plaintiff Sajida Ahad, M.D., a female physician. It

is my understanding that Dr. Ahad brings this action under Title VII of the Civil

Rights Act of 1964, the Federal Equal Pay Act, and the Illinois Equal Pay Act1, and

that she seeks to represent a prospective class of all current or former female

physicians employed as faculty by SIU‐SOM from October 27, 2010 to present

(“Class Period”), who received compensation from SIU‐HC or who received a base

salary from the SIU‐SOM.2 It is also my understanding that Plaintiffs rely on Dr.

D.C. Sharp’s expert report3 (“Sharp Report”) and testimony in support of their

motion for class certification.

2. In connection with Defendant’s opposition to class certification, I was asked to

perform economic and statistical analysis related to Plaintiffs’ allegation that the

Defendant discriminated against Dr. Ahad and the prospective class of female

physicians by paying them less than their male counterparts. I was asked to opine,

from a statistical perspective, whether common issues apply to all prospective class

members. Finally, I was asked to review and evaluate the Sharp Report dated

March 10, 2017.

3. I may supplement or revise this report at a later date if additional relevant

information becomes available to me. I may respond to any additional reports or

1 Case No. 15‐cv‐03308‐SEM‐TSH, Dkt. 31 “Amended Class Action and Collective Action Complaint”,

filed on October 12, 2016, ¶13 2 Dkt. 31 ¶82 3 Expert Report of D.C. Sharp, Ph.D., dated March 10, 2017, for Purposes of Class Certification

3:15-cv-03308-SEM-TSH # 48-4 of 107

opinions by any experts retained by Plaintiffs. If called upon to testify at trial, I may

rely upon certain demonstratives based on my analyses.

II. Qualifications

4. I am a Senior Vice President at Nathan Associates, Inc. (“Nathan”), an economic and

financial consulting firm that provides economic research and analysis to public and

private clients in the United States and abroad.

5. I hold a Ph.D. in economics from the University of Chicago, where one of my fields

of specialization was Labor Economics which is the application of Economics and

Statistics to the study of labor markets, and a B.A. degree in Mathematics and

Economics from Agnes Scott College. Since receiving my Ph.D., I have provided

economic and statistical analyses related to damages and liability in class action

disputes. I have also maintained my connections with the academic community. I

taught a course on Labor Economics and Human Resources at the University of

California, Irvine. I was an Adjunct Assistant Professor at the California State

University, Los Angeles, where I taught a course on Financial Institutions. I have

been an instructor with the Chartered Financial Analyst (CFA) Review program

jointly sponsored by the CFA Institute and the University of Southern California

(USC), for which I have been teaching Risk Management for since 2008. A true and

correct copy of my curriculum vitae is appended to this report as Attachment A.

6. The opinions expressed in this report are based on the analyses that I, or Nathan’s

Staff at my direction and supervision, have conducted. Nathan is compensated for

my time at an hourly rate of $480, with hourly rates for my team ranging from $250

to $350. The amount of Nathan’s professional fees in this matter is not contingent on

the outcome of the litigation.

III. Opinions

7. Based on my analyses, I have reached the following conclusions:

3:15-cv-03308-SEM-TSH # 48-4 of 107

a) It is my opinion that Department is the proper decision‐making unit for

conducting compensation analyses for SIU‐SOM and SIU‐HC. My review of the

voluminous SIU information production shows that it was the Department

Chairs who signed offer letters for new hires in their Departments,

recommended members’ total compensation, and evaluated members’

productivity and performance (with help from Division Chiefs) at their

respective Departments.

b) Analyses of total compensation by Department show no common pattern of

gender pay disparity from 2010 to 2016 for physician faculty members employed

by SIU‐SOM and SIU‐HC among the seven Departments: Family and

Community Medicine, Internal Medicine, Neurology, Obstetrics and Gynecology

(OBGYN), Pediatrics, Psychiatry, and Surgery. I have tested the robustness of my

analyses by varying model specifications, and have reached the same conclusion

regardless of which model was used, that namely, there is no common proof that

female physician faculty members received statistically significant and

unfavorably unequal pay for equal work relative to their male colleagues while

employed by SIU from 2010 to 2016.

c) There was no statistically significant female pay disparity in total compensation

in any of the seven Departments. I have reached this conclusion by performing

statistical analyses on total compensation by Department, controlling for gender‐

neutral factors measuring academic achievements, clinical productivity, and

administrative responsibilities, all of which are compensation determinants

documented in SIU’s Compensation Plan. In reaching this conclusion, I have also

used data from the Association of American Medical College (AAMC), a well‐

established and widely‐used publication of medical school faculty compensation

statistics, as the appropriate market benchmark for my analyses. I have reviewed

scholarly articles on the topic of gender pay equity in the physician faculty labor

3:15-cv-03308-SEM-TSH # 48-4 of 107

market, in order to identify and construct additional gender‐neutral productivity

measures for physician faculty, which may have contributed to the

determination of total compensation at SIU.

d) I have reviewed the report by Plaintiff’s economic expert, Dr. Sharp, and have

examined Dr. Sharp’s statistical analyses, particularly his multiple regression

analyses and the data input into those models. While it is true that a multiple

regression analysis, with accurate data input and correct model specification,

represents a standard and widely‐accepted methodology to study gender pay

disparity, Dr. Sharp’s regression analyses do not properly support Plaintiff’s pay

disparity claims on a class‐wide basis, neither do those analyses propose a

correct methodology to calculate damages on a class‐wide basis:

First, Dr. Sharp’s regression analyses cannot prove the existence of a class‐

wide claim, because SIU’s decision‐making process for physician faculty

members’ total compensation was not adequately modeled. Dr. Sharp’s

model did not attempt to identify, for example, those who were similarly

situated to Dr. Ahad versus those who were not. Dr. Sharp assumed all

members were both similarly situated to Dr. Ahad and among themselves,

without investigating whether this assumption is consistent with

compensation practice at SIU.

Second, Dr. Sharp’s methodology cannot be utilized to calculate class‐wide

damages because female members in decision‐making units with no gender

disparity in pay, and/or female members who were not underpaid by model

prediction, suffered no economic damages.

Third, Dr. Sharp did not analyze total compensation, therefore the question

critical to Plaintiff’s claim of whether a female member was underpaid

relative to her male counterparts cannot be answered.

3:15-cv-03308-SEM-TSH # 48-4 of 107

Fourth, I have found numerous errors in Dr. Sharp’s regression database

construction, so the input data into Dr. Sharp’s regression models are not

accurate.

Fifth, I have identified an econometric application error by Dr. Sharp due to

improper treatment of the variance‐covariance matrix, which results in the

statistical significance of gender coefficient estimates being overstated.

Finally, a major problem with Dr. Sharp’s regression models is that they

omitted several important measures of administrative responsibilities and

clinical productivity. Such omission is not a data limitation because these

measures can be obtained from the available information; instead, the

omission of relevant controls is a modeling oversight.

All of the above criticisms lead to the conclusion that Dr. Sharp’s analyses

cannot be relied upon for the purposes of class certification and damages

calculation.

IV. Plaintiff’s Allegations and Dr. Sharp’s Findings

8. Dr. Ahad alleges in the complaint that she received unequal pay for equal work as

compared to male colleagues during the six years she worked for SIU‐SOM and SIU‐

HC. She further alleges that her experience at SIU was representative of other female

physicians employed as faculty by SIU‐SOM and SIU‐HC during the past decade

from 2007 to 2016. Dr. Ahad believes she was discriminated against in pay because

she has identified a “male replacement” who had only recently completed his

residency yet who was paid $75,000 more than her final salary.4 In support of the

class action and collective claims, Dr. Ahad cites in her complaint, two publications

4 Dkt. 31 ¶2

3:15-cv-03308-SEM-TSH # 48-4 of 107

from the Journal of the American Medical Association (“JAMA”)5, reporting that

female physicians earn less than their male counterparts on average, after

accounting for physician characteristics including specialty, academic rank,

leadership positions, publications, and research time. Dr. Ahad then compares

average compensation between SIU female and male faculty members, combining

the data from all departments and their corresponding divisions, across Assistant

Professor/Associate Professor/Professor ranks, to show the gender pay disparity that

allegedly existed at SIU‐SOM and SIU‐HC from 2007 to 2016.6

9. In support of her motion for class certification, Plaintiff’s expert witness Dr. D.C.

Sharp, filed a report on March 10, 2017, which focused on answering two questions:

First, whether there is a method available to analyze Plaintiff’s claims on a class‐

wide basis, and if so demonstrate how this method may be used; and Second,

whether this method may be utilized to calculate class‐wide damages, and if so,

demonstrate what the method shows regarding pay gaps. Dr. Sharp concluded that

multiple regression analysis is the proper method to analyze Plaintiff’s claim on a

class‐wide basis, and that using this method, he was able to determine that the SIU‐

SOM pay gap for female physicians is between $13,553 per year and $18,318 per

year, and that the SIU‐HC pay gap for female physicians is between $28,097 per year

and $30,559 per year.

10. Similar to Plaintiff’s comparison in the complaint, the multiple regression models

that Dr. Sharp proposed to analyze pay combined all physicians into one large

group regardless of departments and divisions. His models controlled for factors

5 “Gender Differences in the Salaries of Physician Researchers”, Vol. 307 (No. 22) JAMA 2410‐2417 (June

13, 2012), and “Sex Differences in Physician Salary in US Public Medical Schools” (No. 176) JAMA

Internal Medicine1294‐1304 (2016). 6 Amended Class Action and Collective Action Complaint, filed on October 12, 2016, ¶¶8‐10, and Exhibit

8 to Plaintiff’s Memorandum in Support of Her Motion for Conditional Collective Action Certification

and Judicial Notice, filed on October 18, 2016, ¶¶12‐18, 21‐22.

3:15-cv-03308-SEM-TSH # 48-4 of 107

aside from gender, including: specialization7, rank, tenure status, years at SIU,

number of publications and citations, number of residencies and fellowships, years

since medical school graduation, and whether the medical school is ranked in the

top 25. My understanding is that to identify these factors, Dr. Sharp has followed

recent literature which includes the two JAMA studies cited in the complaint.

V. Background on SIU Physician Faculty Hiring and Compensation Process

11. The first step in properly designing a methodology to analyze SIU physician pay is

to understand how pay decisions are made. This section provides relevant

background information on this topic, which I rely upon to conduct economic and

statistical analyses.

A. SIU‐SOM Faculty and SIU‐HC Physician Joint Appointment

12. Ms. Wendy Cox‐Largent, the Associate Provost of SIU‐SOM, filed a declaration on

November, 2016,8 in which she provided a detailed explanation of the organizational

structure, and joint recruitment and compensation process at SIU‐SOM and SIU‐HC:

“The Chief Academic Officer of SIU‐SOM is the Dean and Provost who also serves

as the CEO of SIU‐HC” (Cox‐Largent declaration ¶9, p.4).

7 For the “specialization” control in Dr. Sharp’s multiple regression analyses, he used the “Organization”

column in the Excel file “SIU SOM Physicians Compensation (FY07‐FY16).xlsx” produced by

Defendant, also made available to me on January 11, 2017. The “Organization” column sometimes

contains the name of a department such as “Surgery”, and sometimes contains the name of a

department‐division pair such as “Family and Community Medicine – Carbondale”. Because Dr. Sharp

aggregated physicians from all departments and divisions, and forced all physicians’ data to fit into one

regression under each specification, controlling for specialization only partially and minimally

accounted for variations in physicians’ pay across departments and divisions. His approach has

substantial limitations, which I will explain in more detail in Section VIII of the report where I evaluate

his regression models. 8 Declaration of Wendy Cox‐Largent, and attached exhibits, dated November 29, 2016

3:15-cv-03308-SEM-TSH # 48-4 of 107

13. The hiring of physicians for the SIU‐HC side is a joint decision made together with

SIU‐SOM, through the Master Affiliation Agreement among The Board of Trustees

of SIU, and Memorial Medical Center, and St. John’s Hospital of the Hospital Sisters

of the Third Order of St. Francis (“Master Agreement”).9 As stated in the Master

Agreement:

“Faculty for whom SIU receives support pursuant to Section 7.5 from

HOSPITALS for teaching in SIU residency and fellowship programs must be

eligible for academic faculty appointments at SIU. The Academic and respective

HOSPITAL’S Departments shall recruit faculty jointly in accordance with Section

5.2 of this Agreement. Such appointments will be at the Instructor, Assistant

Professor, Associate Professor, or Professor level in the appropriate track as

defined by the School of Medicine guidelines.” (p. 12, SIU Production 18098,

section 7.1)

14. Conversation with Mr. David Pence, Finance Director for SIU‐HC Administration,

and Ms. Sylvia McDonnough, (Retired) Assistant Director of HR/Payroll for SIU

School of Medicine 10, has further informed me on the topic of joint appointment of

physician faculty between SIU‐SOM and SIU‐HC. According to Mr. Pence, although

SIU‐HC is more business oriented, and may make hiring decisions based on

increased demand for doctors, an important consideration from SIU‐SOM is for the

medical school to have the proper accreditation. Providing residency training is part

of the mission of the SOM, so that residents get exposure to various specialties and

sub‐specialties.11 Therefore the hiring of physician faculty and their compensation at

hire are driven by the demands and needs of both SIU‐SOM and SIU‐HC.

9 SIU Production 18087‐18108, Exhibit 3 to Ms. Cox‐Largent’s Declaration on November 29, 2016 10 Transcribed notes for the conference call on March 21, 2017, produced as backup documents to my

report 11 The mission of the Southern Illinois University School of Medicine is “to assist the people of central and

southern Illinois in meeting their health care needs through education, patient care, research and service

to the community.” (http://www.siumed.edu/pr/about‐siu‐school‐medicine.html, accessed on

05/07/2017)

3:15-cv-03308-SEM-TSH # 48-4 of 107

B. Department Chairs as Decision‐Makers for Physician Faculty Compensation

15. As Ms. Cox‐Largent has explained, for the seven clinical departments at SIU‐SOM,

namely, Family and Community Medicine, Internal Medicine, Neurology, Obstetrics

and Gynecology (OBGYN), Pediatrics, Psychiatry, and Surgery:

“Each of the seven Departments staffed by SIU‐SOM employed faculty is

headed by a Department Chairperson who has overall responsibility for the

supervision of each clinical Department, including recruitment and hiring of

faculty within the Department. Each Department is further subdivided into

Divisions which sub‐division may be based on geographical location or

specialization as described below. Each Division is headed by a Division

Chiefperson.”(¶11, p.5, emphasis added)

…

“Recruitment and hiring of a faculty member is within the responsibility of

the applicable Department Chairperson. Therefore, employment of faculty is

individualized based not only on the individual Department, but who the

Department Chairperson is at the time the hiring decision is made.”(¶14, p.7,

emphases added)

An example was provided by Ms. Cox‐Largent about the hiring process for Plaintiff

Dr. Ahad:

“The position request for a bariatric surgeon was initiated effective November

3, 2006, under the authority of Dr. Gary Dunnington, then the chair of the

Department of Surgery (through his authorized fiscal officer’s signature, Lois

Strom). The request form identifies the position as new (as opposed to a

replacement for an existing faculty member), the specialty as bariatric surgery

under the immediate supervisor Dr. Edward Aflrey (Chair of the Division of

General Surgery), the advertising source (Annals of Surgery), total academic

($40,500) and clinical ($159,500) compensation ($200,000 combined total) as

predicted at the time by the Department Chair, whether a contract with another

party such as hospital would be used to support the academic position (at least

when created in 2006 but which changed by the time of hire as described below),

estimated clinical receipts as compared to the approved clinical compensation to

identify the need for SIU‐HC to provide compensation start‐up support until the

physician establishes a practice, administrative start‐up costs, and a justification

for the position.” (¶14, p.8, emphases added)

3:15-cv-03308-SEM-TSH # 48-4 of 107

I have reviewed and verified the above example described in Ms. Cox‐Largent’s

declaration.12

16. SIU‐SOM’s Guidelines on Faculty Appointments, Promotions, and Tenure13 further

confirm that:

“It is the responsibility of the Department Chairs to maintain an appropriate

balance of faculty commitments to carry out the programs of the Department

within the School of Medicine. These commitments should be reflected in the

faculty appointments in the Department and in the position descriptions of each

faculty member. It is also the responsibility of the Department Chairs to

function within the constraints of their departmental financial base with

respect to faculty appointments.” (B. Descriptions of Faculty Appointments, SIU

Production 8896, emphases added)

17. SIU‐HC produced 542 physician compensation documents for over 500 physicians,

which I understand were produced to Plaintiff’s Counsel. Fifteen (15) of the 542

were referenced in Dr. Sharp’s Report Appendix B. SIU has also produced some

physicians’ personnel files that are maintained by the SIU‐SOM HR department. In

addition to the SIU‐HC compensation documents, I have also reviewed SIU‐SOM

personnel documents for Dr. Ahad, referenced in Dr. Sharp’s Report under

“Appendix B. Documents Considered”. My review of the SIU production shows:

a) Department Chair signed physician faculty’s offer letters only within their

own departments. For example, Gary Dunnington, M.D., Professor and

Chairman of the Department of Surgery in 2008, signed the offer letter to Dr.

Ahad14 offering a compensation package of a) annualized base salary for the

academic appointment, and b) clinical practice income. As another example,

12 SIU Production 11222‐11225 as Exhibits 4 and 5, and SIU Production 16427 (Dkt. # 41‐6) 13 Southern Illinois University School of Medicine. Guidelines on Faculty Appointments, Promotion and

Tenure. Exhibit 4 to Plaintiffʹs Memorandum in support of her motion for conditional collective action

certification and Judicial Notice (filed 10/18/2016). (SIU Production 8893‐8929) 14 Offer letter dated February 18, 2008 (SIU PRODUCTION 11183‐11186)

3:15-cv-03308-SEM-TSH # 48-4 of 107

Gary Dunnington, M.D., signed the offer letter to Dr. John Sutyak in 2001,15

which described all components and contributing factors to the compensation

including: academic rank, annual academic salary, directorship (if any), sources

of funding for clinical income, and additional clinical incentives through earned

work Relative Value Units (or “RVUs”, a concept which I will elaborate later in

this Report when I analyze SIU‐HC compensation).

b) Department Chair recommended physician faculty’s annual compensation. All

physicians typically co‐sign the SIU Physicians & Surgeons, Inc. Compensation

Agreement (also referred to as SIU Healthcare Compensation Agreement in more

recent years) at the beginning of each fiscal year. It was the Chairman at the

Department of Surgery who recommended Dr. Ahad’s total compensation

including a) SIU‐SOM base, b) clinical base, c) other incentives, and d)

preliminary and absolute annual caps.16 For each of the other six SIU‐SOM

clinical departments considered in this matter, respective Department Chair of a

physician’s department recommends that physician’s total estimated annual

earnings in the SIU Healthcare Compensation Agreement.17

c) Department Chair and Division Chief signed off on Dr. Ahad’s duties and

responsibilities. Around the end of every fiscal year Dr. Ahad had to co‐sign a

Position Description, which described her time commitment to categories of

duties and responsibilities in the upcoming fiscal year. The position description

15 Offer letter dated January 16, 2001 (SIU PRODUCTION 17980‐17981) 16 See SIU‐SOM production for fiscal year 2009 (SIU PRODUCTION 11149) and fiscal year 2008 (SIU

PRODUCTION 11148). See also SIU‐HC production for fiscal years 2008 to 2013 (SIU PRODUCTION

11191‐11196). During Dr. Ahad’s tenure at SIU, Dr. Dunnington was the Chair at the Department of

Surgery from fiscal years 2000 to 2013, succeeded by Michael Neumeister, M.D. who became the Chair

around September 2012 (fiscal year 2013). Dr. Ahad resigned on March 21, 2014, three months before

the 2014 fiscal year end. 17 See, for example, SIU‐HC production for Allen Devleschoward, M.D., at Neurology Department,

Neurology Division, for fiscal years 2011 to 2017 (SIU PRODUCTION 18732‐18739).

3:15-cv-03308-SEM-TSH # 48-4 of 107

was co‐signed by the Division Chief at the Division of the General Surgery, and

the corresponding Chair at the Department of Surgery.18

d) Division Chief submitted acknowledgement of performance review to

Department Chair. At each fiscal yearend, Dr. Ahad met with her Division Chief

to discuss her written evaluation and to develop appropriate goals for the

coming year. An acknowledgment form signed by both Dr. Ahad and her

Division Chief was then submitted to the Department Chair.19

18. Based on my knowledge from reviewing all of the above documents, and from

conversations with the persons most knowledgeable (PMKs), I have concluded that

it is the Department Chair who is in charge of: a) Setting physician faculty’s

compensation (both SIU‐SOM compensation and SIU‐HC compensation) at his/her

Department, and b) evaluating their productivity and performance with help from

his/her Division Chiefs. Therefore, the appropriate decision‐making unit for a study

on physician faculty’s compensation is at the Department level, with the

Department Chair being the principal decision‐maker. It is at the Department level

that compensation analysis should be performed.

C. AAMC Physician Compensation Statistics Support Analysis by Department

19. Because of the $187,200 cutoff on annual wages in the Bureau of Labor Statistics

(BLS) Occupational Employment Survey (OES),20 physicians’ or other top earning

18 See “Position Descriptions” for fiscal years 2014 (SIU PRODUCTION 11151‐11152), 2013(SIU

PRODUCTION 11154‐11155), 2012 (SIU PRODUCTION 11156‐11157), 2011 (SIU PRODUCTION 11160‐

11161), 2010 (SIU PRODUCTION 11162‐11163), and 2009 (SIU PRODUCTION 11165‐11166). In the

Department of Surgery, Dr. Ahad belonged to the Division of General Surgery. When Dr. Ahad began

employment with SIU, Dr. Edward Alfrey was the General Surgery Division Chief. In June 2010, Dr.

John Mellinger because the Division Chief. 19 See performance review acknowledgements: June 2013 (SIU PRODUCTION 11187), June 2012 (SIU

PRODUCTION 11153), August 2011 (SIU PRODUCTION 11182), July 2010 (SIU PRODUCTION 11181),

and July 2009 (SIU PRODUCTION 11164). 20 https://www.bls.gov/careeroutlook/2015/article/wage‐differences.htm, accessed on 03/28/2017

3:15-cv-03308-SEM-TSH # 48-4 of 107

professionals’ salary statistics are typically provided by companies specializing in

surveying particular industries, using consistent survey methodology over time.

Those companies often make the industry statistics their propriety research

products, for which annual subscription fees are charged to users for information

access. There are two primary sources of physician faculty data: (i) AAMC Faculty

Salary Report21, which provides total compensation statistics, by faculty rank and

department/specialty, as described above by Ms. Cox‐Largent, and (ii) Medical

Group Management Association (MGMA)’s Provider Compensation module22,

20. The AAMC data are more appropriate for a study of the academic physician

compensation, because it is collected for full‐time medical college faculty. In

contrast, the MGMA data includes not only academic practices, but also hospitals

and private practices23. Additionally AAMC Salary Benchmark represents one of the

factors for physician faculty’s compensation at hire. It is for those reasons that I

have decided to rely on the AAMC data to understand the salary distribution and

the labor market of academic physicians.

21. I visited the Norris Medical Library at the University of Southern California (USC)

and was able to borrow 2001‐2012 and 2014‐2015 AAMC Data Books.24 As general

background, The AAMC Data Book is:

“intended to serve as a convenient source of data for constituents and the public

on a variety of topics related to medical education, such as applicants, students,

faculty, graduate medical education, medical school revenues, and teaching

21https://services.aamc.org/dsportal2/index.cfm?fuseaction=login.login&thread=jump.FSSREPORTS&app

name=FSSREPORTS&frompermissionscheck=true (accessed on 01/30/2017) 22 http://www.mgma.com/industry‐data/mgma‐surveys‐reports/physician‐compensation‐and‐

production‐survey, last accessed on 05/07/2017 23 See http://www.mgma.com/industry‐data/participate‐in‐an‐mgma‐survey. MGMA provides industry

benchmark for physician compensation and production; non‐physician provider compensation and

production; academic compensation and production; medical directorship compensation; on‐call

compensation; and physician placement starting salary. 24 2001 to 2005 AAMC Data Books were published in January, and all subsequent editions were published

in April. 2013 AAMC Data Book was not available at the Norris Medical Library.

3:15-cv-03308-SEM-TSH # 48-4 of 107

hospital operations and finances ... The Data Book tables are derived from

AAMC reports and databases as well as from external sources, such as the

National Institutes of Health, the American Medical Association, the Bureau of

Labor Statistics, and the American Hospital Association.”25

22. Exhibit 1 summarizes the information from the AAMC Faculty Salary Reports,

Table 1126. The 25th‐percentile, the median, and the 75th‐percentile of total

compensation27 are presented by academic rank, from the 2008‐2009 to the 2014‐2015

fiscal periods, by department for each of the seven Departments: Family Medicine,

(Internal) Medicine, Neurology, OBGYN, Pediatrics, Psychiatry, and Surgery.

23. Distributions in AAMC total compensation vary among departments. For example,

75% of the associate professor physician faculty at the Psychiatry Department

earned up to $240,000 in the 2014‐2015 fiscal period. In other words, the top‐25th

percentile associate professors at the Psychiatry Department made at least $240,000

during this period. In contrast, a total compensation of $240,000 would have

belonged to the lowest quartile of the distribution at the Surgery Department where

the ceiling for the associate professor 25th‐percentile was $346,000. To make it to the

top‐25th percentile surgeon, an associate professor would have had to earn at least

$579,000 in total compensation.

24. Exhibit 1 also shows that dispersion of the total compensation distribution can

widely differ among Departments. For example, distribution in total compensation

at the Pediatrics Department is much tighter than that at the Surgery Department.

Generally the top‐25th percentile pediatricians made about 50% more than the

25 See, for example, AAMC Data Book April 2014, Executive Summary. 26 Exhibit 3 to the Deposition of Wendy Cox‐Largent taken on 12/23/2015, and “AAMC Table 11: Faculty

Compensation for All Schools, MDs, Clinical Departments, 2014‐2015” received from Counsel. 27 The total compensation equals the fixed/contractual salary component of total compensation plus the

supplemental earnings components of total compensation (medical practice supplement,

bonus/incentive pay, and known uncontrolled outside earnings) for the 12‐month period spanning a

fiscal year. (Report on Medical School Faculty Salaries, page 1 “Executive Summary”).

3:15-cv-03308-SEM-TSH # 48-4 of 107

bottom‐25th percentile pediatricians, whereas the top‐25th percentile surgeons could

earn 80% more than the bottom‐25th surgeons.

25. Because total compensation distributions vary by Department, it would be

inappropriate to compare total compensation for physician faculty members in

different Departments. Distinct compensation distributions also imply distinct labor

markets for physician faculty recruiting, which is consistent with the fact that SIU

delegates faculty recruiting and total compensation determination to Department

Chairs.

D. Determinants of SIU Physician Faculty Compensation

26. I have established that Departments are the proper decision‐making units to analyze

physician faculty compensation. I have identified Department Chairs to be the

compensation decision‐makers. Next I proceed to investigate determinants of SIU

physician faculty total compensation, by examining factors considered by

Department Chairs.

1. AAMC Salary Benchmark represents one of the factors for physician faculty’s

compensation at hire.

27. Like many major public medical institutions, SIU uses market salary benchmark for

compensation at hire, in order to comport with the market as a whole. Once a need

for hiring a physician faculty is identified and requested by the applicable

Department Chair, a Recruitment Tracking Form is completed. According to Ms.

Cox‐Largent:

“The Recruitment Tracking Form also contains a pre‐printed section where

industry salary data from the Association of American Medical Colleges

(AAMC) can be included. The AAMC is a not for profit association including as

members all 147 accredited U.S. and 17 accredited Canadian medical schools,

almost 400 teaching hospitals, V.A. medical centers and academic societies.

Among its activities, the AAMC collects survey data from its member medical

3:15-cv-03308-SEM-TSH # 48-4 of 107

colleges of faculty salaries. The data comes from all schools, public and private,

from the largest academic urban schools such as Columbia University in NYC to

the smallest community based schools such as SIU in Springfield, Illinois. The

data is categorized by specialty and when available subspecialty, faculty rank

and identifies the number of faculty included at each specialty and rank, and

the mean and 25th, median and 75th percentiles. Industry salary survey is one

of many considerations for each individual faculty hiring decision.” (¶16, p.9,

emphases added)

Ms. Cox‐Largent has provided an example of the Recruitment Tracking Form used

to recruit Dr. Ahad’s husband, Dr. Imran Hassan, where the AAMC statistics were

quoted on the form.28

28. I will use the information collected from the AAMC Data tables as the proper

market salary benchmark to analyze the SIU physician faculty compensation.

2. SIU Compensation Plan Identifies SIU‐SOM Base, Clinical Base, and Clinical

Incentives as Compensation Components.

29. Every SIU Physicians & Surgeons, Inc. Compensation Agreement states (emphases

added):

“Member and SIU P&S hereby agree that Member shall be compensated in

accordance with the SIU P&S Compensation Plan for all services and duties

performed by Member under the auspices of SIU P&S. SIU SM Base shall be paid

by SIU School of Medicine. Clinical Base and Incentives are estimated. Actual

amounts will be calculated pursuant to the SIU P&S Compensation Plan and are

dependent on Member’s measured productivity and availability of funds to SIU

P&S.” (See, for example, SIU PRODUCTION 11192).29

28 SIU Production 17903 as Exhibit 7 to Cox‐Largent’s Declaration on November 29, 2016 29 See, for example, SIU PRODUCTION 18732, for a similar statement in the SIU Healthcare

Compensation Agreement.

3:15-cv-03308-SEM-TSH # 48-4 of 107

30. I have reviewed the Compensation Plan adopted by the Board of Directors of SIU

Physicians & Surgeons, Inc., effective April 16, 1997,30 which identifies three

components of a faculty member’s compensation:

a) SIU SOM Academic Base Salary (“SIU SM Base”);

b) SIU P&S, Inc. Clinical Base Salary (“Clinical Base”) ; and

c) SIU P&S Clinical Incentives (“Clinical Incentives”).31

31. In the subsequent sections, I will outline determinants of each of the three

components of faculty member’s compensation based on my review of the

Compensation Plan.

3. SIU‐SOM relies on various funding sources to cover physician faculty’s SIU‐

SOM Base, and directorship often helps with securing funding

32. For the determination of SIU SM Base, the Compensation Plan states:

SIU SM Base will be provided to all members of SIU P&S.

SIU SM Base may include an administrative component for SIU SM

responsibilities.

SIU SM Base will be paid to faculty members by SIU SM directly, regardless of

the source of funds.

33. Even though all members are paid their academic base by SIU‐SOM regardless of

the source of funds, the actual funding of academic positions does not all come from

SIU SM funds. According to SIU‐SOM mission statement, SIU‐SOM relies on

“relationships and common missions with many community organizations”,

particularly five hospitals which SIU‐SOM has formation affiliation: Carbondale

30 SIU PRODUCTION 17274‐17291 31 Furthermore, Clinical Incentives will be based pursuant to three distribution schemes: Individual‐Level

Clinical Incentives; Department‐Level Clinical Incentives; and Group‐Wide Clinical Incentives.

3:15-cv-03308-SEM-TSH # 48-4 of 107

Memorial Hospital, Blessing Hospital in Quincy, Decatur Memorial Hospital,

Memorial Medical Center in Springfield, and St. John’s Hospital in Springfield.”32

34. For example, Dr. Ahad had a Medical Director Agreement with both the SIU‐SOM

and St. John’s Hospital, which specified that the St. John’s Hospital would pay SIU‐

SOM a contract fee of $125,000 per year, for Dr. Ahad to direct the Bariatric Surgery

Program at St. John’s Hospital. Therefore, St. John’s Hospital was the actual funding

source for Dr. Ahad’s SIU SM Base.33

35. was appointed Associate Professor of Clinical Surgery and

Associate Director of the Southern Illinois Trauma Center. Memorial Health

Systems, St. John’s Hospital, and the School of Medicine each covered $75,000 of

$225,000 SIU‐SOM Base for the first 24 months of his employment

with SIU.34

36. Different hospitals may provide funding for the SIU‐SOM Base because clinical

needs typically drive the decision of SIU SM faculty recruiting. For example, Dr.

Ayman Omar was hired in April 2010 for the Department of Neurology. The hiring

form stated that35:

a) [P]rogram benefits greatly by having a Neuro‐Oncology physician;

b) Expertise does not currently exist in the department/division/practice;

c) Community/market need for service or capacity;

d) This position increases overall clinics offered and patients seen – patients we

would not have the ability to see without this particular expertise; and

32 See http://www.siumed.edu/welcome.html. Also see http://www.siumed.edu/pr/quick‐facts.html for a

current list of affiliated hospitals. 33 See SIU PRODUCTION 17484‐17492 for fiscal year 2009; SIU PRODUCTION 17498‐17507 for fiscal year

2010; SIU PRODUCTION 17512‐17522 for fiscal year 2012; SIU PRODUCTION 17528‐17539 for fiscal

year 2013; SIU PRODUCTION 17545‐17556 for fiscal year 2014 (at $121,250 per year contract amount

instead of $125,000 per year). 34 Offer letter from Dr. Gary Dunnington, Chairman of the Surgery Department, to on

January 16, 2001 (SIU PRODUCTION 17980‐17981) 35 See SIU PRODUCTION 19419‐19420. was approved an SIU SM Base of $120,000 and a

Clinical Base of $50,000 (SIU PRODUCTION 19421, SIU Healthcare Supplemental Hiring Form for the

hiring of ).

Redacted

Redacted

Redacted

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

e) Funding for first year guarantee will be secured by the Department of

Neurology.

37. Medical directorship in a subspecialty with clinical demands is one way of securing

funding from hospitals or from other funding sources. This was the case with Dr.

Ahad and appointments. Medical directorship is also considered an

extra administrative responsibility which contributes to a higher SIU‐SOM Base. For

example, was named Residency Program Director,

Carbondale Family Medicine, effective April 1, 2012.36 His total compensation

package in the new role was $256,000, which consisted of $176,000 in SIU‐SOM Base

(much higher than previous years) and $80,000 in clinical practice income.37

38. At the time of this report, I have not collected all the information on medical

directorship. Similar to being a Department Chair and a Division Chief, medical

directorship information is important for understanding some of the observed

compensation differentials among different physicians in the same

Department/Division, or the same individual with this responsibility versus

without. Currently not all of the directorship information is available to me from

the SIU‐HC personnel files. If at a later date I am able to collect this information, I

will supplement my analyses. For this report, I will limit my discussion on specific

directorship appointments to individual examples.

4. SIU‐HC Clinical Base and Clinical Incentives are largely driven by clinical

revenue, except during start‐up period when clinical income is guaranteed.

Clinical Base

36 See SIU PRODUCTION 30006. 37 See SIU PRODUCTION 29972 and 29971. Also note that “Change of Assignment…” form that is

associated with the offer of the Residency Program Director lists that had a change of salary, a

change of FTE, and an addition of non‐paid assignment. While the assignment is listed as “non‐paid”

there was an effective pay increase associated with increased FTE (SIU PRODUCTION 30008‐30009).

Redacted

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

39. For the determination of Clinical Base, the Compensation Plan states, among other

things (emphases added):

Clinical Base will be provided to the faculty members for a portion of their

clinical activities and for approved SIU P&S administrative responsibilities.

Clinical Base will be calculated pursuant to the Clinical Base Formula, and will

be annually recommended for each faculty member by the relevant Department

chair to the Compensation Committee for its approval.

Clinical Base will be paid at least monthly.

Funding for Clinical Base will be from clinical revenue only.

The portion of Clinical Base paid for administrative duties, if any, will cease to be

paid when a member no longer performs such administrative duties.

40. In summary, Clinical Base is only a portion of the clinical activities, driven primarily

by clinical revenue and subject to yearly cap, with the exception of additional Clinical

Base payment for certain administrative responsibilities.38

Clinical Incentives

41. Compensation for clinical activity is composed of both Clinical Base and Clinical

Incentives. While Clinical Base is calculated and paid as a fixed, budgeted expense

of SIU P&S, Clinical Incentives will be paid (if at all) from the SIU P&S Variable

Funds Pool (the “VFP”).39 A percentage40 of the VFP will be available for the

payment of individual physician‐level Clinical Incentives. Clinical Incentives will only

be paid to the extent that they are earned by faculty.41

38 For example, that annual cap is $114,000 in the Compensation Plan effective April 16, 1997 (SIU

PRODUCTION 17280). 39 SIU PRODUCTION 17281 40 For example, that percentage is 88% in the Compensation Plan effective April 16, 1997 (SIU

PRODUCTION 17283). 41 SIU PRODUCTION 17282

3:15-cv-03308-SEM-TSH # 48-4 of 107

42. A measure used for the calculation of physician‐level Clinical Incentives is called

Relative Value Units (RVUs). RVUs are a measure of value used in the United States

Medicare reimbursement formula for physician services. Before RVUs were used,

Medicare paid for physician services using “usual, customary and reasonable” rate‐

setting which led to payment variability. The Omnibus Budget Reconciliation Act of

1989 enacted a Medicare fee schedule, whereby physician services are classified

under a nomenclature based on the Current Procedural Terminology (CPT) to which

the American Medical Association holds intellectual property rights.42

43. From all physician faculty members’ RVUs in a division, a division conversion factor

is calculated. The Cox‐Largent Declaration has provided an illustration of

conversion factor calculation:

“The conversion factor for a given division is calculated by taking the total

amount of fees received for professional services during a six month period of

time (January 1‐June 30) from all sources (Medicare, Medicaid, Third Party

Private Insurers, and Private Pay) for the work of all division physicians during

that six month time period and dividing the gross receipts by the total number of

RVUs generated by all division physicians during the same period. A percentage

of the resulting number (40%) is then used as the conversion factor for a

succeeding six month period of time. For example, if the total receipts by the

division of General Surgery during an applicable time period are $2,987,873 and

the total RVUs by all division physicians during that time period are 40,289, then

result of $2,987,873/40,289=$74.16 per RVU. Sixty percent of the $74.16 is

retained by SIU‐HC for overhead and expenses, and the remaining 40% or

$29.66 is the amount that each RVU is “worth” to all division physicians for

the next applicable six month period of time.” (¶28, pp.15‐16, emphases added)

44. SIU‐HC makes Clinical Incentive payments monthly based initially on 75% of the

estimated amount due each month according to the formula in the Compensation

Plan43:

42 https://en.wikipedia.org/wiki/Relative_value_unit, accessed on 03/15/2017 43 SIU PRODUCTION 17293

3:15-cv-03308-SEM-TSH # 48-4 of 107

Individual‐Level Clinical Incentive =

Division Conversion Factor × Individual RVUs – Clinical Base

Through a quarterly reconciliation process, an adjustment will be calculated with

respect to the Clinical Incentives paid for the preceding three (3) month period. If

the adjustment is negative, the amount will be repaid over the next three months

through deductions from Clinical Base. Terminated faculty member’s reconciliation

will be conducted on a pro rata basis.44

45. So far the steps described for SIU‐HC Clinical Incentives demonstrate that even

though the calculation process is highly involved, individual‐level Clinical

Incentives are formulaic. The formulas account for various circumstances

surrounding physicians’ practice, such as listed v. unlisted CPTs, patient population,

special contractual agreements, patients’ guarantor, physician group pooling

arrangements. The calculation of Clinical Incentives ultimately ties back to a

physician faculty member’s RVUs which is a pure productivity measure, and is both

meritorious and transparent.

46. Additionally the Compensation Plan does specify that faculty members can be

rewarded additional Clinical Incentives for teaching, scholarly research, and other

contributions which benefit the Department.45 I will incorporate those additional

factors into my study. However I do not believe that those additional factors made

44 SIU PRODUCTION 17287 45 In addition to division‐level RVU‐based Clinical Incentive payouts, faculty members can be rewarded

additionally for other contributions for department‐level Clinical Incentives per the Compensation Plan.

Incentive plans by the Departments reward faculty members for their contribution, including: Teaching

(attendance at teaching functions, student/resident evaluations); Scholarly research (grants,

publications, scholarly speaking engagements); contributions to the Department’s academic, clinical

and/or administrative activities; and charity care and other contributions. In the event that a

Department does not submit a plan or a submitted plan is not approved, funds originally allocated to

the Department will instead be made available for distribution to physician faculty in the form of the

Group‐Wide Clinical Incentives, based upon criteria developed by the Compensation Committee in

conjunction with Department chairs. Group‐Wide Clinical Incentives reward faculty members for

contributions that foster the success of SIU P&S including, but not limited to patient satisfaction,

referring physician satisfaction, and utilization management (SIU PRODUCTION 17285‐17286).

3:15-cv-03308-SEM-TSH # 48-4 of 107

significant contributions to Clinical Incentive payments – this belief is based on a

comprehensive review of SIU Compensation Plan, Cox‐Largent Declaration,

historical division conversion factor calculations for every division46, as well as on

my conversation with Ms. Cox‐Largent, and with Mr. Pence and Ms.

McDonnough.47

47. As Mr. David Pence described, pay from the Practice Plan bears no direct

relationship with rank, tenure, or publications. Every physician has the same

opportunity to earn because pay is dependent on RVUs. Market and practice

building activities such as attending events, meeting with providers all help with

expanding patient population and growing revenue for the practice. Simply put,

physicians are like entrepreneurs, who are free to make the most productive use out

of the platform provided by SIU‐HC, in order to run successful and profitable

practices.

Start‐Up Packages

48. The Compensation Plan contains special provisions for start‐up packages:48

Such packages will be available only to new recruits or faculty members who are

returning after a sabbatical or leave.

The new or returning faculty member may be paid a guaranteed Clinical Salary

in lieu of Clinical Base and/or Clinical Incentive for up to one year before

becoming subject to the Clinical Base or Clinical Incentive formulas.

46 “Ahad S Conversion Factor file w historical CFs 02‐22‐17.xlsx”, received from Counsel on February 23,

2017, describing how division conversion factors were calculated historically from about FY 2000 to

present. 47 See “2017 03 21 Conference call notes.pdf” which includes notes from March 21st, 2017 call with Mr.

David Pence, Finance Director for SIU‐HC Administration, and Ms. Sylvia McDonnough, Retired‐

Assistant Director of HR/Payroll for SIU SOM. 48 SIU PRODUCTION 17287‐17288

3:15-cv-03308-SEM-TSH # 48-4 of 107

This guaranteed Clinical Salary, if not paid first from funds available for

distribution as Department‐Level or Group‐Wide incentive, may be paid as a

Central Practice Expense upon the approval of the Compensation Committee,

In general, such start‐up guaranteed Clinical Salary should be paid for a period

of six (6) months or less.

These payments will be paid in the same manner as Clinical Base.

49. A salary start‐up “allows a new faculty member (hopefully) to build a practice and

become productive. A faculty member on a guarantee also has the opportunity, if

he or she chooses, to terminate the guarantee early and participate in the

compensation plan based on his or her individual RVU production.” (Cox Largent

Declaration ¶ 38, p. 21). According to David Pence49, about 90% of newly hired

physicians have a salary start‐up agreement, which is generally one year long

during which the quarterly reconciliation of Clinical Incentives does not apply. The

salary start‐up is an assurance that physicians will be paid during the initial months

post hire to “put up the shingles’ so to speak, when the focus is on building a clinic

and attracting patients. The remaining 10% of new hired physicians do not have a

salary start‐up agreement, so quarterly reconciliation commences immediately for

them.

50. My review of the SIU‐HC personnel documents shows that physicians at different

departments choose start‐up agreements differently. For example, start‐ups were

signed by over 90% of the newly‐hired surgeons in recent years at the Surgery

Department. In contrast, no physicians signed start‐ups in recent years at the Family

and Community Medicine Department and the Psychiatry Department. For those

who signed start‐up agreements, the 12‐month guarantee arrangement was the most

common duration, followed by 24‐month:

49 See “2017 03 21 Conference call notes.pdf”, notes from March 21st, 2017 call.

3:15-cv-03308-SEM-TSH # 48-4 of 107

“The guarantee takes the faculty member’s total clinical income set forth in the

SIU‐HC compensation agreement during the year and pays that to the faculty

member in 12 equal installments. For Dr. Ahad, who was on the guarantee for

her first two years, she received 1/12 of the $125,000 clinical income from SIU‐HC

each month, irrespective of whether she generated enough RVUs each month to

“earn” that level of clinical income” (Cox‐Largent Declaration ¶38, p. 21).

I have also observed other guarantee durations ranging from 6‐month, 36‐month,

48‐month, or even 60‐month50, in the SIU‐HC personnel documents.

51. As explained in the Cox‐Largent Declaration, “[i]f a faculty member on guarantee is

productive his or her first year and earns more RVUs than paid under the guarantee,

the faculty member can terminate the guarantee at any time, be paid for the

accrued RVUs above what he or she was already paid, and from that point

forward is paid as are other faculty not on guarantee – based on his or her RVU

production times the applicable periodic calculated conversion factor” (Cox‐Largent

Declaration ¶38, pp. 21‐22, emphasis added). For example, both and

voluntarily came off their salary guarantee early.51

52. Because start‐up guarantees eliminate either the upside or the downside in clinical

compensation (no Clinical Incentive component with start‐ups), I will factor the

decision to participate in start‐up agreements and the duration of the guarantee in

my analyses.

50 See – Contract Addendum” on September 7, 1999 which stated “Clinical Incentive

amount is guaranteed by St. John’s Hospital for a period of five years, at which time, actual earnings

will be compared with the amount guaranteed to determine what, if any, action is necessary. The

guarantee will be performed through a written agreement between St. John’s and ” (SIU

PRODUCTION 21381). is with the Department of Pediatrics, Neonatology Division (SIU

PRODUCTION 21380). For other examples of Guarantee Periods duration, see, for example, SIU

PRODUCTION 11206‐11210 (Dr. Sajida Ahad), SIU PRODUCTION 24701‐24705 (Dr. John Mellinger),

and SIU PRODUCTION 21931‐21935 (Dr. Elizabeth Ramsey Unal). 51 See 4/30/2015 email from David Pence to Barbara Thompson, which wrote “Both e and Dr.

voluntarily came off guarantee March 31, 2015” and that “beginning April 2015, these two

physicians should be paid their Ortho trauma call” (SIU PRODUCTION 34590). See also email

exchange on 4/9/2015 between and David Pence, discussion the decision to end

guarantee early (SIU PRODUCTION 29450‐29451).

Redacted

Redacted

Redacted

RedactedRedacted

RedactdRedacted

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

VI. Review of Literature

53. As noted in Section IV, in support of the class action and collective claims, Dr. Ahad

cites in her complaint52, two publications from the Journal of the American Medical

Association (“JAMA”): “Gender Differences in the Salaries of Physician

Researchers”, Vol. 307 (No. 22) JAMA 2410‐2417 (June 13, 2012) (henceforth referred

to as “Gender Differences”), and “Sex Differences in Physician Salary in US Public

Medical Schools” (No. 176) JAMA Internal Medicine1294‐1304 (2016) (henceforth

referred to as “Sex Differences”), reporting that female physicians earn less than

their male counterparts on average, after accounting for physician characteristics

including specialty, academic rank, leadership positions, publications, grant

funding, and research time.

54. I begin this section by summarizing those two articles in order to provide context in

connection with the issues brought up in this case. Additionally, I have performed

my own research reviewing articles on physician labor market and studies of gender

gaps in this market, with the objective of keeping abreast with the more current

academic research on this topic in terms of influential productivity factors that

empirically have strong predictive power on physician pay.

55. I first start with the two articles cited by Plaintiffs. Authors in the “Gender

Differences” (2012) article sent out a US nationwide postal survey in 2009‐2010 to

over 1,700 physicians. The authors ended up with 800 physicians who continued to

practice at US academic institutions and reported their current annual salary. Other

productivity related characteristics were collected for this population. The authors

presented a multivariable model of salary in for this sample and found that being

male was associated with higher salary at the 99% confidence level, after controlling

52 Dkt. 31 ¶s 3 and 5

3:15-cv-03308-SEM-TSH # 48-4 of 107

for degree, academic rank, specialty nature, specialty pay level, publications,

leadership position, and research time.53 In addition to the dataset being constructed

from a mail survey which could potentially have suffered from a self‐selection bias,

certain decisions made by the authors could also be problematic. For example,

instead of using an actual market benchmark by specialty, the authors created four

categories of “specialty pay level” (“low‐paying”, “moderate‐paying”, “high‐

paying”, and “extremely high‐paying”). Furthermore, there was not a good

measure for clinical productivity. The authors included a measure of specialty

nature (“medical specialties”, “clinical specialties for women, children, and

families”, “hospital‐based specialties”, and “surgical specialties”). But this measure

was at best an indirect proxy for clinical productivity, because it did not directly

associate with physicians’ clinical revenue contribution. The authors recognized

other possible explanations to the observed gender difference in their study:

“Still, it remains possible that men and women in our sample did have different

values of made different choices. It is possible that men prioritized

compensation more than women did … As a result, women may have made

trade‐offs in compensation to achieve non‐monetary benefits. For example,

women may have been more likely to choose in situations that successfully

offered lower salaries because of a location in or near a desirable community …

We do not have information on employment status of the respondents’ spouses,

so we cannot ascertain whether spousal employment may have mediated the

salary difference observed” (page 2416).

Those explanations are what a labor economist considers as “supply‐side” factors,

which are factors impacting the labor supply of individuals. Supply‐factors may

reflect an individual’s preference and his/her reservation wage. For example, Dr.

Ahad discussed looking for a job near her husband when she inquired into SIU and

the surrounding areas of Springfield, Illinois, because at the time her husband Dr.

53 “Gender Differences”, page 2415 Table 3

3:15-cv-03308-SEM-TSH # 48-4 of 107

Imran Hassan was already a surgeon at SIU’s Surgery Department.54 Individual

preferences are not “demand‐side” factors such as an employer’s animus for

discrimination against a certain group of employees.

56. Authors in the “Sex Differences” (2016) article assembled a salary database for

academic physicians employed in 24 public medical schools in 12 states. They then

combined the salary database with information on clinical and research productivity

of physicians to analyze sex differences in earnings. Proxies for clinical and research

productivity include age, years of experience, faculty rank, specialty, scientific

authorship, research grant funding, clinical trial participation, and Medicare

reimbursement (proxy for clinical revenue). The authors estimated specialty‐specific

sex differences in salary using a multivariable linear regression model. They found

“substantial heterogeneity across specialties in the size of sex differences in salary”

after controlling for clinical and research productivity. Salary among men exceeded

that of women at the 95% confidence level in 9 of the 18 specialties.”55 They stated as

an implication of their study, that:

“Sex differences in salary varied considerably across specialties and institutions.

Specialties such as orthopedic surgery, surgical subspecialties, obstetrics and

gynecology, and cardiology had the largest absolute sex differences in salary,

whereas radiology, family medicine, and emergency medicine had differences

that were small in magnitude and not statistically significant…which suggests

the potential importance of evaluating specific specialties to understand the

practices associated with improved male‐female equity in academic medicine”

(page E9).

The authors acknowledged a potential concern with their study in that they “lacked

information on faculty track or part‐time status, which could confound sex

differences in salary if women are more likely to enter lower‐paying tracks or work

part‐time. I also find that while informative (and certainly better than not having

54 Deposition of Dr. Sajida Ahad, M.D. (“Ahad Deposition”), 11/11/2016: 33:22‐34:7. 55 “Sex Differences”, page E5‐E6 and Table 3

3:15-cv-03308-SEM-TSH # 48-4 of 107

any control for clinical productivity), Medicare reimbursement is an imperfect proxy

for clinical productivity because physician revenues include private insurance and

other payment sources. Medicare reimbursement is certainly a weaker substitute for

clinical productivity than individual physician RVUs.

57. I have conducted my own literature research on the topic of physician gender pay

gap. I have reviewed articles analyzing compensation data in the labor market for

physicians, in order to identify and construct additional gender‐neutral productivity

measures which may have contributed to the determination of total compensation at

SIU.

58. Baker (1996) examines annual earnings of young physicians under 45 years of age

with less than nine years of practice experience.56 Baker showed that there were no

significant differences in earnings between male and female physicians after

controlling for hours worked, practice specialty, practice setting (e.g., self‐employed,

hospital, or medical school), experience, family status (e.g., married with children, or

not married no children), board status, and malpractice status. He attributed the

narrowing or even elimination of the physician gender pay gap to increases in the

demand for female physicians through social trends, large group practices’

promotion of income equality among their employees, and advances in medical

education enabling female physicians to compete for earnings.

59. Sasser (2005) examined the roles of marriage and parenting and their effects on

physician labor market decisions and compensation.57 She found no gender gap

after controlling for the marital status and the number of children physicians have.

Sasser found that female physicians reduced their number of working hours after

56 Laurence C. Baker, “Differences in Earnings between Male and Female Physicians,” New England

Journal of Medicine, Vol. 334 No. 15 (1996): 960‐964. 57 Alicia C. Sasser, “Gender Differences in Physician Pay: Tradeoffs between Career and Family,” Journal

of Human Resources, Vol. 40, No. 2 (2005): 477‐504.

3:15-cv-03308-SEM-TSH # 48-4 of 107

they get married and have children. However, those physicians did not become less

productive after they return to work. This study concluded that the observed lower

compensation of married female physicians was not due to employer gender

discrimination; instead it was due to physicians’ choices toward spending more

family time. Even though I do not have data on physician’s marital or family status,

I can account for the effects of changes in female physician’s working hours through

physician’s RVUs, as well as possibly through other measures such as grants,

publications, and academic rank. If a female physician spends less time at work as a

result of substituting work time for time to fulfil family responsibilities, her RVUs

(and similarly other possible measures) are more likely to be lower.

60. What I have gathered from the literature review is that data availability and data

quality can drive the empirical outcome of a gender pay study for physicians. It is

often not an easy task to obtain physician compensation determinants ‐ however it is

critical to spend the time and effort to collect and construct productivity measures

when feasible, because without these measures incorrect inferences may be drawn.

VII. Statistical Analysis

61. I have established in the previous sections that at SIU:

Departments are the decision‐making units in setting and reviewing total

compensation;

AAMC Market benchmark is an important factor used by the Department Chair

to set a physician faculty member’s starting pay, and clinical need is another

important factor;

Total SIU Compensation includes SIU‐SM Base, SIU‐HC Clinical Base, SIU‐HC

Clinical Incentives (for employees and years where start‐up agreements do not

apply);

3:15-cv-03308-SEM-TSH # 48-4 of 107

Factors contributing to SIU‐SM Base include: division, faculty rank, SIU tenure,

publications, grants, directorship, administrative responsibilities (Department

Chair, Division Chief);

Factors contributing to SIU‐SM Clinical Base and SIU‐HC Clinical Incentives

include division (conversion factor calculated by division), RVU production, and

start‐up agreements. Relevant but to a lesser extent are directorship,

administrative responsibilities (Department Chair, Division Chief). And even

further to a much lesser extent are grants and publications.

This Section will start with an itemization of information I rely upon, followed by a

description of how I construct the analytical database, and finally a presentation of

my analyses and results.

A. Information Relied Upon

62. Between November 4th, 2016, and May 5th, 2017, I received from counsel PDF

documents including: Pleadings (motions and responses); PMK deposition

transcripts and exhibits; SIU‐SOM personnel documents for Dr. Ahad and selected

faculty in the Neurology department; SIU‐HC compensation PDF documents for

over 500 physicians; data source for SIU P&S preliminary compensation caps fiscal

years 2010 to 2016; Dr. Sharp’s Report, exhibits, backup documents, and deposition

transcripts, and other relevant deposition and hearing transcripts.

63. I have relied upon documents listed in Attachment B, and all the information

sourced in the footnotes of this report. I have directed my team to collect important

compensation‐related information from SIU‐HC compensation PDF documents and

electronic data, and coalesce into additional datasets in order to prepare for the next

step of building an analytical database.

3:15-cv-03308-SEM-TSH # 48-4 of 107

B. Construction of Analytical Database

64. Using the Updated Compensation Data (see Attachment B), I first retrieved rank

from “SIU SOM Physicians Compensation (FY07‐FY16).xlsx”, tenure from “SIU

SOM Physician Academic Base Salary (FY07‐FY16).xlsx”, and merged this

information to the data set “77V1461‐Ahad S SIU Complete Physician List (By

Division) 4‐4‐17.xlsx”. If rank was not available from the Physician Compensation

file, I used the rank that was provided on the Updated Compensation Data.58 I then

filtered data to the relevant employee‐year records by the following steps:

a) Remove non‐physicians, non‐members of SIU‐HC who worked through a FQHC,

fellows, and non‐SIU employees;59

b) Remove partial‐year records corresponding to:

i. First year of hire starting after the beginning of a calendar year,;

ii. Resigned, retired, or deceased before December;60

c) Remove records with calendar years prior to 2010 and limit only to the 2010‐2016

Class Period;

d) Remove records with a rank of “Research Professor”, “Adjunct Professor”,

“Instructor”, or the ambiguous rank of “Assistant/Associate”;

e) Remove records with an organization (Department) of “Executive Assoc” or

“Healthcare”;

f) Remove records with either a zero total compensation, or a zero SIU‐SOM Base,

or a zero total SIU‐HC Clinical compensation; and

g) Identify and investigate a few low “outlier” compensation records:

58 Rank in the Updated Compensation Data only retains the most current rank. Tenure status is not

provided in the Updated Compensation Data. 59 February 27, 2017 letter from Tom H. Wilson to J. Bryan Wood 60 I observe that employees were paid the full calendar‐year compensation when they resigned, retired, or

deceased after December 10th.

3:15-cv-03308-SEM-TSH # 48-4 of 107

i. has been with SIU since

1977.61 He belonged to the Neurology Department, Neurology Division. His

SIU‐SOM Base has been around $1,600. still saw patients,

although his Clinical compensation was on the low end in recent years. I

decided to remove records because of the unusually low

SIU‐SOM Base. It is possible that has retired but the data

did not capture that. Removing his records will not bias my analysis against

female employees in the Neurology Department.

ii. ’s total compensation was $74,649 in

calendar year 2012, $32,721 in 2013, and $37,447 in 2014.62 She belonged to the

Internal Medicine Department, Infectious Disease Division.

appeared to have been given information on Short‐Term Disability benefits63,

but there are no records of her going on disability or any other leave. I

decided to keep records.

iii. ’s SIU‐SOM Base

decreased significant from 2010 to 2013, so did her SIU‐HC compensation.

Nothing else in the data seemed to have changed during this time.

belonged to the Surgery Department, Orthopedic Surgery Division. I checked

personnel records. She was trying to resign in 2011, but rescinded

that decision.64 My initial conjecture was that might have gone on

some part‐time arrangement, because her SIU‐SOM Base was being

significantly reduced in those years. Further conversation with Counsel

confirmed my conjecture: husband

61 “77V1461‐Ahad S SIU Complete Physician List (By Division) 4‐4‐17.xlsx” 62 “77V1461‐Ahad S SIU Complete Physician List (By Division) 4‐4‐17.xlsx” 63 SIU PRODUCTION 23378 64 rescinded her resignation on 1/10/2011 (SIU PRODUCTION 26041). She went off the

“monthly draw” clinical payments, and was receiving SIU‐HC compensation based on actual RVUs

only (SIU PRODUCTION 26038‐26039).

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

was an Assistant Professor of Clinical Surgery at SIU

Surgery Department, Orthopaedic Division.65 resigned in 2010 to

start a new job in Texas. subsequently worked out a part‐time

arrangement with SIU after her husband relocated. I included

records in my analyses.66

iv. Cardiology Division in the Internal Medicine Department was a very small

division. During the Class Period, the Cardiology Division had at most three

physician faculty members working side‐by‐side. One member was excluded

as “retired” (although he still works and is listed on the website and in the

providers’ directory). One member was removed due to his end‐of‐December

termination (partial year), but upon further review I decided his last year’s

records should be brought back since he appeared to have been paid for the

whole year prior to termination. The third member was a more recent hire.

v. had a part‐time appointment with SIU, therefore I

excluded his 2010 full calendar year record.67

vi. was hired on 7/1/2009 as an Instructor.68 Her status

changed to Assistant Professor effective 09/26/2011.69 Her 2011 record should

be excluded since she only spent part of 2011 in the Assistant Professor rank,

which explains why her pay was low in 2011.

65 records from 2006 to June 30th, 2010 are in the file ““77V1461‐Ahad S SIU Complete

Physician List (By Division) 4‐4‐17.xlsx.” 66 Including her records is a conservative approach, because I am including more years of lower

compensation for a female employee. 67 See SIU PRODUCTION 25113‐25118: “Pediatric Endocrinology services are not available in Central &

Southern Illinois since retirement from private practice. The department of Pediatrics will

be searching for a full‐time Pediatric Endocrinologist, with an anticipated 18‐36 month time period. Dr.

has agreed to provide Pediatric Endocrinology services on a part‐time basis to fulfill our patient

care and teaching needs.” (p. 4 Clinical Hire Request for SIU P&S Member, signed in January 2008, SIU

PRODUCTION 25118) 68 SIU PRODUCTION 26947‐48 69 SIU PRODUCTION 26944‐45

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redat d

3:15-cv-03308-SEM-TSH # 48-4 of 107

65. Currently I do not have all the information available to capture part‐time

appointments, disability, FMLA or other leaves of absence in the data. If and when

additional information becomes available to me, I will supplement my analyses.

66. Table 1 shows the number of records excluded at each step. The relevant

population for my analyses is 1,362 employee‐year records.

67. Next I added additional variables or columns to those 1,362 records or rows. Those

are the academic achievements and productivity measures previously described:

indicator for ranks (Assistant, Associate, Full Professor), indicator for tenure status

(“Alternate Track”, “Tenure Track”, and “Tenure Granted”), indicator for serving as

Department Chairs or Division Chiefs, indicator for ever received grants, indicator

for being currently on start‐up, number of RVUs (not applicable to Divisions which

do not report RVUs because of physician pooling arrangements), indicator for

missing RVUs (RVUs should have been reported but not found in the data), and

indicator for rehire. I reviewed Dr. Sharp’s program steps to create the publication

counts from the Scopus datasets, and accepted his calculation of the number of

publications to be added to my dataset. I also reviewed the year of graduation

variable provided by Dr. Sharp in his backup files, and used it in my analyses as

well.

Table 1 ‐ Summary of the Analytical Database Construction Steps

DescriptionNumber of

Observations Remain

Original Data Set 2,892

‐Remove Non Physicians and FQHC 2,855

‐Remove Partial Year Observations 2,217

‐Remove Observations prior to 2010 1,464

‐Remove Ranks other than Assistant, Associate, or Professor 1,461

‐Remove Administrative Departments 1,439

‐Remove Zero Total Compensation 1,439

‐Remove Zero Academic Compensation 1,372

‐Remove Zero HC Compensation 1,372

‐Remove Outliers 1,362

Source: ʺ77V1461‐Ahad S SIU Complete Physician List (By Division) 4‐4‐17.xlsxʺ

3:15-cv-03308-SEM-TSH # 48-4 of 107

68. Subsequent data standardization and validation were performed. Because 2016

compensation data was produced for January through June, I multiplied the 2016

compensation by two as a proxy for the 2016 calendar yearend total. I replaced a

couple of incorrect employee hire dates with the correct ones by reconciling with

their SIU‐HC personnel PDF documents.70

69. I then created an indicator for the M.D. degree, which equals to one if “M.D.” is one

of the degrees obtained, and zero for degrees other than an “M.D.” (“D.O.”,

“Psy.D.”, or “Ph.D.”).71 I calculated years since graduation as a proxy for total

working experience.72 I also created indicators for each calendar year during the

Class Period. Exhibit 2 gives a few example records from the analytical database I

have constructed.

70. Exhibit 3 shows female representation, and by gender: academic rank and tenure

status distribution from 2010 to 2016 for each of the seven Departments. Female

representation was higher (generally in the high 30%, or the 40%, or the 50% ranges)

at Family and Community Medicine, Internal Medicine, (later years at) OBGYN,

Pediatrics, and Psychiatry. Female representation was lower (generally in the high

10%, or the 20% ranges) at Neurology, (earlier years at) OBGYN, and Surgery.

Female faculty members were also predominantly on alternate track, as opposed to

being on tenure track or having been granted tenure.

71. Exhibit 4 compares by gender, average total compensation, average SIU‐HC

compensation, and average RVUs from 2010 to 2016 by Department. In almost

every instance, a higher SIU‐HC compensation (often total compensation too) is 70 For example, I corrected s hire date to be 9/10/2012 (SIU

PRODUCTION 33513). was only receiving HC salary in 2012 and in no subsequent

years. First her hire date was corrected. Then her 2012 partial year record was excluded. As a result all

records for were excluded. 71 missing degree was corrected with “M.D.” See SIU

PRODUCTION 19745. 72 For example, proxy for total experience as of calendar yearend in 2010 = years between 12/31/2010 and

July 1st, year of graduation from medical school.

RedactedRedacted

RedactedRedacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

associated with higher average RVUs. The only exception is 2015 Department of

Pediatrics. Upon further investigation into compensation at the Pediatrics

Department, I noticed that its Division of Neonatology had a pooling agreement in

fiscal year 2010, where all members received an equal share of the pool, with the

exception of two doctors who received different shares.73 P&S Supplemental Hiring

Form for at the Neonatology Division of the Pediatrics

Department confirms the existence of a pooling agreement at that Division in FY10.74

Furthermore, three Critical Care physicians: ,

and signed a pooling agreement in August 2015.75

Additionally Counsel informed me that physicians in the Critical Care Division of

the Pediatrics Department (for example, and .

had a pooling agreement.76 I have not adjusted the RVUs at the

Neonatology Division to their post‐pooling amounts. It is likely that the exception

observed in 2015 is the outcome of a post‐pooling RVU reconfiguration which

altered the direct SIU‐HC/RVU relationship for a subset of the members in the

Neonatology Division.

C. Statistical Analyses and Results

72. To analyze the allegations of gender discrimination in compensation among

physician faculty at SIU from the statistical point of view, I will rely on multiple

regression analyses technique. Multiple regression analysis is a statistical tool to study

73 See SIU PRODUCTION 26464. In addition, St. John’s Hospital agreed to guarantee total compensation

for FY10 at $290,000. The FY11 contract being developed at the time was for both St. John’s Hospital &

Memorial Medical Center to guarantee total compensation at $300,000. 74 SIU PRODUCTION 34580‐34581 75 See SIU PRODUCTION 31931. There was also a pooling agreement dated July 1st, 2016 for the Division

of Pediatric Critical Care (SIU PRODUCTION 31926‐31930). 76 Those physicians were working at the Intensive Care Unit at St. John’s Hospital which required them to

be 24‐hour on call.

Redacted

Redacted Redacted

Redacted Redacted

Redacted Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

the relationship between the variable that has to be explained, called the dependent

variable, and several explanatory variables that are thought to be the factors that

explain the changes in the dependent variable.77 Accounting or controlling for

several factors through the multiple regression analysis allows separating out the

effect of each such factor on the change in the dependent variable, from the effects of

the other factors, or, in other words, holding all other factors constant. Multiple

regression analysis is widely used in a variety of settings, including the allegations

of gender discrimination in compensation. In such framework, compensation is the

variable to be explained, or the dependent variable. After controlling for the effects of

all other legitimate gender‐neutral factors (e.g., productivity) that contribute to

variation in compensation, any remaining portion of unexplained statistically

significant compensation differential is then attributed to gender.

73. I estimated the following model (AAMC Median Model) for each of the following

seven Departments (indexed by j) separately: Family and Community Medicine,

Internal Medicine, Neurology, Obstetrics and Gynecology, Pediatrics, Psychiatry,

and Surgery:

In this specification, i refers to an individual physician faculty member in year t. The

model is set to estimate a set of β parameters, for each of the j departments, in the

presence of εit, a random, person‐ and time‐specific disturbance (or error) term. The

77 See, for example, Daniel L. Rubinfeld, Reference Guide on Multiple Regression. Reference Manual on

Scientific Evidence, 3rd ed., 2011, p. 305

3:15-cv-03308-SEM-TSH # 48-4 of 107

rank‐specific AAMC Median market benchmarks are included at the

division/specialty level k, for each year t.

74. Exhibits 5‐A to 5‐G provide detailed results of the regressions by department, using

AAMC Median as Labor Market Benchmark. Table 2 below presents a summary of

the female coefficient by Department from the regressions.

75. I conclude from Table 2 that there is no statistically significant female coefficient at

the 5% level in any of the Departments, which shows not only that during the Class

Period there was no common proof of gender pay claims on a class‐wide basis, but

also that women were not underpaid in total compensation compared to their male

counterparts in each Department.

76. In reaching these conclusions, I have controlled for gender‐neutral determinants of

pay specified in the equation above. Coefficient estimates on those determinants

generally have the expected signs and statistical significance. Notably there are three

sets of factors that are statistically significantly positive almost across all

Departments, AAMC median, Department Chair indicator, and RVUs (when

applicable). Starting from the AAMC Median or the market benchmark measure

that is unique by year, rank, Department, and Division, coefficient estimates are all

Table 2 ‐ Summary of Female Coefficient Estimates ‐ Regressions with AAMC Median

Total Compensation

Department Estimate t‐value Probability

Employee‐

Year

Count

Female‐

Year

Count

AAMC Model (Median)

Family and Community Medicine ‐$13,698 ‐1.31 0.1953 191 87

Internal Medicine ‐$17,340 ‐1.38 0.1706 374 151

Neurology ‐$32,921 ‐1.26 0.2248 55 3

Obstetrics and Gynecology $6,175 0.13 0.9013 102 31

Pediatrics ‐$3,431 ‐0.20 0.8418 208 97

Psychiatry ‐$823 ‐0.05 0.9592 86 38

Surgery ‐$6,178 ‐0.35 0.7244 346 86

3:15-cv-03308-SEM-TSH # 48-4 of 107

statistically significantly positive across Departments, ranging between the values of

1 and 2 in five of the seven Departments, and is 0.62 for the Surgery Department and

2.69 for the Neurology Department. This is consistent with SIU using AAMC as one

of the considerations to set total compensation at hire.

77. Similarly coefficients on Department Chair indicators all have positive coefficients,

most of them statistically significantly positive. Coefficients on the Division Chief

indicator all have positive coefficients, with the coefficient in the Surgery

Department being statistically significant.78 This is consistent with the Compensation

Plan that administrative responsibilities come with additional remuneration.

78. As specified in the Compensation Plan, RVUs bear a direct relationship with a

physician faculty’s SIU‐HC clinical compensation (except in the situation of a

pooling agreement among physicians as is the case for the Department of Family

and Community Medicine, as well as the Emergency Medicine Division at the

Department of Surgery). A physician’s clinical compensation is formulaically

determined by multiplying his/her division conversion factor by his/her RVUs. As

expected, the RVU coefficients are statistically significantly positive in six

Departments. The RVU coefficient for the Neurology Department is positive but not

statistically significant. In general, the size of RVU coefficients mimics the average

conversion factor amount. For example, the RVU coefficient for the Surgery

Department is 24, which is within the range of the Division conversions factor for

that Department.79 In other words, the RVU coefficient measures the average dollar

compensation for each RVU worked.

78 Neurology Department has four Divisions: Center for Alzheimer’s Diseases and Related Disorders,

Child and Adolescent Neurology/Pediatric Neurology, Neurology/Oncology, and Neurology. The first

three Divisions each had only one employee in any given year. Therefore for Neurology Department, I

only controlled for Department Chair without an additional control for Division Chief. 79 “Ahad S Conversion Factor file w historical CFs 02‐22‐17.xlsx”

3:15-cv-03308-SEM-TSH # 48-4 of 107

79. Coefficients on total experience also have the expected signs with a positive

coefficient on the linear term of experience and a negative coefficient on the

quadratic term of experience, which is consistent with the human capital theory that

pay rises with additional years of experience, but at a diminishing rate.

80. When applicable, coefficients on the current start‐up indicator are positive in each

Department, although not statistically significant. The rehire indicator is populated

in three Departments: Family and Community Medicine (statistically significantly

negative rehire coefficient), Pediatrics (statistically significantly negative rehire

coefficient), and Surgery (statistically significantly positive rehire coefficient). As

discussed later in Section VIII. G. 1., sometimes a rehire was accompanied by a

significant increase in academic salary, while other times just the opposite could

happen, depending on individuals’ circumstances of returning to SIU. Across

Departments, coefficients on grant indicator are not statistically significant.

Coefficients on the number of publications are either not statistically significant, or

statistically significantly positive.

81. To conduct a rigorous analysis, and ensure the robustness of the estimated

coefficients, I additionally ran two variations of the AAMC Median Model. First, I

ran the same specification as above, but used AAMC Mean Total Compensation as

Labor Market Benchmark, instead of the Median:

82. The second model variation is the Year‐Rank‐Division Model, where I specifically

controlled for the individual physician faculty’s academic rank, division, and year:

3:15-cv-03308-SEM-TSH # 48-4 of 107

In this specification, I additionally estimated Associate Professor and Professor

indicators, a set of δ parameters to account for the effects of the different divisions

(specialties) within each department, and a set of γ parameters to control for the

annual market demand effects.

83. Exhibit 6 summarizes the results from those two variations in model specification.

My conclusions remain the same, that there is no common proof or a statistically

significant female pay gap in total compensation for the seven Departments.80

D. Additional Comments on General Surgery Division at Surgery Department

84. While 2008 and 2009 compensation was not analyzed in this report because those

two years are outside the Class Period, some discussion regarding Dr. Ahad’s

compensation upon her initial hire in July 2008 might be helpful in understanding

the unique circumstances surrounding Dr. Ahad’s appointment. Exhibit 7‐A

enumerates faculty members in the General Surgery Division at the Surgery

Department, for years 2008 and 2009.

85. Dr. Ahad began working at SIU on 07/28/2008 in the General Surgery Division at the

Surgery Department. Her SIU‐SOM Base was $125,000 and SIU‐HC compensation

was $125,000. She also signed a start‐up guarantee agreement, placing her on the

salary guarantee period for 24 months following her hire. Dr. Ahad had a Medical

Director Agreement with both the SIU‐SOM and St. John’s Hospital, which specified

80 I have tested the sensitivity of my results with including versus excluding tenure indicators, and

including versus excluding the rehire indicator. None of those variations changes my conclusion.

3:15-cv-03308-SEM-TSH # 48-4 of 107

that the St. John’s Hospital would pay SIU‐SOM a contract fee of $125,000 per year,

for Dr. Ahad to direct the Bariatric Surgery Program at St. John’s Hospital. “When

Dr. Ahad started on July 28, 2008, (and during her entire employment at SIU‐SOM

through March 21, 2014), she was the only fellowship trained bariatric surgeon at

SIU‐SOM” (Cox‐Largent Declaration ¶23, p.12).

86. During the same year, several other physician faculty members joined the General

Surgery Division as Assistant Professors on tenure‐track. None of them was similar

to Dr. Ahad because none of them could have been qualified for Dr. Ahad’s position.

Vice versa, Dr. Ahad was not similar to any of them because she could not have

done their jobs either, as I explain further below.

87. Starting with Drs. Ruth Diane Mayforth, Andreas H. Meier, and David A. Rogers.

While they were listed under the General Surgery Division, they were all affiliated

with the Pediatric General Surgery Division, which was a separate Division from the

General Surgery Division at various times throughout their SIU career.81

Additionally Dr. Meier was a tenured Associate Professor, and Dr. Rogers was a

tenured Professor in 2008 when Dr. Ahad started at SIU. Therefore those three

surgeons were not comparable to Dr. Ahad.

88. Dr. Jan Rakinic has a board certification in Colon and Rectal Surgery as well as

Surgery.82 When Dr. Rakinic was recruited, SIU was searching specifically for a

General Surgeon with a Colorectal specialty.83 Additionally Dr. Rakinic was a

tenured Associate Professor when Dr. Ahad started at SIU, therefore the two of them

were not comparable. According to Wendy Cox‐Largent Declaration, Dr. Imran

Hassan (Dr. Ahad’s husband) was needed because of his specialty in colorectal

surgery and the high demand for colorectal surgery (Cox‐Largent Declaration ¶17,

81 See SIU PRODUCTION 31555, 31572‐31578 for Dr. Mayforth; SIU PRODUCTION 18997 and 19015‐

19018 for Dr. Meier; and SIU PRODUCTION 21219, 21221 for Dr. Rogers. 82 https://www.siuhealthcare.org/doctor/jan‐rakinic.html 83 See SIU PRODUCTION 23734‐23737, clinical hire request in Dr. Rakinic’s personnel file.

3:15-cv-03308-SEM-TSH # 48-4 of 107

p. 10). Dr. Hassan had training and fellowships in Colon and Rectal Surgery.84

Therefore Dr. Hassan was not comparable to Dr. Ahad either.

89. The four trauma surgeons in 2008 and four in 2009: Drs. Jarrod Wall, Tahira Mirza,

Christopher Wohltmann, John Sutyak, and Ibrahim Cetindag, all had appointments

with the Southern Illinois Trauma Center (SITC). All of the SITC positions required

the completion of a trauma or critical care fellowship85, a qualification Dr. Ahad did

not have.86

a) Dr. Wall was hired in August 2008 (but might have started his appointment in

October 2008);

b) Dr. Mirza was hired in February 2008, and then terminated in July of the same

year;

c) “Dr. Sutyak, then an Associate Professor, became Director of SITC in 2004” (Cox‐

Largent Declaration ¶41, p.24);

d) When Dr. Ahad started in July 2008, “ salary was

approximately $123,000 in academic base and $160,000 clinical income… At the

time of his hire, was the Chief of the trauma service at William

Beaumont Army Medical Center and had completed a fellowship in

trauma/critical care surgery” (Cox‐Largent Declaration ¶41, pp.24‐25);

e) “ was hired as an Assistant Professor at the same salary as ,

$175,000 academic base and $150,000 clinical income” (Cox‐Largent Declaration

¶46, p.27)

90. All four doctors had separate trauma RVUs they accumulated in 2008 (except for Dr.

Wall who started in 2008), and 2009 (except for Dr. Cetindag who started in 2009).

and , the two Assistant Professors, had total compensation

84 SIU Production 17918 85 See Cox‐Largent Declaration ¶42, p. 25. See also, for example, job posting for Academic

Trauma/Critical Care Surgeon (SIU PRODUCTION 15669). 86 Ahad deposition 176: 18‐19

Redacted

Redacted

Redacted Redacted

Redacted Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

around $330,000 in 2009. Their total RVUs – combined between General Surgery

and Trauma procedures – were over 5,000 each. Additionally trauma and critical

care surgeons typically were paid a higher base for their relative position compared

to other providers because their contract required that they take a certain number of

trauma calls each month.87 It is for those reasons that trauma and critical care

surgeons had higher total compensation than general surgeons at SIU. As shown in

Exhibit 7‐B, the AAMC statistics median compensation for trauma surgeons was

higher than that of general surgeons too.

91. All the other surgeons at the General Surgery Division all had higher ranks than Dr.

Ahad, and/or tenured, and/or were Department Chair/Division Chief. Therefore

none of them was similar to Dr. Ahad.

E. Additional Comments on Neurology Department

92. Total compensation analysis for the Neurology Department requires additional

consideration due to the relatively small sample of physicians, and most importantly

due to the even smaller number of females who work at this department. Generally

a regression model is an accepted methodology for a sample of 55 employee‐year

records. It is nonetheless prudent to conduct an additional review of the data,

especially for the cohorts of employees hired or worked during the same period. I

proceed next with this additional review.

93. During each full year of the Class Period, there were at least ten physician faculty

members working at the Neurology department, and they were mostly men. One of

them, , was excluded from the analyses because of his low

SIU‐SOM Base compensation for the reason described earlier in Section VII.C. There

was also a relatively high faculty turnover at this Department, resulting in some

87 Ahad S DOL testimony of Dr. Mellinger 1/5/2016, 186:16‐22

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

employee‐year records being excluded by applying the “partial year” exclusion

restriction. For example, Dr. Rita M. Tranquilli, a female physician, was hired in July

of 2011, but terminated in January of 2012, resulting in all of her partial‐year records

being excluded from my regression analyses. Dr. Joni Marie Clark, another female

physician, resigned in 2007, before the Class Period. This leaves only two female

physicians, Drs. Paula K. Rauschkolb , hired on 6/2/2014 and

resigned on 4/15/2016, for calendar year 2015) and Mona Elsayed (

, hired on 10/6/2014, for calendar years 2015 and 2016), who worked at the

Neurology department for at least one full year in calendar years 2015 and 2016.

94. As summarized in Exhibit 8, Dr. Paula K. Rauschkolb was hired in June 2014, but

resigned in April 2016, so that only one record for year 2015 can be analyzed.

was hired in October 2014, and was still active at the time the data

was produced to me. However, it appears that she later resigned in October 2016.88

While I have included both her 2015 and 2016 records in my analyses, only the

record for 2015 is a true full‐year record.

95. This leaves at most three (3) female employee‐year records that can be compared to

fifty‐two (52) male employee‐year records for the analyses. With no full‐year female

physicians working at the Neurology department during 2010‐2014, it is more

meaningful to focus the analyses on the years when female physician were present:

2015‐2016.

96. As Exhibit 8 shows, both Drs. Rauschkolb and Elsayed were Assistant Professors on

the Alternate Track. In 2015, male Assistant Professors on the Alternate Track were

Drs. Fazeel M. Siddiqui, Najib Murr and Sajjad Mueed. All of these physicians’

actual total compensation in 2015 appeared consistent with the additional

administrative or directorship titles they held. For example, and

88 “Appointment.pdf”, p. 1‐2 (“Faculty, Principal Administrative and Civil Service Staff

Resignation/Separation” form and letter of resignation

Redacted

Redacted

Redacted

Redac

Redacted

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

did not have any additional administrative titles, and their total

compensation was at comparable levels of $260,772 and $256,564, respectively. Even

the approximately $4,000 difference (between $260,772 and $256,564) can be

explained by additional years of experience that had, based on the year

of her graduation from the medical school. was a Director of Neuro

Oncology and earned $270,300 in 2015, while was a Residency Director

and earned $267,171 in the same year. earned a higher total

compensation of $373,762, but he served as a Director of the Memorial Medical

Center Stroke Center, and had considerably more experience compare to the other

physician faculty members described here.

97. In 2016, male Assistant Professors on the Alternate Track were

and . Again both men held director positions so their total compensation

was higher than .

98. Based on these examples, there does not appear to be any systematic gender pay

differential among the faculty at the Neurology Department which cannot be

explained by gender‐neutral factors.

VIII. Evaluation of the Sharp Report

99. As explained in Section V. which I reiterate below because of its importance, a

correct method to determine whether all SIU female physicians belong to a single

class is to evaluate whether “similarly situated” female physician faculty employees

have had common adverse experience in pay. This determination can be made by

the following steps:

a) Investigate who determine(s) physician pay at SIU‐SOM and SIU‐HC: Was it the

Dean, department chairs, or division heads? For example, if Department Chairs

determine physician pay, comparison of a surgeon’s pay in the Surgery

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

Department to a pediatrician’s pay in the Pediatrics Department would be

inappropriate because pay decisions were not made by the same individual,

among other reasons of why it would be inappropriate to compare their pay.89

b) Understand how physician pay at SIU‐SOM and SIU‐HC is initially set upon

hire, and how subsequent pay adjustments are made, in order to identify all

available gender‐neutral factors that were considered by the decision‐makers on

physician pay;

c) For each decision‐making unit, apply a standard, widely‐accepted methodology to

analyze physician pay at SIU‐SOM and SIU‐HC, so that contributions from

gender‐neutral factors in pay can be accounted for at the decision‐making unit,

and the question can be answered of whether there remains any gender pay

disparity, for that decision‐making unit;

d) If a common pattern of gender pay disparity is observed across the majority of

the decision‐making units, the pattern is consistent with pay disparity on a class‐

wide basis. If there does not exist a common pattern of gender pay disparity

across decision‐making units, then there is no proof of a class‐wide claim.

A. Dr. Sharp asserted what he needed to show

100. While Dr. Sharp sets out to find a standard, widely‐accepted methodology to

address pay disparity on a class‐wide basis, he has assumed that a class‐wide claim

89 As discussed in Section V. C., market data has shown distinct salary distributions for surgeons and

pediatricians, which helps explaining why decision‐making on physician compensation is delegated to

Department Chairs, rather than to the Dean or the Provost. Another reason, discussed in Section V.C.4.,

is that physicians’ clinical compensation at SIU‐HC is tied to productivity by division via a common

“conversion factor”, or RVU “worth”, as described by Wendy Cox‐Largent’s Declaration ¶ 28. Because

the conversion factor is calculated by division, it varies both among divisions within the same

department and between different departments. What this means is that the same number of RVUs

generated by physicians in different divisions and among different departments are worth different

revenue amounts to SIU‐HC. Consequently physicians in different divisions and departments are

compensated differently even if the physicians’ RVUs were to be the same.

3:15-cv-03308-SEM-TSH # 48-4 of 107

already existed, by combining all prospective members into one group when

applying his multiple regression models. In other words, Dr. Sharp has asserted

what he needs to show. Dr. Sharp is correct that a multiple regression analysis, with

the correct model specification, can be a standard and widely‐accepted methodology

to study gender pay disparity. However without a more rigorous analysis with the

above steps, a mere mechanical application of a regression model to a group of

physicians says nothing about whether the group is “similarly situated” or not.

101. Dr. Sharp admitted in his April 7, 2017 deposition (“Sharp Deposition”) that Dr.

Ahad would not be “similarly situated to a male physician practicing urology”, and

nor would she be “similarly situated to an orthopedic surgeon.”90 When asked what

Dr. Sharp meant by the phrase “similarly situated”, he replied that he meant

“controlling for all of these factors that we can control for … the two genders.

Controlling for all of the factors that we possibly can between them, so that we can

make a comparison between the genders.”91 Factors referred to by Dr. Sharp as

controls include “specialties, the departments they’re in, where they went to school,

the fellowships, the residencies … the year we’re talking about, whether or not

they’re the chair of the department, the number of publications they have.”92

102. Unfortunately Dr. Sharp is confusing two things: i) decision‐making units, versus

ii) factors considered by a decision‐maker to set pay. “[D]epartments they’re in” are

i), decision‐making units used to assign employees to distinct decision‐makers.

Once employees are properly assigned under i), then ii) as a potential set of gender‐

neutral labor market‐related and individual productivity‐related factors, can be used

as controls in a gender pay study. It is entirely possible that different Departments

value the same factors differently, and the effects of the set of control factors (i.e.,

90 Sharp Deposition 66: 24‐67:6 91 Sharp Deposition 67: 13‐67:25 92 Sharp Deposition 66: 14‐19 (italics added)

3:15-cv-03308-SEM-TSH # 48-4 of 107

coefficient estimates) included in ii) may be different for different decision‐making

units in i).

103. The way Dr. Sharp controlled for departments is by constructing a set of

department (or organization) “dummy variables”. For example, the surgery

department dummy variable in Dr. Sharp’s models assigns physicians in that

department a value of one, and others not in that department a value of zero.

Coefficient on a department dummy variable in Dr. Sharp’s regression models was

intended to measure the average difference in compensation between that

department, and the reference department (Alzheimerʹs Center), everything else

being equal (see Sharp Deposition 92:11‐20, question: “are you telling us that the

model suggests that somebody practicing in the department of neurology clinic,

$157,000 less in academic pay than somebody in the Alzheimerʹs Center?” Answer:

“… all else constant, yes…”). This overly simplistic treatment assumes that, among

other things, if one were to compare two otherwise identical physicians (same

medical school, same year of graduation, same year starting at SIU, same rank

progression, same publication accumulation, etc.,), with the only difference being

that one worked at the neurology clinic whereas the other worked at Alzheimer’s

Center, Dr. Sharp’s model93 would predict that every year their SIU academic base

would be different by approximately $157,000, and that difference would neither

shrink nor grow as the years go by. This somewhat extreme estimation from Dr.

Sharp’s regression model results from the problematic treatment of combining

physicians across all departments together in one regression model. Merely

applying “department dummy variables” as controls does not cure the problem. 94

The correct method would have been to analyze department‐by‐department.

93 Sharp Report Exhibit 2 Academic Base 94 As a technical note, if departments are to be used as “controls” rather than as decision‐making units,

the only proper way to do so is to apply a “fully interactive regression model”, whereby the set of

3:15-cv-03308-SEM-TSH # 48-4 of 107

B. Dr. Sharp’s methodology cannot be utilized to calculate class‐wide damages

104. Not performing a rigorous analysis by going through the steps described above

introduces another problem in Dr. Sharp’s empirical findings of alleged pay gaps.

Assuming that his regression models were correctly specified (which they are not),

without taking into account decision‐making units, the average gender pay gap Dr.

Sharp estimated for SIU‐SOM and SIU‐HC may combine units that showed gender

disparity, with units that were either gender neutral in their pay practices or favored

female physicians.

105. Dr. Sharp proposes that multiple regression analysis can be utilized to calculate

class‐wide damages.95 However his report offers no further description of how the

class‐wide damages calculation can be performed. When questioned at deposition

how he would use the multiple regression analysis to calculate damages for Dr.

Ahad, he responded that “it would suggest that every female’s [damages] is the

coefficient on gender, but you could probably fine tune that to get it just for Dr.

Ahad.”96 Subsequently when questioned how he could then use his model to

identify which individual female physicians suffered damages, and whether it

would be all of them or just a subset of them, Dr. Sharp answered that his model

would say all of them suffered the average amount equivalent to the gender

coefficient estimate.97

106. The proposition of calculating class‐wide damages as envisioned by Dr. Sharp is

problematic in that female physician faculty in decision‐making units with either no

department dummy variables “interacts” with all other factors in the regression, but this is not what Dr.

Sharp did. 95 Sharp Report ¶11 and ¶52 96 Sharp Deposition 20:2‐9 97 Sharp Deposition 20:11‐16

3:15-cv-03308-SEM-TSH # 48-4 of 107

gender disparity in pay or higher average pay than male physician faculty members,

suffered no economic damages.

C. Dr. Sharp’s methodology cannot fully address Plaintiff’s claims

107. An issue with Dr. Sharp’s methodology that is related to damages calculation,

but more so to Plaintiff’s claims, is that Dr. Sharp did not analyze total

compensation. When asked about his understanding of the claim that Dr. Ahad is

making, Dr. Sharp confirmed in his testimony that it was total compensation that Dr.

Ahad claiming to have been underpaid, as opposed to either or both academic and

clinical compensation as independent elements.98 To calculate economic damages,

according to Dr. Sharp, a physician faculty “overcompensated in one [compensation

component] and undercompensated in the other …. if that overcompensation and

under‐compensation, if they balance out equally, then there would be net, no

damages.”99

108. Although Dr. Sharp understood that total compensation should have been analyzed,

he did not do so. Instead, he ran separate regression models for SIU‐SOM

compensation and SIU‐HC compensation components. Those models cannot fully

identify which female physician faculty members were underpaid in total

compensation and which ones were not.100

109. As Dr. Sharp explained in deposition, the reason why he did separate analyses

for academic salary and clinical compensation, as opposed to putting them together

and running a regression model on total compensation, is because the academic data

98 Sharp Deposition 71:16‐23 99 Sharp Deposition 26: 2‐10. 100 Even if the SIU‐SOM regression model and the SIU‐HC model were to be correctly specified, a separate

analysis of each compensation component would only identify female physician faculty with pay

disparity in total compensation only if they were to be underpaid in both components. If they were

overpaid in one component but underpaid in the other, one could not determine whether pay disparity

existed in total compensation without performing an analysis on total compensation.

3:15-cv-03308-SEM-TSH # 48-4 of 107

from SIU‐SOM is on a fiscal‐year basis, while the clinical data from SIU‐HC is on a

calendar‐year basis.101 However, the timing gap is no excuse for not analyzing total

compensation. Because fiscal‐year SIU‐SOM Base can be converted into calendar

year SIU‐SOM Base by using the information in the 542 SIU‐HC compensation

documents.102

110. An expert’s analysis is relevant if it directly addresses the claim of the case,

which, as correctly stated by Dr. Sharp’s deposition testimony, is a dispute about

total compensation, not about two separate compensation components. Noting a

potential data constraint in the electronic data is no justification for not attempting

to perform an analysis directly addressing Plaintiff’s claims. One method Dr. Sharp

could have used to overcome this constraint is to identify the SIU‐SOM salary

change between fiscal years and calculate a weighted average of the January‐June

and the July‐December SIU‐SOM Base. Even though the weighted average may not

be perfect, nonetheless it would have been a reasonable approximation of a calendar

yearend SIU‐SOM Base. Not analyzing total compensation limits the application of

Dr. Sharp’s proposed regression methodology to address the issues of class

certification, merit, and damages, all of which are in connection with claims of

gender pay disparity in total compensation.103

D. Dr. Sharp did not properly model multi‐year compensation correlations

111. One of the well‐known empirical facts about salary data is that individuals’

salaries tend to be positively correlated overtime. For example, an employee’s

101 Sharp Deposition 71:8‐15 102 The SIU P&S, Inc. Compensation Agreement, to the extent available in the production, lists physician

faculty’s SIU‐SOM proposed compensation at the beginning of every fiscal year. Because the SIU‐SOM

compensation is paid out every month in equal installments, one could infer calendar‐year end total

SIU‐SOM compensation. 103 SIU’s faculty recruitment tracking form also uses AAMC total compensation as the market benchmark.

This is another reason why total compensation should have analyzed.

3:15-cv-03308-SEM-TSH # 48-4 of 107

previous salary increase is likely accompanied by a current salary increase, followed

by a future salary increase. The regression analysis utilized by Dr. Sharp is an

“ordinary least squares” analysis, which assumes that the error term ɛi,j,k,l,m,t104 is

independently and identically distributed (“i.i.d”), for each physician i at year t. The

physician faculty compensation data is a “panel data”, meaning that each

compensation record is for a particular physician faculty in a given year. A

physician faculty could have only one year of compensation data, or could have six

years of compensation data between 2011 and 2016. Because of the presence of

multi‐year compensation records for the same physician faculty in this panel data

structure, and the fact that individual salaries tend to be positively correlated (so

potentially violating the “i.i.d” assumption by introducing positive covariance

between error terms for the same physician i but in different years t), an adequate

regression model needs to adjust the standard error of the estimate. Without this

adjustment, the statistical significance of the regression coefficients can be grossly

over‐stated.

112. Dr. Sharp did not make this adjustment.105 As I will show in the next sub‐section,

the statistical significance of coefficient estimates in Dr. Sharp’s models is generally

over‐stated. Furthermore in at least some models, the female coefficient estimate

changes from being statistically significantly negative (indicating average

underpayment) to not being statistically significant, at the 5% level.

104 For example, see Sharp Report ¶s 43‐44 for the error term definition in Dr. Sharp’s SIU‐SOM and SIU‐

HC pay models in Exhibit 2. 105 I find it surprising that Dr. Sharp did not make the standard error correction to account for the positive

correlation of compensation over time for the same individuals in the SIU data, even though he must

have been knowledgeable about this topic. This is because his past scholarly publications were based

on empirical research using the Panel Studies of Income Dynamics (PSID) (For example, D.C. Sharp et

al., “But can she cook? Women’s education and housework productivity”, Economics of Education

Review 23 (2004) 605‐614).

3:15-cv-03308-SEM-TSH # 48-4 of 107

E. Partial correction of Dr. Sharp’s regression models

113. After reviewing the data and program production from Dr. Sharp accompanying

his Report, I modified his regression models step‐by‐step to address my criticisms in

Sections VIII.A‐D.

114. First, I applied the standard error adjustment described in Section VIII.D. Exhibit

9 presents the effect of this adjustment, using Dr. Sharp’s Exhibit 2 SIU‐SOM pay

and SIU‐HC pay as an illustration. As expected, statistical significance of the female

coefficient is substantially reduced by this adjustment. While the female coefficient

for SIU‐SOM pay remains the same, the t‐value changes to ‐1.93 which is no longer

statistically significant at the 5% level. Exhibit 10 presents the effect of this

adjustment to Dr. Sharp’s Exhibit 4 SIU‐Academic Base, showing a reduction in the

negative female coefficient to a level that is no longer statistically significant (t‐value

= ‐1.78) at the 5% level.

115. Second, using Dr. Sharp’s data and regression specification, but instead

performing the regressions by Department with standard error adjustments, I find

no common proof of a gender pay disparity across Departments. Exhibit 11

presents the result from a partial modification of Sharp Report Exhibit 2 to

regressions by Department. There is no consistent pattern of female disparity that

can be observed from Exhibit 11 – in fact the majority of the Departments do not

show statistically significant shortfall in female pay.

F. Dr. Sharp’s models generated absurd predicted compensation values

116. In a multiple regression model on compensation, the estimated coefficients for all

the explanatory variables can be used to calculate predicted values of compensation

for each individual physician faculty member. When a particular regression

specification models, or “fits”, the underlying compensation setting process

3:15-cv-03308-SEM-TSH # 48-4 of 107

reasonably well, one would expect that the predicted values of compensation would

be reasonably close to the actual compensation reported in the data. Ideally, the

difference between predicted and actual values would be small, and the ratio of the

predicted‐to‐actual values would be close to one – when that is the case, the

regression model is said to have close to perfect “fit”, and the corresponding

adjusted‐R square, a measure of the “goodness of fit”, is close to 100%.

117. To evaluate Dr. Sharp’s regression model fit, I calculated predicted values of the

compensation based on his estimated regression coefficients based on Dr. Sharp’s

Exhibit 2 model specification.

118. The majority of the calculated ratios fall somewhere between 0.15 and 4.

However, there are some SIU‐SOM predicted‐to‐actual pay ratios as high as 12.

There are also a number of individual records, where Dr. Sharp’s model predicts

SIU‐HC Pay 20 to 30 times, or even 200 times and 3,000 times as high compared to

the actual SIU‐HC Pay observed in the data. There are also some records where Dr.

Sharp’s models predict negative SIU‐SOM Pay and SIU‐HC Pay. I investigated some

of these records in more detail.

119. Exhibit 12‐A lists all available historical compensation records and predicted

values from Dr. Sharp’s model specification based on his Exhibit 2, for Dr. Krishna

Anjali Singh, MD, a female, who was an Assistant Professor of Clinical Surgery in

the Surgery Department from 2006 to 2013, and was terminated in the middle of CY

2013, making her 2013 SIU‐HC Pay record a “partial year” record. Dr. Sharp

analyzed her records for FY2012 and FY2013. Had Dr. Sharp carefully reviewed her

records, he would have noticed that the extremely low SIU‐HC Pay of $57 should

have been excluded. Including this outlier record resulted in his SIU‐HC model

prediction to be 3,194 times higher than the actual SIU‐HC Pay. Dr. Sharp should

have investigated those unusually high predicted values of compensation from his

3:15-cv-03308-SEM-TSH # 48-4 of 107

regressions relative to employees’ actual compensation.106 Furthermore

inappropriately including this extremely low pay record for a female physician in

his analyses of gender differences, Dr. Sharp’s model likely overestimated the

negative effect of gender on compensation.

120. Exhibit 12‐B lists all available historical compensation records and predicted

values from Dr. Sharp’s model specification based on his Exhibit 2, for Dr. Stephen

P. Stone, MD, a male physician and a professor at the Clinical Internal Medicine

Division at the Internal Medicine Department since 2000. Dr. Sharp significantly

overestimated SIU‐SOM pay component for Dr. Stone (as much as 7 or 8 times each

year), while underestimated Dr. Stone’s SIU‐HC pay by about half.

121. Exhibit 12‐C lists all the employees for whom Dr. Sharp’s Exhibit regression

models would have predicted negative SIU‐SOM pay and/or SIU‐HC pay. For

example for Dr. Allen Devleschoward at the Neurology Department, Dr. Sharp

predicted negative SIU‐SOM Pay for FY2012 and FY2013 and negative SIU‐HC Pay,

for 2016. All the SIU‐HC negative predicted pay happened to be for FY2016. Dr.

Sharp may not have realized that the 2016 SIU‐HC actual pay was only for the first

half of the 2016 calendar year. Had he taken the extra step to review his model

predictions, Dr. Sharp could have identified those data issues.107

106 I have performed a thorough review of my model results, including examining predicted values. I was

able to identify an outlier record for ) in 2011. Dr. Rodriguez

was hired on 7/1/2009 as an Instructor (SIU PRODUCTION 26947‐48). Her status changed to Assistant

Professor effective 09/26/2011 (SIU PRODUCTION 26944‐45). Her 2011 record should be excluded since

she only spent part of 2011 in the Assistant Professor rank, which explains why her pay was low in

2011. 107 I identified a part‐time employee, initially through a

vetting process of the predicted values from my regression models. I found out that Dr. Pardo had a

part‐time appointment with SIU (SIU PRODUCTION 25113‐25118): “Pediatric Endocrinology services

are not available in Central & Southern Illinois since Dr. Pardo’s retirement from private practice. The

department of Pediatrics will be searching for a full‐time Pediatric Endocrinologist, with an anticipated

18‐36 month time period. Dr. Pardo has agreed to provide Pediatric Endocrinology services on a part‐

time basis to fulfill our patient care and teaching needs.” (p. 4 Clinical Hire Request for SIU P&S

Member, signed in January 2008, SIU PRODUCTION 25118)

RedactedRedacted

Redacted Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

G. Other database construction and modeling issues in Dr. Sharp’s regressions

122. Dr. Sharp did not carefully review the underlying data. As a result, some of the

variables he constructed for his analyses are erroneous. I enumerate below some

problems found in Dr. Sharp’s analytical data, which impact the validity of his

regression models because these data are model inputs in Exhibits 2‐4 of the Sharp

Report.

1. Erroneous Calculation of SIU Experience (“years_at_siu” Variable)

123. To capture SIU experience, Dr. Sharp calculated the number of years at SIU

(“years_at_siu”) as the difference between the fiscal year starting date and an

employee’s hire date.108 For example, Dr. Wendell W. Becton was hired on March

18th, 2013. For fiscal year 2013 (which started on July 1st, 2012), Dr. Sharp calculated

a negative one (‐1) year of tenure for Dr. Becton (“years_at_siu” = 2012‐2013 = ‐1).

Using this logic, faculty hired from January to June of any year were assigned a

negative one (‐1) year of SIU experience in the first hiring fiscal year. I found 23

such instances in Dr. Sharp’s analytical data, shown in Table 3 below.

108 Specifically, Dr. Sharp relied on the year portion of the fiscal year‐start and employee hire dates.

3:15-cv-03308-SEM-TSH # 48-4 of 107

124. Dr. Sharp assigned the same value of SIU experience to someone hired in

January and December of the same calendar year, because he used only the year

portion of an employee’s hire date. Table 4 below shows two employees hired

almost a year apart but were assigned the same SIU experience by Dr. Sharp. A

better method is to use exact dates also shown in Table 4.

125. Dr. Sharp did not properly account for termination (non‐employment) spells and

subsequent SIU re‐employment in the construction of his “years_at_siu” variable.

Below are a few examples:

Table 3

Records with Negative Years at SIU in Dr. Sharpʹs Regression Dataset

fy emp name hire_date fy_start_date_true years_at_siu

13 Becton, Wendell W 3/18/2013 7/1/2012 ‐1

11 Shea, Laura K 1/24/2011 7/1/2010 ‐1

11 Lobas, Jeffrey G 2/22/2011 7/1/2010 ‐1

11 Loesel, Laura L 4/18/2011 7/1/2010 ‐1

11 Muller, Merle H 6/6/2011 7/1/2010 ‐1

11 Thiruvasahar, Thamilvani 6/27/2011 7/1/2010 ‐1

11 Shrestha, Santosh 6/27/2011 7/1/2010 ‐1

13 McVary, Kevin T 1/14/2013 7/1/2012 ‐1

13 Holder, Larry W 2/4/2013 7/1/2012 ‐1

13 Shad, Mujeeb U 4/15/2013 7/1/2012 ‐1

14 Dixon, William H 1/21/2014 7/1/2013 ‐1

14 Beck, Stephen D 1/27/2014 7/1/2013 ‐1

14 Rauschkolb, Paula K 6/2/2014 7/1/2013 ‐1

14 Couri, Daniel M 6/16/2014 7/1/2013 ‐1

15 Taylor, Funminiyi A 1/5/2015 7/1/2014 ‐1

15 Mathews, Ranjiv I 1/5/2015 7/1/2014 ‐1

15 Bhamidipati, Prasanta L 2/16/2015 7/1/2014 ‐1

15 Flack, John M 5/4/2015 7/1/2014 ‐1

15 Calder, Kevin A 5/18/2015 7/1/2014 ‐1

16 van Ulft, Stephanie A 1/4/2016 7/1/2015 ‐1

16 Tripathy, Shreepada 1/11/2016 7/1/2015 ‐1

16 Crabtree, Traves 6/16/2016 7/1/2015 ‐1

16 Karasis, Michael J 6/6/2016 7/1/2015 ‐1

Note: Fiscal year starts on July 1st of the previous calendar year and ends on June 30th of current calendar year.

Table 4

Dr. Sharpʹs Calculation of Years at SIU Is Inaccurate

Dr. Sharpʹs Calculation Corrected Calculation

fy_start_date_true years_at_siu fy_end_date corrected_exp_siu

12 Nordeman, Linda J 1/1/2010 7/1/2011 1 6/30/2012 2.50

12 Brard, Laurent 12/13/2010 7/1/2011 1 6/30/2012 1.55

fy emp name hire_date

RedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedacted

RedactedRedacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

a) was originally hired on 07/01/2005.109 He

submitted his resignation with the effective termination date of 09/30/2011,

although there was an attempt to rescind it later.110 He was re‐hired effective

10/27/2014.111 Due to 2011 resignation, his years of experience at SIU

should be reduced by approximately three years after he was re‐hired in 2014. As

of the beginning of FY 2016, had almost seven (7) years of actual

tenure with SIU, and not ten (10) as Dr. Sharp mistakenly calculated based only

on the original hire date.

b) was originally hired on 10/13/2008,

terminated on 08/03/2012,112 later re‐hired on 12/01/2014.113 Upon re‐hire, her SIU‐

SOM Base increased from $75,000 to $250,000.114 As of the beginning of FY2016,

had around 5.5 years of actual combined tenure with SIU, and not

just one (1) year as Dr. Sharp calculated based on the most recent hire date.

c) was originally hired on 06/23/2007, terminated

(resigned) on 06/30/2009, and later re‐hired on 07/01/2014.115 Upon re‐hire, his

SIU‐SOM Base increased from $70,000 to $200,000.116 Again Dr. Sharp did not

account for first SIU stint in calculation of the SIU tenure.

126. I understand that the Original Compensation Data (see Attachment B) may not

contain full information on earlier termination and subsequent rehire dates.

However, gaps such as the one presented in case, where the subsequent

adjacent observation to fiscal year 2012 is fiscal year 2016, should have been

109 SIU PRODUCTION 25920‐ 25924 110 SIU PRODUCTION 25936‐ 25938 111 SIU PRODUCTION 25912‐ 25918, and 25930‐ 25931 112 SIU PRODUCTION 31811‐31812 113 SIU PRODUCTION 31847‐31849, and 31878‐31879 114 had a start‐up salary agreement at the time of each hire. 115 SIU PRODUCTION 32459, 32487 and 32489 116 Dr. Sana requested to go off the guaranteed start‐up salary immediately following 2014 re‐hire (SIU

PRODUCTION 32477‐32478).

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

properly vetted. Had Dr. Sharp reviewed SIU‐HC compensation

documents, he would have been able to properly account for years not employed by

SIU.

127. The example of shows that sometimes a rehire was

accompanied by a significant increase in academic salary. This could be because the

physician was recruited back when he/she has become much more established in

their fields. shows the opposite effect of a

rehire: His previous SIU‐SOM Base in 2010 was $95,693, while his SIU‐SOM Base in

2015 after rehire was $86,625. To properly model individual circumstances

surrounding rehires, I have included a rehire indicator in my analyses.

2. Not Accounting for Division Chiefs

128. Division Chiefs are rewarded with additional compensation in SIU‐SOM

Academic Base:

a) was offered a position of Chair of the Division

of Orthopaedics in the Department of Surgery upon hire on July 1, 2009, and

remained in this administrative position until he resigned in September 2015.117

He was offered a total compensation package of $800,000 annually (confirmed by

his actual compensation records).

b) Dr. David Griffen appears to have been a Chair for Emergency Medicine

Division in the Surgery Department from 01/01/2008 to present.118

c) became a Division Chief of Gynecologic

Oncology sometime in 2013.119 His compensation records indicate an increase in

the academic base pay from $250,000 in FY2013 to $374,000 in FY2014.120

117 See SIU PRODUCTION 25744‐47 and 25706. 118 See SIU PRODUCTION 21090‐95 and

http://www.siumed.edu/surgery/emergency_medicine/em_faculty.html

Redacted

Redacted

Redacted

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

d) is currently a Division Chief of Infectious

Diseases121. The Division Chief information was not captured by the electronic

data. However, the data shows that his academic base pay increased from

$77,130 in FY2011 to $145,000 as of January 18, 2011 (and again to $165,000 in

FY2013 when he was promoted to Professor).122

129. Despite the increases in total compensation for new administrative

responsibilities, Dr. Sharp did not use this information in his analyses. Division

Chiefs can be identified by researching the SIU‐SOM and SIU‐HC websites, by

reviewing PDF personnel documents produced by SIU, or through a request to

Plaintiff’s Counsel. I do not know whether Dr. Sharp had inquired Plaintiff’s

Counsel about Division Chiefs, but it does not appear that Dr. Sharp expanded the

extra effort to collect this information from either SIU websites or SIU production of

PDF personnel documents. Consequently his regressions produce biased results,

because the female pay disparity estimated in his regressions are attributable to

extra compensation male Division Chiefs received for taking on administrative

responsibilities.

3. Other Data Issues

130. Dr. Sharp included records that should not have been analyzed. For example, Dr.

Karen E. Broquet was named an Assistant Dean for Graduate Medical

Education at SIU in 2002, and has since been promoted to Associate Dean.123 Dr. Eric

119 See http://www.siumed.edu/ob/brard.html, accessed on 04/14/2017. 120 See SIU PRODUCTION 26482‐26483. 121 See http://www.siuhealthcare.org/doctor/janak‐koirala.html 122 “SIU SOM Physician Academic Base Salary (FY07‐FY16).xlsx” and SIU PRODUCTION 23763, 23766‐

23767 123 Dr. Karen Broquet’s HC Compensation Agreements (SIU PRODUCTION 25200‐25209) do not match

her Academic Base payments in “SIU SOM Physician Academic Base Salary (FY07‐FY16).xlsx”

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

J. Constance was named Associate Dean for students in October 1995124, and is

currently listed as the Associate Dean, Student Affairs and Admissions on the SIU‐

SOM org chart. 125 Neither of them should have been included because their titles

and administrative responsibilities are outside of the rank of assistant, associate, or

full professors.

131. At deposition, Dr. Sharp admitted that some of the academic salaries were

“abnormally small”.126 For example Dr. Sharp included in his analysis

in the Neurology Department. As discussed in

Section VII.B., SIU‐SOM Base has been around $1,600 per fiscal

year, an outlier which should have been removed.

132. Dr. Sharp’s Exhibit 3 SIU‐HC model predicted that every time a physician

faculty’s article got cited, that physician faculty’s SIU‐HC pay would increase by

$14.58.127 Dr. Sharp acknowledged in his deposition that inclusion of citations

“seems to have caused problems”, and that citations is “not an important variable

for physicians clearly.”128 He attributed the problem to “citations are, I think,

probably collinear with publications.”129 When asked intellectually, from an

economic point of view, why Dr. Sharp would expect to see a physician getting

more clinical income which is productivity driven, every time his/her article got

cited, Dr. Sharp did not have a good answer but instead simply conceded that “if

there were really a concern about this variable, you could take it out.”130 This is

another example of how Dr. Sharp’s SIU‐HC regressions did not make any common

(production from 1/11/2017). Dr. Broquet became Assistant Dean in 2002

(http://www.siumed.edu/news/Faculty/Faculty02/Broquet.htm, accessed on 04/27/2017). 124 http://www.siumed.edu/news/Faculty/Constance.htm, accessed on 04/27/2017 125 http://www.siumed.edu/news/orgchart.pdf 126 Sharp Deposition 87:6. 127 Sharp Deposition 57: 15‐19 128 Sharp Deposition 82: 2‐18 129 Sharp Deposition 58: 20‐59:14 130 Sharp Deposition 58: 6‐7

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

sense, because they did not control for the one critical measure that mattered for

SIU‐HC pay, which is the RVU measure.131

133. All of the above model specification errors and data issues make Dr. Sharp’s

results unreliable in assessing the Plaintiff’s claims.

IX. Conclusions

134. HeplerBroom LLC (“Counsel”), counsel for SIU‐SOM and SIU‐HC, asked me to

conduct economic and statistical analyses related to allegations of gender

discrimination in pay from October 27, 2010 to present, brought by Plaintiff Sajida

Ahad, M.D., a female physician. I was asked to evaluate Plaintiffs’ allegation that

the Defendant discriminated against Dr. Ahad and the prospective class of female

physicians by paying them less than their male counterparts, and to opine, from a

statistical perspective, whether common issues apply to all prospective class

members. I was also asked to review and evaluate the Sharp Report dated March 10,

2017.

135. To carry out those tasks, I first determined that Department is the proper

decision‐making unit for conducting gender pay analyses for SIU‐SOM and SIU‐HC,

based on information produced to me and on conversations with SIU PMKs. I then

proceeded to carry out rigorous analyses on the determination of total

compensation. The rigorous analyses started with a thorough validation of the

compensation data Counsel has provided. In addition to productivity variables

already populated in the electronic compensation data, I directed and supervised

my team to search and collect data for AAMC physician labor market benchmark

statistics by year, rank, department, and division; clinical production measure on

RVUs by year and by physician whenever applicable and available, Division

131 Dr. Sharp seemed to understand the link between RVU and SIU‐HC compensation (Sharp Deposition

53: 4‐12). However he did not model this relationship at all.

3:15-cv-03308-SEM-TSH # 48-4 of 107

conversion factors by reconciliation periods, and whether or not physicians signed

start‐up agreements with SIU‐HC and if so, duration of start‐up guarantee. All of

those productivity measures are critical determinants of physician faculty’s SIU total

compensation, without which the econometric models would be improperly

specified.

136. It is only after having carefully examined and understood the underlying data

that I applied the Ordinary Least Squares (OLS) regression technique to the

analytical database I have built, with proper econometric specifications.

137. My analyses of total compensation by Department show no common pattern of

gender pay disparity from 2010 to 2016 for physician faculty members employed by

SIU‐SOM and SIU‐HC among the seven Departments: Family and Community

Medicine, Internal Medicine, Neurology, Obstetrics and Gynecology (OBGYN),

Pediatrics, Psychiatry, and Surgery. I have tested the robustness of my analyses by

varying model specifications, and have reached the same conclusion regardless of

which model was used, that namely, there is no common proof that female

physician faculty members received statistically significant and unfavorably

unequal pay for equal work relative to their male colleagues while employed by SIU

from 2010 to 2016.

138. I have reached this conclusion by performing statistical analyses on total

compensation by Department, controlling for physician labor market benchmarks,

academic achievements, clinical productivity, and administrative responsibilities, all

of which are gender‐neutral market and productivity measures for total

compensation.

139. Finally, I have reviewed the report by Plaintiff’s economic expert, Dr. Sharp, and

have examined Dr. Sharp’s statistical analyses, particularly his multiple regression

analyses and the data input into those models. It is my opinion that Dr. Sharp’s

regression analyses do not support Plaintiff’s pay disparity claims on a class‐wide

3:15-cv-03308-SEM-TSH # 48-4 of 107

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 1 – AAMC Summary Statistics on Total Compensation by Academic RankFrom 2008/2009 to 2014/2015, for Seven Departments: Family Medicine, Medicine, Neurology, OB/GYN, Pediatrics, Psychiatry, and Surgery

Compensation for All Schools with M.D. or Equivalent Degree

Year Count25th

PercentileMedian

75th Percentile

Count25th

PercentileMedian

75th Percentile

Count25th

PercentileMedian

75th Percentile

Family Medicine

2008-2009 1,339 $137,000 $151,000 $171,000 541 $154,000 $170,000 $190,000 284 $169,000 $188,000 $213,0002009-2010 1,281 $138,000 $155,000 $176,000 547 $153,000 $173,000 $194,000 291 $172,000 $190,000 $216,0002010-2011 1,308 $144,000 $159,000 $183,000 545 $157,000 $178,000 $201,000 310 $172,000 $194,000 $219,0002011-2012 1,350 $149,000 $166,000 $195,000 549 $164,000 $182,000 $210,000 335 $180,000 $201,000 $228,0002012-2013 1,416 $153,000 $171,000 $199,000 585 $165,000 $183,000 $209,000 332 $182,000 $204,000 $231,0002013-2014 1,488 $158,000 $178,000 $204,000 547 $171,000 $191,000 $215,000 337 $188,000 $207,000 $236,0002014-2015 1,533 $165,000 $182,000 $211,000 572 $176,000 $195,000 $221,000 349 $195,000 $212,000 $242,000

Medicine

2008-2009 7,296 $138,000 $165,000 $214,000 3,824 $164,000 $199,000 $257,000 3,873 $196,000 $238,000 $303,0002009-2010 7,775 $142,000 $173,000 $225,000 4,066 $167,000 $202,000 $265,000 4,126 $203,000 $247,000 $314,0002010-2011 8,077 $146,000 $180,000 $231,000 4,155 $171,000 $209,000 $270,000 4,125 $208,000 $254,000 $326,0002011-2012 8,772 $153,000 $188,000 $242,000 4,315 $177,000 $217,000 $278,000 4,225 $214,000 $260,000 $335,0002012-2013 9,137 $159,000 $196,000 $251,000 4,459 $182,000 $222,000 $289,000 4,341 $218,000 $267,000 $345,0002013-2014 9,653 $165,000 $202,000 $261,000 4,581 $186,000 $230,000 $298,000 4,415 $223,000 $275,000 $356,0002014-2015 10,193 $172,000 $208,000 $266,000 4,719 $193,000 $237,000 $307,000 4,577 $228,000 $282,000 $367,000

Neurology

2008-2009 981 $129,000 $145,000 $175,000 556 $156,000 $180,000 $210,000 613 $186,000 $216,000 $251,0002009-2010 1,022 $132,000 $150,000 $180,000 575 $159,000 $183,000 $219,000 644 $188,000 $220,000 $256,0002010-2011 1,110 $136,000 $155,000 $188,000 569 $165,000 $189,000 $224,000 667 $191,000 $226,000 $261,0002011-2012 1,167 $143,000 $163,000 $199,000 579 $172,000 $196,000 $230,000 677 $199,000 $236,000 $269,0002012-2013 1,180 $149,000 $170,000 $210,000 595 $177,000 $201,000 $237,000 700 $207,000 $245,000 $280,0002013-2014 1,286 $152,000 $178,000 $214,000 599 $180,000 $209,000 $247,000 699 $205,000 $246,000 $290,0002014-2015 1,369 $160,000 $186,000 $230,000 648 $182,000 $212,000 $256,000 746 $212,000 $251,000 $296,000

ProfessorAssistant Professor Associate Professor

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 1 – AAMC Summary Statistics on Total Compensation by Academic Rank (Cont’d)

Year Count25th

PercentileMedian

75th Percentile

Count25th

PercentileMedian

75th Percentile

Count25th

PercentileMedian

75th Percentile

OB / GYN

2008-2009 1,198 $175,000 $210,000 $264,000 620 $213,000 $251,000 $307,000 407 $231,000 $301,000 $372,0002009-2010 1,281 $182,000 $216,000 $278,000 601 $218,000 $255,000 $325,000 431 $237,000 $304,000 $380,0002010-2011 1,343 $190,000 $228,000 $294,000 588 $219,000 $263,000 $339,000 443 $249,000 $314,000 $375,0002011-2012 1,411 $195,000 $237,000 $300,000 593 $230,000 $276,000 $351,000 443 $257,000 $325,000 $401,0002012-2013 1,536 $200,000 $241,000 $303,000 611 $235,000 $282,000 $361,000 474 $264,000 $332,000 $407,0002013-2014 1,569 $205,000 $246,000 $308,000 639 $239,000 $289,000 $364,000 481 $269,000 $336,000 $423,0002014-2015 1,688 $212,000 $255,000 $325,000 654 $248,000 $301,000 $376,000 495 $276,000 $345,000 $437,000

Pediatrics

2008-2009 4,058 $129,000 $150,000 $180,000 2,013 $152,000 $181,000 $221,000 1,728 $179,000 $214,000 $258,0002009-2010 4,396 $131,000 $153,000 $185,000 2,125 $155,000 $185,000 $225,000 1,836 $185,000 $221,000 $269,0002010-2011 4,632 $133,000 $155,000 $189,000 2,162 $159,000 $190,000 $229,000 1,850 $190,000 $227,000 $278,0002011-2012 4,910 $138,000 $162,000 $196,000 2,218 $162,000 $195,000 $236,000 1,867 $195,000 $232,000 $278,0002012-2013 5,046 $142,000 $166,000 $200,000 2,283 $165,000 $197,000 $240,000 1,967 $199,000 $238,000 $286,0002013-2014 5,254 $145,000 $171,000 $207,000 2,395 $170,000 $203,000 $246,000 1,966 $206,000 $247,000 $299,0002014-2015 5,893 $151,000 $177,000 $214,000 2,561 $176,000 $211,000 $254,000 2,073 $211,000 $254,000 $305,000

Psychiatry

2008-2009 1,523 $139,000 $157,000 $182,000 637 $154,000 $175,000 $199,000 657 $181,000 $213,000 $254,0002009-2010 1,423 $141,000 $162,000 $185,000 614 $158,000 $178,000 $206,000 664 $183,000 $212,000 $260,0002010-2011 1,500 $146,000 $165,000 $190,000 642 $163,000 $186,000 $214,000 666 $186,000 $218,000 $266,0002011-2012 1,538 $150,000 $170,000 $197,000 647 $168,000 $189,000 $218,000 689 $193,000 $222,000 $273,0002012-2013 1,534 $154,000 $175,000 $200,000 632 $173,000 $193,000 $223,000 689 $197,000 $227,000 $280,0002013-2014 1,611 $160,000 $180,000 $210,000 618 $178,000 $199,000 $233,000 694 $200,000 $231,000 $289,0002014-2015 1,660 $165,000 $187,000 $215,000 651 $184,000 $205,000 $240,000 698 $206,000 $242,000 $291,000

Assistant Professor Associate Professor Professor

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 1 – AAMC Summary Statistics on Total Compensation by Academic Rank (Cont’d)

Year Count25th

PercentileMedian

75th Percentile

Count25th

PercentileMedian

75th Percentile

Count25th

PercentileMedian

75th Percentile

Assistant Professor Associate Professor Professor

Surgery

2008-2009 3,315 $224,000 $280,000 $377,000 1,931 $283,000 $360,000 $470,000 1,778 $293,000 $388,000 $500,0002009-2010 3,486 $235,000 $300,000 $400,000 1,978 $293,000 $371,000 $486,000 1,819 $310,000 $406,000 $527,0002010-2011 3,689 $247,000 $308,000 $410,000 2,008 $301,000 $390,000 $505,000 1,885 $321,000 $419,000 $549,0002011-2012 3,892 $252,000 $320,000 $425,000 2,103 $314,000 $408,000 $525,000 1,938 $327,000 $440,000 $574,0002012-2013 4,045 $261,000 $328,000 $438,000 2,183 $320,000 $418,000 $547,000 2,063 $336,000 $452,000 $600,0002013-2014 4,078 $274,000 $339,000 $452,000 2,260 $330,000 $425,000 $567,000 2,105 $350,000 $464,000 $617,0002014-2015 4,263 $283,000 $350,000 $471,000 2,371 $346,000 $435,000 $579,000 2,198 $361,000 $485,000 $650,000

Sources: AAMC Faculty Salary Reports, Table 11, 2008-2009 to 2014-2015.

Notes: Dermatology is not included under AAMC's 'Internal Medicine' (it is a separate department under the AAMC categorization). 'Emergency Medicine', 'Physical Medicine and Rehabilitation', and 'Otolaryngology' are not part of AAMC's Surgery department reported in this exhibit. AAMC included 'Child & Adolescent Psychiatry' division under Pediatrics in 2009-2010.

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 2 - Example Records from Analytical Database

Year Rank Tenure StatusSOM Base Pay

SOM Extra Pay

HC Pay Total Pay RVUsAAMC Mean

AAMC Median

Year of Med School Graduation

(YOG)

Cumulative Publications

Total Experience

Current Startup

Indicator

Rehire Indicator

Department Chair

Indicator

Division Chief

Indicator

Grant Indicator

[a] [b] [c] [e] [f] [i] [k] [l] [m]

Gender=F, Department=Internal Medicine, Division=Infectious Disease, Degree=M.D., Hire Date=09/01/20082010 ASSISTANT Alternate Track $65,000 $1,732 $72,024 $138,756 4,019 $139,800 $134,000 2002 2 8.5 0 0 0 0 12011 ASSISTANT Alternate Track $65,000 $1,583 $64,863 $131,447 3,134 $141,800 $134,000 2002 3 9.5 0 0 0 0 02012 ASSISTANT Alternate Track $71,250 $0 $81,318 $152,568 4,453 $145,900 $138,000 2002 5 10.5 0 0 0 0 02013 ASSISTANT Alternate Track $80,000 $0 $91,857 $171,857 5,183 $148,900 $140,000 2002 5 11.5 0 0 0 0 02014 ASSISTANT Alternate Track $80,000 $0 $86,710 $166,710 4,755 $153,900 $144,000 2002 7 12.5 0 0 0 0 02015 ASSOCIATE Alternate Track $84,167 $0 $78,050 $162,217 3,819 $190,600 $180,000 2002 7 13.5 0 0 0 0 12016 ASSOCIATE Alternate Track $90,000 $0 $94,305 $184,305 4,814 $195,700 $186,000 2002 8 14.5 0 0 0 0 1

Gender=M, Department=Psychiatry, Division=Adult Psychiatry, Degree=M.D., Ph.D., Hire Date=07/14/2008, Term Date=07/31/20132010 ASSISTANT Alternate Track $96,667 $7,699 $73,903 $178,269 3,923 $162,700 $157,000 1992 5 18.5 0 0 0 0 02011 ASSISTANT Alternate Track $114,167 $3,303 $94,155 $211,624 5,223 $166,100 $162,000 1992 5 19.5 0 0 0 0 02012 ASSISTANT Alternate Track $125,808 $0 $81,901 $207,709 4,183 $171,900 $165,000 1992 5 20.5 0 0 0 0 0

Gender=M, Department=Surgery, Division=General Surgery, Degree=M.D., Hire Date=07/24/2006, Term Date=04/04/20132010 ASSISTANT Tenure Track $42,550 $65 $204,219 $246,834 5,643 $261,500 $243,000 1995 41 15.5 0 0 0 0 02011 ASSISTANT Tenure Track $85,050 $387 $196,189 $281,626 5,936 $266,600 $250,000 1995 45 16.5 0 0 0 0 12012 ASSISTANT Tenure Track $72,550 $31 $179,007 $251,588 4,854 $280,800 $262,000 1995 47 17.5 0 0 0 0 0

Gender=F, Department=Surgery, Division=General Surgery, Degree=M.D., Hire Date=07/28/2008, Term Date=03/21/20142010 ASSISTANT Tenure Track $125,000 $0 $117,459 $242,459 3,157 $261,500 $243,000 1998 3 12.5 1 0 0 0 02011 ASSISTANT Tenure Track $125,000 $0 $89,988 $214,988 3,544 $266,600 $250,000 1998 5 13.5 0 0 0 0 02012 ASSISTANT Tenure Track $125,000 $0 $79,139 $204,138 2,921 $280,800 $262,000 1998 6 14.5 0 0 0 0 02013 ASSISTANT Tenure Track $125,000 $0 $110,903 $235,903 4,244 $296,300 $276,000 1998 9 15.5 0 0 0 0 0

Gender=M, Department=Surgery, Division=Plastic Surgery, Degree=M.D., Hire Date=06/16/19972010 PROFFESSOR Tenure Granted $194,144 $1,546 $634,083 $829,772 13,578 $436,100 $409,000 1988 67 22.5 0 0 0 1 12011 PROFFESSOR Tenure Granted $448,244 $584 $560,413 $1,009,241 14,493 $457,600 $450,000 1988 75 23.5 0 0 0 1 12012 PROFFESSOR Tenure Granted $637,500 $12,500 $391,294 $1,041,294 8,902 $447,800 $436,000 1988 77 24.5 0 0 1 1 12013 PROFFESSOR Tenure Granted $750,000 $0 $373,210 $1,123,210 9,970 $501,200 $489,000 1988 86 25.5 0 0 1 1 12014 PROFFESSOR Tenure Granted $750,000 $4 $445,475 $1,195,479 12,196 $496,700 $485,000 1988 101 26.5 0 0 1 1 12015 PROFFESSOR Tenure Granted $750,000 $0 $412,217 $1,162,217 11,536 $530,100 $480,000 1988 105 27.5 0 0 1 1 12016 PROFFESSOR Tenure Granted $750,000 $0 $284,182 $1,034,182 9,034 $534,800 $530,000 1988 109 28.5 0 0 1 1 1

Notes:[a] "77V1461-Ahad S SIU Complete Physician List (By Division) 4-4-17.xlsx" [g] AAMC Faculty Salary Survey Reports[b] "SIU SOM Physicians Compensation (FY07-FY16).xlsx" [h] Variables that were taken from Dr. Sharp's Analysis[c] "SIU SOM Physician Academic Base Salary (FY07-FY16) xlsx" [i] Total Experience = (12 31.Year- 7.1.YOG + 1) / 365

[e] Total Pay =SOM Base + SOM Extra + HC [l] "Division Chair Listing 06_current xlsx"[f] "Ahad S Conversion Factor file w historical CFs 02-22-17 xlsx" [m] "SIU Carbondale grants and contracts funded since 01-01-07.doc", "Copy of Faculty with grants FY07 to FY16.xlsx", and "SIU

Springfield Campus grants and contracts funded 1-1-07 to 2-10-17 pdf"

[d] [g] [h] [j]

[j] HC Personnel Files[k] "76J9761-Ahad S Chair Listing 2006 through 2016 02-13-17 xlsx"

[d] "77V1461-Ahad S SIU Complete Physician List (By Division) 4-4-17.xlsx"; Total Comp was prorated to a full year for 2016 by muliplying by 2

Redacted

Redacted

Redacted

Redacted

Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 3 - Employee Count by Gender, SIU Physician Faculty Female Representation, and Academic Rank and Tenure Status DistributionsCalendar Years 2010-2016, By Department

Female Male

Academic Rank Tenure Status Academic Rank Tenure Status

Year Total % FemaleAssistant Professor

Associate Professor

ProfessorAlternate

TrackTenure Track

Tenure Granted

Assistant Professor

Associate Professor

ProfessorAlternate

TrackTenure Track

Tenure Granted

Family and Community Medicine2010 34 56% 11 4 4 19 5 6 4 13 1 12011 36 53% 11 4 4 19 8 5 4 16 12012 35 51% 10 6 2 18 9 4 4 16 12013 25 44% 6 3 2 11 9 2 3 142014 23 35% 4 2 2 8 11 2 2 152015 22 32% 3 2 2 7 11 2 2 152016 16 31% 3 1 1 5 9 1 1 11

Internal Medicine2010 44 39% 8 8 1 16 1 8 11 8 22 52011 47 36% 9 6 2 17 13 9 8 26 42012 50 38% 12 6 1 19 15 7 9 28 32013 54 39% 12 6 3 21 18 7 8 30 32014 55 40% 13 6 3 22 16 9 8 30 32015 62 45% 17 8 3 28 15 11 8 31 32016 62 44% 18 7 2 27 14 12 9 31 4

Neurology2010 8 2 2 4 5 32011 8 4 2 2 6 22012 9 5 2 2 7 22013 9 4 2 3 6 32014 6 3 2 1 5 12015 9 22% 2 2 3 3 1 6 12016 6 17% 1 1 2 2 1 4 1

Obstetrics and Gynecology2010 13 23% 2 1 3 5 3 2 7 1 22011 13 23% 2 1 3 6 3 1 7 1 22012 15 27% 3 1 4 6 4 1 7 2 22013 15 33% 4 1 5 6 4 8 1 12014 16 31% 3 2 5 5 5 1 7 3 12015 15 33% 3 2 5 4 5 1 6 3 12016 15 40% 4 2 6 4 4 1 5 3 1

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 3 - Employee Count by Gender, SIU Physician Faculty Female Representation, and Academic Rank and Tenure Status Distributions(Cont’d)

Female Male

Academic Rank Tenure Status Academic Rank Tenure Status

Year Total % FemaleAssistant Professor

Associate Professor

ProfessorAlternate

TrackTenure Track

Tenure Granted

Assistant Professor

Associate Professor

ProfessorAlternate

TrackTenure Track

Tenure Granted

Pediatrics2010 26 31% 6 2 8 10 3 5 15 32011 29 45% 11 2 13 9 2 5 14 22012 28 46% 11 1 1 13 6 5 4 13 22013 30 50% 10 4 1 15 6 6 3 13 22014 27 52% 9 4 1 14 8 5 132015 32 50% 11 4 1 16 9 5 2 15 12016 36 50% 12 5 1 18 11 5 2 17 1

Psychiatry2010 14 43% 4 2 6 7 1 82011 12 50% 4 2 6 5 1 62012 13 46% 3 3 6 6 1 72013 10 30% 1 2 3 6 1 72014 11 36% 2 2 4 5 1 1 72015 13 46% 4 2 6 4 2 1 72016 13 54% 5 2 7 4 2 6

Surgery2010 56 27% 9 3 3 3 7 5 11 16 14 10 11 202011 59 24% 10 2 2 4 7 3 13 17 15 11 13 212012 52 23% 7 3 2 4 5 3 15 13 12 10 14 162013 42 24% 5 3 2 3 4 3 10 12 10 9 9 142014 43 23% 6 3 1 4 3 3 13 10 10 10 11 122015 47 28% 8 4 1 2 7 4 14 12 8 8 14 122016 47 26% 8 3 1 2 7 3 16 10 9 8 15 12

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 4 – SIU Physician Faculty Average Total Compensation, Average SIU-HC Compensation, and Average RVUs by Gender

Calendar Years 2010-2016, By Department

Female Male

YearTotal Male

Total Female

Average Tot. Comp.

Average HC Comp.

AverageRVUs

Average Tot. Comp.

Average HC Comp.

AverageRVUs

Internal Medicine2010 27 17 $175,816 $68,755 2,870 $306,936 $134,067 5,646

2011 30 17 $180,835 $63,403 2,481 $298,431 $123,957 5,282

2012 31 19 $193,468 $70,145 2,723 $312,594 $130,012 5,539

2013 33 21 $205,853 $71,362 2,853 $313,065 $132,695 5,905

2014 33 20 $215,111 $65,976 2,718 $307,591 $119,442 5,289

2015 34 22 $238,064 $93,926 3,230 $304,931 $106,939 4,404

2016 31 23 $220,492 $84,905 3,136 $327,017 $120,086 4,609

Neurology2010 8 $221,411 $91,137 3,717

2011 8 $221,082 $83,745 3,409

2012 9 $252,906 $120,781 3,938

2013 9 $290,359 $119,302 4,111

2014 5 $285,988 $83,250 3,172

2015 5 1 $270,300 $80,300 1,637 $271,364 $79,276 2,473

2016 5 1 $322,491 $202,491 7,657 $348,189 $90,597 3,676

Obstetrics and Gynecology2010 10 3 $365,998 $138,069 3,166 $406,100 $206,129 4,235

2011 10 3 $369,748 $141,403 4,329 $384,526 $189,227 4,679

2012 11 4 $332,010 $152,445 4,698 $349,393 $182,304 5,104

2013 10 5 $337,498 $161,911 5,461 $407,937 $220,165 6,251

2014 10 5 $339,676 $163,255 4,035 $442,951 $250,785 6,510

2015 9 5 $356,202 $180,395 4,965 $480,117 $242,766 6,553

2016 9 5 $375,791 $197,627 4,764 $540,718 $295,291 6,805

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 4 – SIU Physician Faculty Average Total Compensation, Average SIU-HC Compensation, and Average RVUs by Gender

(Cont’d)

Female Male

YearTotal Male

Total Female

Average Tot. Comp.

Average HC Comp.

AverageRVUs

Average Tot. Comp.

Average HC Comp.

AverageRVUs

Pediatrics2010 18 7 $227,840 $92,752 3,362 $252,504 $90,736 3,409

2011 16 13 $240,214 $110,753 4,284 $274,719 $88,458 3,677

2012 15 13 $247,858 $107,121 4,227 $280,717 $87,014 3,612

2013 15 15 $249,173 $94,520 4,043 $287,160 $94,729 4,229

2014 12 13 $269,806 $110,428 4,600 $259,408 $116,899 4,776

2015 12 14 $266,075 $114,605 4,768 $265,138 $123,449 4,542

2016 15 12 $274,409 $118,520 4,784 $272,903 $106,373 4,414

Psychiatry2010 8 6 $161,553 $42,303 1,780 $176,029 $46,932 2,420

2011 6 5 $178,815 $43,104 2,157 $200,097 $56,826 3,044

2012 7 6 $165,777 $35,307 1,674 $213,108 $56,162 2,648

2013 7 3 $221,491 $39,467 2,071 $224,475 $67,790 3,305

2014 7 3 $234,639 $40,918 2,134 $259,276 $81,472 4,043

2015 7 4 $200,935 $41,179 2,013 $281,040 $86,550 4,119

2016 6 6 $176,015 $36,212 1,865 $279,723 $104,317 4,940

Surgery2010 37 12 $264,457 $147,085 4,274 $435,092 $251,815 7,402

2011 40 11 $278,278 $143,640 4,777 $443,468 $221,308 7,119

2012 35 10 $299,895 $157,856 4,948 $453,256 $229,891 7,255

2013 28 8 $366,582 $204,029 6,645 $486,618 $217,030 7,268

2014 26 7 $419,545 $216,787 7,044 $528,675 $221,007 7,430

2015 26 8 $424,518 $208,321 5,732 $501,129 $218,786 7,304

2016 25 9 $455,993 $228,771 6,256 $528,362 $234,900 8,216

Notes: Family and Community Medicine Department's RVUs are not used in the analyses due to the pooling arrangement. RVUs for the physician faculty in the Emergency Medicine Division of the Surgery Department were not available.

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 5-A - Family And Community Medicine

Total Compensation Regression Results – Using AAMC Median as the Labor Market Benchmark

Variable Name Estimate t-value ProbabilityFemale -13,698 -1.31 0.1953AAMC Median 1.93 4.53 0.0000 *MD Indicator 21,389 2.00 0.0518Division Chief 44,936 1.72 0.0915Department Chair 75,264 2.39 0.0209 *Total Experience 2,146 0.88 0.3816Total Experience Squared -38 -0.64 0.5231Tenure Track 40,560 3.28 0.0020 *Tenure Granted -4,524 -0.18 0.8586RVUs --- --- ---RVU Missing Indicator --- --- ---Current Startup Indicator --- --- ---Grant Indicator 21,727 1.32 0.1918Cumulative Publications 54 0.05 0.9595Rehire Indicator -61,105 -6.68 0.0000 *Constant -163,552 -2.33 0.0243 *

R-squared 0.621Adj R-squared 0.595N of obs 191

Notes: * statistically significant at 5%. Standard errors are clustered by employee ID.

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 5-B - Internal Medicine


Variable Name Estimate t-value ProbabilityFemale -17,340 -1.38 0.1706AAMC Median 1.32 6.53 0.0000 *MD Indicator --- --- ---Division Chief 38,304 1.40 0.1658Department Chair 142,167 3.27 0.0015 *Total Experience 6,526 1.96 0.0526Total Experience Squared -171 -2.18 0.0319 *Tenure Track --- --- ---Tenure Granted 28,939 0.71 0.4813RVUs 23 7.55 0.0000 *RVU Missing Indicator 101,634 4.27 0.0000 *Current Startup Indicator 11,335 1.14 0.2557Grant Indicator 15,168 0.51 0.6079Cumulative Publications -3 -0.01 0.9960Rehire Indicator --- --- ---Constant -165,060 -3.39 0.0010 *



3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 5-C - Neurology


Variable Name Estimate t-value ProbabilityFemale -32,921 -1.26 0.2248AAMC Median 2.69 4.38 0.0004 *MD Indicator 104,257 2.12 0.0487 *Division Chief --- --- ---Department Chair 112,160 1.70 0.1068Total Experience 6,884 0.75 0.4660Total Experience Squared -231 -1.23 0.2369Tenure Track --- --- ---Tenure Granted -33,035 -0.31 0.7626RVUs 9 0.93 0.3657RVU Missing Indicator -24,399 -0.46 0.6501Current Startup Indicator 39,049 0.65 0.5230Grant Indicator -22,858 -0.97 0.3437Cumulative Publications -845 -0.72 0.4843Rehire Indicator --- --- ---Constant -336,551 -2.63 0.0176 *



3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 5-D - Obstetrics And Gynecology


Variable Name Estimate t-value ProbabilityFemale 6,175 0.13 0.9013AAMC Median 1.18 2.64 0.0160 *MD Indicator --- --- ---Division Chief 62,443 1.32 0.2041Department Chair 71,679 0.43 0.6695Total Experience 11,931 1.16 0.2622Total Experience Squared -360 -1.58 0.1309Tenure Track 9,298 0.27 0.7909Tenure Granted 272,280 1.85 0.0799RVUs 19 2.30 0.0332 *RVU Missing Indicator 49,916 0.84 0.4127Current Startup Indicator 19,131 0.47 0.6443Grant Indicator -55,260 -1.19 0.2495Cumulative Publications 1,365 3.00 0.0074 *Rehire Indicator --- --- ---Constant -144,993 -1.16 0.2602



3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 5-E - Pediatrics


Variable Name Estimate t-value ProbabilityFemale -3,431 -0.20 0.8418AAMC Median 1.78 6.82 0.0000 *MD Indicator 188,256 3.59 0.0008 *Division Chief 88,306 1.94 0.0576Department Chair 177,819 3.77 0.0004 *Total Experience 5,115 1.78 0.0805Total Experience Squared -195 -3.09 0.0033 *Tenure Track --- --- ---Tenure Granted -17,442 -0.34 0.7343RVUs 10 3.75 0.0005 *RVU Missing Indicator 210 0.01 0.9899Current Startup Indicator 2,518 0.21 0.8345Grant Indicator -30,920 -1.27 0.2092Cumulative Publications 1,471 1.80 0.0775Rehire Indicator -167,221 -7.26 0.0000 *Constant -334,700 -4.52 0.0000 *



3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 5-F - Psychiatry


Variable Name Estimate t-value ProbabilityFemale -823 -0.05 0.9592AAMC Median 1.49 4.58 0.0001 *MD Indicator 89,883 3.86 0.0008 *Division Chief 13,310 0.95 0.3513Department Chair 45,393 2.16 0.0417 *Total Experience 2,145 0.68 0.5010Total Experience Squared -67 -0.84 0.4088Tenure Track --- --- ---Tenure Granted --- --- ---RVUs 17 3.94 0.0006 *RVU Missing Indicator -220 -0.01 0.9926Current Startup Indicator --- --- ---Grant Indicator 202 0.03 0.9787Cumulative Publications 2,460 1.26 0.2213Rehire Indicator --- --- ---Constant -206,307 -3.05 0.0057 *



3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 5-G - Surgery


Variable Name Estimate t-value ProbabilityFemale -6,178 -0.35 0.7244AAMC Median .62 4.99 0.0000 *MD Indicator 11,630 0.51 0.6097Division Chief 136,100 4.60 0.0000 *Department Chair 358,202 16.29 0.0000 *Total Experience 3,467 0.97 0.3349Total Experience Squared -163 -2.41 0.0178 *Tenure Track -4,932 -0.20 0.8413Tenure Granted -23,833 -0.94 0.3489RVUs 24 8.52 0.0000 *RVU Missing Indicator 143,501 5.49 0.0000 *Current Startup Indicator 23,746 1.72 0.0884Grant Indicator 20,512 1.59 0.1159Cumulative Publications 904 2.63 0.0101 *Rehire Indicator 98,963 5.14 0.0000 *Constant -18,592 -0.29 0.7710



3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 6 – Regression Analyses by Department Show No Female Pay Disparity That Was Statistically Significant at the 5% Level

Results Are Robust By Three Model Specifications with Different Market Benchmark Measures: AAMC Median, AAMC Mean, and Rank-Division-Year Controls

Total Compensation

Department Estimate t-value ProbabilityEmployee-Year Count

Female-Year

Count

AAMC Median

Family and Community Medicine -$13,698 -1.31 0.1953 191 87Internal Medicine -$17,340 -1.38 0.1706 374 151Neurology -$32,921 -1.26 0.2248 55 3Obstetrics and Gynecology $6,175 0.13 0.9013 102 31Pediatrics -$3,431 -0.20 0.8418 208 97Psychiatry -$823 -0.05 0.9592 86 38Surgery -$6,178 -0.35 0.7244 346 86

AAMC Mean

Family and Community Medicine -$12,575 -1.17 0.2489 191 87Internal Medicine -$19,925 -1.53 0.1301 374 151Neurology -$31,928 -1.18 0.2558 55 3Obstetrics and Gynecology $10,814 0.22 0.8275 102 31Pediatrics -$4,856 -0.27 0.7856 208 97Psychiatry -$1,523 -0.10 0.9245 86 38Surgery -$11,993 -0.70 0.4835 346 86

Rank-Division-Year

Family and Community Medicine -$16,007 -1.46 0.1504 191 87Internal Medicine -$15,000 -1.39 0.1679 374 151Neurology -$46,653 -1.59 0.1313 55 3Obstetrics and Gynecology -$623 -0.02 0.9858 102 31Pediatrics $2,567 0.15 0.8789 208 97Psychiatry -$15,533 -0.61 0.5466 86 38Surgery -$23,115 -1.64 0.1042 346 86

Notes: Standard errors are clustered by employee ID.

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 7-A - Faculty Members in the General Surgery Division at the Surgery DepartmentCalendar Years 2008 and 2009, the First Two Years of Dr. Ahad’s Employment with SIU

Partial Department Division Actual Calendar Year Pay

EmpID Name GenderYear

IndicatorChair

IndicatorChief

Indicator Hire DateTermination

DateRank Tenure Status SOM Base

SOM Extra

HC TotalStartup

GuaranteeRVUs

Trauma RVUs

Total RVUs

General Surgery 2008M 1 0 0 08/11/2008 ASSISTANT Tenure Track $54,167 $0 $27,336 $81,503 Yes 0 0 F 1 0 0 02/25/2008 07/15/2008 ASSISTANT Tenure Track $39,424 $0 $47,850 $87,273 Yes 423 1,020 1,443

60365 Ahad, Sajida F 1 0 0 07/28/2008 03/21/2014 ASSISTANT Tenure Track $43,478 $0 $44,692 $88,171 Yes 153 153

A F 1 0 0 06/30/2008 PROFFESSOR Tenure Granted $63,095 $0 $39,373 $102,469 Yes 3 3 F 0 0 0 07/01/1981 12/31/2014 ASSISTANT Alternate Track $54,808 $0 $77,512 $132,321 3,878 3,878 M 1 0 0 08/04/2008 06/30/2012 ASSOCIATE Tenure Granted $98,810 $0 $46,987 $145,796 Yes 161 161 M 0 0 0 07/24/2006 04/04/2013 ASSISTANT Tenure Track $41,921 $0 $197,258 $239,179 Yes - R 3,682 3,682 F 0 0 0 08/01/2001 04/30/2010 ASSISTANT Alternate Track $182,045 $0 $76,943 $258,988 2,554 2,554 M 0 0 0 08/01/2003 10/27/2015 ASSISTANT Tenure Track $140,952 $111 $153,608 $294,671 1,532 3,005 4,537 F 0 0 0 07/24/2000 10/15/2011 ASSOCIATE Tenure Granted $86,213 $0 $220,450 $306,663 5,138 5,138 F 0 0 0 10/29/2001 ASSOCIATE Tenure Granted $115,241 $0 $200,429 $315,670 4,898 4,898 M 0 0 0 07/01/2001 ASSOCIATE Tenure Granted $174,367 $1,307 $198,860 $374,535 2,007 4,082 6,089 M 0 0 0 08/01/2000 07/09/2012 PROFFESSOR Tenure Granted $284,990 $0 $110,442 $395,432 3,519 3,519 M 0 0 1 08/08/2005 03/31/2011 PROFFESSOR Tenure Granted $236,035 $0 $198,575 $434,610 5,963 5,963 M 0 1 0 09/01/1997 08/31/2012 PROFFESSOR Tenure Granted $524,163 $0 $158,516 $682,679 4,046 4,046

General Surgery 2009M 1 0 0 08/31/2009 06/02/2015 ASSISTANT Tenure Track $44,444 $0 $36,409 $80,853 Yes 0 0 F 0 0 0 07/01/1981 12/31/2014 ASSISTANT Alternate Track $55,631 $0 $42,114 $97,744 2,931 2,931 F 0 0 0 06/30/2008 PROFFESSOR Tenure Granted $150,000 $0 $89,887 $239,887 Yes 50 50 M 0 0 0 07/24/2006 04/04/2013 ASSISTANT Tenure Track $42,550 $0 $202,897 $245,447 5,308 5,308 F 0 0 0 08/01/2001 04/30/2010 ASSISTANT Alternate Track $184,776 $0 $66,221 $250,997 2,204 2,204 F 0 0 0 07/24/2000 10/15/2011 ASSOCIATE Tenure Granted $93,278 $0 $160,259 $253,536 3,854 3,854

60365 Ahad, Sajida F 0 0 0 07/28/2008 03/21/2014 ASSISTANT Tenure Track $125,000 $0 $159,318 $284,318 Yes 4,042 4,042

F 0 0 0 10/29/2001 ASSOCIATE Tenure Granted $116,970 $0 $185,723 $302,693 4,839 4,839 D M 0 0 0 08/01/2003 10/27/2015 ASSISTANT Tenure Track $178,171 $0 $152,559 $330,730 2,261 2,845 5,106

H M 0 0 0 08/11/2008 ASSISTANT Tenure Track $175,000 $0 $158,128 $333,128 Yes 2,529 2,889 5,418 M 0 0 0 07/01/2001 ASSOCIATE Tenure Granted $228,813 $0 $158,054 $386,867 2,308 3,121 5,429 M 0 0 1 08/08/2005 03/31/2011 PROFFESSOR Tenure Granted $229,758 $0 $171,163 $400,921 3,916 3,916 M 0 0 0 08/04/2008 06/30/2012 ASSOCIATE Tenure Granted $300,000 $0 $144,778 $444,778 Yes 2,333 2,333 M 0 1 0 09/01/1997 08/31/2012 PROFFESSOR Tenure Granted $532,027 $0 $124,949 $656,976 3,708 3,708

Notes: RVUs for 2008 are recorded for October 2007 - September 2008 period. RVUs for 2009 are recorded for October 2008 - September 2009 period."Yes-R" under "Startup Guarantee" stands for the year when a physician goes off a Guarantee Period, which is followed by the Reconciliation Period.

while listed under the General Surgery Division, were affiliated with the Pediatric General Surgery at various times throughout their SIU career.(See SIU PRODUCTION 31555, 31572-31578 for ; SIU PRODUCTION 18997 and 19015-19018 for and SIU PRODUCTION 21219, 21221 for ).

RedactedRedacted

RedactedReRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRedactedRe

ReReRedactedRedactedRedactedRedacted

RedactedRedactedRedactedRedactedRedactedRedactedRe

RedactedRedact Reda Redac

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 7-B - AAMC Median Compensation for General Surgery and Trauma/Critical Care Surgery DivisionsYears 2008/2009 - 2013/2014

General SurgeryTrauma/Critical Care Surgery



2008-2009 $243,000 $257,000 $302,000 $321,000 $340,000 $366,0002009-2010 $250,000 $261,000 $310,000 $325,000 $354,000 $360,0002010-2011 $262,000 $281,000 $316,000 $351,000 $358,000 $369,0002011-2012 $276,000 $292,000 $340,000 $347,000 $369,000 $380,0002012-2013 $290,000 $303,000 $350,000 $354,000 $373,000 $393,0002013-2014 $292,000 $320,000 $357,000 $368,000 $390,000 $419,000

Source: AAMC Faculty Salary Survey Reports, Table 11 (2008/2009-2013/2014)

Assistant Professor ProfessorAssociate ProfessorYear

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 8 - Faculty Members at the Neurology DepartmentCalendar Years 2015 and 2016, the Only Two Years during the 2010-2016 Period with Female Neurologists

Actual Calendar Year Pay

EmpID Name Gender Hire DateTermination

DateRank Tenure Status SOM Base

SOM Extra

HC TotalYear of

GraduationAdmin Title

Neurology 2015M 12/11/1984 12/31/2015 ASSOCIATE Alternate Track $68,858 $0 $37,733 $106,590 1977

rew M 9/13/1999 ASSOCIATE Alternate Track $153,457 $600 $42,735 $196,793 1981 Chief, Alzheimer's CenterM 9/9/2013 6/21/2016 ASSISTANT Alternate Track $102,000 $11,761 $142,804 $256,564 2004F 10/6/2014 ASSISTANT Alternate Track $120,000 $0 $140,772 $260,772 1996

M 10/20/2014 ASSISTANT Alternate Track $170,000 $4,500 $92,671 $267,171 2004 Residency DirectorK F 6/2/2014 4/15/2016 ASSISTANT Alternate Track $190,000 $0 $80,300 $270,300 2007 Director, Neuro Oncology

M 10/20/2014 ASSOCIATE Alternate Track $300,000 $0 $61,211 $361,211 1996 Director, Neurocritical CareM 9/13/2010 ASSISTANT Alternate Track $250,000 $11,761 $112,002 $373,762 1987 Medical Director, Stroke Center at MMCM 11/19/2012 PROFFESSOR Tenure Granted $362,000 $0 $61,108 $423,108 1979 Chair, Neurology Department

Neurology 2016M 9/13/1999 ASSOCIATE Alternate Track $153,457 $600 $39,110 $193,168 1981 Chief, Alzheimer's CenterF 10/6/2014 ASSISTANT Alternate Track $120,000 $0 $202,491 $322,491 1996

M 9/13/2010 ASSISTANT Alternate Track $250,000 $0 $100,559 $350,559 1987 Medical Director, Stroke Center at MMCM 10/20/2014 ASSISTANT Alternate Track $170,000 $44,519 $136,754 $351,273 2004 Residency DirectorM 11/19/2012 PROFFESSOR Tenure Granted $362,000 $0 $60,643 $422,643 1979 Chair, Neurology DepartmentM 10/20/2014 ASSOCIATE Alternate Track $307,381 $0 $115,921 $423,302 1996 Director, Neurocritical Care

Notes:Admin Titles are taken from SIU-SOM Personnel files received from Counsel on 5/1/2017Year of Graduation was taken from Dr. Sharp's production ("Audited AMA.xlsx")

RRedRedRedRedRedRedRedRed

RedRedRedRedRedRed

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 9 - Rebuttal to Sharp Report Exhibit 2Standard Error Adjustment Substantially Reduces the Statistical Significance of the Negative Female Coefficient in Dr. Sharp’s Misspecified Regression Model

Coef SIU-SOM Pay - Dr. Sharp's Model SIU-SOM Pay - Cluster Standard Errors SIU-HC Pay - Dr. Sharp's Model SIU-HC Pay - Clustered Standard ErrorsVariable Name Label Estimate St. Error t-value Probability Estimate St. Error t-value Probability Estimate St. Error t-value Probability Estimate St. Error t-value Probability

Constant b0 54,641 36,885 1.48 0.1388 54,641 44,397 1 23 0 2193 63,233 30,803 2.05 0.0403 * 63,233 26,649 2 37 0.0183 *Dept Chair b1 137,216 17,668 7.77 0.0000 * 137,216 47,405 2.89 0.0041 * 20,294 15,329 1 32 0.1858 20,294 32,921 0.62 0 5380Female b2 -17,890 5,752 -3.11 0.0019 * -17,890 9,290 -1.93 0.0550 -30,000 4,853 -6.18 0.0000 * -30,000 8,752 -3.43 0.0007 *Associate Profesor b3 44,456 7,676 5.79 0.0000 * 44,456 13,772 3 23 0.0014 * -8,752 6,524 -1 34 0.1800 -8,752 10,968 -0.80 0.4255Professor b4 131,208 9,933 13 21 0.0000 * 131,208 29,477 4.45 0.0000 * -13,124 8,350 -1 57 0.1163 -13,124 14,892 -0.88 0 3789MD Degree b5 47,416 17,304 2.74 0.0062 * 47,416 22,958 2.07 0.0397 * -6,917 14,790 -0.47 0.6401 -6,917 21,709 -0 32 0.7502MD and DVM Degrees b6 83,672 43,178 1 94 0.0529 83,672 24,121 3.47 0.0006 * -49,361 35,941 -1 37 0.1699 -49,361 24,050 -2.05 0.0410 *MD and JD Degrees b7 11,500 39,819 0 29 0.7728 11,500 27,486 0.42 0.6760 -1,390 33,182 -0.04 0 9666 -1,390 25,055 -0.06 0 9558MD and MPH Degrees b8 113,874 91,086 1 25 0 2115 113,874 49,719 2 29 0.0227 * -45,437 75,747 -0.60 0 5487 -45,437 40,700 -1.12 0 2651MD and PhD Degrees b9 63,725 19,865 3 21 0.0014 * 63,725 29,682 2.15 0.0326 * -866 16,842 -0.05 0 9590 -866 26,695 -0.03 0 9741Family & Community Medicine CORE b10 48,759 40,887 1.19 0 2333 48,759 38,126 1 28 0 2019 60,034 38,452 1 56 0.1188 60,034 46,306 1 30 0.1958Family & Community Medicine/Carbondale b11 20,952 34,335 0.61 0 5418 20,952 38,418 0 55 0 5859 40,038 28,488 1.41 0.1602 40,038 16,151 2.48 0.0137 *Family & Community Medicine/Decatur b12 27,092 35,452 0.76 0.4449 27,092 39,228 0.69 0.4903 45,758 29,398 1 56 0.1199 45,758 14,527 3.15 0.0018 *Family & Community Medicine/Quincy b13 -6,521 34,055 -0.19 0.8482 -6,521 37,777 -0.17 0.8631 74,984 28,301 2.65 0.0082 * 74,984 14,836 5.05 0.0000 *Family & Community Medicine/Springfield b14 55,660 34,909 1 59 0.1111 55,660 37,185 1 50 0.1354 26,294 32,290 0.81 0.4157 26,294 15,542 1.69 0.0917Internal Medicine Clinic b15 -26,115 93,301 -0 28 0.7796 -26,115 37,601 -0.69 0.4879 1,324 77,303 0.02 0 9863 1,324 13,816 0.10 0 9237Internal Medicine-SMS b16 23,465 31,956 0.73 0.4629 23,465 37,804 0.62 0 5352 58,593 26,498 2 21 0.0272 * 58,593 14,323 4.09 0.0001 *Medical Humanities b17 -74,659 71,132 -1.05 0 2941 -74,659 54,713 -1 36 0.1734 -14,632 59,027 -0 25 0.8043 -14,632 29,340 -0 50 0.6183Neurology Clinic b18 -150,920 54,110 -2.79 0.0054 * -150,920 52,992 -2.85 0.0047 * 25,539 45,047 0 57 0 5709 25,539 30,440 0.84 0.4021Neurology b19 -2,741 34,884 -0.08 0 9374 -2,741 42,256 -0.06 0 9483 67,030 28,914 2 32 0.0206 * 67,030 24,852 2.70 0.0074 *Obstetrics & Gynecology Clinic b20 277,237 60,002 4.62 0.0000 * 277,237 37,825 7 33 0.0000 * 66,622 49,728 1 34 0.1806 66,622 14,317 4.65 0.0000 *Obstetrics & Gynecology b21 41,999 33,036 1 27 0 2039 41,999 49,681 0.85 0 3985 154,331 27,378 5.64 0.0000 * 154,331 19,355 7 97 0.0000 *Pediatrics b22 33,403 32,284 1.03 0 3010 33,403 39,438 0.85 0 3976 57,451 26,835 2.14 0.0325 * 57,451 15,144 3.79 0.0002 *Plastic Surgery b23 112,299 40,469 2.77 0.0056 * 112,299 70,060 1.60 0.1100 329,784 33,539 9.83 0.0000 * 329,784 46,927 7.03 0.0000 *Psychiatry b24 36,865 33,673 1.09 0 2738 36,865 38,134 0 97 0 3344 12,212 27,912 0.44 0.6618 12,212 13,235 0 92 0 3569Surgery b25 51,827 32,542 1 59 0.1115 51,827 41,752 1 24 0 2154 143,851 26,967 5 33 0.0000 * 143,851 23,558 6.11 0.0000 *Tenure Track b26 16,353 11,830 1 38 0.1671 16,353 25,609 0.64 0 5236 -28,008 9,836 -2.85 0.0045 * -28,008 20,802 -1 35 0.1791Tenure Granted b27 47,334 10,957 4 32 0.0000 * 47,334 28,574 1.66 0.0986 -6,912 9,126 -0.76 0.4490 -6,912 20,671 -0 33 0.7383Years at SIU b28 -4,544 1,223 -3.72 0.0002 * -4,544 2,417 -1.88 0.0610 1,905 1,043 1.83 0.0680 1,905 1,660 1.15 0 2519Years at SIU Squared b29 37 42 0.88 0 3790 37 73 0 50 0.6190 -84 35 -2 38 0.0174 * -84 55 -1 53 0.12722012 b30 14,696 8,760 1.68 0.0937 14,696 3,825 3.84 0.0001 * -1,614 7,294 -0 22 0.8249 -1,614 3,783 -0.43 0.66992013 b31 26,645 8,898 2 99 0.0028 * 26,645 5,166 5.16 0.0000 * 2,954 7,544 0 39 0.6955 2,954 5,090 0 58 0 56212014 b32 37,141 9,055 4.10 0.0000 * 37,141 6,328 5.87 0.0000 * 5,201 7,646 0.68 0.4965 5,201 6,016 0.86 0 38802015 b33 46,511 9,094 5.11 0.0000 * 46,511 7,263 6.40 0.0000 * 6,452 7,657 0.84 0 3996 6,452 6,685 0 97 0 33522016 b34 52,261 9,041 5.78 0.0000 * 52,261 7,466 7.00 0.0000 * -57,048 7,634 -7.47 0.0000 * -57,048 6,493 -8.79 0.0000 *

R-squared 0.406 0.406 0 389 0 389Adj R-squared 0 388 0 388 0 370 0 370N of obs 1,156 1,156 1,119 1,119

Notes: * statistically significant at 5%.

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 10 - Rebuttal to Sharp Report Exhibit 4Standard Error Adjustment Eliminates the Statistical Significance of the Negative Female Coefficient in Dr. Sharp’s Misspecified Regression Mode

Coef Academic Base Pay - Dr. Sharp's Model Academic Base Pay - Clustered Standard ErrorsVariable Name Label Estimate St. Error t-value Probability Estimate St. Error t-value Probability

Constant b0 13,972 32,779 0.43 0.6700 13,972 26,989 0.52 0.6050Dept Chair b1 265,722 34,923 7.61 0.0000 * 265,722 33,950 7.83 0.0000 *Female b2 -13,682 3,992 -3.43 0.0006 * -13,682 7,693 -1.78 0.0763Associate Profesor b3 39,050 5,795 6.74 0.0000 * 39,050 10,950 3.57 0.0004 *MD Degree b4 31,073 14,740 2.11 0.0352 * 31,073 7,771 4.00 0.0001 *MD and DVM Degrees b5 81,270 30,134 2.70 0.0071 * 81,270 18,240 4.46 0.0000 *MD and JD Degrees b6 4,476 29,723 0.15 0.8803 4,476 13,052 0.34 0.7319MD and PhD Degrees b7 52,153 16,275 3.20 0.0014 * 52,153 18,069 2.89 0.0042 *Family & Community Medicine CORE b8 31,538 34,890 0.90 0.3662 31,538 20,713 1.52 0.1289Family & Community Medicine/Carbondale b9 8,605 29,144 0.30 0.7678 8,605 21,704 0.40 0.6920Family & Community Medicine/Decatur b10 -1,097 29,591 -0.04 0.9704 -1,097 21,371 -0.05 0.9591Family & Community Medicine/Quincy b11 -15,012 29,086 -0.52 0.6059 -15,012 21,788 -0.69 0.4913Family & Community Medicine/Springfield b12 19,943 29,190 0.68 0.4946 19,943 23,893 0.83 0.4046Internal Medicine Clinic b13 -59,790 68,728 -0.87 0.3845 -59,790 22,477 -2.66 0.0082 *Internal Medicine-SMS b14 170 27,521 0.01 0.9951 170 21,158 0.01 0.9936Medical Humanities b15 -242,325 62,332 -3.89 0.0001 * -242,325 33,777 -7.17 0.0000 *Neurology b16 -18,833 29,449 -0.64 0.5226 -18,833 28,659 -0.66 0.5116Obstetrics & Gynecology Clinic b17 272,170 41,830 6.51 0.0000 * 272,170 23,659 11.50 0.0000 *Obstetrics & Gynecology b18 21,918 27,696 0.79 0.4289 21,918 28,275 0.78 0.4388Pediatrics b19 10,562 27,735 0.38 0.7034 10,562 24,151 0.44 0.6622Plastic Surgery b20 25,170 33,809 0.74 0.4567 25,170 28,187 0.89 0.3726Psychiatry b21 19,408 28,351 0.68 0.4938 19,408 22,005 0.88 0.3785Surgery b22 -4,794 27,868 -0.17 0.8634 -4,794 23,003 -0.21 0.8350Tenure Track b23 42,469 8,685 4.89 0.0000 * 42,469 18,518 2.29 0.0225 *Tenure Granted b24 14,640 11,161 1.31 0.1899 14,640 27,450 0.53 0.5942Years at SIU b25 -4,207 1,038 -4.05 0.0001 * -4,207 1,879 -2.24 0.0259 *Years at SIU Squared b26 91 42 2.16 0.0313 * 91 75 1.22 0.22502012 b27 5,893 6,423 0.92 0.3590 5,893 2,805 2.10 0.0365 *2013 b28 10,329 6,423 1.61 0.1081 10,329 3,715 2.78 0.0058 *2014 b29 19,333 6,501 2.97 0.0030 * 19,333 4,918 3.93 0.0001 *2015 b30 32,960 6,406 5.15 0.0000 * 32,960 5,308 6.21 0.0000 *2016 b31 35,201 6,471 5.44 0.0000 * 35,201 5,844 6.02 0.0000 *Cumulative Publications b32 1,077 198 5.43 0.0000 * 1,077 975 1.10 0.2700Cumulative Citations b33 -12 9 -1.35 0.1762 -12 24 -0.52 0.6021Grad Top 25 Med School b34 -11,906 7,343 -1.62 0.1052 -11,906 18,338 -0.65 0.5167Number of Residencies b35 19,890 4,349 4.57 0.0000 * 19,890 9,191 2.16 0.0312 *Number of Fellowships b36 13,631 3,045 4.48 0.0000 * 13,631 5,927 2.30 0.0221 *Yrs Since Med School Graduation b37 5,168 1,135 4.55 0.0000 * 5,168 1,825 2.83 0.0049 *Yrs Since Med School Graduation Squared b38 -123 27 -4.55 0.0000 * -123 43 -2.86 0.0045 *

R-squared 0.374 0.374Adj R-squared 0.352 0.352N of obs 1,159 1,159

Notes: * statistically significant at 5%.

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 11 - Estimates for "FEMALE" CoefficientPartial Correction to Dr. Sharp’s Exhibits 2-4 by: 1) Adjusting Standard Errors; and 2) Analyzing Compensation by Decision-Making Units

Demonstrates a Lack of Common Proof of Gender Pay Disparity

SIU-SOM Pay SIU-HC PayDepartment Estimate St. Error t-value Probability Total N Female N Estimate St. Error t-value Probability Total N Female N

Regressions with Dr. Sharp's Exhibit 2 Controls

Family and Community Medicine -$12,883 12,338 -1.04 0.3022 180 84 -$14,428 10,989 -1.31 0.1965 153 84Internal Medicine -$39,412 16,671 -2.36 0.0203 * 316 128 -$46,588 18,146 -2.57 0.0119 * 316 128Neurology -$18,973 50,880 -0.37 0.7138 49 2 -$19,568 35,994 -0.54 0.5938 49 2Obstetrics and Gynecology $50,159 84,651 0.59 0.5609 81 24 -$86,702 25,717 -3.37 0.0034 * 81 24Pediatrics -$8,540 27,650 -0.31 0.7588 183 84 -$5,056 20,303 -0.25 0.8045 173 84Psychiatry $4,305 10,029 0.43 0.6738 57 22 -$29,916 14,581 -2.05 0.0581 57 22Surgery -$11,202 17,824 -0.63 0.5314 288 82 -$27,570 25,259 -1.09 0.2782 288 82

Regressions with Dr. Sharp's Exhibit 3 Controls

Family and Community Medicine -$11,161 13,046 -0.86 0.3970 180 84 -$8,526 9,984 -0.85 0.3981 153 84Internal Medicine -$29,452 13,690 -2.15 0.0342 * 316 128 -$35,671 16,719 -2.13 0.0357 * 316 128Neurology -$107,680 34,933 -3.08 0.0068 * 49 2 $116,289 57,924 2.01 0.0608 49 2Obstetrics and Gynecology $126,569 60,101 2.11 0.0495 * 81 24 -$96,683 20,898 -4.63 0.0002 * 81 24Pediatrics $15,777 32,941 0.48 0.6342 183 84 -$13,728 21,564 -0.64 0.5275 173 84Psychiatry $13,540 10,394 1.30 0.2123 57 22 -$40,671 14,901 -2.73 0.0155 * 57 22Surgery -$21,378 17,134 -1.25 0.2156 288 82 -$30,420 24,111 -1.26 0.2106 288 82

Regressions with Dr. Sharp's Exhibit 4 Controls (Excluding Professors)

Family and Community Medicine -$16,338 11,923 -1.37 0.1793 149 70 -$2,880 9,882 -0.29 0.7726 124 70Internal Medicine -$31,495 13,265 -2.37 0.0200 * 254 113 -$30,083 16,413 -1.83 0.0706 254 113Neurology -$127,649 62,133 -2.05 0.0645 33 2 $213,122 82,578 2.58 0.0256 * 33 2Obstetrics and Gynecology $140,179 56,353 2.49 0.0243 * 76 24 -$91,592 20,420 -4.49 0.0004 * 76 24Pediatrics $10,984 31,242 0.35 0.7271 160 80 -$20,396 21,139 -0.96 0.3409 154 80Psychiatry $13,071 9,202 1.42 0.1774 52 22 -$40,184 14,671 -2.74 0.0160 * 52 22Surgery -$8,616 14,062 -0.61 0.5423 211 66 -$8,222 25,999 -0.32 0.7528 211 66

Notes: * statistically significant at 5%. Analyses are performed based on Dr. Sharp's model specification, but without controlling for the "organization" variables in the by-department regressions.

3:15-cv-03308-SEM-TSH # 48-4 of 107

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 12-CHistorical Compensation Records and Predicted Values from Dr. Sharp's Model Specification (Based on Dr. Sharp's Exhibit 2 Parameters)

For Employees with Negative Predicted Values

SIU SOM Pay SIU HC Pay

FY EmpID Name Gender Hire DateAcademic

BaseActual Predicted

Predicted / Actual

Actual PredictedPredicted /

Actual

07 M 9/13/1999 $107,028 $107,028 - - $53,699 - -08 M 9/13/1999 $109,597 $142,907 - - $55,218 - -09 M 9/13/1999 $153,457 $153,457 - - $58,409 - -10 M 9/13/1999 $153,457 $153,457 - - $62,237 - -11 M 9/13/1999 $153,457 $154,007 $100,953 0.66 $54,581 $58,331 1.0712 M 9/13/1999 $153,457 $154,057 $111,946 0.73 $52,107 $56,685 1.0913 M 9/13/1999 $153,457 $154,057 $120,265 0.78 $56,418 $61,053 1.0814 M 9/13/1999 $153,457 $154,057 $127,203 0.83 $51,584 $62,932 1.2215 M 9/13/1999 $153,457 $154,057 $133,089 0.86 $42,735 $63,646 1.4916 M 9/13/1999 $153,457 $154,057 $135,428 0.88 $19,555 -$559 -0.03

07 M 7/1/1977 $1,507 $0 - - $0 - -08 M 7/1/1977 $1,543 $1,543 - - $0 - -09 M 7/1/1977 $1,583 $1,583 - - $0 - -12 M 7/1/1977 $1,583 $1,596 -$15,216 -9.53 $4,437 $34,553 7.7913 M 7/1/1977 $1,609 $1,625 -$5,289 -3.25 $14,652 $35,216 2.4014 M 7/1/1977 $1,625 $1,658 $3,256 1.96 $32,819 $33,390 1.0215 M 7/1/1977 $1,658 $1,658 $10,749 6.48 $36,535 $30,400 0.8316 M 7/1/1977 $1,658 $1,658 $14,695 8.86 $7,606 -$37,511 -4.93

07 F 9/5/2006 $120,321 $99,048 - - $47,875 - -08 F 9/5/2006 $122,880 $122,880 - - $26,323 - -09 F 9/5/2006 $126,075 $126,075 - - $25,769 - -10 F 9/5/2006 $126,075 $126,075 - - $59,218 - -11 F 9/5/2006 $109,000 $109,000 $103,442 0.95 $45,187 $44,801 0.9912 F 9/5/2006 $112,000 $112,000 $113,923 1.02 $53,613 $44,334 0.8313 F 9/5/2006 $125,009 $125,009 $121,731 0.97 $63,252 $49,881 0.7914 F 9/5/2006 $159,000 $159,000 $128,158 0.81 $66,918 $52,939 0.7915 F 9/5/2006 $159,000 $165,667 $133,532 0.81 $70,501 $54,832 0.7816 F 9/5/2006 $169,000 $169,000 $135,359 0.80 $26,989 -$8,195 -0.30

15 F 7/28/2014 $161,000 $149,917 - - $42,858 - -16 F 7/28/2014 $161,000 $161,000 $168,786 1.05 $20,488 -$16,700 -0.82

RedRedRedRedRedRedRedRed

Redt

Red

ReReReReReRe

ReRedRedRedRedRedRedRedRedRedRedRed

tRedRed

3:15-cv-03308-SEM-TSH # 48-4 of 107

Exhibit 12-CHistorical Compensation Records and Predicted Values from Dr. Sharp's Model Specification (Based on Dr. Sharp's Exhibit 2 Parameters)

For Employees with Negative Predicted Values(Cont’d)

SIU SOM Pay SIU HC Pay

FY EmpID Name Gender Hire DateAcademic

BaseActual Predicted

Predicted / Actual

Actual PredictedPredicted /

Actual

09 F 10/27/2008 $160,000 $109,565 - - $10,477 - -10 F 10/27/2008 $160,000 $186,667 - - $12,315 - -11 F 10/27/2008 $200,000 $200,000 $156,547 0.78 $7,351 $33,249 4.5212 F 10/27/2008 $200,000 $204,617 $166,882 0.82 $8,794 $33,119 3.7713 F 10/27/2008 $205,540 $212,325 $174,544 0.82 $11,405 $39,002 3.4214 F 10/27/2008 $211,706 $229,100 $180,824 0.79 $12,435 $42,397 3.4115 F 10/27/2008 $231,000 $221,442 - - $24,671 - -16 F 10/27/2008 $221,000 $221,000 $187,734 0.85 $8,063 -$18,063 -2.24

07 F 8/1/1987 $140,847 $140,847 - - $84,665 - -08 F 8/1/1987 $144,228 $144,228 - - $90,505 - -09 F 8/1/1987 $147,978 $152,978 - - $95,497 - -10 F 8/1/1987 $147,978 $147,978 - - $91,354 - -11 F 8/1/1987 $147,978 $147,978 $64,395 0.44 $90,122 $56,876 0.6312 F 8/1/1987 $147,978 $147,978 $76,264 0.52 $88,401 $53,210 0.6013 F 8/1/1987 $156,478 $156,478 $85,460 0.55 $107,590 $55,557 0.5214 F 8/1/1987 $181,478 $181,478 $180,027 0.99 $100,045 $51,044 0.5115 F 8/1/1987 $181,478 $181,478 $186,789 1.03 $113,913 $49,737 0.4416 F 8/1/1987 $181,478 $181,478 $190,005 1.05 $53,933 -$16,489 -0.31

07 M 7/1/1978 $144,085 $144,085 - - $7,858 - -08 M 7/1/1978 $291,627 $291,627 - - $20,450 - -09 M 7/1/1978 $299,210 $299,210 - - $19,929 - -10 M 7/1/1978 $299,210 $299,210 - - $28,606 - -11 M 7/1/1978 $299,210 $299,210 $333,289 1.11 $23,466 $89,908 3.8312 M 7/1/1978 $299,210 $299,210 $345,816 1.16 $26,674 $84,726 3.1813 M 7/1/1978 $299,210 $299,210 $355,669 1.19 $18,279 $85,558 4.6814 M 7/1/1978 $299,210 $168,851 - - $24,204 - -15 M 7/1/1978 $157,000 $158,238 $234,345 1.48 $23,959 $60,784 2.5416 M 7/1/1978 $157,000 $162,716 $238,218 1.46 $9,119 -$6,958 -0.76

Notes:· Dr. Sharp excluded all records prior to FY 2011 from his analysis. · All employees listed in this Exhibit were active as of the date when the data were produced for Dr. Sharp’s analyses. · FY10 and FY11 records for are not included in Dr. Sharp’s analyses, although they were available in the datasets produced.· FY15 record for , and FY14 record for were excluded from Dr. Sharp’s SIU-SOM Pay and SIU-HC Pay analyses due to his “full year” restriction.

Red

RedRed

tRed

RedRed

tRedRed

ReRe

ReRedReReReReReRedReReReReReReReReReRe

RedactedRedacted Redacted

3:15-cv-03308-SEM-TSH # 48-4 of 107

C H E N S O N G , P h . D . , C F A

Phone: (949) 474-4938 3 Park PlazaFax: (949) 474-4944 Suite [email protected] Irvine, CA 92614

CURRENT POSITIONS

Senior Vice President, Southern California Office, Nathan Associates Inc. Instructor for the Chartered Financial Analyst (CFA) Review Program Level III Risk Management

EDUCATION

Ph.D., economics, University of Chicago, 2000 M.A., economics, University of Chicago, 1995 B.A., mathematics and economics, Agnes Scott College, 1993

SPECIALIZED EXPERIENCE, RESEARCH, OR INTEREST

Labor and Employment Disputes; Securities Disputes; Statistics and Econometrics; Finance

PAST POSITIONS

2006-2010 Navigant Consulting, Inc., Director Los Angeles, CA 2000-2006 Resolution Economics, LLC., Senior Economist Beverly Hills, CA 2008-Current CFA Review Program Level III Risk Management, Instructor Los Angeles, CA 2007-2008 California State University Los Angeles, Adjunct Assistant

Professor Los Angeles, CA 2000-2006 Center for Population Economics, University of Chicago,

Senior Investigator Chicago, IL 2004 Fall University of California Irvine, Lecturer Irvine, CA 1992-1993 Federal Reserve Bank of Atlanta, Intern Atlanta, GA

COURSES TAUGHT

Labor Economics and Human Resources Financial Institutions

Price Theory Risk Management

Valuation of Options and Other Derivatives

EXPERIENCE SUMMARY

Dr. Song is an economist, a CFA charter holder, and an expert witness. Since receiving her Ph.D. in economics from the University of Chicago, she has provided consulting services and expert testimony in class action discrimination disputes and class action wage and hour disputes. Dr. Song

ATTACHMENT A3:15-cv-03308-SEM-TSH # 48-4 of 107


2

has studied issues related to definitions of labor markets, pay, promotion and separation, executive compensation, employee classification and alleged time reporting violations. In addition, Dr. Song has maintained a strong affiliation with the academic community. She lectured on Labor Economics and Human Resources at the University of California, Irvine. Dr. Song was also an adjunct assistant professor with the California State University, Los Angeles where she taught a course on Financial Institutions. Currently, she is an instructor with the Chartered Financial Analyst (CFA) review program and teaches a course on risk management.

LITIGATION CONSULTING EXPERIENCE

Labor Employment

Assignments representative of Dr. Chen Song’s experience in labor employment litigation consulting matters include: Expert witness services for defendant (a health club) in a case involving meal break violations

for non-managerial hourly employees. Services included analyzing electronic timeclock and payroll records, and submitting a declaration to rebut plaintiffs’ expert’s analysis as well as to assist counsel with opposition to class certification.

Consulting services for defendant (a lumber company) in a case involving meal break violations and unpaid overtime. Services included designing and supervising data entry of punch records in order to reliably capture all relevant information, analyzing punch records and payroll data, and preparing a summary of risk assessment to assist counsel with mediation.

Consulting services for defendant (a home healthcare agency) in a case involving unpaid overtime claims. Services included analyzing a sample of time-keeping records and payroll data in order to assist counsel with mediation.

Consulting services for defendant (a large grower of herbs) in a case involving meal break violations. Services included reviewing and analyzing time-keeping and payroll files from various time-keeping and payroll systems, in order to assist counsel with mediation.

Consulting services for plaintiff (a public company specializing in virtualization, cloud computing and enterprise solutions) in a trade secrets and contract dispute matter. Services included calculating lost profits and unjust enrichment in connection with the alleged unlawful and unfair solicitation, raiding of employees and theft of trade secrets.

Expert witness services for defendant (a restaurant chain) in a class action discrimination case. Services included construction analytical databases from human resources and payroll data, modeling the promotional process, performing statistical analyses to examine plaintiffs’ claim of discrimination in promotion and pay, and assisting counsel with class certification defense.

Expert witness services for defendant (a hospital) in a single-plaintiff discrimination claim. Services included rebutting opposing expert’s lost earnings analysis and estimating economic damages.

3:15-cv-03308-SEM-TSH # 48-4 of 107


3

Consulting services for defendant (a pharmaceutical company) in a single-plaintiff wrongful

termination claim. Services included rebutting opposing expert’s calculation of lost employee stock options awards.

Consulting services for defendant (a restaurant chain) in a case alleging unpaid time related to the donning and doffing of uniforms. Services included culling over 200 store manager declarations to build an analytical database, summarizing ranges of the donning and doffing time by uniform types, studying variations in the donning and doffing time by store and by shift, and performing statistical analysis to identify factors potentially impacting the donning and doffing time.

Consulting services for defendant (a regional social services provider) in a case involving meal and rest break violations. Services included designing an observational or time-in-motion study to evaluate plaintiffs' claims that service professionals were not provided opportunities for break during their work days, and assisting counsel in defeating class-certification.

Consulting and expert witness services for defendant (a global defense, security and aerospace

systems company) in a wrongful termination matter alleging race discrimination. Services included evaluating the economic impact of the plaintiff’s termination from his employer.

Consulting and expert witness services for defendant (a restaurant chain) in a case involving meal and rest violations. Services included analyzing large time clock and transactional databases, designing statistical tests to assess the merit of the plaintiffs’ claim of a company-wide policy for meal and rest violations, and filing a rebuttal report to assist counsel with a motion for class de-certification.

Consulting services for defendant in a case involving wrongful termination allegation by several

managers in an entertainment software company. Services included stock options valuation using a modified version of the Black-Scholes Option Pricing Framework. The modification incorporates a marketability discount arising from sales restrictions and separation risk.

Consulting services in connection with reduction in force (RIF). Services included performing adverse impact analysis on gender, race and age, and critiquing the statistical analysis conducted by an EEOC expert.

Consulting services in connection with an age discrimination claim brought forth by a sales rep

at a large investment bank. Services included analyzing compensation, sales and performance rating information for liability assessment, calculating lost earnings, and rebutting opposing expert’s damage models.

Consulting services in connection with NASD arbitration on a breach of contract claim from a

portfolio manager. Services included evaluating returns to portfolios and calculating lost earnings.

Consulting services for defendant in a case involving breach of contract allegation by a high-level executive in the gaming and high-tech industry. Services included estimating damages using information from analyst reports issued by major investment banks, and studying academic peer-reviewed journal articles that formed the basis of opinions on executive compensations in the high-tech industry.

3:15-cv-03308-SEM-TSH # 48-4 of 107


4

Consulting services for defendant’s counsel in a wage and hours matter alleging uncompensated

overtime and missed meal periods for a prospective class of hourly employees at a large transportation company. Services included designing and developing a daily punch-in/punch-out database and hours worked from a large number of electronic records, performing workweek calculation, estimating exposure, penalties, the value of per hour regular pay, per hour overtime pay and per hourly double-time pay, and assisting counsel with mediation.

Consulting services in connection with research on gender discrimination in pay and promotion

at a university. Services included estimating the difference in pay and merit increases by gender, identifying inputs into the compensation and academic award structure, and presenting findings to the faculty.

Consulting services for the board of trustees in a university system in response to a labor union

allegation of gender and racial discrimination in the faculty award policy. Services included statistical analyses of award records.

Consulting services in several wrongful termination matters. Services included use of detailed

census and other data to estimate labor market availabilities by geographic location, evaluating job search efforts and mitigation opportunities, constructing financial models to compute economic damages under a variety of scenarios, rebutting opposing expert’s analysis, and assisting counsel with cross-examination.

Consulting and expert witness services for defendant (a large institution for national security) in

a case involving an EEOC investigation of gender discrimination in pay and promotion. Services included creating analytical databases from human resource files, employee job application and performance review paper documents, performing statistical tests and regression analyses to study the relationship between productivity-related variables, pay, and gender, and filing declarations to assist counsel with liability assessment.

Consulting and expert witness services for defendant (a healthcare provider) in a case involving

time-shaving, meal and rest violations. Services included studying daily activity patterns of four types of healthcare workers, designing statistical tests to assess the merit of the plaintiffs’ claims that the defendant had adopted a company-wide policy to systematically prune overtime entitled by its employees, estimating the extent of unpaid meal and rest breaks from daily punch records, and co-authoring a rebuttal report to assist counsel with a motion for class de-certification.

Consulting services for defendant in connection with an allegation of age discrimination in

promotion in a large biotech company. Services included processing large job history data and conducting statistical analyses of employment records.

Securities Class Action and Financial Modeling

Assignments representative of Dr. Chen Song’s experience in litigation consulting matters involving securities class action and financial modeling include: Consulting services for defendant (a large broker-dealer company) in connection with a

derivative class action suit in which the plaintiffs alleged Section 10b-5 and other violations. Services included employing econometric approach to estimate a range of potential damages per

3:15-cv-03308-SEM-TSH # 48-4 of 107


5

share by taking into account the timing of information releases to the capital markets, and developing a multi-trader model to estimate the number of shares potentially damaged. The multi-trader model was used to account for shareholders’ varying propensities to trade, correct for trading volume overstatements caused by market maker and specialist activities, and adjust shares outstanding for short interest, company insiders’ beneficial ownership positions, and institutional shares not traded during the alleged fraud period.

Consulting services for defendant (a high-tech company) in connection with an SEC investigation of options backdating practices. Services included conducting a series of statistical tests on the company stock option grant dates to identify those dates where the short-term return was unusually favorable.

Consulting services for defendant (a state security entity) in a lawsuit brought forth by a

securities trading company alleging $500 million loss due to reduction in business capacity as a result of the 1993 bombing of the World Trade Center. Services included performing time series regression models to address the issue of loss causation, calculating damages during the alleged business interruption period, and testing the sensitivity of model outcome by studying various time-to-maturity segments of the traded fixed income securities.

Consumer Class Action

Consulting services for defendant in a consumer class action against a large bank. Plaintiffs in

the prospective class alleged that the bank's practice of processing debit card transactions from highest to lowest dollar amount resulted in a higher number of overdraft fees assessed on customers' accounts than if debit card transactions were posted in the chronological order that they occurred. Services included critiquing Plaintiffs’ expert’s data and damage analyses, running alternative but-for scenarios for sampled customers’ accounts and estimating damages. Construction Defects

Consulting services for plaintiff in a construction defects claim against a builder. Services

included estimating, using a hazard model, the life span of pipes installed in certain residential neighborhoods during the class period.

EXPERT REPORTS AND TESTIMONIES 2013-2017

Scott Kurtz v. Dignity Health and Mercy Hospital of Bakersfield, Case No. S-1500-CV-282231,

DRL, related to damages calculation in a disability discrimination claim. Deposition testimony provided on September 1, 2015. Trial testimony provided on October 1, 2015.

Harry Boon v. Canon Business Solutions, Inc., Case No. CV11-08206 R (Cwx), related to data analysis of meal and rest claims. Report filed on August 31, 2015.

T.J. Simers v. Tribune Company, et al., Case No. BC 524 471, related to rebuttal damage calculation for a wrongful termination claim. Deposition testimony provided on October 23, 2014.

3:15-cv-03308-SEM-TSH # 48-4 of 107


6

Appointed by the Court as the neutral technical advisor in an employment class action Sibley, et al. v. Sprint Nextel Corp., no. 08-CV-2063 (District of Kansas), effective October 6, 2014.

Herminia Rodriguez and Hilaria Martinez, individually, and on behalf of others similarly situated current and former employees vs. Equinox Fitness Century City, Inc., et al, Case No. BC437283, related to data and statistical analysis of meal breaks and time clock rounding. Declaration filed on May 14, 2014.

KB Home v. K&L Gates, LLP, et al. Case No. BC484090, related to statistical sampling and damages calculations in a malpractice claim in connection with class certification. Report filed on November 25, 2013. Supplemental report filed on January 10, 2014. Deposition testimony provided on January 15, 2014. Rebuttal report filed on February 14, 2014.

Catherine Zulfer v. Playboy Enterprises, Inc., Case No. CV12-08263-BRO(SHx), related to mitigation efforts in a wrongful termination claim filed by a former corporate controller. Report filed on November 15, 2013. Deposition testimony provided on December 17, 2013.

Claude Riley v. Waste Management, Case No. 3:12-cv-02897-dms-dhb, related to damages calculation for a wrongful termination claim. Report filed on November 4, 2013.

Christina Caputo v. Prada USA Corp., et al., Case No. CV12-03244 FMO (RZx), related to damages calculation for promissory estoppel and wrongful termination claims. Deposition testimony provided on May 16, 2013.

Kimberly Patterson v. JPMorgan Chase Bank National Association, et al., Case No. CV12-03237 DDP (SHx), related to damages calculation for a wrongful termination claim. Report filed on May 14, 2013.

Juana Lambaren v. U-Haul Co. of California, AAA Case No. 72 160 01173 10 JOG3, related to

damages calculation for a wrongful termination claim. Summary of Opinions filed on February 11, 2013. Arbitration testimony provided on February 21, 2013.

PUBLICATIONS 2007-2017

Chen Song and David Sharp, “Rethinking the Unemployment Status of Job Applicants”,

Law360 Expert Analysis, July 19, 2012.

Will Carrington and Chen Song, “Seeding Accounts? Detecting Discrimination Among Brokers”, LexisNexis Mealey's Litigation Report Employment Law, Volume 7, Issue #1, August 2010.

Joao Dos Santos and Chen Song, “Analysis of the Wealth Effects of Shareholder Proposals”, July 22, 2008, and “Analysis of the Wealth Effects of Shareholder Proposals – Volume II", May 18, 2009, Research Paper Released by the U.S. Chamber of Commerce. Presentation at the Conference “Shareholder Rights, the 2009 Proxy Season, and the Impact of Shareholder Activism”, Hosted by the Center for Capital Markets Competitiveness on June 23, 2009.

3:15-cv-03308-SEM-TSH # 48-4 of 107


7

AWARDS AND RESEARCH GRANTS

Phi Beta Kappa, Elected to Membership 1993

Far East Funding, 1994-97

National Institute on Aging, Program Project Grant (Award # P01 AG10120), “Early Indicators of Later Work Levels, Disease and Death”, Robert W. Fogel, Principal Investigator. Senior investigator (September 2001 to August 2006)

PROFESSIONAL MEMBERSHIPS

American Economic Association American Bar Association Los Angeles Society of Chartered Financial Analysts

COMMUNITY ACTIVITIES

Member, Board of Directors of the Los Angeles Center for law and Justice, a legal-aid organization which provides education and direct representation in housing and family law for low-income communities in East Los Angeles

3:15-cv-03308-SEM-TSH # 48-4 of 107

Attachment B – Documents Received and Reviewed

Electronic data I relied upon are referenced in my report according to names in bold:

a) Original Compensation Data in Excel format, which I understand to be the same

Excel files analyzed by Dr. Sharp:

i. “SIU SOM Physician Academic Base Salary (FY07‐FY16).xlsx”; and

ii. “SIU SOM Physicians Compensation (FY07‐FY16).xlsx”;

b) Conversion Factor File with historical conversion factors by division:

i. “Ahad S Conversion Factor file w historical CFs 02‐22‐17.xlsx”;

c) A one‐month sample (May 2016) of provider charge encounters posted into the

billing system, for the month RVUs are being calculated, or “PPO1” Monthly

Report:

i. “Copy of MAY 2016 ‐ PPO1 File ‐ Used for June 2016 Payroll File.xlsx”;

d) Faculty Grant Data (see Attachment C – Grant Data)

i. “SIU Carbondale grants and contracts funded since 01‐01‐07.docx”;

ii. “SIU Springfield Campus grants and contracts funded 1‐1‐07 to 2‐10‐17.pdf”;

iii. “Copy of Faculty with grants FY07 to FY16.xlsx”;

e) Updated Compensation Data in Excel format prepared by SIU and counsel, with

principal updates of adding the “Division” column and synchronizing all

compensation components (SOM, SOM Extra, and HC) to be actual payments to

physician faculty as of calendar yearend:

i. “77V1461‐Ahad S SIU Complete Physician List (By Division) 4‐4‐17.xlsx”;

f) Department Chair List:

i. “76J9761‐Ahad S Chair Listing 2006 through 2016 02‐13‐17.xlsx”;

g) Division Chief List:

i. “Division Chief Listing 06_current.xlsx”;

1

3:15-cv-03308-SEM-TSH # 48-4 of 107

h) RVUs ‐ When applicable1, collect from the Conversion Factor File, by division

from 2010 to 2016, for every October to April period:

i. Physicians’ RVUs (i.e., total RVUs from October 1st of the previous calendar

year to September 30th of the current calendar year);

ii. Division conversion factors;

iii. Exhibit 10 to the 11/29/2016 Declaration of Wendy Cox‐Largent, SIU

PRODUCTION 17580‐17586; and

iv. Dr. Holly Brockman’s Compensation personnel file, SIU PRODUCTION

23173;

i) Startups – Collect from SIU‐HC compensation PDF documents:

i. Whether or not physicians signed a startup agreement;

ii. Startup guarantee begin date;

iii. Duration of guaranteed salary (i.e., 6‐month, 12‐month, 24‐month, 36‐month,

48‐month, or 60‐month); and

iv. Date physicians voluntarily terminated startup, if applicable.

j) AAMC Total Compensation Benchmarks – Collect from AAMC Faculty Salary

Survey Reports and AAMC Data Book:

i. Mean and median total compensation by department and by sub‐specialty

(division) from the periods of 2008‐2009 to 2014‐2015.

1 It is my understanding that physicians at Family and Community Medicine (FCM) Department pool

RVUs so their SIU‐HC compensation does not depend on RVUs. According to Mr. Nelson Weichold,

Chief Operating and Financial Officer of SIU‐HC, physicians at FCM are rewarded for offering all kinds

of medical services to the community. Therefore there is no direct correlation between SIU‐HC

compensation and RVUs at FCM (conversation with Mr. Weichold on April 28, 2017). Furthermore, it is

my understanding that the Emergency Medicine Division at the Surgery Department does not record

RVUs.

2

3:15-cv-03308-SEM-TSH # 48-4 of 107

Attachment B ‐ Documents Received and Reviewed

Document DescriptionDocument

Type

Case Filings

• Answer and Affirmative Defenses to Amended Class Action and Collective Action Complaint (filed

11/02/2016)PDF

• Plaintiffʹs Motion for Conditional Collective Action Certification and Judicial Notice (filed 10/18/2016) PDF

• Plaintiffʹs Memorandum in support of her motion for conditional collective action certification and Judicial

Notice (filed 10/18/2016) [with exhibits]PDF

• Defendantsʹ Response to Plaintiffʹs Motion for Conditional Collective Action Certification and Judicial Notice

(filed 12/02/2016) [with exhibits]PDF

• Declaration of Wendy Cox‐Largent (dated 11/29/2016) [with exhibits] PDF

• Exhibit B to Declaration of Angela M. Buhl (faculty listing) PDF

• Plaintiffʹs Reply to Defendantsʹ Response to Plaintiffʹs Motion for Conditional Collective Action Certification

and Judicial Notice (filed 12/21/2016) [with exhibit]PDF

• Deposition of Sajida Ahad, M.D., taken on 11/11/2016 [with exhibits] PDF

• Deposition of Wendy Cox‐Largent, taken on 12/23/2015 [with exhibits] PDF

• Deposition of Nelson Weichold, taken on 08/25/2016 [with exhibits] PDF

• DOL Testimony of Dr. Mellinger, 01/05/2016 PDF

• Deposition of David Justin Rea, M.D., taken on 12/17/2015 PDF

Plaintiffʹs Expert Report and Supporting Documentation

• Plaintiffʹs Rule 26(a)(2) Class Certification Expert Witness Disclosures PDF

• Expert Report of D.C. Sharp, Ph.D. ‐ For Purposes of Class Certification (dated 03/10/2017) PDF

• Backup Documentation for Dr. Sharpʹs Expert Report (including Exhibits, Literature, SAS programs, input

and output, and Web Docs)

• SIU Medicine website information for physicians‐Specialties, education, publications, etc. PDF

• Advertisement for Dr. Cagleʹs position (Academic Orthopaedic Surgeons position) PDF

• Dr. Sajida Ahad HR Personnel File (SIU PRODUCTION 11145‐11187) PDF

• Dr. Sajida Ahad SIUHC Personnel File (SIU PRODUCTION 11188‐11236) PDF

• Faculty and PAA Request to Offer in Lieu of Recruitment PDF

• Imran Hassan, M.D. CV PDF

• Hiring Audit Form for Christopher Wohltmann PDF

• Hire Offer Comparison for Christopher Wohltmann PDF

• Letter to John Sutyak from Gary Dunnington dated 01/16/2001 PDF

• RVUs for Aamir Zakaria for August 2005‐October 2006 PDF

• Search Waver Tracking Form PDF

• Hiring document for Dr. Cetindag (Section III to be Completed Before Hire) PDF

• SIUʹs 3rd suppl Rule 26 disclosure documents, including ʺSIU School of Medicine Chairperson Listing 2006‐

currentʺPDF

• SIU Healthcare Supplemental Hiring Form PDF

• Letter from Thomas H. Wilson to J. Bryan Wood, dated 02/27/2017 PDF

• Deposition of D.C. Sharp, Ph.D., taken on 04/07/2017 PDF

Other documents

• Email from David Pence to Wendy Cox‐Largent dated 03/30/2017 Email

• Transcribed notes for the conference call on March 21, 2017, between SIU, HeplerBroom LLC and Nathan

Associates, Inc.

• SIU P&S preliminary compensation caps ‐ Data Sources FY10 ‐ FY 16 contracts Word/PDF

• AAMC Table 11: Faculty Compensation for All Schools, MDs, Clinical Departments, 2012‐2013 mht

• AAMC Table 11: Faculty Compensation for All Schools, MDs, Clinical Departments, 2013‐2014 PDF

• AAMC Table 11: Faculty Compensation for All Schools, MDs, Clinical Departments, 2014‐2015 PDF

3

3:15-cv-03308-SEM-TSH # 48-4 of 107

Attachment B ‐ Documents Received and Reviewed

Document DescriptionDocument

Type

SIU Compensation and Personnel Files

• Physician Compensation Files (542 files) PDF

• Personnel records for Abdelhak, Tamer zip (PDF)

• Personnel records for Ala, Thomas Andrew PDF

• Personnel records for Elsayed, Mona zip (PDF)

• Personnel records for Gilchrist, James M PDF

• Personnel records for Mueed, Sajjad PDF

• Personnel records for Murr, Najib PDF

• Personnel records for Rauschkolb, Paula K zip (PDF)

• Personnel records for Siddiqui, Fazeel M zip (PDF)

• Personnel records for Zec, Ronald Francis zip (PDF)

• General Surgery Salary Info as of 07/28/2008 PDF

Literature

• Jagsi, R, et al. (2012). Gender Differences in the Salaries of Physician Researchers. JAMA, 307(22): 2410 ‐2417.

• Jena, A.B., Olenski, A.R., and Blumenthal, D.M. (2016). Sex Differences in Physician Salary in US Public

Medical Schools. JAMA Internal Medicine, 176(9): 1294‐1304.

• Baker, L.C. (1996). Differences in Earnings between Male and Female Physicians. New England Journal of

Medicine, 334(15): 960‐964.

• Sasser, A.C. (2005). Gender Differences in Physician Pay: Tradeoffs between Career and Family. Journal of

Human Resources, 40(2): 477‐504.

• Sharp, D.C., et al. (2004). But can she cook? Womenʹs education and housework productivity. Economics of

Education Review, 23: 605‐614.

• Daniel L. Rubinfeld, Reference Guide on Multiple Regression. Reference Manual on Scientific Evidence, 3rd

ed., 2011.

4

3:15-cv-03308-SEM-TSH # 48-4 of 107

Attachment C – Grant Data

I have taken the following steps to incorporate information on grants received by

physicians into my analyses:

1. Three files were received from council on 2/23/2017 relating to physician grant activity:

a) “SIU Springfield Campus grants and contracts funded 1‐1‐07 to 2‐10‐17.pdf”;

b) “SIU Carbondale grants and contracts funded since 01‐01‐07.docx”; and c) “Copy of Faculty with grants FY07 to FY16.xlsx”.

2. In order to get the PDF and Word document files in a workable format to combine with

the “Grants FY07_FY16” tab from “Copy of Faculty with grants FY07 to FY16.xlsx”, the

information provided in: a) “SIU Springfield Campus grants and contracts funded 1‐1‐

07 to 2‐10‐17.pdf”; and b) “SIU Carbondale grants and contracts funded since 01‐01‐

07.docx” was transferred into separate Excel files. The Excel versions of these three files

were combined and used for analysis in STATA.

3. In order to eliminate any potential overlapping observations between these three files,

grants were matched based on the physician’s first name, last name, BP/Account

Number, and start date of the grant.

a) Those records with the same physician first and last name (taken from the “PI

Name”), the same BP/Account Number, and the same start date were marked as

duplicates and were counted as only one grant.

b) Those observations with the same physician first and last name, the same

BP/Account Number, and start dates within 35 days of each other were also marked

as duplicates and were counted as only one grant. In these cases, the PDF record

was retained over the Excel or Word document records, and the Excel record was

retained over the Word document record.

c) Those records with the same physician first and last name, the same BP/Account

Number, but start dates greater than 35 days apart were counted as two separate

grants.

d) Some adjustments were made to first and last names taken from the “PI NAME”

from “Copy of Faculty with grants FY07 to FY16.xlsx” in order to accurately identify

duplicate grant observations between files and eventually merge these to

compensation related information.

4. Once these grant files were combined and duplicate grants were removed, the

information was used to identify whether physician received at least one grant during a

calendar year. This information was then merged onto the Physician Compensation

dataset.

1

3:15-cv-03308-SEM-TSH # 48-4 of 107