Siddharth Hariharan, et al v. Adobe Systems, Inc., et al
Filing
1
FILED ON 11/07/2013 PETITION FOR PERMISSION TO APPEAL PURSUANT TO RULE 23(f). SERVED ON 11/07/2013. [8856405] (HC)
Case No. __________
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
IN RE HIGH-TECH EMPLOYEE ANTITRUST LITIGATION
Petition for permission to appeal from the United States District Court
Northern District of California
The Honorable Lucy H. Koh, Presiding
Case No. 5:11-2509-LHK
DEFENDANT-PETITIONERS’ EXCERPTS OF RECORD
VOLUME II OF VIII
ROBERT A. VAN NEST, #84065
DANIEL PURCELL, #191424
EUGENE M. PAIGE, #202849
JUSTINA SESSIONS, #270914
KEKER & VAN NEST LLP
633 Battery Street
San Francisco, CA 94111-1809
Telephone: 415 391 5400
Facsimile: 415 397 7188
Attorneys for Defendant and Petitioner
Google Inc.
789556
EXCERPTS OF RECORD
N.D.CAL.
DOCKET #
DOCUMENT
PAGE
Volume I of VIII
(District Court Orders—Public Versions)
1.
531
Oct. 24, 2013 Order Granting Plaintiffs’
Supplemental Motion for Class Certification
(public redacted version)
0001
2.
382
April 15, 2013 Order Granting in Part and Denying in
Part Plaintiffs’ Motion for Class Certification
(public redacted version)
0087
Volume II of VIII
(Expert Reports – Public Versions)
3.
518-2 518-4
Expert Report of Professor Kevin M. Murphy
(public redacted version)
0140
4.
424-2
Supplemental Expert Report of Edward E. Leamer,
Ph.D.
(public redacted version)
0340
Volume III of VIII
(Expert Reports – Public Versions)
5.
440
Supplemental Expert Report of Professor Kevin M.
Murphy
(public redacted version)
0402
6.
442
Expert Report of Kathryn M. Shaw, Ph.D.
(public redacted version)
0570
Volume IV of VIII
(Depositions and Declarations – Public Versions)
7.
308-1,
445-2
Deposition of Edward Leamer
0676
8.
538-8 538-11
Declaration of Danny McKell in Support of
Defendants’ Opposition to Plaintiffs’ Motion for
Class Certification
(public redacted version)
0691
1
789556
N.D.CAL.
DOCKET #
9.
516-6
10.
DOCUMENT
PAGE
Declaration of Frank Wagner in Support of
Defendants’ Opposition to Plaintiffs’ Motion for
Class Certification
(public redacted version)
0713
District Court Docket Report
0725
Volume V of VIII
(District Court Orders FILED UNDER SEAL)
11.
Oct. 24, 2013 Order Granting Plaintiffs’ Motion for
Class Certification
(under seal version)
0804
12. 383
April 15, 2013 Order Granting in Part and Denying in
Part Plaintiffs’ Motion for Class Certification
(under seal version)
0890
Volume VI of VIII
(Expert Reports FILED UNDER SEAL)
13.
Expert Report of Professor Kevin M. Murphy
(under seal version)
0944
14.
Supplemental Expert Report of Edward E. Leamer,
Ph.D.
(under seal version)
1144
Volume VII of VIII
(Expert Reports FILED UNDER SEAL)
15.
Supplemental Expert Report of Professor Kevin M.
Murphy
(under seal version)
1180
16.
Expert Report of Kathryn M. Shaw, Ph.D.
(under seal version)
1348
17.
Expert Witness Report of Kevin F. Hallock, Figure 7
1454
2
N.D.CAL.
DOCKET #
DOCUMENT
PAGE
Volume VIII of VIII
(Depositions and Declarations
FILED UNDER SEAL)
18.
Deposition of Michael Devine
1455
19.
Declaration of Danny McKell in Support of
Defendants’ Opposition to Plaintiffs’ Motion for
Class Certification
(under seal version)
1458
20.
Declaration of Frank Wagner in Support of
Defendants’ Opposition to Plaintiffs’ Motion for
Class Certification
(under seal version)
1480
21.
Exhibit 24 to Declaration of Lin W. Kahn in Support 1492
of Defendants’ Opposition to Plaintiffs’ Supplemental
Motion for Class Certification
3
3
Case5 11 cv 02509 LHK
Document518
2
Filed10 07 13
Page1
of
81
UNITED STATES DISTRICT COURT
NORTHERN DISTRICT OF CALIFORNIA
SAN JOSE DIVISION
RE HIGH TECH EMPLOYEE
ANTITRUST LITIGATION
IN
Master Docket
No 11 CV2509 LHK
EXPERT REPORT OF
PROFESSOR KEVIN
MURPHY
THIS
M
DOCUMENT RELATES TO
Date
ALL ACTIONS
Time
January
17
130 pm
8 4th Floor
2013
Courtroom
Judge
Honorable Lucy
H Koh
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
2
Page2
of
81
Table of Contents
I
CREDENTIALS
1
II
ASSIGNMENT AND SUMMARY OF CONCLUSIONS
2
III
BACKGROUND ON THE DEFENDANTS
13
IV ECONOMIC THEORY AND EMPIRICAL EVIDENCESHOW THAT
INDIVIDUAL FACTORS PREDOMINATE OVER ANY COMMON FACTORS IN
DETERMINING WHETHER AND BY HOW MUCH ANY MEMBER OF THE
PROPOSED CLASS WAS INJURED BY THE CHALLENGED CONDUCT
A
The Challenged Agreements Would Not Meaningfully Reduce
the Supply of
17
Information
1
Shows
Evidence
Source
Employees
that
Restrictions
C
The Alleged Conspiracy Would
D
on
Harmed
if
Methods Would Not
Recruiting
Benefit
There
any
is
Affect
Market Compensation
Some Members
of
22
Employees
and Therefore Determining
Individualized
were Injured
and By
How Much
Would Require
25
Analysis
Tremendous
The Composition
Variation
in
Compensation
of Total Compensation
Paid to Individual
Differs Across Employers
Employees
26
V DR LEAMER PROVIDES
NO ECONOMIC SUPPORT FOR PLAINTIFFS
CLASS CERTIFICATION REQUEST
Summary
B
Economic
of
Dr Leamers
Analysis
28
28
Opinions
Does not Support
Dr Leamers
Claim that
Agreements would Reduce Information Flows Limit
Compensation
1
Evidence
Dr
Shows
that the
30
Flow
Leamer Wrongly Claims
Discovery
of Information and thus
the Challenged
Compensation Data Shows
3
the Challenged
Price Discovery or Affect
Generally
Not be Reduced by
2
25
and
Employees
A
20
Class Even
the Proposed
Others
Individualized
1
2
an Important
18
Employee Compensation is Highly
Which
of Other Defendants are not
and Hires
of Recruits
B
if it
15
Price
Discovery Would
30
Agreements
that His EmpiricalAnalysis of
that Restricting
Defendants
Cold Calling Impedes the
Price
38
Process
Data do not Support
Dr
Leamers
Agreements Prevented Increased
Class that Otherwise Would
Claim that
the
Timing
of the
Compensation to Members
Have Accompanied Economic
NonCompete
of the Proposed
Expansion
40
Case5 11 cv 02509 LHK
C
Googles
Big
Document518
Bang Does
Filed10 07 13
2
Not Support Plaintiffs and
Page3
Dr Leamers
81
of
Claim of
Class Wide Evidence
D
42
Economic Theory and Empirical Evidence
a Rigid Salary
Compensation
at
Adjustment
Dr
Dr
Dr
that in
Some Employees Results
of
in
47
of Others
Some Common Factors
then Only
Individuals Compensation
if
Common
is
Factors Potentially
Compensation
Leamers
Process
5
Compensation
44
Structure
Defendants Necessitate
Several
Analysis Wrongly Assumes that
Leamers
Affect
4
Practices
Circumstances Increased
Reduced Compensation
Affected by
Claim that
43
Economic Theory Does not Support
Some
3
Dr Leamers
Rigid Compensation Structures
Defendants have
1
2
Refute
49
Model Does not Demonstrate
Because
Leamers
it
his Hypothesized
Price Discovery
Cannot Explain Compensation Changes
Constant
Ranking
Compensation
Attribute
54
Analysis
is
55
Misleading
6
One
Cannot
Conclude
that Because
Some
Defendants had Policies and Even
Formulas for Annual Compensation
Adjustments
Move
E
Dr Leamers
Dr
Dr
Model
Econometric
Impact Because
1
2
Common
Exercise
Impact across Defendants
Framework Underlying
Statistical
5
Dr
Show Common
57
59
is
Impact
Assumed Not
61
his Regression
Leamer Does not Report any
Whether
Fails to
in its Implementation
No Common
Demonstrates
The
of
56
Undercompensation
of
in
Dr
that a Limited
Flawed Both Conceptually and
Warm up
Leamers
Leamers
is
it
Demonstrated
3
4
Number
the Structure
Cold Calls Would
Additional
his Results
are Robust
Leamers Regression
Dr Leamers
Analysis
Sensitivity Tests from which
is
Improper
70
or Fragile
Model Does Not Explain Changes
in
Compensation
Over Time
6
Dr
72
Leamers
Conduct
Variable
Cannot Capture
the Impact of the Challenged
74
Agreements
7
Estimated Persistence
Discovery
66
to Evaluate
Model and
Effects
are Inconsistent
with
his Claim that Defendants
Dr Leamers
Price
had Rigid Compensation
Structures
8
77
Summary
78
Case5 11 cv 02509 LHK
I
My
Filed10 07 13
2
Page4
81
of
CREDENTIALS
1
Document518
Professor
name
I
of
earned a doctorate
my bachelors
Angeles
in
At
I
I
degree
degree
the University
theory empirical
have taught
in
from
also in
of
labor
economics from
the University
of
Chicago
Economics
I
I
teach economics
in
My
Of
teach
economics and
graduate level courses
the economics
the incentives
of public
of
my teaching
focuses
focus
in
both research
particular relevance
across
wage
industries
to
structure
labor economics
by age
of
the
Political
education
wages and
age
have authored or
economics Those
including
price
issues In these courses
policy
to
how
and the
to apply economic
economic
I
the
use the tools of economics
has been on integrating
articles
by
skill
Several
of
and gender
My
to
analysis
principles
and
on
my papers
My
structure
work
in labor economics
labor
has
as well as the determination of relative
level
have focused on the determinants
work on wage determination
labor mobility I teach
with a focus on the wage
groups differentiated by
I
microeconomics
in
the issues in this matter I have published extensively
and occupations
addressed the links between
5
I
Los
of California
the impacts of regulation
on two things how
and teaching
addressed the market determinants of wage
of the
1986
that motivate firms and individuals
markets and the determinants of wages and compensation
wages
in
analysis
empirical
4
Economics at
of
both the Booth School of Business
markets the determinants of market prices and
of
system Most
data
Service
of Chicago
the University
understand the behavior of individuals firms and markets and
to
Distinguished
1983
since
economics
cover a wide range of topics including
legal
J Stigler
George
1981
and the Department
operation
am the
in the Booth School of Business and the Department
Chicago where
received
3
M Murphy
Kevin
Economics
of
the University
2
is
PhD level
courses
on
also has
empirical
and the determinants of relative wages across
education and skill
co authored
more
than sixtyfive articles
have been published
American Economic Review
in leading
the Journal of
Economy
1
scholarly
in
a
variety
of areas in
and professional
Law and Economics
journals
and the Journal
Case5 11 cv 02509 LHK
6
I
am a
Fellow of the Econometric
Arts and Sciences
Economic
In
1997
under the age of forty
five year fellowship
7
In addition
regulatory
matters
In
the John
was named
I
my position
to
matters
joint
2005
who show
to individuals
consulting
member of
Bates Clark
the
exceptional
Medal
81
of
which the American
American economist
outstanding
merit
of
American Academy
MacArthur Fellow an award
a
I
the University
at
firm that specializes
on a
have consulted
economic and
involving
practices
was awarded
a
Page5
and promise
that provides a
for continued
and
work
creative
Economics a
I
and
Society
Filed10 07 13
2
awarded once every two years to an
Association
1
enhanced
Document518
ventures
and
of
in the application
variety
I
am
of economics
mergers class
of anticompetitive
at Navigant
also a Principal
of antitrust intellectual
legal issues such as
allegations
Chicago
law and
to
and other
property
damages
certification
exclusionary access
labor
tying price
fixing and price discrimination
8
I
bodies
and
US
I
have submitted expert
of the reports I have
attached
per hour for
reports in
Trade Commission and
Federal
my CV
Court
have submitted testimony in Federal
as
Appendix
on
A
Navigant
I
US Senate
numerous cases
have consulted
and the testimony
filed
my work
I
the
for the
I
and to
have
state
testified
US Department
on
behalf
of the
of Justice
have given over the past four years
Economics is being compensated
regulatory
at
is
a
A list
provided
rate of
in
1,250
this matter
II
ASSIGNMENT AND SUMMARY OF CONCLUSIONS
9
I
Google
have been asked by Counsel for Adobe Systems
Google Intel
Inc
Lucasfilm and
by
1
individual
Pixar
Intel
Corporation
collectively
and representative
Defendants
2
Inc
to provide
Adobe
Apple Inc
biennially
until
2009
but
it
Apple
Intuit Lucasfilm Ltd
an economic
Plaintiffs that an alleged
plaintiffs
The John Bates Clark Medal was awarded
Intuit
Inc
now
is
analysis of claims
conspiracy
awarded
annually
among
See
http www aeaweborghonors awards clarkmedal php
2
In
Re High Tech
Certification
Employee
Antitrust
and Memorandum of Law
Litigation
in
Support
Plaintiffs
Notice of Motion and Motion for Class
Motion
2
October
1 2012 p 1
Case5 11 cv 02509 LHK
Defendants to
and
fix
the compensation
suppress
impact and would be susceptible
a
class of
ll
natural persons
Document518
to class
employees
of their
wide proof
employed on
Filed10 07 13
2
Class
AllSalaried Employee
or
Court to
certify
a
creative
and or
research
United
States
Class
Technical
by one
excluded categories
or
and development
potential
10
two
the
alternative
when
classes only
calling each others
that cold calling
impact
essential to their claim of antitrust
who
limited to the particular individuals
Consolidated
4
Amended
Complaint
define the All Salaried
Plaintiffs
in
Re
receive
Employee Class
from March 2005 through December
Lucasfilm
2009
and senior
6
7
8
is
cold
the Defendants
from the Class are
executives
50
between
the
two
2009
c Google
e Intuit
Ph
Motion
D
ways and
Plaintiffs
claim that due to the
Litigation
persons employed
Complaint
on a salaried
1
basis in the
2005 through December 2009
b
2009
December 2009 f
from March 2005 through December
g Pixar
corporate
d
from January 2005 through December
officers
p1
October
of
effects of cold calling are not
from June 2007 through
or
method 6
7
a variety
in
8
Antitrust
recruiting
cold
members of
the
boards of directors
1 2012
46
Complaint
distinguish
42
Complaint
E Leamer
calls
All natural
as
employees
of all Defendants
Expert Report of Edward
Complaint
retail
same
The
to refrain from
a particularly effective
a Apple from March
from January 2005 through December 2009
Excluded
5
Report
do not
the
of their claims
In support
also distinguish
the compensation
that
Adobe from May 2005 through December 2009
5
among
High Tech Employee
United States by one or more of the following
Intel
I
cold calling has a significant impact on employee compensation
that
3
my report
in the technical
and with
periods
typically
and in
more
evidence specific to one or the other
present
employees
same time
Dr Leamer
and
or
a
a salaried basis in the
Leamer
claim that there was a conspiracy
Plaintiffs
employed on
E Leamer PhD
class definitions
I
who work
AllSalaried Employee Class
Plaintiffs
certify
have asked the
Plaintiffs
natural persons
the
wide
2005 through December 2009
alternative
fields that are
Edward
by
arguments and evidence provided
between
ll
more of the Defendants during
offer the Expert Report of
Plaintiffs
a
as
a class
by one
States
4
defined
as identified for the
would have
a salaried basis in the United
Class As an
81
of
have asked the Court to
Plaintiffs
of the Defendants during part or all of the period from January
the
3
Page6
including
According
to
Dr Leamer
orally in writing telephonically
otherwise applied
for
a job opening
ColdCalling
or electronically
refers to
Leamer Report Footnote
3
communicating
with another
3 adopting
firms
directly
employee
essentially the
in
who
same
any manner
has not
definition
as in
Case5 11 cv 02509 LHK
conspiracy
alleged
the resulting
and mobility had a cumulative
Document518
on
Filed10 07 13
9
Class members resulting in
all
from Defendants than they otherwise would have received
11
Plaintiffs
acknowledge
of a small number
consisted
challenged agreements or
Defendants to not cold
conspiracy
do
impact on a class that includes
DNCC
employees
identified
virtually all
among the
which
11
claim that the small number of
Plaintiffs
during the periods
agreements
not cold call
each others
call
US
I
by employees who
did not receive
reduced the compensation of
engineers
to
agreements
job
workers
cafeteria
41
Complaint
Plaintiffs
bilateral
in
general
My
application
if
if
those
potential
in cold calls
a
Defendant
those employees
potential
all
do not
cold call restrictions
to
received
but also
from
firms as well
by
not prohibited
the
challenged
at
issue in this
at
case
is
words
that
they generally
a firmwith which
employment with
in other
if
or potential
before filing an
information
emailing employees
in general
have received
about a particular job opening
had expressed no interest in exploring
new employment opportunities
was
that
simply were gathering
or
According
or almost all employees in their proposed
conduct
include
12
pairs of
the compensation
of the Defendant
by potential applicants
from calling
of the alleged
between
value
their
that they might otherwise
applicants
certain pairs of
agreements had a class wide
not only affected
claim that
would
This definition
understanding of the
were intended to prevent
exploring
Plaintiffs
as the
salaried employees at all seven Defendants
allegedly
calls
my report
Despite the limited nature
other salaried employees
such as responding to inquiries
openings
agreement
all
cold
lower compensation
agreements between
Defendants reduced the information available to employees about
Plaintiffs the reduced flow of information
81
seven Defendants
refer to in
claim that the reduction
Plaintiffs
of
10
conspiracy
that the alleged
of bilateral
Page7
of competition and suppression of compensation
elimination
effect
2
the
it
had an
Defendant
or in
they were a totally passive
candidate
9
10
11
Complaint
123
Dr Leamer
says that he
Report
12
110 emphasis added
Complaint
According
who
are
s that
Defendants
entered
into several
Plaintiffs
leverage
claim that
that
the
e outreach
cold calling
i
seeking employment and other active
compensation
negotiating
by restricting
to Plaintiffs
not actively
depressed
showing
understand
additional
agreements Leamer
22
by impairing
information
to solicit applications
competition
for
flow about compensation
and job offers reducing
of employees and minimizing movement of employees
Dr Leamer
Defendants
describes
agreements
preventing class members from discovering
abundant
evidence
would tend
the true
to
of
their
between
to all Class
suppress employee
value
4
common
work
from candidates
employees the agreements
firms Motion
members capable
compensation
Motion
p 16
p3
of
generally by
Case5 11 cv 02509 LHK
classes
were
by an amount
affected
conventional methods and
12
In his report
Dr
common
Leamer
Document518
that can
Filed10 07 13
2
be measured on a class wide
Page8
81
of
basis using
evidence
1 whether
two questions
addresses
there is
common
proof
to
Classwide
Non Compete
each proposed class capable of showing that the
the competition
sic compensation
13
of its
members
or formulaic method capable of quantifying
suffered
by each class
that
or nearly
all
14
Dr Leamer
all members
Dr Leamers
2 whether
and
amount
of both classes
had
their
reliably using
materially reduce the information available
to
there is
reliable
yes and
is
suppressed
economic methods
First the challenged
to Defendants
a
both questions
compensation
standard
analysis has three essential steps
reduced
artificially
of suppressed compensation
concludes that the answer
aggregate amount that can be quantified
13
the
Agreements
by an
15
agreements must
employees Second
that reduction
informationmust cause the salaries of individual
employees to be reduced Third the
rigid compensation
must cause the reductions
some employees
of
to
reduce compensation
First the labor markets from
recruiting practices of
number
would
of employees
Defendant
the information
15
wide basis Economic theory and
provided
common
to others
compensation
of the Defendants
Leamer Report
at
10 see
would
not
at
be
to one
11
5
effects
would be
common
at
the second
Finally
if
highly
In particular
employee likely would
an impact on one
21 57
the
directly affect only a small
of the proposed class
This would stop the chain
would
enormous and diverse and
Second any
members
across
are not rigid then
Leamer Depo
hire are
is critically
the information levels of employees at any
the first step
Ibid
Leamer Report
that
that reduced the informationprovided
impact on individual
14
not meaningfully affect
and would not be
same conduct
structures
of employers
This would break the chain
individualized
13
a class
which Defendants
a small number
the
on
somewhat
in the compensation
evidence demonstrate that his analysis of each of these essential steps
empirical
flawed
structures of the Defendants
in
increase
step since the
the compensation
individuals
compensation
Case5 11 cv 02509 LHK
would
14
Thus
None
based on
my analysis
I
with economic
Opinion
1 The
implausibility
suppressed
as
logic
by
which
I
explain
was dwarfed by
to
the
11
percent
number
Opinion
by
estimated
2 Empirical
challenged
during the
Dr Leamer
recruiting
employee turnover
movements between
turnover
agreements
Given
to recruit
that
from
method
their
16
DNCC
recruiting activity
See Table 1 below
to
of Defendants
subject
and from other Defendants
to
will
were
tend to overstate
6
new
actual
their
that
implausible
employees by the large
that
the
employees
whether resulting from cold calling
one
was
this
percent
2001
to
of Defendants
2011 Employee
means that during
by
unaffected
1A
still
and
1B
available
DNCC agreement the
agreements
was
agreements was even lower
and see Exhibits
a
It is
Defendants
to
hires and separations over the period
of potential
And
the
class wide impact on the information
for only about
Defendants that had
opinions
VB1 below
other Defendants
accounted
that
of over 8,000
hiring activities demonstrates
forms of recruiting of other than cold calling
employees
movements
or
empirical
at these firms
they received
as a proxy for underlying recruiting activity
about 99 percent
16
to
opinions
my core
are
combined workforces
economically significant
Employee movements
Dr Leamers
an average
hired
undercompensated
see Part
at
class period demonstrates
compensation
of their
evidence of Defendants
had no
conduct
or another
81
three
not surprising
is
it
about labor market opportunities and compensation available
a
of
detail in this report
of applications
for five years these firms consistently
amount
in
2005 and 2009 Defendants
equal
all
The following
evidence
empirical
of Dr Leamers claim that average
Dr Leamer and Plaintiffs claim
workers per year
alone
to their allegations
high level of hiring by Defendants
Collectively between
hiring
Page9
by documentary and
theory and contradicted
Plaintiffs claims are not supported
Plaintiffs allegations
let
conclude that Plaintiffs claims and
Given the lack of economic
regarding
hold
links in the chain
of the required
are both inconsistent
evidence
Filed10 07 13
2
compensation for other members of the class This would stop the chain
not increase
the third step
Document518
fraction
the actual
Using
this
period
the challenged
for details
17,18
Given the
and used during the class period
of
turnover
accounted
for
importance of the challenged
by the
Case5 11cv 02509LHK
relative unimportance
outside
effect
of
the class period
cross
hires
separations
before and
and
any
in that
restriction
hires
cross
separations
movement would not have a
represented
after
The
the class period
Defendants
1.1 percent
just
1.2 percent
the Defendants
of
on
counterfactually
1
As such
the information available
material
collective
hires
total
hires and
shares from
clearly demonstrate that
employee
of the labor market
changes in those flows would have no
to Defendants
those flows and the associated
81
Defendants
of their total
account for a minute fraction
activity for employees of these firms
effect
data in Table
of
Defendants both within and
the class period
represented
Page10
a share that is not materially different than the corresponding
movements between
substantial
Filed1007 13
2
employee movement between
on compensation Moreover during
from other Defendants
hires
Document518
employees even
recruiting activity represented
if
the only
source of information available to employees
17
For purpose
According
of this discussion
to Plaintiffs
Google Intel began
in
I
use the period from 2005 to 2009 to approximate the
agreements
2005
Together
annual employment at the Defendants
18
This analysis
the
uses
involving
these
hires
and separations
slightly
one year windows
between
out
four
companies
from 2001 to
to identify
previous employer and the hire date
at
Defendants
due to the two different windows
the
of the seven Defendant
four
accounted
class
period
Adobe Apple
about 92 percent
of total average
2011
hires
looking back
new employer and
in a
for
companies
given year the numbers
used
7
one year between
separations
of
hires
the separation
and separations
date at
year For
may differ
looking ahead one
Case5 11cv 02509LHK
Document518
Filed1007 13
2
Page11
81
of
Table 1
Summary of
Hires
and Separations
Defendant Companies
at
Annual Average
2001
From Other Defendant Companies
2010
2001
2009
2011
2011
5,795
Overall Hires
2005
2004
8,814
11,435
8,193
35
From Other DNCC Defendant Companies
95
159
85
28
69
123
64
From Other Defendant Companies
0.6
1.1
1.4
1.0
From Other DNCC Defendant Companies
0.5
0.8
1.1
0.8
12,182
15,985
16,525
14,700
Overall Hires and Separations
FromTo
Other Defendant Companies
71
FromTo
From Other DNCC Defendant Companies
Source
b
Based on
analysis
There were
many
in Exhibit
sources
Defendants averaged
hires
total
1 averaged
suggests
20
of
percent
market and the associated
During
looking
for
Defendants
labor market
to
work
of
and employees
total
agreement was in place
1A
leaving
at
Therefore
in a typical year
co workers
some
This
with the labor
Other sources
of
at other firms dedicated
whom may
internet
of
have
sites
based advertising as well as cold calling from
employers and from Defendants where
Plaintiffs theory
8
new
of Defendants
during the year
employees had direct contact
friends working
Under
workforce
about 11.3 percent
workforce
total
informationfrom
non Defendant
Defendants
roughly 7,200 a year based on Table
information
such as job boards and media and internet
number
hires
the class period
1 averaged
of Defendants
labor market information include
DNCC
1.1
0.9
during the year while separations
that about
the very large
1.8
1.4
about 78,000 employees a year see Exhibit
about 9.2 percent
been actively
1.2
0.9
Defendants
including
roughly 8,800 a year based on Table
workforce
127
of labor market information available
non Defendants
to
168
239
1A and 1B
employees other than cold calling
Defendants to go
305
139
0.6
0.5
Other Defendant Companies
191
57
From Other DNCC Defendant Companies
no
of the spread of information
Case5 11cv 02509LHK
Document518
informationfrom these other sources
Filed1007 13
2
which
any
vastly exceeds
Page12
reduction
of
81
in information
resulting from the challenged
Defendants employees even
c
agreements would have been widely disseminated among
if
Yearto year
fluctuations
that might have
Defendants
widely from
varied
roughly one percent
the hypothesized
fluctuations
in
overall
in cold
There was no reduction
DNCC
2005
for
by Defendants
as a
whole
Thus
of 4,100 in
Opinion
3 A reduction
because
there are
a
Market
agreements
alleged
a all other
in inter Defendant cold calling
including
by which Defendants
the price employers
agreements affected
words they
affected
available
fill
to
compensation
b
As a
matter
of
neither the
those jobs
VB1 below
see Part
Defendants or
2005 2009
with
b Defendants
as during the
Dr Leamers
economic theory
would
would not
recruit
2001
central
employee
number
there is
wide harm
for labor
number
no reason why they would
see Part
conspiracy
when
not lower compensation
demand
jobs nor the
of the class generally
in
other
of employees
affect market
IVB below
to restrict
a small number of
approaching employees at one or a
on a class wide basis
agreements were not commitments to reduce salaries or
9
in class
determined by supply and demand for labor
is
of available
the alleged
result
employees
neither the supply of nor the
Therefore
or compensation
other firms
large
pay for labor and thus the compensation
employers from using a single recruiting tool
few
Such
Defendants during the class period The
the data are inconsistent
earned by members of the proposed class
The
a
Defendants
many channels
price
2009
would be caused by
premise that the agreements reduced informationflows and consequently
movements between
hiring
with economically significant
agreements was essentially the same during
2011 period
low
that
calling due to the challenged
hires between
cross
the class period
to a
conduct on class wide compensation
in
any hiring changes
aggregate changes dwarf any changes in the
percentage of Defendants hiring from either
with
Over
agreements
hiring activity are inconsistent
effects of the challenged
pairs of Defendants
hiring activity vastly exceed
of total hires accounted
reduction
calling between
a high of 12,700 in
of over 8,500 hires These
difference
no cold
resulted from the challenged
by Defendants
d
in
there were
restrict
The challenged
employment and would
Case5 11cv 02509LHK
not have
changed
Document518
Defendants had to
positions
demand
the supply of or
a
certain employees through
The
fill
for labor overall
or the
Page13
number
agreements only affected
alleged
method
particular
Filed1007 13
2
cold
Even
calling
if
of
81
of job
recruiting of
the agreements
reduced recruiting of certain employees from particular employers and potentially
affected
certain individuals
a
new employee for
Adobe
rather than from
class benefitted
c
As a
matter
rather than
potential
of
the person hired
see Part
Defendants had
This reduction
Plaintiffs theory of information
employees as well which
fact that a reduction
could be positive
would
Opinion
a
is
If
such as Microsoft
member of
to the extent
less cold
the proposed
has an effect could raise
it
number
calling reduced the
would reduce
it
could increase
the
or
flow
this
the opposite
would
the pool of potential
amount
increase
of the effect
negative Moreover
for
of
of compensation
hires
that the
employees from the smaller resulting labor pool
some
compensation
by
hypothesized
available
to
members
The
both sides of the
in
is
ambiguous
cold calling
and reduce demand for others
across
Under
of other
Plaintiffs
the fact that the reduction
individuals
on wages would not be common
implies
of the proposed class
that
see
IVB below
4 Defendants
Defendants had
highly
between
below
compensation structures are not rigid
and exercised substantial
employees
individual
was
calling
in cold calling affects the options
demand
increase
the impact
Part
a
Apple
firms and workers means that any overall impact on compensation
market
it
cold
by Defendants
contacted
offer to attract
to
non Defendant
from Microsoft
reduce average compensation
for those Defendants
Adobe
those
IVC below
economics reduced
candidates
by Defendants through
agreement with
from a
an open position
recruiting through
to increase
benefit those hired
For example if as a result of an alleged
channels
recruited
which would
channels
other unrestricted
would be
the impact
as a result
Dr Leamers own
similar individuals
in every year and for each
compensation
model
with large variations
individualized
observationally
flexibility
unexplained
by
Defendant
in setting compensation
implies
that
employee compensation
even within particular job categories
see Part
IVD below
there is substantial
Dr Leamers model Dr Leamer
10
of
As
I
and
demonstrate
dispersion
in
employee
has shown that different
Case5 11cv 02509LHK
jobs have
Document518
different average compensation
compensation
Filed1007 13
2
but not that increases
Page14
of
81
in an individuals
resulting from a cold call results in higher compensation
for other
employees
b Dr Leamers
premise
a change
not imply that
also
in
A rigid
flawed
compensation
wage
increase
for all
structure
even
if
one existed would
more employees would
for one or
because the cost of increasing
entire structure
enormous an
is
compensation for one employee
employees and would be
Thus Dr
resisted
the
shift
would be
Leamers
theory makes no economic sense
c
Finally
Dr Leamers
alternative
analysis cannot distinguish
that compensation
hypothesis
the impact
of Defendants
he hypothesizes from an
employees
is
broadly
determined by competition in a vast labor market and that adjustments for individual
employees unique circumstances
individualized
Opinion
methodological
outside
offer are highly
VD3 below
see Part
5 Dr Leamers
such as an attractive
conduct
and
regressions suffer from severe conceptual
flaws and are completely unreliable and thus uninformative His regression
methodology provides evidence that
is
with his conclusion
inconsistent
of class
wide impact
and damages
a
Given the nature
of Plaintiffs allegations
common
challenged conduct was
there is class
Dr Leamer
during the conduct period were
methodology
values of his
some
suggests
so called
Dr Leamers own
common
whether the impact of the
is critical
effect
is
not
actually
and
substantial
variation
overcompensated
across
Indeed
common
Specifically
positive
statistical
across
a class
see Part
whether
wide
changes
application
of his
the estimated
Defendants
and for
VE 2 below
methods which
I
Thus
critique
Defendants in the estimated
as the result of the challenged
11
on
that compensation
conduct effects vary substantially
the
understanding
Defendants
across
were
to
can be measured
to demonstrate
regression specification
below show
some employees
the impact
fail
that the changes
of the Defendants
further
Defendants
across
wide impact and whether
Data analyzed by
basis
the question
conduct
impact with
Case5 11cv 02509LHK
b Dr Leamers
data
theory
how
members
When
firm
level compensation The
shows
the conduct and
theory
effect
estimated
Defendants
economic
SP
500
existence
My
economic
their
analysis
evidence to support
on members
fails
of these factors invalidates
should not change
periods
see Part
as follows
recruiting
IV I
hiring
show
Moreover
benefitted
there is
these
same
facts
that he fails to support
his findings
of class
if
Plaintiffs
conduct
effect the
members for
all
as measured
by changes
in
the
or even
In Part III I provide background
and compensation
practices that is relevant
that there is neither
is
information on the
economic
logic
many
my
nor empirical
conduct would have a
only one of
to
common
impact
recruiting tools and other
These facts together
refute
imply that some members of the proposed class will have
basis to certify
the
VE 5
agreements would reduce compensation on a class wide
from the same conduct that
no economic
see Part
VE 5 below
Cold calling
Plaintiffs claim that the challenged
his statistical
undercompensation
smaller
Plaintiffs claims that the challenged
of the class overall
determinants
Similarly simply controlling for changes in overall
yields substantially
estimates
In Part
regression
For example limiting his regression analysis
Defendants are not an important source of hires for any Defendant
basis
conduct
to account for important
and financial market performance
report is organized
Defendants and
Leamers
to
but also a major error
conduct effects are unreliable
VE 4 below
index
highly
not only contrary
is
determined
implying overcompensation
positive
conditions
overcompensation
15
is
is
effects also are highly unstable reflecting the
estimated
post conduct
see Part
stock
This
Yet doing so completely changes his estimated
correct
is
81
VE3 below
Dr Leamer
imprecision with which they are estimated
to
is
Dr
estimated
see Part
that his claimed
below Dr Leamers
of
evidence that the challenged agreements reduced compensation
of the proposed class
analysis and
Page15
and obvious feature of his
a critical
compensation
properly
In his conduct regression analysis
of
Filed1007 13
not independent
are correlated
an individuals
no meaningful
provides
c
of
inference
in statistical
of
only because he ignores
significant
that his observations
own
his
2
impact of the challenged agreements on compensation
estimated
statistically
Document518
Plaintiffs
a class
allege
Part
harmed
other
employees which means
V critiques Dr Leamers
economic requirements for
12
class certification
analysis and explains
Case5 11cv 02509LHK
Document518
2
Filed1007 13
Page16
of
81
Case5 11cv 02509LHK
Document518
2
Filed1007 13
Page17
of
81
Case5 11cv 02509LHK
21
Third
bonus
to
on employee referrals
current employees
important channels are
new
who
website
LinkedIn
and dice
com
changed
over time
throughout
22
graduates
is
LinkedIn
for
28
example
for
by
implications First
It
implies
that
the use
a
utilize
81
use of
in
companys
to the
networking
firm Other
importance
may have
mid2000s
since the
job
such as
sites
of these different channels
by Defendants
of
many
but the
different recruiting
and
in the use of one channel can
or at least reliance
that both employers
implies
applications
the
the recruiting practices of these firms
reduction
information on hiring and compensation
members that
of
more
and
increased
it
unsolicited
increased
by
ultimately are hired
The importance
different channels has characterized
important
compensated
who
27
com professional
and job fairs
the past decade
Page18
have formal referral programs that provide a
refer individuals
From an economic standpoint
channels
Many
such as monster
job boards
many
26
university
applicant
use of
Filed1007 13
2
use a large variety of channels for recruiting employees They tend
the companies
rely heavily
to
Document518
on
other channels
it
these other channels will have
implies
that individuals
expanded
opportunities
be
This has two
and employees have alternative
Second
will
sources
including
critical
of
class
as a result of the
reduced cold calling
IV
ECONOMIC THEORY AND EMPIRICAL EVIDENCE SHOW THAT
INDIVIDUAL FACTORS PREDOMINATE OVER ANY COMMON FACTORS
IN DETERMINING WHETHER AND BY HOW MUCH ANY MEMBER OF THE
PROPOSED CLASS WAS INJURED BY THE CHALLENGED CONDUCT
23
The
allegations
pairs of Defendants
Declaration
of
Steven
to
in this matter concern the impact of the challenged
cold calling on compensation
eliminate
Burmeister in Support of Defendants
received
Opposition
For example employee
of
Fichtner
28
referral
new hires See
Opposition to Plaintiffs
27
the
Defendants
Motion
to Plaintiffs
26
percent
by
for Class
p3
Certification
40
agreements between
Motion
for
Depo at 177 210 See
See Declaration
and Declaration
of Jeff
is
the
most important recruiting
Declaration
Vijungco
of Tina Evangelista
of
Jeff
Vijungco
Class Certification
also
pp
Declaration
of
of
method
Declaration
of Jeff
Tina Evangelista
23 Declaration
of
p3
15
for
Adobe
Adobe Systems Inc
Frank
Wagner
in
accounting for about
35
Support of Defendants
Vijungco
p2
p 12
p 10 Declaration
of Chris Galy
p2
Case5 11cv 02509LHK
Document518
employees The challenged agreements
salaried
24
The
who
five individual
certified
at
issues
in
and that there
each class
member without
Plaintiffsclaims
compensation
many
members
other recruiting channels
claim to represent
salaried
in
things both that
the
of quantifying
common
how
that analysis explains
if
2 do not
2 that
had a
issues
predominate
in
to
analysis can
1 do not
agreements that
another
3 do not
compensation Thus
labor markets operate
at
damages owed
of
An economic
evaluations and
reference
or
at
restrict
the relevant
Defendants
economic
an agreement that potentially limited one of
one Defendant might have been made aware
Defendant would reduce compensation
theory has three essential elements
reduction
relative to what
in
gentlemens
a Pixar employee
in
of
by
received
under his theory
In particular
by members
informationmust lead to a reduction
they would have received
of the compensation
class wide reduction
Pixar
of the seven
affect direct determinants of an employees
or performance
opportunities
amount
the
analyses
agreements to affect compensation received
rigid nature
29
persons
all
order to have such a class
compensation
August
would
2 2012
notify
at
at the defendant
that
Lucasfilm
it
if it
in compensation
for internal equity
would not
counter
123 16 126 15 134 23 135 6 148 3
16
if
and
5
3
a
This sequence
Lucasfilm
offer to
1
for those
firms must then generate
Pixarmade an
order
by Defendants
absent the challenged agreements
through the pressure
agreement with Lucasfilm
and that Pixar
see McAdams Depo
structures
in
of the proposed class
those agreements must materially reduce the level of informationpossessed
individuals
or fix
of the proposed class
for the alleged
employees
virtually
employees
technical
understand that
I
individualized
methods by which employees
Dr Leamers
25
on
way
cause class wide changes
employment
81
many employees could be hired
in this lawsuit
among other
reasonable
relying
only
whether given how
recruiting
specific
a
such as promotions
hiring nevertheless
issue is
is
compensation
relate directly to
of
determining whether class members have been injured by the alleged
conspiracy
support
Page19
29
2005 and 2009
must demonstrate
Plaintiffs
did not restrict
or in the alternative
any time between
over individual
named
plaintiffs
were salaried employees
Defendants
all
compensation
or any other element of
Filed1007 13
limit how
prohibit hiring employees of other Defendants
wages
2
made an
a Lucasfilm
offer to
employee
Case5 11cv 02509LHK
which underlies
Dr Leamers
and inconsistent
with economics
A
The
Document518
price discovery
and
Filed1007 13
2
and
internal
evidence
empirical
equity frameworks
of
81
speculative
is
show below
as I
Agreements Would Not Meaningfully
Challenged
Page20
Reduce the Supply of
Information
26
As a
matter
of
economic theory
the impact
source will depend on the size of the supply
the extent
source
is
which supply
to
reduced Here
challenged agreements
provides
leads
to
is
the supply of
Dr
In
Plaintiffs
Dr
and
at
Classwide impact
A would
Defendant
period
A has
a
depend on
DNCC
25
Dr Leamer
31
Cold calling
available
is
any specific
fraction
recruited
32
at
my discussions
A has DNCC
is possible
or almost all class
to
employees
importance
of cold calling
32
recruiting
If
outside
As hires while employees
agreements account for one percent
in the
with
either
Defendants
them
that
than reduces
compensation
above
or
and the evidence
does not disagree
that there
that
is
not
tracked
data
with this but he simply claims that his conduct
However
as
by
directly
the
to
managers
explain
I
do
not
in their
have
channels
does not account
that
other
for
many employees
a very
were
referrals
was an impact which means that
40 425 183 223 413 21 414 7
of
an assumption he
Therefore
cold calling generally
from the Defendants recruiting
such as employee
of other
below market
data and their recruiting
this generally
and interviews
the class
gained by cold calling could reveal
information
is
with which
of other Defendants
measures of the importance of cold calling relative to other recruiting
of recruiting
pudding
that
compensation
provided in declarations
through other channels
Dr Leamer
in the
It
not clearly identified
and in
than information
large
nor explains
or employees that current
declarations
1 the
importance
assumes that all price discovery raises rather
acknowledges
recruiters
of
of Defendant
percent
for all
the
that cold calling
the information
hires then the share of Defendant As hiring potentially affected
Defendant As
neither
was reduced by
agreements on informationpossessed by employees
2 the
and
when supply from one
less information available
combined impact
the
Defendants with which Defendant
30
in
or
recruiting activity with information flow
agreement as a source of potential
calling accounts for
cold
in
of the challenged
relative to other recruiting channels31
Defendant
the reduction
Leamer equate
and claim that reduced cold calling results
27
increases
less price discovery and lower compensation
In effect
of market supply
the elasticity
information that allegedly
model
Leamers
30
members
and
restriction
to the market from other sources
it
supply to the market from one
of eliminating
I
explain
does not demonstrate impact
17
this
combined
below
his
effect
conduct
regression
provides the
must be large
regression
is
Leamer
proof
Depo
so flawed that
it
Case5 11cv 02509LHK
agreements during the class period
important
from firms in general
A did
As
heavily
Even
expand
not
companys employee
a
no cold
A had
Defendant
1
DNCC
has
was
the impact
hired
of Recruits
Defendants
Defendants
new
33
hires
no
was
of
agreement s
because
it
as
assumes that
more
likely that another
A did
with which Defendant
of the Defendants
recruited
Other Defendants
a
without
with which
call
cold
a
B would
cold
loss of
the top
not
single firm
just
for
of other
I
if
20 previous employers
34
by Defendant
of
in a
from an agreement between
impact on the information
available
cross hires between
a very small fraction
the Defendant
Defendants employees
summarized the former employer of
opportunities
call
not have a meaningful
have accounted
3 shows
are not an Important
employee compensation by
agreements affected
the relative importance
based on recruiting data provided
that
more
one
or even
earned by employees of either company
Exhibit
as
cold calling employees of
a Defendant
Using Defendants data
The
A and B could
compensation
it
81
and Hires
on
recruiting efforts
Defendant
make
is to
non Defendant
that the challenged
likelihood
reducing informationdepends
Defendants
A avoids
someone from
Evidence Shows that Employees
The
29
either
agreement a
call
agreements
is
other recruiting channels
utilizing
Defendant
if
the effect
of
lost information is replaced
Source
28
it
by
A has DNCC
Defendant
overestimate
recruiting efforts
Page21
assuming that cold calling
a cold call agreement if that employee
by which
process
its
would
Filed1007 13
2
only 0.25 percent
likely
this
below
explained
Defendants with which
not have
is
hiring from the Defendants with which
in
Defendant
Document518
of their total
new
A striking
and
to
A and
Defendant
hiring
anyway
hires at each of the Defendants
observation
from this exhibit
firms accounts for more than six percent
is
of hires at
any Defendant and that the top 20 firms combined typically account for less than 20 percent
of
33
Despite the availability of this information
to Plaintiffs
deposition
hires
34
First Set of
that
he had no
by a Defendant
My
Defendants
way to
for
which
understand was provided by the Defendants
I
Document Production
quantify
See Leamer
staff standardized
field in the
Requests
Depo
the
at
October
importance of other
Defendants
in
response
claimed during his
as the former employer of new
45 18 48 25
employer names in the recruiting databases
data typically
3 2011 Dr Leamer
was self reported
to the extent
by the applicant
18
possible
and was entered
the prior
employer
as free form text
Case5 11cv 02509LHK
Document518
Filed1007 13
2
Defendant and
hires
total
collectively other Defendants
Thus
even a policy
stronger than the limitation on
periods
class
that Plaintiffs
members
30
typically accounted
that eliminated
a
would
In fact measured
over
all
for less than three percent
hiring from other Defendants
all
single recruiting channel
here
challenge
only about one percent
from
certain Defendants
not meaningfully affect
Defendants
numbers could mask
a narrower
time period
by
low
of employees
of other Defendants
by one Defendant
that any reduction
were an effect for individual
of the importance
to
31
The number
initiated
to
identify
by cold
number
as a fraction
be lost by
of hires
4B shows
from these exhibits
these
separations
same
the total
by one Defendant
even outside
the class
agreements would
Thus
even
if
calling
among Defendants
Defendants versus between
year At
the
from another
and
Exhibit
at
4A
shows
the
the seven Defendants
RD
class
out of the Defendant
same time movements between
As
can be seen
firms is large
the defendants are
the class period
If
hiring
Defendant were economically important in the
price discovery process then employee movement between
19
a
Defendants and other firms
Creative and
of employees in
provides
price discovery that could
regardless if one looks before during or after
of employees
analysis
transitions that might have been
transitions to and from Defendants
figures for the Technical
to
there
any damages they suffered
as a percentage of total employees
movement
and highly variable from year
miniscule by comparison
and thus
But the extremely
the proposed class
labor market
of all employee
on cold
restrictions
and
on
conduct
of the challenged
information and potential
of
broken out by movements between
Exhibit
quantify
Defendant to Defendant
of
summarize the amount
conceivably
effect
or narrower
and of cold calling as a recruiting channel would be
and
those individuals
calls
accounts for
employees or small groups of employees an individualized
of other Defendants
needed
the alleged
in cold calling because
any significant class wide economic
not have
during certain
share of hires
for a substantial
been affected
implies
much
the flow of information to
particular employees that might have
period
is
of
of total hires
Theoretically these aggregate
level of hiring
which
hiring from other Defendants
group of employees where other Defendants accounted
total
81
of
hires Other Defendants account for at most three of the top 20 former employers at any
total
way
Page22
Defendants should account for a
Case5 11cv 02509LHK
substantial
part of the overall
opposite
true Even
calls
is
amount
the
members
if
all
of the proposed class
B
32
Market
price
separations
including
agreements affected
affected
jobs
neither the
Therefore
compensation
33
Even
there is
no reason why they would
if contrary
a meaningful
class wide basis
While
some employees
that
that
upward
Plaintiffs
Affect
Market Compensation
demand
number
for labor
affect market
compensation
pressure
The
decline
in
overall
a reduction
same
in
reduction
above
the decline
recruiting efforts
The
reduction
on compensation
fact that the reduction
in this
that effect
has been
reduced
in potential
at
hires
required
Defendants
to
would
fill
raise
Vijungco
would
not
on
a
hires for those
the level of recruiting of
the open positions which
would
of the effect hypothesized
the opposite
in cold calling affects the options
available
to
by
both
ambiguous even
if it
were
discovery process given that information on compensation
to
a small
only at the later stages
group
that
then
is
pp
56 and
Declaration
of the recruiting
interviewed
and Intuit clearly state that they do not discuss compensation
of Jeff
was
scenario the fact that there would be more demand for some
most commonly provided to candidates
Declaration
those
fill
or
in cold calling
the pool of potential
in recruiting reduces
Hiring should be a reasonable proxy for the price
candidates
to
they
cold calling would reduce the number of firms contacting
and the level of compensation
material Moreover
is
The
words
in other
of employees available
employers and employees makes the overall impact on compensation
35
for
more intensively
compensation and certainly would not reduce compensation
same amount
other individuals
put
the evidence presented
to
necessarily reduce overall
by
into account the incentive
and
of the class generally
sufficient to cause
firms
by
received
determined by supply and demand for labor
jobs nor the
of available
by cold
initiated
pay for labor and thus the compensation
neither the supply of nor the
number
the
limited both in terms of its magnitude
the price employers
is
81
of
that exactly
impact on compensation
Would Not
Methods
Page23
shows
exhibit
other recruiting channels
earned by members of the proposed class
alleged
The
even before taking
fluctuations
by using
Filed1007 13
and from Defendants were
to
would be extremely
on Recruiting
Restrictions
workers
of
lost and the potential
of information
recruiters to compensate
and
2
35
movement
hires
relative to other market level
Document518
pp
process
once the number of
a job or job opening
until the later stages
of Chris Galy
20
for
34
Both Adobe
of the recruiting
process See
Case5 11cv 02509LHK
and
individuals
common
hires
less
members
across
would stand
34
One
demand
types
employees
of
employment
are widespread
in
of software
engineers
in the United
employment
of software
engineers
in the industries
35
The economics
Employees
when they
are
Exhibits
leaving
1A
and
1B show
about 11.3 percent
of the
performed this comparison
conservative
37
Robert
Details
the Defendants
reason
why
of several
in
jobs
of
or less of
of
operate
36
compensation
of
Defendants
and
to
relocate
Defendants are very
of employees hired at
for
of employees at Defendants
engage demonstrates
and Michael
by Defendant
2
1992
are
a
Google
Intel
in
hires at Defendants
during a year
of workers
Defendants averaged
participants
of the average
9.2 percent
in the job market
software engineers and limited the analysis
that
e employees
i
year This means that on average about 20
were active
Defendants
P Ward
p 440
shown
new
the class period annual
focusing on software engineers
calculation
Journal of Economics
38
percent
willing to change
based on CapIQ information for the Defendants
compete
H Topel
which
percent
leaving
and restricting total employment to firms engaged
Defendants
two
Defendants average number
e workers
Defendants combined workforces
I
in
more
other
with any single
States and only about 10 percent
6
During
employed during
of workers
Defendants
industries
for
and the labor forces
that turnover
38
i
while annual separations
number
and
and
or joining is substantial
represented
36
37
For example as shown in Exhibit
and Pixar were
36
young
are
of both proposed
engineers
meaningfully by changes
will not be influenced
recruiting and hiring practices
young
across
of labor mobility provides an additional
Defendants employees
geographically
for software
accounted
employment
would be harmed
accounting for only a small fraction
collectively
5 Defendants
Exhibit
81
of
in the pool of potential
and a large portion
and
geographically
the seven Defendants
As shown
are left out potentially
opportunities
Page24
on compensation would not be
Workers who remain
to all Defendants
Employment
software engineers
employer or even
who
Filed1007 13
2
that the impact
implies
while those
common
job category
is
for others
of the proposed class
to benefit
classes
Document518
in
account
Job Mobility
Appendices
because that
1A
is
the profession
businesses of the type
for
21
that
of
in a general
in
2D
Men
in
which
even this
the
named
sense
in
Plaintiffs
which
only a small share of job opportunities
and the Careers of Young
through
of
in a typical year
to industries
simply to show
percent
107 The Quarterly
Case5 11cv 02509LHK
the sense that they
had sought
out a
Document518
new
This extensive
source of information on market compensation
also
shows
that
employee movements
extremely small fraction
37
Thus
data
of total
show
Filed1007 13
by
job or been recruited
informationabout compensation
obtained
2
to
or cross
across
is
for compensation
employers because
of employees
means
that Defendants
The
if
38
it
same
the labor markets in which
skills
Defendants
a large
a
to
number
of other
firms than others but the
and experience
these firms recruit
and
to
equalize
hire is broad
of employees into and out of Defendants
employees have access
for an
and
and other firms
vast flow of informationabout market
and compensation
opportunities
C
at
for employees against
of employees with the
employees are mobile The movement
thereby
provides a natural
Defendants accounted
companies The competition may be more immediate with some
tendency
81
employer and
another
employee turnover
hires between
compete
of
both employees and employers The exhibit
hires and separations
that Defendants
Page25
Alleged Conspiracy
Would
Benefit
Some Members
of the
Proposed Class Even
Harmed Others
Even
if
the alleged
the proposed class because
cold call the necessary
conspiracy
reduced the compensation
they did not receive
corollary is that
it
information or a job opportunity
of
because of a
lost
compensation of other members of the
increased
proposed class by opening up opportunities
by some members
received
that they otherwise
would
not have
received
or
under Plaintiffs theory providing them with information that they otherwise would not have
obtained
Thus
Plaintiffs
uniform across class
and
Dr
membersnor
at least
in principle
Defendants generally follow the same process
open position
39
arguments imply that the impact
Adobe
Intuit
reaches
out to the
and Intel
all
processes See Declaration
of
describe
Jeff
companys
the roles
Vijungco
p 2 Nearly
defendants use
Jeff Vijungco pp 23 Declaration of Frank
Tina Evangelista
p3
Evangelista
all
is
neither
even harmful to all but rather a mix of benefits to some
caused by the same conduct that could
39
own
Leamers
for filling
open jobs
recruiting department
of managers sourcers
p3
have injured others
Declaration
multiple
Wagner
to a
39
talent
22
Declaration
with an
manager or
and recruiters as key to their recruiting
of Chris Galy
p 2 and
Declaration
sources to find potential candidates
p 10
A manager
of Chris Galy
of Tina
See Declaration
p 2 and
Declaration
of
of
Case5 11cv 02509LHK
Document518
Filed1007 13
2
Page26
81
of
thirdparty
who
recruiter
either before or at the
of
pursue a variety
who
is
of avenues
responsible
candidates
inappropriate
interviews
methods
The manager
Plaintiffs claims
and
the recruiters role
the decision
is
how
affected
Plaintiffs
to
fill
fill
that job is led
only the methods
alleged
if
agreements
including
DNCC
would be
job
by
DNCC
or hired through
See Bates Document
initiate
is
institutional
cold
open
positions
from
fill
fill
some open
a
also use
contact
on
networking
sourcer
with potential
and
or
for
important in evaluating
Dr Leamer
the impact that
the recruiter
cold
framework
The
in person
recruiters
would
23
the
decision
manager and
Since the role of a
find candidates
calling employees at certain companies
companies such
The
at
would be
recruited
through
by
the
as Defendants
firms with which they
net effect of the challenged
channels other than cold calling employees of a
76550DOC000014
claims to
agreements challenged by
or recruiting employees
that candidates
whether
The
made by
is
alleged
calling employees at other
agreement
the likelihood
and
to hire generally
agreements through other channels
to increase
In order to
used to find qualified job candidates
they were constrained
with which there was no
have
to
by
fairs
may
then generally use phone interviews
incentives
who
and selection of
To add
to hire
given the relevant
recruiter is to identify candidates
other channels
identify
to
make economic sense and whether
a job to
marketing
including
that are unqualified uninterested
who
and
candidates
the recruiter
a small group of promising candidates
ultimately decides
has a logical magnitude
whether there
to obtain
of finding
employees
current
candidates
Recruiters
applicants
and on
an employee makes a
efforts
recruiter
by
referred
efforts
40
way
to fill some Defendants
difficult
active
to eliminate
for the position
Understanding
estimate
making more
any
posted internally
jobs website
important
and sponsoring job
through cold calling
including
other screening
40
for
more
An
is
the friend then submits an application
to identify additional
typically those that are
job opening
companys
the
not require
and candidates
such as LinkedIn and attending
positions
40
may
an employment opportunity and
the pool of internal applicants
websites
on
that it is posted
through employee referrals which
friend aware
to
same time
such as monstercom and LinkedIn com
websites
is
The
will assist with the search for candidates
agreements
interviewed
DNCC
offered
Defendant
a
Case5 11cv 02509LHK
cold
including
would
who
denied to the employees
100,000 inquiries a month on
receives
begin
applications
41
while
The consequence
would
the class
common
a
receives
that
already
salary
working
by doing
hardware engineers
so
for the
reduction
probability that
in potential
person hired
off
was
If
able to obtain
Declaration
See
harm
The
class
one
company who
of
Jeff
the
class
call
but where
cold
worked
Vijungco
to
a
accepted
a
to Plaintiffs
only because
of
is
Fichtner
not be
In Plaintiffs
compensation
but for world
for the position
hired might be
an
different external candidate
employs
etc
fairs
etc
employees The
non Defendant
all
that recruiters
called or hired increases
means
software
including
that he
the
use
43
with any
fact that the
was made
that person
better
becomes a
claims has been injured even though he
the challenged
conduct
44
Thus
under the
p3
him it
this
person would not have been hired for
that
he would not have received would have been higher
24
done
claim harmed the
html
Depo at 144 521 147 12 19
44
of
gets hired
employees have access to
the job
all
some members
member who
http online wsj com article SB10001424052970203750404577173031991814896
43
which
two could
professionals
a Defendants
employed by
thus according
a better position
calls
is
example
jobs and potentially
switches
hire might be
of those other candidates
hires through
and
Instead
member who
human resource
accountants
he was previously
of the class
41
42
The
that
million resumes in 2011.42
of the hundreds or thousands of firms that
wherever he previously
by doing so
member
two
received
for
the portal through
site
employee referrals LinkedIn monstercom job
other methods
Adobe
and distinguishing
Or the new
Defendant firms where recruiters do not cold
The
career
Dr Leamer
the cold call
81
value allegedly
the Plaintiffs logic from the conduct that Plaintiffs
to
did not receive
someone
Through
their
hires is vast
not be class wide
some were harmed
if
someone who was employed at one
42
Google
evidence such as that offered by
higher
engineers
adobe com
that there could
benefit even
member who
internal hire
its
of
and possibly the benefit of a better job
the cold call
of such potential
reported
is
it
is
has benefited according
class
The pool
Page27
Defendants
information about
the
did not receive
with one of the Defendants
with
obtain
Filed1007 13
2
non DNCC
or other channels directed at
calls
process these other individuals
41
Document518
is
irrelevant
whether
Case5 11cv 02509LHK
Document518
Filed1007 13
2
Page28
very theory put forward by Plaintiffs the challenged conduct would benefit
even
harmed others There
if it
D
is
Employee Compensation
Which
if
any
no class wide harm even
if
Highly Individualized
is
Employees were Injured and By
some
and
some class members
are injured
individuals
Therefore
81
of
Determining
How Much Would
Require
Individualized Analysis
1
43
There
is
Tremendous
The tremendous
Variation in Compensation Paid to Individual Employees
variation
in annual compensation
7A
each Defendant shown in Exhibits
at
theory
that a rigid compensation
and
structure
employees resulting from cold
individual
each year the range of
at odds
is
would be
changes
of the proposed classes
with a central tenet of Plaintiffs
that changes
necessitates
calls
compensation
total
7B
members
for
transmitted
in compensation
substantially
differs
the class
across
for
45,46
In
Defendants
across
47
44
Thus
compensation
during the alleged
percent
45
The
the
is
conspiracy
Dr Leamer
by
presented
analysis
below
his data
no
promoted
cannot
I
he analyzes
be moved
discuss
in
48
Changes
in
Appendix
example
tenure
sex
base salaries also show
4A
through
Company specific
for
job
Pixars
4D
14A
Dr
very large
the distribution
performance
bonuses
that
also
of
affects
are tied to the
48
In
a particular Defendant
same conclusion
differences
Even
and
this understates
3
14B
Leamer
takes
variations
in
point
out in
any
year
a 10
received
in
response
is
employee
of
not
Dr
in
pay variation
to external
since
e
pressure
one of
there
g
explained by individual
and Appendix
also
compensation
films
critique
in his regression
3A
Appendix
levels which
my
even for individuals
the level of
account
into
success of individual
25
I
this variation
see
compensation
changes
As
compensation
in
across job classifications
below and show in Exhibits
age
at
the Defendants
This implies that under Plaintiffs theory the
from Software Engineer 2 to Software Engineer
characteristics
47
this
substantial
46
As
raise
supports
show
over 4,000 narrow job categories
no reason individuals
in lock step across
period some employees
raise while others received
data
Leamers
move
does not
show
at
analyses
3B
substantial
I
also
show
variations
a particular Defendant
see McAdams Depo
at
42 243
3
Case5 11cv 02509LHK
propensity
for salary changes
coworkers
would vary
terms
employees received
across
for an individual
substantially
show
these data
Document518
that the requirement of
internal
would have
understand
2
any employees compensation
affected
first
equity and
increases
why
one employee received
The Composition of
a much
In
of
his or her
Dr Leamers
annually
differ substantially
for each Defendant
In order to understand whether
and
if
81
the degree to which
His theory would have to be tested and evaluated
separately to understand the source of the variation
across
of the proposed class
similar percentage compensation
Defendants
Page29
employee to be propagated
members
across
Filed1007 13
2
so by how much it
is
a cold call
necessary to
larger raise than the other
Total Compensation Differs Across Employers and
Employees
45
at
8A
Exhibits
the Defendants
components
of
and
8B summarize the
These exhibits
show
that Defendants
differ
in their relative reliance
employee compensation base salary bonus and
example during
conspiracy
the alleged
equity
likely varies across
internal equity
Plaintiffs
Defendants as well
evaluated
46
49
For
compensation
structure
which compensation
the degree to
theories apply
is
which
would vary
across
This implies that the validity of the Plaintiffs theory would have to be
The composition
Exhibits
9A
Apple and Google that
50
to
the three types of compensation
rigid
options
three
separately for each of the Defendants
Defendant
49
and
or
on
period a
Since the degree
individualized
by employees
composition of compensation received
Leamer Depo
at
of compensation
and
9B show
Dr Leamer
also varies substantially
the composition of total
analyzes in his Figures
across
are in
Appendices
15 through
26
5A
titles
within each
compensation for the jobs at
17.50
278 25 282 22
Corresponding exhibits for other Defendants
job
through
5E
Case5 11cv 02509LHK
Document518
2
Filed1007 13
Page30
of
This variation
the validity of Plaintiffs
employees
47
individual
level as
Dr
and
Plaintiffs
The impact
theories
would need
Leamer have
of differences
well The
in
be
in
job offer will depend
on
his preference
riskaverse employee
or
one who
the share of compensation
on
may
options and
stock
Thus
salary
an offer of substantially
company
another
will affect
how
The
and
all its
stock
match
if
the outside
on
stock
will vest
an employee
offer is heavily
compensation
vest
offered to an
when
the
that are unvested
employee without
51
same compensation
Consequently
employer wants to match
of
may place
may
receives
is
in
another
how
All else equal
company where
273 1
4
the equity
because
I
value
of base
to
from
offered
to
by
offer that he
at
same
outside
his current employer
an employee
that
would
a potential
interest
that holds substantial
the employees
inquiry
individualized
soon an employees
the
current employer
options
if
the
offer also will likely differ
For example Plaintiff Siddharth Hariharan testified that while employed by
employment elsewhere
little
base compensation
creates
stock options
the response
the outside
A highly
be worth less
an outside
allowing them to be exercised in the near future will be more likely to
hire than
51
he holds
options
forms
weighted toward stock options then
by some Defendants
options
and how many
different
in
resulting
a corresponding amount
because the impact of the challenged agreements will depend on
options
and any potential
than lower compensation
options
that offer might require only a small increase
reliance
the
compensation in base salary These same factors
an employer might respond when
asks his employer to
matching
almost
that provides
less than
to
provided in base salary
employers frequently
much
extends
greater expected total compensation
consists of a large expected bonus
if it
48
to change
value expected bonus
call
compensation
for receiving
expects
that
do so
failed to
the composition of total compensation
difference
implies
separately for each group of
established
and equity means that the value to an employee of a cold
bonus
him
to
81
was
still
vesting
was growing again and
I
at
Zynga
I
would have to
27
Zynga
was not looking
start all over
to
again
he turned
work
for
down
an offer
another
Hariharan
Depo
at
Case5 11cv 02509LHK
V
Document518
Filed1007 13
2
Page31
81
of
DR LEAMER PROVIDES NO ECONOMIC SUPPORT FOR PLAINTIFFS
CLASS CERTIFICATION REQUEST
49
support
Plaintiffs
class certification motion
their
with the Leamer Report which they
claim demonstrates that the challenged agreements suppressed the compensation of
all
Class members52
Leamers
and so provides
reduction
in compensation
of the proposed class were
A
for proving impact
Summaryof Dr
Dr Leamer
50
evidence
types
divides
capable
is
generally
54
of
or the
link between
showing
amount
2 Defendants
of
that the
to
of cold
52
price discovery
to
Motion
53
In his
asked
at
or nearly all
has not provided
or
members
a class
wide
First he argues that
agreements suppressed compensation
mean that on average members
which he says
negotiating
of the proposed class
In support
leverage and movement
3 empirical
that raises compensation
of employees
their
evidence that job
supports
He
offers three
about compensation
documents which he claims demonstrate
stayers which he claims
he
supported by economic literature of a
is
b informationflows
calling and
class wide
his conclusions
and
between
concern about the
movers
receive
that cold calling
explained at his deposition
that he also
p 23
report
Dr
Leamer does not provide a
in
his deposition
and when
Greater than 50 percent
Leamer Report
clear
definition
of
what
he means by
all
or nearly all
When
Dr Leamer what percentage are you confident class
that
my opinion is that most members of each class were
your expert opinion
members were undercompensated
undercompensated
54
all
was an average
damages
non compete
impact of cold calling on compensation and
leads
Dr Leamer
that
because of the challenged conduct
internal
higher compensation than
members nor
Dr
Opinions
1 economic theory
a the
53
neither that there
or nearly
However
for class certification
his analysis into three parts
job opportunities employees
firms
amount
Leamers
by which he appears
of evidence
of class
undercompensated
lower compensation
received
support
analysis and the evidence he offers demonstrate
generalized
method
the required
all
asked
he replied
what
Leamer Depo
65 heading IV A
percentage
is
322033 10
emphasis
added
28
most
and asked to provide a
range
he responded
Case5 11cv 02509LHK
regards
his conduct regression
agreements was to suppress
Document518
Filed1007 13
2
his claim that the impact
as supporting
Page32
of
81
of the challenged
55
compensation
generally
noncompete
51
Dr Leamer
Second
claims that
class wide evidence
agreements suppressed the compensation of
employee class and
class
technical
that he claims to establish through
members
nearly all individual
support
part of his
this
56
which
I interpret
argument he again
more
that Defendants
Agreements
52
55
don’t
was
would
Dr Leamer
the aggregate
not
at
require
have the
made
information
I
80725
a
that tells you
Leamer Depo
56
57
58
opines that
amount
Q How
I
members
consequence
of the agreement
done
to
it
answer
that
It’s
determined
is
through that damage
the
impact of these
at
agreements
100heading IV B
And
And
Leamer Report
or
some members To
1
results in
economic theory
somewhat
rigid
I
by common
that
factors and
NonCompete
analysis are capable of
I
don’t
Apple and Adobe
between
A Well
have the cold calling agreements
secondly
I
to translate that into some measure of
you can
haven’t
Q Your
largely
claims shows
thing
that
101
added
29
do So
had a database
I
tell
regression
101
Leamer Report
to
all
suppression to members of the proposed
model because
emphasis
harm
of
cold calling that was made and all the cold calling that
the
all
97 19 100 12
Leamer Report
average impact
58
was suppressed
no simple
question
of the allsalaried
broad effects on compensation of the
indicated
I already
have the information on
that the
undercompensation
such that one would expect
information
have
that the
forms of econometric
standard
much
don’t
members
regression analysis that he
of compensation
going to be very difficult
econometric exercise
have indirectly
e the
rigid salary structures
data set that
don’t
as a
is
3 multiple
class
showing
documents which he claims confirm concern with
have widespread effects on compensation
to
Leamer Depo
that
maintained
Finally
computing
and
earned by individual
compensation
internal
of
of evidence
equity
internal
all
opinion
reflects
offers three types
specifically demonstrat
Non Compete Agreements 57
an
as
of the proposed class and not just
2 Defendants
and
internal equity
capable
or nearly
his first set of analyses
that he claims demonstrates that concerns with
salary structures
all
is
that
you need to carry
would allow me
you that before or after
analysis
A The
to
out an
do it But
I
during comparisons
regression
yeah
See also
Case5 11cv 02509LHK
class
caused by the
model
statistical
53
As
I
Non Compete Agreements
2
Filed1007 13
59
words he claims
In other
explain none of
Dr
Leamers
opinions
are supported
analysis Empiricalevidence
including
evidence he ignores
interpretation of evidence that
he offers
contradicts
members have
members
B
Analysis Does not Support
Compensation
Dr Leamers
is
available
Dr Leamers
all
or nearly
all
economic theory does
by many ways
characterized
to employees
Price Discovery or Affect
Generally
on
available
of recruiting
jobs
of the labor market at issue here
not fit the facts
employees
a vast
amount
useful an economic
fit
the key characteristics
Dr
modeled and
model must
Leamer has
not attempted
theory of compensation impact
the competitive
nature
to
of the environment
to match
in
economic frameworks
and that these frameworks
can suppress
worker
Leamer Report
Leamer Depo
explain
compensation
135 emphasis
at
price discovery
various
In order to be
framework
and
their
and thus
employees operate
Price Discovery
to
link the challenged
useful
He
in evaluating
agreements
claims that
to
there
the
are three
the impact of the agreements
mechanisms by which anti Cold Calling
generally
Would
Agreements
claimed by Plaintiffs
that are particularly
Defendants
or market that is being
of the industry
of Information
on economic theory
widespread effect on compensation
movement
which Defendants and
Not be Reduced by the Challenged
relies
his
where
evidence on the amount of available informationand
available
Evidence Shows that the Flow
Dr Leamer
of information
and market compensation mobile employees and
employment and employee
60
to
Claim that the Challenged
Flows Limit
of
59
60
his opinions and demonstrates that class
account for only a small fraction
55
to provide a
as well as proper analysis and
tremendous density of employers and employees in small geographic areas
1
81
of the proposed class
Economic
one that
of
by proper economic
generally and that there has been no harm
not been injured
Agreements would Reduce Information
54
Page33
aggregate damages and demonstrating causality
for calculating
now
Document518
He
added
24825 16
30
focuses
primarily
on
the
agreements
market price
Case5 11cv 02509LHK
framework
discovery
workers
down
slows
occur
transactions
Calling
an
is
many
Calling
uninformed
logic
channel of information
labor contracts
employees
price discovery
63
are negotiated
by members
is
inconsistent
a
and valuing
62
According
he concludes
claims
is
many
Dr Leamer Cold
to
opportunities and
unequal bargains
and
the unique features of
Dr Leamer
Consequently
levels
81
of
absent
Cold
informed and
between
that the challenged
agreements
of the proposed class and cause employees to be
Dr Leamers
that he claims supports
compensation
61
Page34
very sluggish price discovery
about outside
in
The consequence
However
undercompensated
process
prices far from equilibrium
important
Filed1007 13
of uncovering
task
the price discovery
at
2
that labor markets can have
arguing
expensive and time consuming
that the
restrict
Document518
argument about
price discovery
is
invalid
and
the
reduced cold calling and class wide reduced
a link between
with assertions that he makes to support
Dr Leamer Exaggerates
that link
Information from the Challenged
the Loss of
Conduct
56
Dr Leamer
how much
provides no evidence of and concedes
informationmight have been lost
from the challenged agreements
Defendants plays
in
available
57
between
and
62
63
64
65
calling
his logic
and
4B
at
among
Defendants
for
I
discussed
above
Ibid
75
Leamer Report
Leamer Depo
at
80423
Leamer Depo
at
members
in Part
47248 14
31
the
employees
calling
by
other
amount
A1 showed
IV
after
the impact
evaluate
to
of information
of the proposed Class
before and
know
he has done no
that
roughly one percent
during the periods
73
Leamer Report
of
be used
would have on
a very small fraction
including
he claims cold
data that can
the Defendants
and does not
potential
he provides and concedes
on compensation
which
Defendants accounted
employee flows
61
by
4A
Exhibits
the central role that
recruiting and hiring flows
on cold
thus
Despite
Defendants employees and
the price discovery process
analysis of Defendants
that restrictions
64
to
he has not studied
65
movements
of the overall
the challenged
Case5 11cv 02509LHK
Document518
Filed1007 13
2
Page35
81
of
toyear
Thus
conduct
any reduction
informationflow from the challenged agreements would be
in
extremely small relative to the level of the overall flow of information or to the natural year
fluctuation
58
and
Plaintiffs
employees
the flow of
in
Dr Leamer
allege that the informationflow that
data
show
that these were
through
become aware
obtain
indicator
of the importance
extremely small
mean that
This does not
a cold call
Some
of this only because
employees
compensation However even
greater
compensation
for
Plaintiffs theory
some employees
that
would
it
same conduct would
if
may
not create
and
information
opportunities
may
compensation
then use the informationgained to
the loss of cold calls
benefit
flows and
under certain
may have below market
of a cold call and
of the
of cross
of such reduced information
an employee could not benefit from additional
circumstances
obtained
a good
is
members
lost to
The number
proposed class occurs as a part of the hiring and recruiting process
Defendant hires
was
resulted in lower
harm Indeed
class wide
some members
under
of the proposed class as I
explained above
Dr Leamer
59
also ignores
employees to compensate
the informationflow through
restriction
on
calling
among the
number
the vast
would
Google
at
including
60
software
engineers
candidates
engineers
in
cold
5 shows
Exhibit
these exhibits
calls
to
the
ability
including
also
to
accounted
targeted
for
two
Defendants
operate
Consistent
demonstrate that the reduced information
32
so
total
the
use of cold
did not cold
call
Defendants
on
the restriction
good candidates
would be expected
would have been
the Defendants
for the
economics
other
if
hire
and
cold calling to employees
to
at
They
cold
calling
Other firms
change
their
behavior
employees
percent
States and only about 10 percent
in which
suitable
recruits
their
non Defendants
in the United
skills
recruiters at Intel
job boards more intensively
that the Defendants
the industries
with
eg
Intel
meaningfully restricted
both Defendants and
and make additional
if
recruiters
one channel by
Defendants need not even reduce the
of other firms from which
Google
at
their
a matter of
during certain periods they likely increased
also use other channels
employees
As
other channels
For example
calling in their recruiting processes
employees
and
for restrictions in informationflowing through
increasing
cold
of both employers
the incentive
of
or less of
employment
many other
with
employment
results
of
of software
employers had
established
above
flow through a limited channel would
Case5 11cv 02509LHK
not have
impact on the
a meaningful
through
efforts
by
provided
1A
Exhibit
Page36
still
would have access
firms other than the Defendants
as well as to
of
available
to
81
to
the recruiting
Defendants
recruiting
other channels
A simple exercise
61
Filed1007 13
2
flow of labor market information
total
Indeed employees
employees of the Defendants
opportunities
Document518
the realities
illustrates
hiring from and
movements
of total hires over the class period
As shown
of this marketplace
to other Defendants accounted
A conservative
one
for roughly
to help understand
calculation
Table 1 and
in
percent
how much
information potentially could have been reduced by the challenged agreements could use the
2010 2011
higher post period
lost hires as the difference
rate of 1.4 percent
between
as a base of
hire level in the post and class periods
the cross
Defendant cross hires were lower by only about 0.3 percent
an annual difference
with
Defendants
each of the 30 additional
if
provided
information
would be 60
provided
of cross hires of roughly
of one percent
employees
movements
seeking work elsewhere
material
economic
all
sources
effect
on
left
available
to
total
labor turnover
and the firm to which
In other
it
moved
there
to add to the total bits of information
8,800 7,200
Defendants employees
sources friends
etc
would be even smaller Such
overall
compared
year
from one Defendant to another
through other channels
internet
Inter
of about 7,200 per year at
of 16,000 bits of information
in the bits of information
informationfrom
who moved
bits of information annually
by employee
30 employees per
of Defendants
both the firm that employee
information on market conditions
actively
in
to
additional
of 0.38 percent
obtain
twotenths
or less than
66
the
of total hires during the class period
hires of about 8,800 per year and total departures
total
words
number
in the
comparison and measure
such as
new
the actual
67
or an increase
Since employees
hires
co workers
percentage reduction
a small difference
would have no
compensation
66
This calculation
as
a benchmark
slightly
Even
conservative
is
because
as
I
discuss
share
this
employee
of
above using both
no change in aggregate crosshiring among
higher cross hiring in the post class period
increased
67
is
there
may reflect
the
the
pre class and post class
Defendants
periods
during the class period
growth of Defendants
and thus their
employment overall
percentage
is
movement and
much too large because
ignores
information
it
assumes
all
information
obtained by employees
33
in other
acquisition
ways
comes through
The
Case5 11cv 02509LHK
62
This change
also is de minimis
Over
recruiting activity
12,700
in
2005
That range
is
Document518
low
to a
compared with yearto year
by Defendants
the class period hiring
2009
of roughly 4,100 in
more than 200 times
Filed1007 13
2
the variation
the challenged
Leamers
agreements
marketplace
implied by the calculations
the evidence presented
small
of the relevant
Even
Plaintiffs
if
there were
employee compensation
Technical
are poster
children
by
characterized
facilitate
opportunities
Dr
to other forces
simply untenable
24
22 and
given the
do not
and at
wide impact
is
discovery that apply
in
Amounts of
to
lost
with
not grounded
in
the proposed class
Available
some markets with
such a group
a large portion
The
labor force
Information
limited access
to
of the
In particular
members
of the
AllSalaried Employee
labor market in which
they participate
class
is
use of internet and other channels by both employees and employers
and information flows
contacts
allegedly
so that incremental cold calling might affect
represent
least
such as monstercom dice
and
of the information
analysis ignore and are inconsistent
Vast
of employees
compensation
an informed
and
Have
and Defendants
Plaintiffs
extensive
current employees
friendships
is
his claim of class
dynamics of information
Employee
for
mobility
market and
some groups
informationabout appropriate
Fichtner
due
above demonstrates that any such effect
His theory and empirical
labor
of the specific
b
proposed
flows that
provides no evidence of the importance
vanishingly
consideration
68
in cold calling due
reduction
in hiring
roughly 30
realities
Dr Leamer
the nature
to
above
based on his Figures
of total compensation
incremental information
to
because of the agreements
64
from a high of almost
claims that the impact of the challenged conduct was economically significant
imply a sensitivity
would be
81
a range of roughly 8,500 employees per year
Given the degree of fluctuation
implying effects as large as 20 percent
63
of
in hiring and
fluctuations
varied
employees per year of what would be caused by the hypothesized
to
Page37
university
many publicly
com
graduates
with fellow students
salarycom
who
available
etc
are recruited
and colleagues
Depo at 45 946 4
34
data sources
and networking
by technology
68
on
salaries and
among both
firms
and maintain
Case5 11cv 02509LHK
Dr Leamer
65
implies
Defendants employees
Document518
that the labor market
2
Filed1007 13
from which Defendants
when
Page38
81
of
and where
hire
watercooler
compensation
obtain
information that they can use
consists of hapless
by
weakly with employers who
market
of technical
centers
contradicts
Dr Leamers
hiring
and separations
annual hires and separations
percent
that
70
compensation
at
of average annual
lacks
changing number of
Valley and other geographic
implication
71
firms to
But this characterization
in Silicon
The high
Defendants shown in Exhibits
as a fraction
consulting
members
that
high costs for transactions
involve
on
of information which
scraps
large and constantly
unsupported
their
rely mostly
often hire private
time money and personal dislocation
services including
obtain
to
and other employees located
proposed class are employed in jobs that
turnover
sources
The agglomeration of Defendants and a
other employers
technology
69
Internet
provide aggregated informationabout
credibility
who
and poorly informed employees
talk perhaps supplemented
they then use to bargaining
they negotiate
of the
involving
labor
rates of employee
4A and
4B
with the
employment between
sum of
10 and 25
during the conduct period demonstrates the substantial flow of information of the type
Dr Leamer
Dr Leamers
claims was restricted into and out of these firms and contradicts
claims that these employees were immobile
c
Lost Information
will not
have Class wide Impact
if it
is
Unique to
Individual Employees
66
valuing
According
to
Dr Leamer
expensive and timeconsuming task of uncovering
the
the unique features of workers
that cold calling helps uncover
slows
unique
down
the price discovery
features of potential
process
employees
is
72
and
But his claim
inconsistent
with his
claim that there would be class wide impact from reduced cold calling through the price
discovery process Indeed he stated
69
70
71
72
Leamer Report
75
Leamer Report
74
Leamer Report
73 emphasis
his deposition
75
Leamer Report
at
added
35
that
employers including
the Defendants
Case5 11cv 02509LHK
often differentiate compensation
information would
lost
would
Filed1007 13
2
the price discovery
process
employee specific
d
is
employee specific
unique
well accepted
economic
Dr Leamer
Claims about
Lost Information
provides
economic models
the labor
which
Leamer
to
lower
market but
cites
all
prices might rise faster
cites
but the
rockets
for understanding
limitations
that paper does not
why
prices rise quickly
in
by
itself
information
by Joseph
show
socalled
than they
fall in
Stiglitz
many
demonstrate that
imperfections
The
limitations
other three
and feathers phenomenon
rockets
is
or claim that
75
as a result of information
the
framework
cites
markets with imperfect
information
according
76
Dr
in the labor market leads
of information
and feathers model does not imply that a restriction of
information in the labor market would cause a reduction
only
Price Discovery
nor the broader
the cited literature
of his claim that restriction
these in support
wages
Neither
and incorporate
involve
and
claims One paper he
neoclassical
that acknowledge
economic papers he
that
Literature
model has
for his
undercompensated
employees are
to
74
literature
support
argues that the fullinformation
markets including
The
cold called and
the extent
flow and price discovery
claims that his information
in the economics
literature
73
There would be no class wide impact
Dr Leamers
Finally
81
features
not
To
features
of
then the effect of reducing cold calling also
are not Supported by the Economic
67
who was
to the unique features of the worker
relate
Page39
unique
employees to account for such
across
any impact on employees without those
not have
will be
Document518
in
wages
Rather
response to positive information but
fall
these models explain
slowly in response to
unfavorable information
73
I
Leamer Depo
know
dont
occur
are
are
Leamer Report
Joseph Stiglitz
Review
76
The
460 June
cited
people
for
who
promotions
thinking
result in the price
these
internal
turned out to be extraordinarily
A
discovery process
promotions
the big
bumps up
that
good
Information and the Change in the Paradigm
in
Economics 92
American
Economic
2002
paper by Green et al
Ye 2006
internal
66
major over the counter
and
Q Dont
would be relevant Because Im
it
compensation
74
75
316 16 317 17
at
that
develop
financial
theoretical
2010
is an empirical
market not
models
paper documenting
asymmetric
price
a labor
market The
cited
papers by Tappata
explaining
asymmetric
price
responses
36
adjustment in a
2006
and Yang
Case5 11cv 02509LHK
68
The economic
closely
to
on
and in turn
Samuelson made an
results
Dr Leamer
to
parties have
would
knows
prevail
with
more
is
77
information
78
some mutually
Similarly more recent economic
Dr
beneficial
Leamer
Thus
are forgone
asserts in his report that lack
employers and employees
the available
supported by the economic
by
William Samuelson
In
February
Ausubel Lawrence
80
81
Science
full
when
similar results
J and
BV
beneficial
model
Samuelsons
party
who
when
does not generally
to
information but instead
the
demonstrates
79
there is asymmetric
information
of information would disadvantage both
Leamers
empirical
his claim that the challenged
price discovery
framework
1986
Ibid
Motty Perry
study
unique
features of individual
agreements had a class wide impact
52
July
81
1984
bargaining game with asymmetric
to calculate
from the behavior
Sequential
not
evidence of Defendants recruiting and
a sequential
In order
is
correctly
of his
his
informed
Bargaining under
payoff
the
opponent
uninformed player
p 1004
Cramton and Raymond J Deneckere
Bargaining with Incomplete
Game Theory Aumann Robert J and Sergiu Hart eds Vol 3 Amsterdam
Peter
of
chapter
Leamer Report
6870
Leamer Report
50 2002
73
37
And
Asymmetric Information Academic
2 1986
M
Information Handbook
Elsevier
or
and draw the proper inferences
Grossman Sanford
Press revised
79
literature
a similar vein Grossman and Perry
must anticipate
literature
Bargaining Under Asymmetric Information Econometrica
information and obtain
see
However
Further his argument that cold calling uncovers
employees contradicts
78
some mutually
evidence shows that the challenged agreements would not
Dr
77
that
80
meaningfully affect information flows
hiring practices
81
of the proposed class
information is more favorable
with asymmetric
trades
of
to the informed party than the price
favorable
informed party than the price that would prevail with
69
members
that the informed party has superior information will take this into account
establish that the price that prevails
Indeed
of
information Rather he explains that an uninformed
full
formulating his strategy
that
Page40
informationcorresponds more
literature showing
asymmetric
does not establish that the resulting price
that
this
Filed1007 13
hypothesizes that reduced cold calling affects
under compensation
early contribution
when
are foregone
in
2
with asymmetric
bargaining
mechanism by which
the
negotiations
trades
literature
Document518
Case5 11cv 02509LHK
2 Dr Leamer Wrongly
Document518
Filed1007 13
2
Claims that His Empirical Analysis
Page41
81
of
Defendants
of
Compensation Data Shows that Restricting Cold Calling Impedes the Price
Discovery
70
In addition
Process
wrongly claiming
to
economic theory and economic
that
show
literature
that reduced cold calling limited informationflows and price discovery and thereby
ed
suppress
82
employee compensation on a widespread basis
analysis that he characterizes
Cold Calling would
discovery process
83
Dr Leamer
compensation
He
Defendant85
this
However
claims that
packages
a
show
6 and
symptom
empirical
evidence capable of showing that
instead
restricting
the price
no common evidence
that there is
of price discovery at
who moved
for those
7 which
between
compare
Defendants than for those
of
it
does not support
this
process
his conclusion
better
who stayed
84
compensation
total
Defendants and employees that
2001 and 2011 demonstrate
were meaningful
work would be
median base and
the
move between
each year between
comparison
provides
employee compensation by impeding
his data
of employees that
in
common
Leamer
compensation
says that his Figures
respectively
additional
suppress
artificially
suppressed employee
71
as
Dr
86
stay
However
at
a
even
if
that the challenged
agreements impaired informationflows and price discovery
72
First the economics
generally
receive
atypically large
occur generally and
82
83
84
85
wage increases so
the pattern
89 emphasis
Leamer Report
shows
shown by
that job changers
Dr Leamer
would
Economic theory and evidence imply
not evidence of disequilibrium
Leamer Report
mobility
80
91
Dr Leamer
There
are
thousands
and the
of
between
those
who stayed
at
a Defendant
91
but his
stayers category also
stayers category only excludes from the Defendants total employees
the individuals
Defendants
very few
stayers
movers in
Dr
his data
Leamer has
only between
not controlled
26
to
178 movers
in a
for potential differences
year compared
in
mix between
with tens of
the
few movers
many stayers
Topel Robert
Journal
added
stayers as
describes
new hires His
who moved
87
on between employer
87
is
Leamer Report
includes
86
literature
of
H
and Michael
Economics
May 1992
P Ward
Bartel
Job Mobility
Ann
P
and the Careers of Young
and George
38
J
Borjas
Men
The Quarterly
Middle Age Job Mobility
Its
Case5 11cv 02509LHK
that an
employee who moves
costs
However he
portion
of their higher
Second
the process
movers are
desirable to another
their
compensation for
same compensation
movers do
employee
increase
not increase
firm matching
if
stayers and
they had
the compensation
moving costs and what
of such
attractive
to
means that ex ante
unique
features
because
it
In other
words
The
because they are
Dr Leamer says but
as
nothing about
reveals
not evidence that stayers
of stayers
uniquely
will be
and were chosen
exactly this conclusion
supports
moving
and
generally
moved
that the relevant
movers
are selected
their
is
employers must
in
promoted
not affect compensation
than observably
must net out movers
Employees who move on average
in effect getting
81
of
for this rather than reflects a disequilibrium
movers
by which
recognizes
acknowledges
the magnitude
compensates
firm perhaps because of
movement does
appropriate
88
Dr Leamer
suppressed
is
no evidence about
compensation
stayers are not equivalent
the hiring firm
First as
Page42
in compensation
a larger increase
whether compensation
provides
Filed1007 13
2
moving Dr Leamer
employees for the cost of
disparity for evaluating
earnings
likely will obtain
employees for two reasons
similar incumbent
compensate
Document518
would have
compensation increases
economic
vast
received
but does not rely
literature
on
the
of
on
disequilibrium
or
undercompensation
73
their
Second and
movement
compares
critically evidence that
affects the compensation
the compensation
that the compensation
of
of
affects
argument requires that the change
Without
stayers
this
concerns that converts
unproven
one
Determinants and Consequences
Borjas
George
J
Topel Robert
Journal
of
compensation
for
which he wrongly
Working Paper
stayers is not evidence that
comparison
year
Yet
but provides
No 161 NBER
is
closed by
employees
no evidence
the compensation
there is
paper
May 1992
Determinants and Consequences
Bartel
Job Mobility and the Careers of Young
Ann
P and
Working Paper
George J Borjas
No
161
39
NBER
of
internal equity
Working Paper Series
Cycle Working
it
his price discovery
movers also changes
claims
is static
No
no
support
for
January 1977
233 NBER Working
1978
H and Michael P Ward
Economics
of stayers
raise into raises for all
Job Mobility and Earnings Over the Life
Paper Series February
88
stayers in a particular
in compensation
link
persons
Dr Leamers
of stayers
movers and
movers
movers earn more than
Men
The
Quarterly
Middle Age Job Mobility
Its
Working Paper Series January 1977
Case5 11cv 02509LHK
on
his claim that a restriction
Document518
information
employees of other Defendants affected
74
Thus
not support
Dr Leamers
his claim that
compensation
3
According
2005
because
its
expansion period began
we
2004
in
in
his Figure
9 he
uses
agreements was motivated
expansion with employees
average
provides
data
He
was
peremployee
support
expansion
after
from Apple
by Defendants
incentive
89
90
91
revenue
at
Proposed
of the
Expansion
these agreements were
90
members
He
put in place
of expansion
period
of informed
in
would
that
claims that this
of the proposed class
referencing
his Figure
if
8
91
his claim that the timing of the
sharing
improved
Apple
during the
profits
NonCompete
the risk of sharing
the gains
of
Apple with average per employee compensation in his Figure 9
Dr
Leamer appears
employees
to
increased
DNCC
94
Leamer Report
Members
Non Compete
a threat to the firms high levels of profits and that the comparison
434 17 18
Leamer Report
of the
Economic
Apple revenues surged and when
for this conclusion
them
at
to
rise
to
desire to avoid
to enter into information restricting
Leamer Depo
conduct or that the
with periods
to support
agreements as evidence that the Defendants incentive
that the cost to
the alleged
do
of labor markets and
claims that Apple data show that the
went into effect when
with the workforce
to
compensation than would be expected absent the challenged agreements
less equity
Agreements
why
and that subsequently
use 2010 and 2011 as the relevant
Then
members
of class
calls
members generally
Claim that the Timing
surprise
81
equilibrium
the timing of the agreements coincided
otherwise have caused compensation
received
a
of
through cold
obtained
Have Accompanied
not
Page43
operation
Compensation
Increased
Dr Leamer
to
the normal
from reaching
Class that Otherwise Would
89
of class
was impaired by
Dr Leamers
Data do not Support
Agreements Prevented
75
reflect
price discovery
conduct prevented compensation
alleged
would have been
that
6 and 7
Figures
Filed1007 13
2
98
40
to view the timing of the challenged
reduce informationflows
when
their
profits
equivalently
increased
agreements also increased
at
so their
that
time
Case5 11cv 02509LHK
Document518
2
Filed1007 13
Page44
of
81
Case5 11cv 02509LHK
Document518
2
Filed1007 13
Page45
of
81
Case5 11cv 02509LHK
79
The
change
in
heavily
influenced
Google responded
fact that
compensation
by
Dr Leamers
Thus
Economic
of Big
the
Bang
uncommon nature
is
offers three additional
class
101
a
were
points in
time
common evidence
Dr Leamers
Facebook
Claim that
Rigid Compensation Structures
Dr Leamers
analysis in support
evidence that he claims demonstrates
the link between
specific
at
81
instituting
of evidence related to factors such as
Refute
of
at Defendants
not only fails to demonstrate
AllEmployee Class102 To supplement evidence
showed
outcomes
to different Defendants
employee compensation would have been widespread
the
Page46
from Facebook by
that compensation
specific
essential element of
motion
Filed1007 13
to a unique challenge
view
factors that were
have
The second
certification
the
2
Theory and Empirical Evidence
Defendants
80
supports
discussion
but in fact highlights
D
Document518
that the artificial suppression of
extending
which
I
to
all
or nearly
above
discussed
of analysis that he says support
his claim of
members
all
that
agreements and suppressed compensation
the challenged
types
of Plaintiffs class
of
he claimed
generally he
wide spread
impact on
members
See
http
for
example
techcrunch
a
November 2007
online
com 2007 11 21 facebook
Facebook
article
Googlers At
Stealing
stealing googlers at an alarming
rate
a
An Alarming Rate
May 2008 online article
Google Finds That Perks Cant Keep Some Employees From Leaving
http
www dailytech com Google Finds That Perks Cant KeepSomeEmployeesFromLeaving article
11794 htm
at
President
of global
communications
and
public
affairs Elliot
Schrage
jumped ship to work
week Just two months prior Sheryl Sandberg had left to become
Facebook and a May 2009 Wall Street Journal online article Google Searches
Facebook
this last
the
number two
for Staffing
at
executive
Answers
startups
http
recent
online wsj com article SB124269038041932531
weeks amid
display advertising
the departures
chief
of top
102
Facebook
Leamer Report
html
including
David Rosenblatt Meanwhile
engineering director Steve Horowitz
like
executives
Inc and
Twitter
101 emphasis
Concerns
advertising
about a talent exodus
sales
midlevel employees
and search quality chief
Inc
added
43
Santosh
like lead
Jayaram
have revived
boss Tim Armstrong
designer
in
and
Doug Bowman
continue to decamp
to
hot
Case5 11cv 02509LHK
Economic
theory
principles
Defendants
Multiple
As
I
now
through a
the impact
regression
common
of the proposed class though
maintaining a
discuss
somewhat rigid
compensation
considerable
structure
difference
objective
individual
do
105
the sparse
106
evidence
is
employ
to
evidence
faculty
Journal
by
of the
rapidly
105
a
or circumstances
more
in
flexible
There
hand and
performance
and
does not
is
seniority and other
the one
in
broad labor
the
and other
my review
106
Based on
of declarations
support of this claim in detail
its
employees
market for members of the proposed
times and the unimportance
price
2003
However
by Facebook
discovery
find that
I
and
class with
of cold calling among
process
procedural
equity
is
more important than
sample
Do New York Basic Books 1984
World Can Unions and Industrial Competition
vol 22
Winter 2008 pp 153 176
What Do Unions
a Dynamic
Economic Perspectives
in
about cold calling of
at various
and Honoree
in their
B and James L Medoff
Sluggish Institutions
of
Dr Leamer
concern
information flows and the
Terpstra
T
at each Defendant
by
However he
compensation and compensation changes
offered
hiring
work
Freeman Richard
Coexist
that
would
loyalty
contributions
compensation on
in setting
goals
loyalty than
unionized workforces
to the university
For example in their
Hirsch Barry
to maintain worker
relative to other compensation
he cites relate to Googles
in contributing
equity
Dr
with
structure
to assure internal equity
salary structure
to determine
employers such as Facebook
internal
104
101
not address
Defendants
compensation
firms have incentives
of workers
by another Defendant This
new
rigid
with compensation managers
note that documents
not
a
promotes greater worker
between
characteristics
Leamer Report
I
to
internal equity
evidence are inconsistent
systems of the Defendants that rely on individual
my interviews
to
agreements on compensation
that emphasizes and rewards individual
characteristics
compensation
104
Dr Leamer
salary structure
rigid
by adhering
103
Theory Does not Support a Rigid Salary Structure
the strength of this incentive
which a
103
to
loyalty
81
of
evidence demonstrates a widespread impact on compensation of
members
According
salary structure
theory and empirical
claim that
81
maintain worker
Page47
analysis
Leamers
Economic
rigid
of the challenged
explain both economic
1
to
that he says reflect adherence
internal documents
and
principles
Filed1007 13
firm incentives
implicating
of internal equity
2
Document518
1
44
p 135
Case5 11cv 02509LHK
Document518
2
Filed1007 13
Page48
of
81
Case5 11cv 02509LHK
importance
Alexandre
Document518
of this effect or even that
Mas111
to support
his
is
it
performance
argument but Mas does not
than the one
role of
at
issue
demand and
this article
new
acrossthe board wage cuts
of
employee
translates
Dr Leamer
no
adjust compensation
hiring
call
their
was reduced by
of incumbent
the challenged
theory in
less
Microsoft
those challenged
Yahoo
or
at
any
Adobe
internal
no
decline
in
not explicitly considered
information
113
or analyzed
made
Leamer Report Footnote 126
Quarterly
Leamer Report Footnote 126
Labor Economics 243
at
referring
See
faculty
to
of
time
Defendants with
Albert
Rees
by a
pressures
employment
that
in
resulting
DNCC
to
or
place
Since
Dr
the
according
However
from which
it
if
new
The Role of
1993
87 101 18 4132 20
46
Apple did
hires from
employees
obtained
same adjustments to maintain
Leamer has
Mas Pay
2006
to
agreements than
called or otherwise recruited
Apple and
783
but
but simply that Defendants agreed not to cold
cold
Alexandre
112
received
any hypothesized
not
shown
and indeed has
whether the challenged agreements resulted
Journal of Economics
the
the information that results
and the available evidence demonstrates that there was not
Performance 121
Leamer Depo
to be
to
describes
claims
do not claim
agreements were not
but instead
who
maintain internal equity there was
to
of the hundreds of other companies
equity would have had
members job
adjustments for other
hires transmit
Plaintiffs
agreements
then the same information was transmitted
113
new
employees
employee movement between
if
by
higher compensation
Dr Leamers
employees of certain other Defendants for certain periods
not cold call employees
of
how
at all into compensation
to maintain internal equity
would have occurred
112
inherent in
in this source of informationand thus
reduction
81
the dispersion
in
salary structures of university
the fact that if
ignores
adjustments to compensation
in
on
forces
if
police union
of
article
changes
by Albert Rees
article
fairness which does not support
employees to preserve
Finally
an
also cites
the impact of market
on
an
cites
consider
the effect
Page49
a very different labor market and different event
simply emphasizes the uncertainty
or incumbent
84
Dr Leamer
here
He
context
in this
wages within an employer Rather he examines
of
111
present
Filed1007 13
2
Reference
Fairness
in
it
is
in
any
loss of
irrelevant
Points and Police
Wage Determination 11
Journal
Case5 11cv 02509LHK
Document518
2
Filed1007 13
Page50
of
81
Case5 11cv 02509LHK
informationconsidered
in
employees For example
Each manager
overall
setting an overall
in
2004
Intel
the poorest
the top performance
87
Assume
rating averaged
Assume
that the
implication
receiving
Leamers
the managers
to
much
smaller
based heavily
Intel
on
compensation
Based on my
with similar
budget that year
increases
discussion
filed in this
matter
employees
willingness
I
the fact that the total
a cold call
percent
is
116
from
salary increase
who
This
an
is
not only that the individual
five percent
generally
is
too
he manages 25 employees
market adjustment then his other 20
Thus
the exact
the budgeting
a system
where
amount depends
process
granting
that other employees receive
implied by
Dr
Leamers
amount budgeted
that drives
above average
below average
theory
for salary increases
generally
by thirdparty companies
surveys performed
on
like
76583DOC001487pptx
with compensation
understand that Defendants
to
the
and responsibilities
skills
than five percent
average increases
data from the compensation
data
a 20
market compensation
that
who
received
a cold call
received
provide information
calls
among those 20 employees
ignores
to increase
that salary increase both for the individual
some employees may require
Dr Leamer
who
for
115
ability
a job offer with
received
claim that cold
increases rather than the above
116
match
salaries
guidelines
Intels
of those employees received
give five of his employees a 20 percent
to
employee
employees were that employees
salary increase
changes at Defendants essentially creates
salary increases
115
to
salary increase
the salary distribution
compensation
2005
In
learns that one of his engineers
undercompensated but
a job offer is
employees will receive
88
when one
that the engineer
manager wants
Dr
of
and he wants
is
policies
across
over 10 percent
the job offer and for other engineers
received
If
at one Defendant
81
companies
performance ratings should get no salary increase while those
Defendant and
another
low
salary increase
for a group of employees
manager
a
in
of
to allocate her pro rata share of that budget in salary
Given this system a manager would have only limited
compensation
Page51
for the
change
a four percent
some cases by company
distributing the aggregate budgeted
received
Filed1007 13
2
budget for salary increases
budgeted
how
then decided
to her staff guided in
increases
Document518
pursue outside
reasons and would remain so even
if
managers
at the
infrequently
opportunities
as
evidence
he received an increase
48
Defendants
and
counter outside
in
that
pay
the
my
review of Declarations
offers
employee
because they consider an
is
disaffected
for other
Case5 11cv 02509LHK
Radford and
employees
on
not
117
The informationeach Defendant
that information is specific
positions
at
but his ability to
least
in
process
the short
obtain
to
much
may even
81
of
hires and separated
and similar firms is
broader than the Defendant firms and
Defendant which then
by
a raise
will not provide a basis
threatening
on which
are matched
move
to
the
company Due
compensation increases
lead to smaller
to another
would decide
the Defendant
salaries throughout
in favor of increasing
it
new
from Radford
receives
the
at
from
Page52
with the
An
which the Defendant benchmark employee compensation
might be able
do so
market intelligence
budgeting
types of jobs
the firms against
employee
individual
to
Filed1007 13
2
information obtained
idiosyncratic
from data on salaries paid in a marketplace
derived
same
Document518
for other
to
Defendant
to ignore
the fixed
employees
at
run
3 Dr Leamers
Affected by
Analysis Wrongly
Some Common
Assumes that
Factors then Only
if
Individuals
Common
Compensation
Factors Potentially
is
Affect
Compensation
Dr Leamer
89
who would
claims that changes in compensation of a small number of employees
have received
a cold call
would have class wide impact
increase
compensation
most in demand
common
at
across
any
to
of
let
However
alternative
117
Indeed
current
Plaintiff
in
part
To
support
by common
individuals
factors
provides
such as experience
because they received
that the level of compensation
Hariharan testified that he did not share
a
new
job
Hariharan
120
49
on
paid by Defendants
His analysis cannot distinguish
employers when considering
Leamer Report
claim he
this
that compensation
185 16
118
and related practices would be expected
his analysis does not demonstrate that changes
class wide
hypothesis
time
show
alone a small number of
compensation
Defendant but for the challenged conduct
the board rather than be narrowly focused
118
factors analysis
another
Cold Calling
that
point in
employees can be explained
education
from
those
the impact
compensation
at
an
to
that are
empirical
individual
job
title
compensation
a cold call
of Defendants
Depo
in
the skills
to
would
and
of a subset
affect
he hypothesizes from an
employees
information
is
broadly
with prospective
104 105 18 136 24 137 12 184 5
or
Case5 11cv 02509LHK
Document518
Filed1007 13
2
Page53
of
81
determined by competition in a vast labor market for similar employees and that adjustments for
unique circumstances
Dr Leamers
90
members tended
to
of particular employees are highly
individualized
claim to be able to demonstrate that generally
move
claim of class wide impact
119
time
over
together
Such movement
is
is
neither
the compensation
nor informative
surprising
of class
about his
As
the hallmark of a competitive marketplace
I
explained above the labor market in which Defendants compete for employees and in which
members
of the proposed class seek
extensive
flow of informationthrough a variety of channels and high employee
it
is
not surprising
common factors
that
employers and jobs
across
the market
would have
compensation
more or
pay variation
An
attracting
less attractive
11A
a
claims that
119
120
variation
Leamer Report
The degree of
across
to
121
Exhibits
Google
jobs
123
that
If
several
with
makes
Leamer never even examines
he had he would have discovered
grades and
He
many
a regression
variation
it
with job categories
will understate
ignores
and
11B show
Dr Leamer
the ability of
that both the level
within his job categories
122
titles
titles
and that
which
analysis that
see
left
built
relatively little scope
about
that
the distribution
across
of annual changes in total compensation
analyzes in his Figures
job
categories
companies
for the
pay
in
for the
using the same algorithm that
121
122
50
Apple
17 Appendix 6A through 6E show
major jobs at the other five Defendants
15 through
jobs at Apple and Google
Leamer Report
employers to differentiate
events
major at these
Leamer Report
the ability of
Dr Leamer
on
high level
he claims shows
employers to move individuals
of annual changes in total compensation
123
Dr
that
Defendants had highly structured compensation systems
identify
122
unique characteristics
vary greatly across individuals
distribution
the
not competitive
and keeping good employees At the same time
provides
cold calls or other
11A
is
130
individuals since
response
Thus
mobility
in average compensation
compensation that
ranges of salaries for grades and
established
for individual
offered
of the variation
rapid and
11B120,121
and
two dimensional matrix with
management
much
any given employer
within these job categories
Dr Leamer
91
to
by
broad and characterized
employees because each possesses
and rate of growth of compensation
Exhibits
is
explain
who
employer
difficulty
varies across
that individual
employment
uses to identify
and
the
I
the top ten
Case5 11cv 02509LHK
90
percent
age
of the variability
in a class
tenure gender location job
Leamer concludes
Document518
124
and employer
title
time can be explained by a
common
he bases this conclusion
the
a
that is explained
by
81
of
in each year
Based on his regression results
125
characteristics
that measures
statistic
The
statistic
the fraction
by
Dr
salaries within each firm at each point
in
set of employee
Rsquared
is
Page54
members compensation can be explained
almost the entire variation
that
Filed1007 13
2
in
on which
of the variance
Rsquared
in
the dependent
variable
power and
zero no explanatory
means
of 0.9 or so
variables
do
an
individuals
good job
of explaining
12 compares
92
perfect explanatory
that the regression
which
job title
Exhibit
a
one
the independent
is
power As used
has a good
equation
employer
variables and that lies between
an
fit and the independent
age
specific
Dr Leamer
by
tenure gender and location
compensation
the persons
Rsquareds
reported
in
Dr
Leamers
Figures
11 and 13 with
the
employeespecific
when
values
only the employerspecific job
age
factors of
same with
not demonstrated
claims that he has
as he
he would acknowledge
compensation
among employees
is
The
exhibit
employee
or without the
affect
are included
variables
tenure gender and location
regression is almost the
that even
title
but not the
shows
specific
an employees
largely explained
compensation
but only that variation
by employerspecific job
compensation
average pay across job categories
fact that job titles explain
mean
124
Regression analysis
many
the
126
identifying
a statistical tool to
different
individuals
the
measure
in
with different
The wide
of the Plaintiffs
because
variation
titles in
dependent
or
time periods
variables and crop yields
the
here age
by using data for
with different
dependent
to the
here a class
variables
variables
the
does not
addition
variable
control
on the dependent
and or different
However
in compensation
within job
independent
126
values for
the
amount
variables
129
variation
classes
in
different levels of average
impact on a
variables
titles
can be used to understand the relationship between
sunshine and fertilizer the independent
Leamer Report
the
one or more
characteristics
analysis
firm
wide
in compensation
impact of the independent
variables For example regression
proposed
of the
variation
annual compensation of changes
etc
of rainfall
125
is
in
a large fraction
that there is not substantial
members
tenure
wide variation
Leamer has
controlled for important employee specific factors
job titles have
is
Dr
variables
employees with different employerspecific
and there
fit of the
that the
in
pay across
Hence Dr
class rather
job
Leamers
categories
finding
is
a
consequence
of a high
than any homogeneity
51
R squared
of the broad definitions
to
a large extent
of the Plaintiffs
reflects the heterogeneity
Case5 11cv 02509LHK
variation
across
Document518
In fact the data used in
titles
Filed1007 13
2
Dr Leamers
Page55
regression analysis
of
show
81
exactly
the
opposite
A simple test
93
employees
individual
12 and 14
Figures
of the ability
shown
is
drawn from
schedule
compensation
The
2008 and 2009
in
which
is
that remains unexplained
Leamers
Figure
compensation
Defendant
example
actual
explain
in Exhibits
12
that
Leamers
there is substantial
in
2007
and predicted
between
regression
dispersion
127
Exhibit
14B shows
between
plus or
minus
of
compensation
rigid
15.4 and 1.7 percent
more
8.4 percent
predicted
their
shows that while an
dispersion
accounting
and actual
an individuals
in the unexplained
15 13
In the
same year
differences
based on
52
title
of his
based on the results in
compensation
Dr
and the
In each year and for each
portion
9
job
individuals
of compensation
the percentage difference
5 11 5 and
resulting
employee specific factors
actual
127
predicts
Plaintiffs
remains in the portion
for
the distribution
model
named
of
percent
or
for
Dr Leamers
or
Figure
For
between
more for half
Apple Google
the percentage difference
compensation was about plus or minus 25 percent
the distribution
for
there is considerable
compensation was about plus or minus 10 percent
Intuit Lucasfilm and Pixar respectively
and predicted
after
in the middle of his conduct period
Adobes employees and
actual
even
which shows
of the difference
Dr
the
Second
compensation
wide
his compensation
14A
Brandon Marshall
models
2006 than
respectively
factors regression analysis
compensation
is illustrated
in
were overcompensated
Plaintiffs
Dr Leamers
and actual
predicted
helps
to
from his
with a rigid compensation system
and employer
This
named
in the difference
inconsistent
between
difference
of
conclusions can be
2006 and 2007 respecively and 37.9 and
years and individuals
Dr Leamers common
from
in
Two
while Daniel Stover received
predicts
compensation
use regression estimates
of the Defendants
greater compensation
than predicted
than predicted
across
based on
compensation
I
to explain
would have been earned by each named
that
he was employed by one
Dr Leamer
by
asserted
compensation
94
regressions
13A and 13B
in Exhibits
17.4 percent
received
less compensation
variation
Leamers
this table First for the most part the
relative to their predicted
example
Dr
to predict the compensation
in each year that
Plaintiff
of
more
14
for
10
of
Intel
between
percent
of
Case5 11cv 02509LHK
Adobes employees and
plus or
Document518
minus
44 40 14 29 22
128
Intel Intuit Lucasfilm and Pixar respectively
imply that 10 percent
his regression estimates
Filed1007 13
2
and 30 percent
Based on
of Googles
scope to
pay
differentiate
129
Therefore
studies
structure
employees
of over 4,000 specific
to
Dr Leamers
model demonstrates
who
earned by employees
his regression which
employees even
across
contrary
his regression
by
predicted
in
2007
salaries
accounting
percent
for
of the
demonstrates that there was wide
job classifications
he
claims of a relatively rigid compensation
that there is substantial
year included
in
in compensation
variation
have the identical values of the characteristics
job titles in a typical
Apple Google
more than 40
or
the narrow
within
for
81
of
the 5th and 95th percentile
approximately 900 classmembers earned less than 40 percent
average compensation
Page56
including
Dr Leamers
being
regression
in
model
one
130
This evidence shows that the Defendants did not have the type of formulaic compensation
structure
would
that
challenged conduct
95
Even
unexplained
based on
a given job
Dr
Leamers
Dr Leamers
compensation
128
there is large variation
regression
and the compensation that
which was
The
title
132
In
2007
Exhibits
model
amount
the
in
15A
the
of Software
Developer
the actual
predicts for the top ten
Engineer 3 employees at
Dr Leamer
job based on the algorithm
simply because observations
the distribution
between
model
regression
of compensation
15B show
and
of the differences
Dr Leamers
ten percent
the top ranking Apple
fits Intel data better
regression
model
Figure 12 regression
Apple and Google jobs
Apple
would be class wide impact from
131
within
by
Plaintiffs claim that there
support
from Intels employees
used in his
such a
constitute
highperforming
large
129
about 60 percent
portion
This
of
course ignores
of the regression
the firms ability
and otherwise potentially
data
to differentiate
undercompensated
compensation
employees
across
into
new
employees
jobs
by moving
with higher average
compensation
130
According
over 100,000
60,000
131
As
employees
I
noted
compensation
across
132
The
shown
Dr
job
Leamers
employees
in
2005 and 2009
about 7,000 different
the
proposed
All Salaried
job titles and the proposed
Employee
Class
Technical Class includes
includes
over
about 2,400 job titles
above even
in
response
these
figures understate
to external
the flexibility that
the
Defendants
pressure because they ignore Defendants
had to differentiate
ability to
move
individuals
titles
distributions
in
data between
in
Appendices
of the of differences
7A
through
in actual
and
predicted
7E
53
compensation
for the other
five
defendants are
Case5 11cv 02509LHK
Dr
by
predicted
again
necessary
to
model
job
by
predicted
Together
assumed
and
because of other factors
were not harmed by
his
which any
Model
VD
minus
percent
that individualized
was under
individual
analysis
would be
or overcompensated
from the
of undercompensation
the conclusions
compensation
that
Does not Demonstrate
discuss
Dr Leamer
in
his Hypothesized
Price Discovery
detail the regression analysis
which
offers as evidence that the challenged
and damages allegedly
Dr Leamer
suffered
draws from it are inconsistent
that allows
structure
him
to
infer that the
by
refer to as
I
conduct affected
price discovery to affect
all
members
his conduct regression estimates I simulate
with his claim that there
assumed loss
identical individuals
based on the empirical
year for each Defendant
133
I
person so
in
that
subsequent
into
account
2004
the
I
result
only difference
years
compensation
and thereby
the persistence
for individuals
otherwise similar people
compensation can
diverge
shows how
Assume
that
in their
predict
effects
I
identical
in all characteristics
54
Using
compensation in each
two
compensation can diverge
who
individuals
are
comparable
in their residuals
compensation
each subsequent
based on
2004 The
over time
common way
is the difference
in their
from the prior two years
were identical in
equity causes
unexplained
performed the same experiment 50,000
who
internal
over
in all
unexplained portion of their compensation for each
2005 compensation
the difference
rigid
over time of otherwise
dramatically
there are
or the
a
in compensation
distribution of
randomly draw residuals
randomly drawn residuals
of
The
perform the following experiment
characteristics
in
133
is
of informationand
of the proposed class in a
the change
the
the Class That analysis and
reduced price discovery combined with Defendants commitment to
localized
who
challenged agreements
aggregate compensation of members of the proposed class and that he uses to estimate
amount
of the
Cannot Explain Compensation Changes
it
conduct regression that
employees at Google
III
paying damages to members of the proposed class
my report I
of
Engineer
because of the challenged conduct rather than
structure
to avoid
Process Because
In Section
to
81
of
2007
and indeed could have benefited
4 Dr Leamers
97
in
on compensation means
wage
rigid
model
Page57
or less than the compensation
of the employees earned plus or
Dr Leamers
the evidence
more
Similarly for Software
percent
determine the extent
relative to the
Filed1007 13
2
earned approximately 42 percent
Leamers
the top ranking
compensation
96
17
15 through
Figures
Document518
that
Dr
in
Leamers
times to obtain
resulting
distributions
Dr Leamer
model and
I
do
the
same
year taking
the
a distribution
new
of
resulting
show how compensation
claims explain
an individuals
Case5 11cv 02509LHK
Document518
due
time for otherwise comparable individuals
Dr
by
that remains unexplained
Leamers
Filed1007 13
2
to the portion
conduct
of
an
Page58
individuals
81
of
compensation
which
regression each year the effect of
cumulates over time
98
Compare for example two
characteristic
controlled
tenure and location
Dr Leamers
for
Dr
Leamer
model
implies
e 2004 and
that these
By 2009
by 24 percent
demonstrate that otherwise
Thus Dr
compensation
compensation
schedule
134
Such
conduct
own model
His
somewhat rigid wage
structure
results
regression
estimates
results provide
the proposed class could
no
the compensation
shown
model
different
a
his claim of
contradicts
imply that he has no basis
to
salaries
17
end up with tremendously
rigid
conclude that
class wide effects through his claimed
to
in both the levels
As such Dr
even within job categories
for Plaintiffs claims that the
support
By 2006
two
of the
16 and
in Exhibits
Empirical evidence shows wide variation
and rates of growth of employee compensation
Leamers
e 2003 compensation
between
changes in compensation would translate
individual
age gender company
i
prior
identical employees can rapidly
Leamers
are identical in every
two employees would on average have
the difference
would be around 37 percent
individuals
2004
of
in his conduct regression
i
as well as current
conduct
that differed
by
who as
employees
amount
of
harm
members
to
of
be determined on a class wide basis
5 Dr Leamers
Attribute Compensation
Constant
Ranking
Analysis
is
Misleading
99
compensation
I
excluded
and by
Lucasfilm
of the pairs
at
least
86
18B show
and Pixar
If
differences
would
the
likely
evidence
masks
years
in compensation
Exhibit
exhibits
situated
the
17 shows
The
55
variation
two firms
90th percentile
the
I
have converted
the analysis to include
Exhibit
16 shows
figures indicate
by at least 56 percent
the entire distribution
were promoted
be even larger
relatively stable
and expanded
that differed
numbers assume
employees
of
substantial
because of missing data for those
after five
identically
titles
would have compensation
of employees
percent
two
Dr Leamers
both by firm and overall
differences
from 2005 through 2010 In both
each year
that
into annual changes
and the 90th percentile
percent
and
within and across job
trends
15 17
his Figures
134
18A
Exhibits
two employees
at different
of differences
the
that
after
in
mean
ten
two
years
each year
have the same job
rates
the
title
then the compensation
in
Case5 11cv 02509LHK
25 job
top
titles
Dr Leamers
fashion
groups
linked
across
Dr Leamer
structure
15 17
The
at
his deposition
136
not rigid
18A
Exhibits
that such evidence
Moreover
and
link of the
form
required
relevant
than compensation
changes
in
compensation
by
Dr
in
18A
within Defendants
constant
and
the overall
upward
theory Changes
Indeed
was
substantial
compensation
compensation
in
Dr Leamers
relatively stable then one
variation
across
titles
in a year and
mask
that Because
does not
are
would be
and
from year
claims that
If
Dr
this
Some
across
101
Dr Leamers
logic
level that substantially
to
his Figure
135
I
136
show
exceeds
7 ripples through
charts
Q Could
implies
for the
Would Move
that
year The
and
scale of his
variation
had
Defendants
of
that of current employees
the rigid compensation
a nonrigid wage structure
8A
forces
Policies
and Even
of
the Structure
a small number
top 10 jobs in Appendices
that
job titles
Formulas for Annual Compensation Adjustments that a Limited Number
Additional Cold Calls
Leamer
a given year
large positive
to
more
would expect
similar in
Defendants
are
which would be driven by market
trend in compensation
One Cannot Conclude
in
compensation changes with
attribute
a wage
stable ordering
changes in compensation for others
titles
independent of any internal equity concerns
6
are in any
true would not be sufficient to establish a
if
for different job titles also
changes in compensation across
negative
figures
in
even
Leamers
for one group drive
18B show
and
across
alone causally related across groups
let
levels because they more closely proxy
average compensation
But Exhibits
a roughly
135
parallel
consistent with
limited claim that changes
which
were correct that compensation across job
changes
would be
the maintenance of
18B examine the
similar for the different job categories
causal
in a
of compensation
fact that the ordering
even mean that changes in compensation are correlated
100
and Google
job titles for Apple
imply that compensation increased
81
of
through concerns about internal equity or other forces
job categories
was
10
Page59
stable over time does not imply that changes in compensation
admitted
that
Filed1007 13
2
rather than just the top
highly misleading
is
job titles is relatively
way
Defendant
for each
claim that his Figures
across
Document518
and
as you’ve defined
Depo 283 23 25
56
movers
eg 20
hired at a compensation
or
30
percent
structures of Defendants
higher according
to
cause an
8B
it
lead to parallel lines
A Yes
it
could
Leamer
Case5 11cv 02509LHK
or similarincrease
equivalent
unreasonable to expect
earned by a
amount
cold
compensation
not
who
the person
compensation
new
102
hire and not
One way
in
if
compensation
would
it
is
Because
is
by
75,000 and a senior software
engineer receives
110,000
The
with
an outside
firm can respond
perfectly
rigid
compensation
by promoting
impact
E
let
causing a
of all
employees
employee receives
to a position
if
an
outside
higher
that provides
on compensation
within
wage
the firms
engineer
125,000 and
a job title
structure
position
a junior
Dr
to
of
senior software
ripple effect on compensation
Leamers
Rsquared would be 1
Dr Leamers
The
third
Model
Econometric
it
is
between
of
no
compensation
the challenged
of
would show a
regression
yet there is
Undercompensation
Flawed Both Conceptually and
issue addressed
by
Dr
Leamer
is
whether
be used to demonstrate that the challenged agreements
of
any
if
ripple effect
of junior
and
conduct had any
alone a class wide impact
Impact Because
103
a valued
no informationabout whether
provides
higher
directly gain information
the junior software engineer
and the rigid structure reflected in the fixed relationship
senior software engineers
job but
job titles a junior software
Using data such as this
his
the
25
as the result of a cold call with compensation
offer
structure
who
engineer position that pays
125,000 in compensation without
other junior software engineers
substantially
employees only
which occurs even
two
by a
in compensation
increase
regression focuses
firm has only
required
it
cost the firmnot only
current
with a promotion
countering
not be able to identify this type of effect
that pays
engineer
substantial
Dr Leamers
rigid To explain assume a
software
a
required
or even
than
employer would be willing to offer a
only employees
to
which an employer can respond when
offer of higher compensation
total
it
An
hire if
performed
compared with
increase
make the
percent
previously
the proposed class
in
impact were limited to similar employees or
from the
by 25
greater compensation
employee would
called
economically
is
it
81
of
whether to offer employment to a
employee would
cold
Page60
employees But
percent
of all salaried employees
for all employees
a substantial
of current
who demands 25
call
In effect hiring the
Filed1007 13
2
A firm considering
current
more than was earned by
percent
mover
a
attribute
compensation
increasing
lesser
constant
in compensation
to occur
this
identified through
candidate
Document518
members
of the proposed class
He
first
presents
57
in its
standard
Fails to
Show Common
Implementation
econometric
analysis can
generally suppressed the compensation
a simple analysis of the change
in total
Case5 11cv 02509LHK
compensation
an
for Defendants
claims that this
he
analysis
later
Filed1007 13
2
described
Page61
warm up
as a
evidence that there was undercompensation
suggestive
is
Document518
81
of
He
exercise
during the class
period
Dr Leamer
then performs a
undercompensation
from the alleged
104
conduct regression
He
agreements
to
attempt
indicated
at
the aggregate
to quantify
his deposition
that his
conduct regression also had a central role in making up for the lack of direct supporting
evidence for other parts of his theory
to
and
evaluate
agreements
quantify
amount
105
Thus
Dr Leamers
forms of econometric
conduct
process
if
the
amount
of computing
his
conceptual framework
other independent empirical
logic
links the
it
of information reduction
evidence
informationand thereby
hampered
structure
106
to
each
However
or almost
and
Dr Leamers
and
conduct
137
Leamer Depo
at
413 21 414 7
cold calling declined
identified
138
139
in his
price
period
framework
at
Non
support
for
which he has no
the lynchpin in his chain of
on compensation
somewhat rigid compensation
the proposed class
common
not
that
in multiple
common impact
to all Defendants
he does not have evidence
but stating that the regression
is actually
there
135
Leamer Report
Leamer Depo
discovery
is
acknowledging
during the conduct
is
for
standard
of compensation
but also as empirical
price discovery
regression is also flawed
effect he estimates
amount
to a measurable impact
implementation masks the fact that he has no evidence of
undercompensation
139
through his claimed
each member of
demonstrate that
agreements reduced cold calling and
price discovery
both on average for the proposed class
substantial
Employee Class caused by the
His conduct regression
that the challenged
possibility
to
the aggregate
aggregate damages
and to measure
was
137
suppression to the AllEmployee Class and Technical
Compete Agreements
though he lacked evidence
of information lost
regression is not only offered
analysis are capable
138
said that even
because his conduct regression would show
during the conduct period
to affect the price discovery
he
information that was lost because of the challenged
of
he did not need this evidence
undercompensation
enough
the
In particular
24 825 36
58
will
ways
First his
but rather that the
When
on whether
disaggregated
the
decide whether
amount
of
the effect
Case5 11cv 02509LHK
by Defendant
overcompensated
independence
fails
his
model
by each
implies
that his
are demonstrably
of
with his claim of a
simple sensitivity tests
107
Before presenting
his Figure
compensation
calculated
illustrating
how
by contrasting
the before
for the
US economy
he can
illustrate
employs
in
Defendants
the
relevant
140
141
142
143
144
Motion
and
after
Demonstrates
econometric
NonCompete
the
after
warm
up
works
143
2001 and 2011
if
comparing
in relevant periods
before
No Common
analysis
his review
144
at
379 910
Leamer Depo
at
380 21
140
59
that
Impact
in
on employee
when
analysis
of
provides
the Agreements
141
142
a
At
growth cycle
18 he
his
way
of
periods
concludes that
damages which he
for measuring
and after
showing
of
Dr Leamer
Agreements
the regression
company
no
Members
the average change
136
fragile and
is
140
in compensation
at
also
the
during the conduct period with
He
selects
periods because he concludes that these years
Leamer Depo
Leamer Report
to
of
regression provides
summarized in his Figure
they were a single
lack
NonCompete Agreements
Based on
p3
Leamer Report
conduct
model
compensation during the periods
before and
model by
as
The
structure
regression
or virtually all Class
before and after methodology
pooled together
non conduct
Leamers
standard
analysis as a
between
his regression
what was happening
all
of the effect of the
this
determined independently which
class wide methods and evidence are capable
his proposed
were in effect with compensation
deposition he referred to
Leamers
Dr
For these reasons
statistical
imprecise and are not reliable
Dr
Warm up Exercise
an estimate
19
are highly
impact Third
suppression of compensation affected
1 Dr Leamers
is
somewhat rigid compensation
estimates
Dr Leamers
Second
false In particular his analysis assumes that
a Defendants employees
model
for Plaintiffs claims that
support
81
of
employees during the conduct period
or proof of class wide
estimates
Page62
conduct regression model would show that some Defendants
received
inconsistent
Filed1007 13
own
assumptions regarding
is
2
their
his
compensation
Document518
2004 and 2011
reveal
the kind of
as the
Case5 11cv 02509LHK
compensation
Document518
that occur in expansion periods
increases
Based on this comparison he calculates
for all Defendants
108
as a
whole
Dr Leamer
While
19
19
I
model
warm
up
by an expanded
version
Dr
Leamers
defendant by defendant
implies
that the
and
theory
and
as
Dr
implies
Defendant that allegedly
109
However
substantial
as
at
same
he
the judgment
type of effect
in Exhibit
underlying differences
according to his methodology
undercompensated
overcompensated
by more
Leamers
individual
145
146
the
147
148
than
27
their
percent
that
147
19 Dr Leamers
Defendants
was
better to
because the
firms
namely undercompensation
at
Plaintiffs claim of
each individual
Figure 19 methodology
undercompensation
in estimated
Intel
it
his
the conspiracy
at each
and Google are the only two companies
masks
Defendant
Indeed
that
employees during the conduct period while the other five Defendants
employees Apple employees
and Pixars
he
estimates
for
example were
employees by more than 70 percent
Figure 19 shows an overall average
result
an average
that blends together
negative
overcompensated
This suggests
opposite
effect only because
two
effects
that
Dr
at
out of seven
140
Leamer Report
Dr
their
in
rather than
examined
model
more efficient
In
did in his Figure
individually
a minimum suppression of compensation
participated
shown
145
2005 2009
19 analysis
during the conduct period should be found for each Defendant individually
common impact
for
Leamer
his claim that
made
then
pool across firms in order to create a more coherent
148
81
problem with his
of his Figure
that I perform the analysis for each Defendant
are sufficiently similar
of
2005 2007
percentages
a fundamental
and after comparisons
before
them together
simply pooling
regression
is
similar to
underpayment
estimated
Figure 19 as a
uses
perform the same
The only change
that were
Page63
146
regression analysis can be illustrated
Exhibit
Filed1007 13
2
Leamer assumes
that there
would have been no change
weak economy
Leamer Depo
at
364 24 365 16
Leamer Depo
at
364 20 365 7
60
in
compensation
in
2008 and 2009 because
of
Case5 11cv 02509LHK
Defendants show a negative
110
for
impact and one
of the employees in the
percent
warm
This
Dr Leamer
whether there
up
him pause about
of the
when
Class
of
81
Intel accounts for about 60
by company should have
large magnitudes
of his methodology
Page64
149
whether his regression
to consider
rather than reflecting the impact
Filed1007 13
two Defendants
disaggregated
common impact The
the ability
2
AllSalaried Employee
exercise
and caused him
is
Document518
is
raised a red flag
capable of determining
of the effects also should have
conduct
to identify the effects of the challenged
of other factors that differ between
the conduct
and
given
non conduct
periods
2 Dr Leamers Common
Demonstrated
111
After
using Figure 19 as an illustration
analysis which he says
is
defendants as well as for
explain
model
Figures
112
total
Impact across Defendants
a
better approach
employees
150
Dr Leamer
because
He
he calculates annual
his Figure
conduct had
150
151
at
See
Dr
model
among
that attempts
20 and
his Figures
percentages
because
it
23
151
to
Using this
by Defendant and year
his
is
Leamers
not
common
impact
Defendants
Figure
Leamer Report
at
Dr Leamers
regression
lower compensation
Once
disaggregated
common
at all
Defendants
by Defendant
and for
all
Dr Leamers
impact and implies instead
that the
Defendants
3
141
Dr
across
underlying
assumes rather than establishes or demonstrates that
regression analysis completely fails to demonstrate
149
a regression
19 analysis the approach
common
or virtually all employees of the
impact
conduct regression
his
allows for differences
employees
undercompensation
analysis is fundamentally flawed
alleged
presents
24
As suggested by
the challenged
it
constructs
annual compensation of individual
22 and
Assumed Not
is
in his Regression
Leamer
said
his deposition
that
total
compensation
121 821
61
is
the relevant
variable
for
analysis Leamer
Depo
Case5 11cv 02509LHK
a
Summaryof Dr
Dr Leamers
113
indicator
variable
and
variable
variable
Leamers
model uses
regression
year as the dependent
An
Document518
includes
when
for
This variable
is
the interaction153 between
linger over
152
He
includes
employees within
Dr Leamers
Because
variable
to
in those
In
effect
purport
two years
154
of the
two
The hiring
for
variables
rate variable
Leamer Report
allegedly
variables
on
participated
in
that represent
and age squared and
age
reflect
how
a time trend and employer
normal
variation
in
over time and across Defendants
on
these variables
the effects
to measure
the extent
Dr Leamer
compensation
not full
across
155
with age and hiring
along with average employee
to which
then combines
indicator
caused by the challenged agreements
initial annual undercompensation
age and hiring
by company
undercompensation
in
one
his initial undercompensation
conduct years he assigns a
on
is
interaction
the
of
independent
measured
two explanatory
value
Since his
year
estimates
of 0.5 and 0.25 to the conduct
variables
measures
the multiplicative
or
joint
variable
as the log
of the ratio of
firm in the previous year
155
conduct
the
respectively
model an
a regression
also includes
which he claims
effects
undercompensation
calculate
2005 and 2009 were
153
effect
variable that is turned
that Defendant
and employee
characteristics
these to control
estimates
company
variables
industry
the industry
remains in subsequent years
152
zero one
in
Defendant 154
the immediate
uses the coefficient
persistence
variables
claims that his conduct variable alone and interacted
identify
rate at a given
or
in each
time
variables
He
Leamer
the conduct variable
characteristics
rate together
Dr
81
of
of each employee
agreements were
dummy
essentially a
Page65
Compensation
the following independent
or lagged compensation
Employee
114
of
real annual compensation
the challenged
any of the challenged agreements
Persistence
Model
during the period when
for a particular Defendant
the hiring rate at a given
Filed1007 13
2
142
62
new
hires to the
number of employees
at the
Case5 11cv 02509LHK
and the persistence
22 and
24
115
At
measures
Document518
Filed1007 13
2
his deposition
Dr Leamer
the average impact
for
acknowledged
of the conduct across
Defendant
that the conduct variable
Thus
Defendant
Defendants
all
measurement
specific
he claims that
but claimed
of the impact
However
conduct will differ at the two Defendants
his
model
same
even
if
assumed
same
for
those individuals
for a Defendants
provision
still
the impact to be the
forces
the hire rate is the
makes no
firms is
of the effect across
unique
Defendant
another
at
at
individuals
employed by
characteristics
each
of
any
to note that
by
his
same age
of the
his
are younger or
rather than demonstrated
two
are
that the
the aggregate impact
important
is
it
model
because he interacted
that employees at one Defendant
the extent
to
employer has a slower hiring rate than
commonality
since
his Figures
in his
conduct variable with variables measuring the age of employees and hiring rate
alleged
81
156
regression allowed
their
of
by company
annual undercompensation
effects to calculate total
Page66
model
as long as
He
different Defendants
to affect the potential
impact of the
challenged conduct even though his theory says that he should
b
Once Disaggregated Dr Leamers
Undercompensation
116
Given the nature
challenged conduct was
of Plaintiffs allegations
common
across
Defendants
used the regression framework offered by
have
to estimate
done
of individual
criteria
156
157
is critical
observations
for belonging
Figure 22 presents
technical
and
22
employee
The
Dr
as
for
to the
Leamer acknowledged
a
particular Defendant
class some
his results for the
class I will focus
central
conclusions
Dr Leamer
Leamer Depo
at
on
a class
to address
this
allsalaried
my
are the
analysis
same
employee
on
365 816
63
wide basis Thus
question
which
annual data for
class
all
are estimated
Dr
One way
employee
putative
class
I
the large
employees
who
is
have
to
do so
Leamer claims
to
number
fit
the
only because he can
while Figure 24 presents
his results for that
for the technical
whether there
in his deposition despite
of his coefficients
results differ
157
understanding
to
separately for each Defendant
the regression
however
whether the impact of the
the question
class wide impact and whether the impact can be measured
would be
Does Not Show
Regression
for All Defendants
class
though
his results for the
his exhibits
some
Figures 20
of the individual
Case5 11cv 02509LHK
combine
158
Defendants
data across
Document518
Therefore
in order to test whether
challenged conduct was similar across Defendants
conduct variables for each Defendant
separate
estimate
117
conduct effect
of the aggregate
Defendant
20
estimates
I
show
22 and 24
Figures
I
a combination
reflects
In stark contrast to
for all Defendants
common
undercompensation but
throughout
The
during the period
agreements according
magnitude
also
to
and
Plaintiffs
Dr
overcompensation
show
Leamers
Dr
and the sign of the estimated
2005 and 2009
conduct
for
Apple
generally
are
positive
two
for one or
hub
much
years
of the network
of
Dr
smaller than
the conduct regression
undercompensation
reported
In fact for
for most years while
the
from
results
the
Lucasfilm and
estimates
overcompensation
Leamer
impact
undercompensation
overcompensation
estimates
from Dr Leamers
Leamers
Class two Defendants
once disaggregated by Defendant
substantially
Dr
impact from the alleged
instead
undercompensation
Thus
results
not only differ
Dr
specific
of these disaggregated
in every year between
Google shows
the period
Adobe and Intel
other companies
Leamers
that includes
Defendant
of the regression analysis
nature
both the AllSalaried Employee Class and Technical
two
separate
81
of the
of his regression
and also includes
159
results
disaggregated analysis does not suggest
percentages
the effect
of
from the disaggregated model and compare them with
results
percentages
negative
Pixar show no
use a version
Page67
estimates
specific
In Exhibit
Leamers
By
with age and hiring rate
interactions
Filed1007 13
2
but also vary greatly across
in both
results
companies
and over
160
time
158
Dr
sector
the
Leamers
model becomes
period
159
2003 2011
regression
includes
overspecified
2006
160
Results
unavailable
As a
multiplying
outputs
a Pixar specific
which were
result
when
many
its
to
variables
Dr
is
for
that
only vary by year
example
the
new
for
hire ratio
using annual data from a single
effect
Leamer
I
See
was
example change
variable
company
As
for the
in
IT
a result
nine year
estimated
from the disaggregated model are provided in Appendices
conduct
2006 revenue
the reported
have included in
Pixar
reported
revenues
for
my
regression
2005
Pixars
9A and 9B
revenue
data
In order
after
2005
2011 xlsx Pixar was acquired by Disney
only nine months
I
annualized the 2006 number by
number by 12 9
of disaggregating
by Defendant
regression In Appendices
10A to 10C
an
specification
aggregated
estimated
over which the regression
Detailed regression
to estimate
in
conduct
employment in San Jose or by company by year
regression
I
can also be illustrated using a simplified
provide regression
that includes
details
a single
64
and
conduct
version
undercompensation
of
Dr Leamers
estimates
variable and excludes interactions
from
Case5 11cv 02509LHK
Dr Leamer
118
Document518
Filed1007 13
2
should not be surprised by the finding
Page68
uncommon impact
of
Defendants if as he believes the challenged agreements had an actual impact
spurious
or unrelated
differences
theory
These include
firms that
differed
119
At
so
He
efficiency
Dr Leamer
e
i
to
offered
no
conduct
Leamers
the potential
shown
the
in
hiring
with the single
11C
Appendix
four Defendants
undercompensation
162
163
rate
the level
of the
varies across
a firmand that pressure
information was reduced
the
and
quantify
would have been reduced
Comparison
11A
model except
conduct
Pixar Lucasfilm
to
still
and
to
11C
I
Adobe
at
at
the impact
However
gain from
of the
of the ability
to analyze
Figures 22 and 24 shows that the
the results differ
somewhat from Dr
undercompensation
specification
will
details
and
for all Defendants
be informative
as to the
undercompensation
and even the sign of the estimated
increase compensation
and
The
differ substantially
364 8365 1
Leamer Depo
of the efficiency
had no undercompensation
257 814
Leamer Depo
163
results across
in
the expense
provide regression
percentages
at
and decided that the
now disaggregate
the model by interacting
company
now there are seven Defendant specific conduct variables
vary greatly in magnitude
Leamer Depo
shows
using the simplified
variation in the size
was
While
he had done
that I
so
variable
there is large
differences
Dr Leamers
with
Employee Class
version
is at
if
the
was reasonable
wrong conclusion about
efficiency
by
results disaggregated
of the data to identify
he weighed the importance
that results
In Appendices
impact of the challenged agreements
161
how
the gain in
This suggests
from the simplified
seven Defendants
other
to which
data for all Defendants
Technical Class the simplified
impact of disaggregation
indicators
at
structure
other than acknowledging
for reaching
When
and age and
throughout the period
estimates
the ability
very similar for the All Salaried
for the
wage
said that he did not provide
combining
specifics
challenged agreements
results are
for internal equity
examined the models defendant by defendant
have
and no explanation
pooling against
between
would depend on
demand
and that the extent
impact of the challenged agreements
gain from pooling
Defendants
Leamer acknowledged
firms
claimed
Dr Leamer
and was not just
161
Defendant because efficiency
162
Dr
of the impact
pressure will affect the general
his deposition
hypothesized
his deposition
the fact that the magnitude
the situation each firmfaces
across
At
for internal equity and the fact that
outside
on
depends
agreements
across
Defendants that would translate into different conduct effects even under his
across
demand
firms
to those
81
of
365 3366 2
65
but rather
estimated
from
negative
Dr Leamers
effects
As
Three
the estimated
impacts at the
of
Case5 11cv 02509LHK
the question
at
companies
namely whether
issue
economic
consistent with the
were more
estimates
of
Thus
further
below show
substantial
3
The
Statistical
significance
variation
of his estimated
However
data
an individuals
122
the
amount
case
the
Even
reflect
if
compensation
is
that the efficiency
to conclude
methods which
that his
I
critique
impact with some
164
Analysis
Improper
is
t values
errors and
to test the
thus to test his hypothesis
Dr Leamer
these values
is
more
of those of other
that the
assumes that
employees
then the
whether the regression
is
insufficient
is
especially
a major
is
a
under his
identifying
and
inference
the larger
the data although
observations
misleading measure
in
this
of the
an underlying relationship
in the
would differ by job type
unsubstantiated
if
of individual
highly
theory of
reliable estimates
However
variation
own
error in statistical
statistically
number
to capture
because that effect
and obvious feature of his
a critical
are estimated
years for the Defendants
informative only under the
not consistent
ignores
This
provides
the coefficients
impact
system that imposes
an assumption that
and
not independent
determined
model
by Defendant
were any
are
Dr Leamers
In calculating
common impact
statistically
there
Defendants
a
employee
the disaggregation
agreements
across
of
to evaluate
ability
164
number
is
a regression
of data with which
voluminous largely
conduct
reports standard
Dr Leamer
are correlated
compensation
All else equal
Leamer
statistical
of the challenged
coefficients
only because
significant
that his observations
how
he did
few
at a
impact of the challenged agreements on compensation are highly
the estimated
statistically
and
employee is independent
of each individual
Dr
81
or indeed any impact
Defendants in the estimated
Underlying
22 Dr Leamer
as
of
were not meaningful
his estimates
across
as the result
Page69
few employees
to claim
senseless
regression specification
conduct reduced employee compensation
the compensation
is
call a
Doing so simply allowed
Framework
20 and
In his Figures
common impact
to cold
failure
it
precise even though
overcompensated
employees
how
Filed1007 13
2
or not there is
harm
appropriate
is
Dr Leamers own
120
statistical
theory
translates into class wide
gain from pooling data
121
Document518
wrong
impact of the challenged
Analyses disaggregated
assumption of a rigid
formulaic adjustments across all types of jobs locations etc within a firm
with the evidence
66
Case5 11cv 02509LHK
between
and the labor empirical
Dr Leamers
123
sample contains
and
167
labor economics
employeryear so
a
his statistical
in
literature
Page70
the econometrics
81
literature
Dr Leamers
problem
of this
of
165,166
particular
in
severe version
observations but fewer than 60 unique combinations
thus effectively fewer than
This means that
Filed1007 13
2
widely recognized
over 500,000 individual
conduct variable
Dr
60
from which to estimate
observations
Leamer has almost 10,000
analysis greatly overstates
his
per group
observations
of
per
sample size and the
the effective
of his estimates
resulting precision
Dr Leamer
treats
each Defendants
employees
as
an employer But the supposed
rigid compensation
he
or she provides
completely
determined at
is
and thus lack of independence
structure
economic framework on which he
feature of the
if
about the underlying structure by which compensation
independent information
critical
is
regression suffers from
employer and year
124
This problem
the variables
generally
Document518
for his conclusion
relies
is
a
that the
challenged agreements to reduce cold calling reduced price discovery which rippled through the
members
compensation
of all
compensation
system He
influence
their
This problem
for all Pixar
is
well known
and James MacKinnon
within a given dataset
ignore
. is
it
and use
when
166
Dr
of ignoring
thought that the within group
OLS standard
errors
OLS
are
many
H Econometric
D and Jrn
Angrist Joshua
Leamers
because he lacks
University
conduct
revenue
firm are affected
movie
In
their
correlation
per
Press
regression
statistical
between
it
that
in large and
econometrics textbook
. is small
that
commonfactors
demand
in the
test
for
PCs and
Russell
across
the
Davidson
observations
may be tempting
This can be a serious
to
mistake unless
can be drastic underestimates even with small values of .
sample
group Davidson
size
is
per group
much
Pischke
is
smaller
Russell and James
large
The
correlation
than the actual
G MacKinnon
sample
of
size
Econometric
p 305
Analysis 6th Edition Chapter 9.3.3
Steffen
by
rigid
Pixar can result
at
the correlation
severe when the number of observations
observations
Defendants
performing his
covariance matrix
and Methods Oxford University Press Inc 2004
Jersey Princeton
167
the
particularly
is
Greene William
2008
it is
a
econometric literature
the potential error
terms within groups means that the effective
there
Theory
If
at
of the
or a short term reduction
estimation with the usual
zero since
actually
The problem
the error
OLS
employees
in the
describe
employees
eg a highly successful
compensation
unusual bonuses
when
failed to take into account
from the challenged agreements
aside
165
of the proposed class because
Mostly Harmless
New
Jersey Pearson
Econometrics
Prentice
Chapter 8.2
Hall
New
2009
includes
data for Lucasfilm
data for seven companies
over nine years
and Pixar for some years his regression
employer years
67
2003 2011 However
includes
only 55
Case5 11cv 02509LHK
Document518
microprocessors that power them can cause a decline
Intel
to
impose
a
wage
Dr Leamers
125
compensation
independence
move
together
statistical
inferences
informing his estimates
estimates
must take
A generally
factors
this
Put differently
groups
estimated
the price per ounce
information
the dairy
and
Dr
errors
or present
Leamer
the underlying nature
a regression
to
explain
package
pints as
sizes at
a
in a single
within a
for in
estimates
are
that
from the regression
of his data
well and
is
a
the price of all package
spike
to
price distance
commonly
is
at
they provided
the cost of gasoline
The
this
or
but he did not even
tstatistics
used for
by
state
It
is
were
as
if
Dr
using data on
grocery stores in each
state
completely independent
recognize that a store that
rises
sells
high priced
at that store
gallons
say because
it
needed to deliver the milk to the
power
from a dairy store quality etc
68
year
to affect compensation
and gallons sold
sizes will increase
common
certain
were not meaningful
the price of gallons
if
in
chance
if
used to
not only failed to implement
the price of milk per ounce
store as
would have
by
errors and resulting
standard
to result from
unlikely
and there
the impact of gasoline
receives
the information
inference
that are affected
impacts of variables hypothesized
A proper analysis
likely sells high priced
store then
company
of pints quarts half gallons
and treating the various
from
of observations
a particular
to address
the estimated
significant
regression
on employee compensation
statistical
who
be accounted
that must
Dr Leamers
affect
employees
to take into account the fact that observations
in his report that his reported
testing whether
is far
method
contains
any other methodology
although
observations
clustering the standard
referred to as
Leamer had
would
calling
such as
into account
such as those affecting
statistically
cold
way
in a
of the
Dr Leamer a shock
of individual
much more limited and any
is
accepted
a regression
acknowledge
and
through
obtained
are related or correlated
based on over 500,000 individual
estimate
to Plaintiffs
members
of all
generally and the impact would not be limited to only the employee
Defendant and within a year
making
with his claims of a rigid
inconsistent
call Statistically this means that compensation
the cold
126
According
81
of
2009
is
in informationabout compensation
compensation
is
assumption
Page71
revenue and profitability and lead
in Intels
and with Plaintiffs claim that compensation
structure
proposed class would
an increase
such as occurred in
freeze
Filed1007 13
2
of the regression
is
not enhanced
to identify
by
Case5 11cv 02509LHK
including
because
individual
127
Dr Leamer
cases
such as this
standard
that
it
errors
recognizes that
168
was
When
standard
for
shows
that
factors unrelated
compensation
169
170
of
171
at
Angrist Joshua
University
Standard
counting
about whether his
your wealth
New
in
describe
to
assumption in
to cluster his
failure
Dr
small change
having
commonly
used
170
in
lots
Leamer admitted
of individuals
but
In Exhibits
21A
which are used
Dr Leamers
determine
to
estimates
no meaningful
estimates
show
coefficient
once properly
line
in the table is
22A and 22B
statistical
I
significance
22 and
in his Figures
estimates
4
24.171
This
for any employer
computed
or
standard
are completely consistent with there
conduct and his estimates
Thus
I
variable
errors In Exhibits
undercompensation
Leamers
21B
and
Figure 20 and 23 regressions except that the
undercompensation
Dr
studies of labor markets and widely
on employer year The conduct
p values
conduct
members
of
Leamer Depo
Chapter 9.3.3
his deposition
Dr Leamers
from
imply that
to that
Leamer Depo 374
Princeton
at
significant at conventional levels under the properly
regression provides
81
sizes at a particular store
improper to rely on the independence
such as this
being no true effect of the desired
168
is
errors is
clustered
none
p values
of
firm
at a
Leamers
is statistically
errors The
it
significant under these proper standard
Dr
Page72
169
show t statistics and
calculated
exhibit
details
now
errors are
not statistically
further
to
as necessary in analyses
and other
prices of four different package
use of that language
Clustering standard
estimates
year
asked
equivalent
only having one experiment
accepted
Filed1007 13
2
same underlying information
seems like appropriate
128
on
observations
the
all reflect
Document518
resulting entirely from
analyzed
Dr
Leamers
random
conduct
evidence that the challenged agreements reduced
of the proposed class
618
375 19 376 5
D
and Jrn
Steffen
Pischke
Mostly Harmless
Jersey
errors for the
annual
undercompensation
estimates
69
Chapter 8.2
Econometrics
pp 308 315 and Greene William
Pearson Prentice Hall 2008 p 188
Press 2009
H Econometric
are calculated
New
Jersey
Analysis 6th Edition
using
a bootstrap
method
Case5 11cv 02509LHK
4 Dr Leamer Does
Whether
129
Dr Leamer
several
it
he wrote
analyses
sensitivity
reporting
130
and
with
Consistent
Q Would
we
would have
to
is
see in the
I’ve
done
document
sturdy but
there’s
models One
As
to
the
at
or
a consequence
perform their
own
complete and more honest
172
as follows
someone who
the conduct regression
on
judgment
is
evaluating
its reliability
until
saw
that evaluator
analysis
would help
to
determine reliability
your regression analysis
know
record of econometrics
something
practiced
responded at his deposition
Dr Leamer that
sensitive
First
much more
is
purposes
researchers
think that the sensitivity
I
81
of
related to the regression
analyses
A would
Q And how
A
you need
it
of statistical
for individual
inferences
to reserve
as
art
for reporting
view Dr Leamer
this
econometric
perhaps thousands
ought to be demanding
you agree
sensitivity
the
that
are selected
efficient
of claimed
of the fragility
analysis
finds pleasing
much more
is
many
fitting
that the researcher
Page73
Robust or Fragile
many years ago
involves
Filed1007 13
2
not Report any Sensitivity Tests from which to Evaluate
his Results are
wrote
computer terminal
Document518
did not carry out a complete sensitivity analysis
I
how
that discusses
But
I
have estimated
and there
is
this
should be carried out and
more than one model more than one
some dimensions
some dimensions
I
of variability
in
which
in which
it’s
not sensitive
the changes
have a
this isn’t
that
and
you
it’s
can be
173
substantial
He
testified
further that
there are
conclusions with change
the coherence
the
model
of the
which damages can be
172
Edward
E Leamer
that are
this
Lets
Take
the
I
174
Leamer Depo
at
my suggested
at
of variability in which
model
Con
Out of Econometrics
358 818
70
the
and economic judgments about
that are implied
that demonstrates the
73 The American
by
method by
174
356 120
Leamer Depo
directions
the accuracy of the estimates
1983
173
some
made econometric
produced
one as
computed
variability
And
substantially
models
and selected
some
Economic Review 1
Case5 11cv 02509LHK
131
Dr Leamers
provided
models
Dr
Leamer
tried
177
to
fit
of any sensitivity
to the data
He
also said that
He
Thus
those other directions
is
it
analyses
175
a
acknowledged
176
one
but its
unclear
what
it
substantial
alternative
is
of
81
the
which has
178
he
but
models
models
sic
sensitive
what alternative
unclear
specification
examin ing
and
Page74
and computer programs he
Thus
out an extensive
Ive carried
not a complete
different directions
Filed1007 13
2
data documentation
with data by a defendant
with disaggregation
defendant
and the backup
report
no evidence
contain
Document518
to
do
defendant by
analysis in several
any
did not describe
Dr Leamer
to
tried
fit
of
to the
data
132
The
results
Defendants
reported
provided
I
clearly
show
above
allowing the impact of the conduct to
the fragility
of the single regression specification
common way
Another
conclusions that can be
drawn
period for which the regression
of testing the robustness
is
estimated
Dr Leamer
to changes
bases his conclusion
challenged agreements reduced compensation on a regression that compares
during the class period
and
2004 and
after
essentially
2010 and
2011 An
only the before period or only the
regression
and
test
2005 2009
after
to
the
alternative
combined periods
specification
period as the
that
across
Dr Leamer
and of the
of a regression specification
verify that the results are robust
is to
differ
that the
compensation
before effectively
to test robustness
control or benchmark
whether the challenged agreements affected
in the time
period
is
2003
to use
in the
compensation of members of the
proposed class
133
Exhibit
differences
23 shows
in the estimation
Leamers conduct
regression
undercompensation
175
176
177
178
179
Dr Leamers
that
period
179
implies
model
at
generally
at
at
at
22 and 24
365 1516
Leamer Depo
in his Figures
360 1618
Leamer Depo
366 79
Detailed regression
outputs
are
it
is
but very different
substantial
359 6
Leamer Depo
a test of whether
robust
Using only the pre period as the benchmark
percentages than reported
Leamer Depo
fails
provided in Appendices12A
71
to
12D
to
Dr
estimated
almost twice as large
Case5 11cv 02509LHK
Defendants but substantially
for several
Dr Leamers
benchmark
instead
conduct
overcompensation
Another sensitivity
using data outside
compensation
of
virtually
same magnitude
Dr Leamers
during the conduct period
as Exhibit
and
compensation
If
81
of
as the
no undercompensation
though opposite
model
is
to first estimate
Dr Leamers
higher
24 shows the
least
some
model
than actual
in
but
sign from
the
is
to
predict
robust one would expect the
compensation during the conduct
overcompensation
of the years
his conduct regression
estimates
compensation
predicted
therefore implying
years and five Defendants in at
sensitivity
Using only the post period
his conduct periods and then use the coefficient
However
period
implies
Page75
24
compensation to be generally
predicted
actual
regression
22 and
test
Filed1007 13
2
lower for Intel
of roughly the
effects he reports in Figures
134
Document518
180
levels are in fact lower than
two Defendants
at
Dr Leamers
model again
in all
fails
the
test
5 Dr Leamers
Regression
Model Does Not Explain Changes
in
Compensation
Over Time
135
from
The
analysis presented
Dr Leamers
correlated
regression
of his data
nature
firm
level compensation
relies
on comparison
predict
obtaining
showed
above
model
that are not accounted
level
a reliable prediction
critical
136
Exhibits
the impact
to estimating
25A
and
if
implies
we
that can be
for in his
of compensation
model
factors that drive
Given that his methodology
to the level
that his
model would
absent the challenged agreements
that the factors that
Leamer does
not account for are
by company and year between
compensation
earned by a firms employees and the average level of compensation
180
conduct
Detailed regression
plot the difference
Dr
important
Dr Leamers
regression
outputs
are
Figures
is
those agreements
quantitatively
I
drawn
account for the
that there are important
of compensation
any of
25B show
conclusions
are fundamentally different once
That correlation
of the actual
that the statistical
20 and
23
provided in Appendices
72
The
13A
and
exhibit
13B
shows
the average
predicted
that these prediction
by
Case5 11cv 02509LHK
errors are substantial
in
Document518
Dr Leamers
an
virtually all the factors that explain
Dr Leamer
Rather
the
is
terms economically
years and Defendants
distributed across
some
predicts very poorly in
years for
I
and predicted
actual
in
compensation
2003 and 2004 and
performed
a standard
firm
level compensation
Pixar in
statistical
test
be
are too large to
explained
for whether
solely
errors
A consequence
Leamers
181
Dr
estimated
Leamer used
evaluating
residuals
results
the
year level
at
model
companies
in
between
model
This
test
by company and year
the
regression
and asks whether those average residuals
test
factors
resoundingly rejects the
and
other
in
establishes the need to use
ways
statistically significant
Critically the
which
that contrary
implies
for important factors that determine
important
compensation
determinants of firm
level compensation
conduct effects will capture
the
term
economically
estimates
the impact
numerous
significant
one should focus on economic
value predicted
in his
just
not controlled
significance
Employee Class model Figure
ranges from 39 percent
provide accurate
residual
his
or correct for that correlation
of omitting
in his All Salaried
compensation
above
which means that
at
time
the Defendants over
138
Defendant
182
standard
claims Dr Leamer has
his
by
to
there are important factors explaining
by sampling error The
average residuals are economically not
to
one of which according
with the most extreme examples
Dr Leamers
no such omitted firm
specific
that there are
clustered
not explained
capturing
2004
that are omitted from
in compensation
hypothesis
model were
Large average differences
are evident
81
of
are not evenly
different Defendant
at
essentially examines the average residuals from his regression
variation
the
some companies
outcomes
different years have been left out of the regression
137
and
whether the Defendant had a challenged agreement with another
model
being Google
if
compensation
individuals
Page76
significant181
would be expected
as
important factors that are unique to compensation
employees
Filed1007 13
2
by
his
of the
below
regression
the
A model that
model is 16.5 percent
roughly plus or minus
indicating
33
percent
economically
by Dr Leamers
predicts
variable
that his
times in his deposition
20 are
overall
The
root
regression
compensation
mean square
The average
Actual real total
significant
to
49 percent
so poorly cannot
error
of the average
at
using a 95 percent confidence level
The test Ftest results are F39 5047711319.6 for Dr Leamers
Figure 20 regression and
F39,292367 832.09 for his Figure 23 regression Pvalues for both tests are virtually zero
73
Dr
and stated that in
model measures average compensation
182
that
other than the
and not just statistical significance
value predicted
impact of his conduct
of variables
is
the
company
Case5 11cv 02509LHK
challenged agreements that
To
the potential
illustrate
Document518
between
systematically
differ
problem
I
Filed1007 13
2
Page77
non conduct
the conduct and
considered what would happen
if I
81
of
periods
simply add a variable
measuring the performance of the stock market from his regression which potentially would
measure
general
acknowledges
shows
economic and
financial performance
from adding the change
the results
conduct regression
184
results the addition
P 500 index
Leamers
much
Dr Leamer
that
overcompensation
183
26
Exhibit
as an explanatory variable
Figure 22 and 24
smaller
2005 2009 Thus
Class throughout
economy
S
in the
of this variable yields
AllSalaried Employee Class and
the Technical
Dr
In contrast to
the
see his Figure 8 and related discussion
compensation
likely affect
in
in his
undercompensation
undercompensation
for all Defendants
estimates
except
for the
Google
for
of economically significant
the existence
classwide
by
factors not captured
be unreliable
measures
his
of
Dr
model causes
damages and
Figure 20 and 23 regression estimates
Leamers
unreliable
as
a method
to
of demonstrating common
impact
6 Dr Leamers
Conduct
Variable Cannot Capture the Impact of the Challenged
Agreements
Dr Leamers
139
conduct
variable
challenged agreements between
reflects
refers to these throughout
Plaintiffs
do
not claim
in their
non compete agreements
his report as
Complaint
or in their
Motion
was
183
184
the
independent
185
14A
Appendices
and 14B show
expected
See Declaration
allege
with
Intuit
through a cold call
Defendant
Evidence
I
that these
185
positive
detailed
outputs
regression
The
sign and is statistically significant
of Jeff Vijungco
I
am
Apple See
p6
on
unaware
of
coefficient
under
Adobe and Apple continued
any requisitions that went
also Declaration
of Chris
has agreed to not cold call employees
have never been instructed
presented
Dr
estimate
Leamers
basis
above
on the change in
SP
assumption
to refrain
at
Galy
Google
from making cold
pp
45
to recruit
unfilled
I
no cold
as
any other company and
call
understand that plaintiffs in this case
To my knowledge
calls to
and hire from each other
because of the
Google employees
no such agreement
To the contrary I have made cold calls to Google employees
am aware that other recruiters at Intuit have also done
so
74
exists
I
and have never given any
such instruction to anyone else at Intuit
same
as long as
observations
during the Class Period
agreement
understand that
98
Leamer Report
500 shows
not identified or recruited
I
for Class Certification
agreements prevented a Defendant from hiring applicants from another
that applicant
Dr
employees for a period of time Although
Defendants to avoid cold calling each others
Leamer
pairs of
on the
Case5 11cv 02509LHK
Dr Leamers
and that underlies
Document518
Filed1007 13
2
demonstrates that during the period of the challenged agreements
of other Defendants even
amount
when
cold calling were an important
if
and
hiring from other Defendants
by
the challenged
variable
conduct that he
is
call
way
Defendants hired employees
during the conduct period
eliminated
would have
as
of recruiting from other Defendants
calling activity were
can identify and measure
leading
or substantially
some
restrictions
extended to other firms as well
186
of the challenged
to
reduced
agreements only
the impact of those agreements properly
of which
were
typically
the impact
Dr Leamer
cites
not limited to the other Defendant
the
the
reasons
1 shows
his Figure
support
to
if
represents
His conduct variable cannot do so for several
trying to evaluate
First evidence I reviewed
cold
cold
his regression that represents
in
7
in his Figure
agreements
Dr Leamer
140
81
of
they had agreed not to cold call those employees and that the
of hiring from other Defendants did not decline
occurred
movers
earned by
analysis of compensation
Page78
identified in Figure
that
1 but
In part this reflects the fact that the motivation for these
ofinterest
agreements generally does not appear to be holding
undercompensating
potential
senior executives
calling
on
or perceived
of another
employees
the willingness
agreements
commercial
a
but not during the
include
the impact
non Defendant
corresponding
See
e
non conduct
If
DNCC
unchallenged
same
period
biasing
his estimated
effect
he can
separate
their
Dr Leamer
upwards
If
as
he can measure
impacts from the
policies
Yahoo EBay
PayPal
and Bizrate See
176 177 178 179 182 183
Bentley
of cold
agreements or
by
firms lead to undercompensation
if
of
as the challenged
periods then the effect estimated
agreements only
impact of unchallenged
g 231APPLE041662
conflicts
or concerns about the impact
For example Google had restrictions on hiring from Genentech
Geshuri Exhibits
or
from concerns about
existed during the
of those other policies
the impact of the challenged
187
arose
arrangements
of partners to collaborate
involving
compensation of
from membership on one companys Board of Directors
Leamer claims DNCC agreements between
186
but instead
187
unilateral policies
would
a firms
down
Depo 61 21622 251618
75
Lambert
Depo 21 78
Dr
Case5 11cv 02509LHK
141
The Consent Decrees
made
clear that restrictions
circumstances
such as
Document518
that the Defendants
on
recruiting other
when they
including
are
development technology
joint
signed with the
the function
integration
in Section
not to consider
IV
from employees
applications
ventures
projects
hire employees of another
person provided
any other person adopt enforce
142
of
the Plaintiffs theory
legal alternative
to
the challenged
how
not
permissible
an individual
where
restrictions
understand what collaborations
188
190
in
Final Judgment
Final Judgment
in
in
still
there
of
190
same way
agreements
and
for
Thus
affected
the
affected
Dr Leamers
by
would only be
the
by
affected
legal restrictions
which employees were involved
legal restrictions
the price
in
analysis to
calling
etc
a unilateral policy
Directors
is
to
and
Proposed
and
Proposed
avoid cold calling another
CEO
of
a competitor
of
non
discovery process would have been affected
and the impact and loss from the challenged
amount above
76
that
and
which employees were involved
on cold
Board of
and
would have been
employees
during the Class period to not cold call employees
challenged agreements
the incremental
the
and
world for
v Adobe Systems Inc et al 317 2011 p 7
v Lucasfilm Ltd 592011 p 6
theory
pressuring
opportunities
but for
This requires individualized
example when a member of
that
hiring prohibitions
United States of America
employees
was
from
v Adobe Systems Inc et al 317 2011 p 6
v Lucasfilm Ltd 592011 p 5
United States of America
then under
it
requesting
United States of America
were unilateral policies at Defendants
Defendants
would have
existed during the class period
United States of America
in
from
or
recruit
any less restrictive
limiting cold calling
Similarly nothing prevents an employer from implementing
Defendants
determined
recruiting the other Defendants
would have been
that there
Final Judgment
Final Judgment
189
on
as on balance procompetitive
the likelihood
is
some employees
determination
call
cold
policy
policy
of employees involved in those collaborations
requiring
collaborations
eg
collaborations
purposes of measuring impact and loss has
others
solicit
to
a
189
compensation
agreements
only to employees involved in actual
compensation
person or
further that
state
to adopt
or maintain such a policy and are prohibited
or maintain such a
including
The Consent Decrees
that Defendants are prohibited
any other person to adopt enforce
Under
joint
from unilaterally deciding
of another
of Justice
agreement
collaboration
188
81
of
are legal under certain
of a legitimate
joint
shall prohibit a Defendant
Page79
US Department
companies employees
teaming agreements and the shared use of facilities
n othing
Filed1007 13
2
caused by legal agreements
in the
If
Case5 11cv 02509LHK
143
Dr Leamers
Second
cold calling but
treats
conduct
variable
any agreement between
agreements between
multiple
Document518
a Defendant
Filed1007 13
2
cannot measure
Apple
pairs of Defendants
participated
Leamers
in
according
challenged agreements with several
to
as having
the
better and
in
Plaintiffs
and
Dr Leamer
Apple
Dr
that is
Rather such models would show
more
as
in a single challenged
other Defendants simultaneously
of price discovery
they apply that more information results
on
same impact
price discovery framework does not imply that the amount of information
restricted is irrelevant to the process
81
For example Adobe and Apple
and other Defendants
while
of
the intensity of restrictions
were assigned the same conduct values every year but Adobe engaged
agreement with one Defendant
Page80
where
rapid price discovery than less
agreements should have a larger impact than a single
information and thus that multiple
agreement
7
Estimated
Persistence Effects
Discovery
Model and
his
are Inconsistent with
Claim
Defendants
that
Dr Leamers
had
Price
Rigid Compensation
Structures
144
Dr Leamer
The
describes
percent
in
some year
on
for
finding
is
that effect lingers
frameworks
this
meaning when
very persistent effects
indicates
However
on which he
relies
that
many
makes him or
all
years
inconsistent
or virtually
all
levels gained
Defendant affects
only a weak
191
basis for
why
all
gets
a bump up
in
coworkers
If this
past compensation
144
77
agreements and
a significant and
wide spread
claims that economic theory combined
employees
employees compensation
the challenged
would cause
members Dr Leamer
by one Defendants
persistence between
Leamer Report
year and the
with the price discovery and internal equity
as the theoretical
class
a worker
her better off than comparable
with Defendants rigid compensation structures shows that
compensation
in the previous
191
resulting reduced flow of information to employees
impact on
as follows
that two for each employer The fact that these numbers sum to around 90
compensation
145
estimates
variables are the levels of total compensation
persistence
year before
his persistence
new
information on appropriate
from cold
calls
were true however
levels and current
from another
then there should
ones
a
mover who
be
is
Case5 11cv 02509LHK
induced to
move
a 30
receives
to Defendant
percent
Dr Leamers
Defendant
irrespective
is
determined
that any increase
in
persistent increase
B by
a
cold
his
model
implies
restored more than 95 percent
remain for
Adobe Apple
information in
of
damages imply
reconcile
with the
employment
per
that any effect
and
a highly
is
slow
as well as with his
own
it
is
of
members
of the
flow has been fully
the flow of information is
after
has the lowest estimated
not just
of the reduced
employees have obtained
after
on compensation would
of reduced information
rates of hiring at these firms
year
the firm generates
in
the information
after
that even five years
Google
model
Yet
compensation
and Pixar employees while more than 60 percent
Dr Leamers
As
individuals
no reason why aggregate compensation
of the impact
remain for Google employees
A and
compensation
proposed class would remain depressed so long
resumed Indeed
twoyears
compensation relative to others
his estimates
provides
81
in all Defendant
by demonstrating that an
informationflow persists strongly for an extended time even
more information Yet he
of
employees previous compensation
of those other
in that individuals
model and
Page81
call from Defendant
any year largely by his previous
in
an individuals
Dr Leamers
Filed1007 13
2
should cause a change
regression rules out such an impact
compensation
146
A from
in compensation
increase
employees compensation
Document518
persistence
glacial
averaging
of the effect
level The flow
Such slow adjustment
over
would
percent
is
of
hard
to
of average
theory
price discovery
8 Summary
147
Dr Leamers
undercompensation
are
made
regression
model
of undercompensation
and
percentages by company and year are invalid
to permit a test of his theory
there is
no evidence
of
his derived
When
common
estimates
of annual
necessary corrections
impact from the
challenged conduct no evidence of average impact across members of the proposed class and no
basis for his estimates
of undercompensation
Kevin
M Murphy
November
78
12
2012
Case5 11 cv 02509 LHK
Filed10 07 13
3
Exhibit
Hires
Document518
1A
and Separations at Defendant Companies
Hires
Page1
of
46
From Other Defendants vs
To
Separations
Hires
Overall
Separations
Year
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2001 2004 Avg
2005 2009 Avg
2010 2011 Avg
2001 2011 Avg
2001 2004
Total
2005 2009
Total
2010 2011
Total
2001 2011
Total
Notes
This analysis
excludes hires indicated as acquisitions hires showing the same
the hiring and separations
employment
Source
Dr
in
that appear as immediately
each year
Leamer’s employee
data
rehired by the same
defendant
defendant
company
company
as their immediate previous
within one year
Number
employer within one year of
of employees
is
calculated
as average
Case5 11 cv 02509 LHK
3
Exhibit
Hires
Document518
1B
and Separations at Defendant Companies
Hires
Filed10 07 13
From
To
Page2
of
46
Other DNCC Defendants vs Overall
Separations
Hires
Separations
Year
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
20012004 Avg
20052009 Avg
20102011 Avg
20012011 Avg
2001 2004
Total
2005 2009
Total
2010 2011
Total
2001 2011
Total
Notes
This analysis
excludes
the hiring and separations
employment
Source
in
hires
indicated as acquisitions
that appear as immediately
each year
Dr Leamer’s
employee
data
hires
showing the same defendant
rehired by the
same defendant company
company
as their immediate previous
within one year
Number
employer within one year of
of employees
is
calculated
as average
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
3
Page3
of
46
Exhibit 2A
Number of Employees by Defendant and
All
Adobe
Apple
Salaried Employee Class
Google
2001
2,503
5,096
2002
2,226
5,255
2003
2,291
5,424
1,329
2004
2,508
5,684
2005
3,791
6,474
2006
3,663
6,993
2007
3,951
2008
4,203
2009
Year
Intel
Intuit
Lucasfilm
Pixar
All
Defendants
210
3,169
66,242
542
3,982
63,569
4,311
62,439
2,346
4,247
64,172
4,117
4,418
73,556
6,873
4,498
74,045
7,951
8,768
5,069
73,247
9,135
10,983
5,081
75,205
4,928
10,005
11,175
4,683
75,166
2010
5,010
11,655
13,988
4,605
80,193
2011
5,385
13,226
18,179
4,770
90,070
Source Dr Leamer’s
backup
data
and materials
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
3
Page4
of
46
Exhibit 2B
Number of Employees by Defendant and
Technical Creative and
Adobe
Apple
Google
Intel
RD
Intuit
Year
Class
Lucasfilm
Pixar
All
Defendants
2001
1,582
2,670
101
1,557
34,484
2002
1,441
2,866
207
1,977
33,881
2003
1,450
2,954
509
1,907
33,517
2004
1,579
2,942
1,026
1,829
33,592
2005
2,205
3,358
2,258
1,814
40,479
2006
2,218
3,677
3,776
1,863
41,216
2007
2,277
4,248
5,290
2,244
42,550
2008
2,400
4,950
6,388
2,349
44,243
2009
2,552
5,589
6,825
2,237
45,453
2010
2,489
6,663
8,693
2,308
48,994
2011
2,639
7,582
11,139
2,457
55,338
Source Dr Leamer’s
backup
data
and materials
Case5 11 cv 02509 LHK
Document518
Exhibit
3
Filed10 07 13
Page5
of
46
3
Top 20 Previous Employersof Hires by Defendant Companies
Adobe
Note
Hires through
transfers
acquisitions
are excluded
This analysis uses
Adobe’s
compensation data and may not include
all internal
Case5 11 cv 02509 LHK
Document518
Exhibit
3
Filed10 07 13
Page6
of
46
3
Top 20 Previous Employersof Hires by Defendant Companies
Apple
Note
Analysis restricted
to
hires for
job codes provided
in
the compensation data
Case5 11 cv 02509 LHK
Document518
Exhibit
3
Filed10 07 13
Page7
of
46
3
Top 20 Previous Employersof Hires by Defendant Companies
Google
Case5 11 cv 02509 LHK
Document518
Exhibit
3
Filed10 07 13
Page8
of
46
3
Top 20 Previous Employersof Hires by Defendant Companies
Intel
Case5 11 cv 02509 LHK
Document518
Exhibit
3
Filed10 07 13
Page9
of
46
3
Top 20 Previous Employersof Hires by Defendant Companies
Intuit
Case5 11cv 02509LHK
Filed1007 13
Document5183
Exhibit
Page10
of
46
3
Top 20 Previous Employersof Hires by Defendant Companies
Lucasfilm
Number
Rank
Previous
of Hires
Percentage
of
Total Hires
2008Q2 2012Q1
26
Employer
LUCASFILM
2008Q2 2012Q1
7.1
1
ELECTRONIC ARTS
20
5.5
2
IMAGEMOVERS
8
2.2
3
WALT
6
1.6
4
ACTIVISION
5
1.4
5
ORPHANAGE
5
1.4
6
2K
4
1.1
7
CBS
4
8
DIGITAL
1.1
1.1
9
DIGITAL
DISNEY
INC
GAMES
DOMAIN
4
PDI
4
10
SONY
4
1.1
1.1
11
APPLE
3
0.8
12
DOUBLE
3
0.8
13
DREAMWORKS
3
0.8
14
MICROSOFT
3
0.8
15
PIXAR
3
0.8
16
ZYNGA
3
17
CRYSTAL DYNAMICS
2
0.8
0.5
18
MUNKYFUN
2
0.5
19
ADOBE
1
0.3
20
EBAY
1
0.3
3
0.8
61
16.7
187
51.2
0
0.0
7
1.9
365
100
Self
FINE
PRODUCTIONS
INC
EmployedUnemployed
Unknown
Other
NonDefendants
Other Defendants
All
Lucasfilm
Defendants excluding
Total
Lucasfilm
Case5 11cv 02509LHK
Document5183
Exhibit
Filed1007 13
Page11
of
46
3
Top 20 Previous Employersof Hires by Defendant Companies
Pixar
Number
of Hires
Percentage
of
Total Hires
2001 2012Q2
2001 2012Q2
PIXAR
5
0.6
1
LUCASFILM
22
2.5
2
BLUE SKY STUDIO
18
2.1
3
WALT
16
1.8
4
PDI
10
1.1
5
TIPPETT
10
1.1
6
APPLE
8
0.9
7
DREAMWORKS
RHYTHM
HUES
6
8
0.7
0.7
Rank
Previous
Employer
DISNEY
6
9
UC BERKELEY
5
10
WDFA
5
0.6
0.6
11
ELECTRONIC ARTS
4
0.5
12
ESC ENTERTAINMENT
4
0.5
13
MICROSOFT
4
0.5
14
SONY
4
0.5
15
BRIGHAM
3
0.3
16
FRAMESTORE
3
17
GOOGLE
3
0.3
0.3
18
TAMU
3
0.3
19
WARNER BRO
3
0.3
ACTIVISION
2
0.2
7
0.8
420
48.2
294
33.7
7
0.8
40
4.6
872
100
20
Self
YOUNG UNIV
EmployedUnemployed
Unknown
Other
NonDefendants
Other Defendants
All
Defendants excluding
Pixar
Pixar Total
Note The
lengths
of
the periods
analyzed vary by
Sources Recruiting data from Apple Google
company based on
Intel Intuit Lucasfilm
data availability
and Pixar
Compensation data from Adobe and Apple
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page12
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page13
of
46
Case5 11cv 02509LHK
Filed1007 13
Document5183
Page14
of
46
Exhibit 5
Employment
of Software Engineers
of Industries of Defendant Companies
Industries of
Defendant
Adobe
Year
Apple
Google
Intel
Intuit
LucasFilm
Pixar
Defendant
Defendant
Companies
Companies
Companies
79,910
2002
1,165
1,263
8,065
2003
1,167
1,228
7,811
101,470
2004
1,258
1,207
8,317
105,160
2005
1,694
1,336
10,656
106,890
2006
1,728
1,333
11,742
2007
1,880
1,411
13,907
108,650
2008
1,958
1,425
15,404
122,130
2009
1,984
1,282
16,301
127,860
2010
1,865
1,361
18,728
124,910
2011
1,939
1,475
22,318
134,150
96,440
Adobe
Apple
Google
Intel
Intuit
LucasFilm
Pixar
1.5
1.2
1.6
1.2
1.2
1.6
1.1
1.2
1.8
1.7
1.6
1.6
1.4
1.3
1.2
1.0
12.2
1.5
1.4
1.1
1.1
15.0
10.1
7.7
7.9
10.0
12.8
12.6
12.7
16.6
2002 2004 Average
8.6
2005 2009 Average
12.1
2010 2011 Average
15.8
of All Industries
Defendant
Industries
Adobe
584,020
0.2
0.2
0.2
0.2
1.4
1.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
1.2
1.4
1.5
1.7
868,210
All
0.2
0.2
0.2
0.2
0.2
0.2
1.8
1.9
2.2
921,500
0.2
0.2
2.4
651,740
717,420
758,050
Source Defendant
employment
numbers
are based on
data as well as classification of software engineers
Employment
Industry
of industries of
Specific
Defendant
Data for the following
Dr Leamer’s employee
performed by
companies
my
834,850
staff
based on BLS OES National
NAICS codes
764,430
based on CapIQ company
information
851,850
852,670
Apple
Google
Intel
Intuit
LucasFilm
Pixar
334100
Computer
519100
Other Information
334400
Semiconductor
511200
Software Publishers
2005 2009 Average
512100
Motion Picture
2010 2011 Average
and Peripheral
Equipment
Manufacturing
Companies
Services
and Other Electronic
and Video Industries
Component Manufacturing
2002 2004 Average
1.2
1.7
2.3
Case5 11cv 02509LHK
Document518
Filed1007 13
3
Page15
of
46
Exhibit 6
Age
Distribution of
New
Hires
2001 through 2011
Adobe
Apple
Google
All
31 to 35
36 to 40
41 and over
7
6
17
24
22
30
24
30
22
17
36 to 40
41 and over
Source Dr Leamer’s
RD
8
6
7
17
26
22
29
27
33
21
12
backup
data
and materials
19
30
24
14
13
Class
20
24
21
27
25 and under
31 to 35
Creative and
Pixar
Class
7
Technical
26 to 30
Employee
Lucasfilm
19
24
22
28
25 and under
26 to 30
Salaried
Intuit
Intel
18
32
24
15
10
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page16
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page17
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page18
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page19
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page20
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page21
of
46
Case5 11 cv 02509 LHK
Document518
3
Filed10 07 13
Page22
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page23
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page24
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page25
of
46
Exhibit 12
RSquareds
in
Dr Leamer’s
Compensation Structure Regressions Are Mostly
Attributable to
AllSalaried
R Squareds
in
Dr Leamer’s
Figure
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Source Dr Leamer’s
11
95
94
94
93
93
92
91
92
92
90
92
Figure 11
Employee
Employer and Job Indicators
Technical
Class
Including Only
Excluding
Employer and Job
Employer and Job
Indicators
Indicators
94
93
93
93
92
92
91
91
92
90
91
and 13 regressions
21
21
22
19
20
21
21
20
20
22
24
RSquareds
Dr
in
Leamer’s
Figure 13
89
89
88
88
88
87
85
86
88
84
88
Creative
and
RD
Class
Including Only
Excluding
Employer and Job
Employer and Job
Indicators
Indicators
89
88
88
88
87
87
85
86
88
84
87
15
16
16
18
16
19
17
19
17
18
21
Case5 11cv 02509LHK
Document518
Filed1007 13
3
Page26
46
of
Exhibit 13A
Named
PlaintiffsActual Total Compensation
by Dr Leamer’s
Comp
Actual Total
Named
Plaintiff
Employer
Year
Predicted
by Dr
Comp
Leamer’s
Model
1
Brandon Marshall
Predictions
Model
Figure 12
Total
vs
2
Difference
3
ADOBE
2006
73,895
61,035
12
Difference
31
12,860
17.4
Michael
Devine
ADOBE
2006
131,222
124,424
6,798
Michael
Devine
ADOBE
2007
146,540
135,001
11,539
5.2
7.9
Mark Fichtner
INTEL
2001
151,712
133,620
18,091
11.9
Mark Fichtner
INTEL
2002
124,426
120,980
3,446
Mark Fichtner
INTEL
2003
109,352
109,349
Mark Fichtner
INTEL
2004
123,374
120,221
3,153
Mark Fichtner
INTEL
2005
133,431
135,403
1,972
Mark Fichtner
INTEL
2008
122,013
133,469
11,456
Mark Fichtner
INTEL
2009
138,501
139,125
Mark Fichtner
INTEL
2010
152,238
141,816
Daniel Stover
INTUIT
2006
79,129
Daniel Stover
INTUIT
2007
Daniel Stover
INTUIT
Daniel Stover
INTUIT
Siddharth
Hariharan
LUCASFILM
Dr Leamer’s
Figure 12 regressions
Source
10,422
2.8
0.0
2.6
1.5
9.4
0.5
6.8
91,136
12,007
15.2
103,265
105,061
1,796
1.7
2008
175,177
108,817
66,361
37.9
2009
132,553
121,416
11,137
8.4
2007
102,000
90,819
11,182
11.0
3
624
Case5 11 cv 02509 LHK
Filed10 07 13
Document5183
Page27
of
46
Exhibit 13B
Named
Plaintiffs
by
Actual Total Compensation
Dr Leamer’s
Actual
Named
Plaintiff
Employer
Year
Total
Comp
1
Brandon Marshall
Predictions
Model
Figure 14
Total
vs
Comp
Predicted
by Dr
Leamer’s
Model
2
Difference
3
ADOBE
2006
73,895
60,754
12
Difference
31
13,141
17.8
Michael
Devine
ADOBE
2006
131,222
124,661
6,561
Michael
Devine
ADOBE
2007
146,540
134,724
11,816
5.0
8.1
Mark Fichtner
INTEL
2001
151,712
135,177
16,534
10.9
Mark Fichtner
INTEL
2002
124,426
121,965
2,461
2.0
Mark Fichtner
INTEL
2003
109,352
109,866
Mark Fichtner
INTEL
2004
123,374
119,152
Mark Fichtner
INTEL
2005
133,431
134,261
Mark Fichtner
INTEL
2008
122,013
132,988
Mark Fichtner
INTEL
2009
138,501
139,074
Mark Fichtner
INTEL
2010
152,238
141,186
Daniel Stover
INTUIT
2007
103,265
105,025
1,760
1.7
Daniel Stover
INTUIT
2008
175,177
108,866
66,311
37.9
Daniel Stover
INTUIT
2009
132,553
122,644
9,909
7.5
Siddharth
LUCASFILM
2007
102,000
89,439
12,561
12.3
Source
Hariharan
Dr
Leamer’s Figure 14 regressions
514
4,222
830
10,974
573
11,052
0.5
3.4
0.6
9.0
0.4
7.3
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page28
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page29
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page30
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page31
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page32
of
46
Exhibit 16
Dr Leamer’s Model
Implies
Very Large Differences Over Time
Individuals with Identical Characteristics and Starting
Simulations Based
Adobe
on Dr
Leamer’s
15
32
90th Percentile
31
67
Compensation Levels
Intel
Difference in Compensation after
Average
46
Two
Intuit
All Firms
Years
11
22
100
the Compensation of
Conduct Regression
Google
Apple
in
16
33
24
56
22
46
37
86
Difference in Compensation after Five Years
29
61
Average
90th Percentile
53
62
111
16
34
135
Notes
1
2
3
4
Compensation
Percent
Based
using
on 50,000 simulations of compensation
Lucasfilm
Source
differences are constructed
differences are defined as differences
Dr
and Pixar are excluded because
Leamer’s
backup
data
in
and
residuals
from Dr Leamer’s
Figure 20 regression
logs
growth
there
and materials
coefficients
from 2004 through
is insufficient
data to
2009
for
each firm
do simulations
in all
years
model
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page33
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page34
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page35
of
46
Case5 11 cv 02509 LHK
3
Document518
Exhibit
Filed10 07 13
Page36
of
46
19
Average Percent Change in Total Compensation
Dr Leamer’s Figure
19 Disaggregated by Company
vs
Dr Leamer’s
Figure
Average Change
Year
Adobe
2002
Compensation
27.8
2003
in Total
Apple
Google
Lucasfilm
Pixar
Pooled
Intel
Intuit
27.2
0.6
1.5
2.1
5.1
8.5
4.7
2.3
13.1
8.3
10.3
9.8
6.9
1.3
2006
10.6
5.6
13.9
0.5
9.1
2007
11.2
4.5
8.8
2008
6.9
12.0
8.8
2009
7.5
2.9
0.1
7.4
6.8
7.4
2010
3.0
11.1
7.9
8.7
12.7
2011
2004
2005
EstimatedOverpayment Underpayment
Year
Adobe
Apple
Google
2005
3.4
0.6
4.2
8.8
4.9
0.0
14.5
16.4
2008
0.0
2009
0.0
0.0
2006
2007
6.5
9.7
1.8
Initial
Lucasfilm
Pixar
Pooled
Intel
Intuit
8.7
12.2
0.6
0.4
8.9
2.8
8.5
35.6
17.2
26.8
9.5
0.9
0.0
6.4
0.0
3.8
0.0
3.8
0.0
9.0
0.0
2.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Pixar
Pooled
EstimatedOverpayment Underpayment
Cumulative
Year
Adobe
Apple
Google
Intel
Intuit
2005
3.4
4.2
8.7
12.2
0.6
2.8
35.6
9.5
2006
13.0
25.9
12.5
9.5
11.4
62.3
10.3
27.5
13.3
15.1
71.4
12.9
27.5
9.5
9.5
18.9
2008
4.0
8.9
8.9
18.9
13.3
15.1
71.4
12.9
2009
8.9
27.5
9.5
18.9
13.3
15.1
71.4
12.9
2007
Note
This analysis
follows
Defendants even though
Source
Dr
for
Lucasfilm
Leamer’s methodology in his Figure 19 of treating 2005 as the first year of the agreements for
Intuit Lucasfilm and Pixar the first alleged agreements started in other years
Leamer Report backup data and programs
all
19
Case5 11cv 02509LHK
Document518
Exhibit
Undercompensation
Defendant
Estimates Using
Conduct Variables and Other Defendant
Effects
in
Dr
Specific
Adobe
Apple
Google
2005
1.82
2.54
12.73
2006
4.37
0.72
26.90
0.68
2.65
19.16
6.26
6.45
2008
2.19
4.06
5.70
8.01
10.24
2009
20.26
1.53
5.43
8.96
10.02
RD
Apple
Google
2005
1.92
2.01
11.08
1.71
2006
5.82
2.95
22.47
0.62
2007
0.05
5.23
13.12
3.03
6.93
2008
1.29
7.33
0.88
3.44
8.59
2009
22.60
6.28
10.56
4.67
7.47
Leamer Figure 20 and 23 regressions including
indicators
Pixar revenue
and
Dr Leamer’s
vs
conduct age and
data after 2005 are included
Figures
22 and 24
Lucasfilm
Pixar
Year
Adobe
Apple
Google
25.47
2005
1.61
1.59
1.78
1.67
30.64
2006
4.28
4.43
4.44
4.70
14.63
12.44
13.95
28.52
2007
6.64
6.94
6.39
7.46
3.24
17.24
14.28
14.15
36.96
2008
9.08
9.56
8.40
10.05
5.64
19.94
15.76
13.79
31.11
2009
9.15
9.73
7.51
9.95
5.70
20.12
14.65
Intel
Technical Creative and
Class
Adobe
company
Undercompensation Estimates in Dr Leamer’s
Interactive
9.59
Year
Intel
46
1.70
Intuit
0.51
Technical Creative and
of
AllSalaried Employee Class
1.89
2007
Intel
Page37
20
Employee Class
Year
Source
Filed1007 13
Specific
Leamer’s Regression
AllSalaried
3
Intuit
Lucasfilm
12.13
RD
Pixar
10.56
Class
hiring
Pixar
Year
Adobe
Apple
Google
28.18
2005
1.56
1.90
3.07
1.64
17.23
interactions
Lucasfilm
6.60
Intuit
30.70
2006
4.29
4.96
7.23
3.06
14.77
10.47
23.38
36.34
2007
6.48
7.79
9.36
3.38
3.41
18.08
10.61
24.38
34.92
2008
8.80
10.64
11.20
4.76
5.21
20.44
11.87
24.05
28.33
2009
8.44
10.51
9.00
4.19
4.96
20.54
9.62
between
rate variables
Intel
Intuit
Lucasfilm
10.80
Pixar
9.28
Case5 11 cv 02509 LHK
Document518
Exhibit
Dr
Page38
of
46
21A
Leamer’s Figure 20 RegressionUsing Corrected Standard Errors
Employee Class
AllSalaried
Dependant
Filed10 07 13
3
Variable
Log Total Annual Compensation CPI
Variable
Estimate
St
Error
T Value
Conduct
Age
0.0067
0.0031
2.18
Conduct
Age 2
0.0001
0.0000
2.45
Conduct
Log Number
0.0028
0.0247
0.12
0.1647
0.1269
1.30
0.6949
0.0608
11.42
0.7404
0.0587
12.62
0.4945
0.0530
9.33
0.6690
0.0351
19.06
0.7090
0.0458
15.48
0.6944
0.1840
3.77
0.8131
0.1069
7.61
0.2963
0.0461
6.43
0.2610
0.0407
6.41
0.3732
0.0453
8.25
0.3001
0.0389
7.71
0.2551
0.0433
5.89
0.1983
0.0780
2.54
0.1779
0.0979
1.82
Log Age Years
0.3591
0.1799
2.00
Log Age 2
0.0394
0.0233
1.69
Log Company Tenure Months
0.0107
0.0415
0.26
Log Company Tenure
0.0012
0.0043
0.28
0.0027
0.0020
1.37
San Jose
1.4353
0.3827
3.75
Among Defendants
0.0961
0.0456
2.11
0.0038
0.0076
0.50
0.0154
0.0214
0.72
0.2485
0.0568
4.37
0.1070
0.0785
1.36
of
New
Hires In the
Firm
Number
of
Employees
1
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
INTEL
Log Total Annual Compensation CPI
INTUIT
Log Total Annual Compensation CPI
PIXAR
1
1
1
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
INTEL
Log Total Annual Compensation CPI
INTUIT
Log Total Annual Compensation CPI
PIXAR
2
2
2
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
DLog
Information Sector Employment
Log Total Number
Year
of
Transfers
in
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog Firm Revenue
Firm
Number
of
Employees
1
Hires
1
EmployeeCPI 1
EmployeeCPI
0.2170
0.0814
2.67
APPLE
0.0627
0.2642
0.24
GOOGLE
1.0364
0.3351
3.09
INTEL
0.1522
0.2431
0.63
INTUIT
0.1462
0.2151
0.68
PIXAR
0.7251
0.6673
1.09
LUCASFILM
0.1352
0.2762
0.49
State
Location
Per
YES
Indicators
Constant
YES
RSquare
0.926
Observations
Note
Source
Significant
Dr Leamer’s
504,897
at
1
level
Significant
at
backup data and materials
5
level
Standard
Significant
at
10 level
errors clustered on employer year
Case5 11 cv 02509 LHK
Document518
Exhibit
Dr
Filed10 07 13
3
of
46
21B
Leamer’s Figure 23 RegressionUsing Corrected Standard Errors
Technical Creative and
Dependant
Page39
Variable
RD
Class
Log Total Annual Compensation CPI
Variable
Estimate
St
Error
T Value
Conduct
Age
0.0079
0.0033
2.38
Conduct
Age 2
0.0001
0.0000
2.71
Conduct
Log Number
0.0121
0.0281
0.43
0.2196
0.1362
1.61
0.6744
0.0650
10.38
0.7234
0.0570
12.70
0.4367
0.0672
6.50
0.6401
0.0325
19.67
0.6703
0.0486
13.81
0.6491
0.2295
2.83
0.8462
0.0911
9.29
0.3053
0.0523
5.83
0.2538
0.0391
6.49
0.3659
0.0476
7.68
0.3179
0.0353
9.00
0.2857
0.0439
6.51
0.1045
0.0896
1.17
0.1448
0.0805
1.80
Log Age Years
0.5894
0.1877
3.14
Log Age 2
0.0696
0.0239
2.92
Log Company Tenure Months
0.0297
0.0477
0.62
Log Company Tenure
0.0025
0.0049
0.52
0.0065
0.0024
2.64
San Jose
1.4378
0.4146
3.47
Among Defendants
0.0973
0.0493
1.98
0.0008
0.0080
0.10
0.0240
0.0241
0.99
0.2720
0.0617
4.41
0.0661
0.0853
0.78
of
New
Hires In the
Firm
Number
of
Employees
1
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
INTEL
Log Total Annual Compensation CPI
INTUIT
Log Total Annual Compensation CPI
PIXAR
1
1
1
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
INTEL
Log Total Annual Compensation CPI
INTUIT
Log Total Annual Compensation CPI
PIXAR
2
2
2
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
DLog
Information Sector Employment
Log Total Number
Year
of
Transfers
in
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog Firm Revenue
Firm
Number
of
Employees
1
Hires
1
EmployeeCPI 1
EmployeeCPI
0.2068
0.0869
2.38
APPLE
0.1220
0.2718
0.45
GOOGLE
1.3682
0.4309
3.18
INTEL
0.1569
0.2761
0.57
INTUIT
0.1393
0.2268
0.61
PIXAR
1.5864
1.0458
1.52
LUCASFILM
0.0127
0.3184
0.04
State
Location
Per
YES
Indicators
Constant
YES
RSquare
0.874
Observations
Note
Source
Significant
Dr Leamer’s
292,489
at
1
level
Significant
at
backup data and materials
5
level
Standard
Significant
at
10 level
errors clustered on employer year
Case5 11cv 02509LHK
Document518
Filed1007 13
3
Page40
of
46
Exhibit 22A
Dr Leamer’s
Estimates of Undercompensation Are Not StatisticallySignificant
AllSalaried Employee
Adobe
Google
Apple
Dr
Class
Intel
Intuit
Lucasfilm
Pixar
Leamer’s Annual Undercompensation Estimates Figure 22
2005
1.61
1.59
1.78
1.67
12.13
10.56
2006
4.28
4.43
4.44
4.70
14.63
12.44
2007
6.64
6.94
6.39
7.46
17.24
14.28
2008
9.08
9.56
8.40
10.05
5.64
19.94
15.76
2009
9.15
9.73
7.51
9.95
5.70
20.12
14.65
T Statistics
for
3.24
Annual Undercompensation Estimates
2005
0.94
0.74
0.47
0.96
1.17
0.91
2006
0.88
0.81
0.49
1.49
0.98
0.86
2007
0.90
0.80
0.55
1.62
0.86
0.93
0.88
2008
0.90
0.80
0.60
1.63
0.99
0.95
0.79
2009
0.94
0.82
0.64
1.62
1.04
0.96
0.72
P Values
for
Annual Undercompensation Estimates
2005
35.3
46.5
64.1
34.0
24.9
36.8
2006
38.2
42.3
62.7
14.2
33.0
39.3
2007
37.1
42.6
58.7
11.1
39.4
35.5
38.4
2008
37.0
42.6
55.1
10.8
32.6
34.4
43.2
2009
35.0
41.7
52.3
11.2
30.1
34.3
47.7
Notes
1
2
Estimates with
Standard
tstatistics
below
errors are clustered
1.96
in
absolute value
on employer
Source Dr Leamer’s Figure 20 regression data
and
year
or equivalently
with
pvalues
greater than
5
are not
statistically
significant
at
the
95 level
Case5 11cv 02509LHK
Document518
Filed1007 13
3
Page41
of
46
Exhibit 22B
Dr Leamer’s
Estimates of Undercompensation Are Not StatisticallySignificant
Technical Creative and
Adobe
Google
Apple
Dr
RD
Class
Intel
Intuit
Lucasfilm
Pixar
Leamer’s Annual Undercompensation Estimates Figure 24
2005
1.56
1.90
3.07
1.64
10.80
9.28
2006
4.29
4.96
7.23
3.06
14.77
10.47
2007
6.48
7.79
9.36
3.38
3.41
18.08
10.61
2008
8.80
10.64
11.20
4.76
5.21
20.44
11.87
2009
8.44
10.51
9.00
4.19
4.96
20.54
9.62
T Statistics
for
Annual Undercompensation Estimates
2005
0.81
0.77
0.71
0.83
0.91
0.78
2006
0.78
0.79
0.72
0.94
0.85
0.72
2007
0.79
0.80
0.75
0.76
0.79
0.83
0.67
2008
0.79
0.80
0.77
0.81
0.83
0.83
0.61
2009
0.79
0.81
0.80
0.72
0.84
0.83
0.49
P Values
for
Annual Undercompensation Estimates
2005
42.4
44.7
48.2
40.8
2006
43.7
43.0
47.5
35.0
39.9
47.4
2007
43.6
43.0
45.6
44.8
43.1
41.0
50.7
2008
43.5
42.8
44.3
42.4
40.9
41.0
54.1
2009
43.1
42.4
42.8
47.8
40.4
41.2
62.7
Notes
1
2
Estimates with
Standard
tstatistics
below
errors are clustered
1.96
in
absolute value
on employer
Source Dr Leamer’s Figure 23 regression data
and
year
or equivalently
with
pvalues
greater than
36.8
5
are not
statistically
significant
44.1
at
the
95 level
Case5 11cv 02509LHK
Document518
Exhibit
Undercompensation
as
Benchmark
Estimates
Using
in
Dr
AllSalaried
Apple
2005
2.71
3.61
6.33
2006
7.94
9.12
15.64
2007
12.15
14.47
20.77
1.56
2008
16.55
19.95
25.25
2009
15.87
19.92
22.16
Apple
Google
of
46
23
Undercompensation
as
Benchmark
Estimates
Using Post Conduct Period
in
Dr
Leamer’s Regression
AllSalaried Employee Class
Lucasfilm
Pixar
Year
Adobe
Apple
Google
Intel
2.81
14.56
16.52
2005
2.35
2.55
2.76
3.65
22.11
19.53
2006
6.66
6.74
6.80
6.18
27.43
19.88
2007
10.43
10.54
9.43
6.72
2.74
9.00
30.44
23.69
2008
14.40
14.43
11.85
1.37
8.34
30.04
20.65
2009
14.55
14.49
10.20
Intel
Technical Creative and
Adobe
Page42
Employee Class
Adobe
Year
Filed1007 13
vs
Leamer’s Regression
Year
Google
PreConduct Period
3
Intel
Intuit
RD
Pixar
2.29
14.80
12.66
5.08
19.72
15.17
4.83
24.07
16.81
9.43
8.35
27.74
19.25
9.05
8.51
28.06
17.56
Technical Creative and
Class
Intuit
Lucasfilm
Lucasfilm
Pixar
Year
Adobe
Apple
Google
Intel
Intuit
RD
Class
Intuit
Lucasfilm
Pixar
2005
3.46
4.70
8.39
3.54
16.57
18.91
2005
2.33
2.26
1.81
2.25
16.28
11.56
2006
10.10
11.69
20.04
3.90
25.84
21.64
2006
6.47
6.08
4.52
5.96
20.36
13.40
2007
15.29
20.74
18.40
25.15
25.38
7.90
10.96
31.64
34.10
2007
24.35
2008
10.17
14.00
9.38
12.71
6.50
8.46
9.12
12.50
4.58
8.08
24.38
29.55
0.43
1.63
20.55
2008
28.54
14.99
16.28
2009
19.53
24.64
23.64
0.33
9.96
32.41
19.40
2009
14.25
12.62
7.12
12.37
8.24
29.30
14.15
Source
Leamer Figure 20 and 23 regressions estimated
and pre conduct
period
data only
using conduct
Source
Leamer Figure 20 and 23 regressions estimated
and post conduct
period
data only
using conduct
Case5 11cv 02509LHK
Document518
Exhibit
Undercompensation
Estimates Predicted Using
Conduct Period Data
in
AllSalaried
Dr
Apple
Google
Intel
vs
2005
5.01
0.84
0.72
2006
2.65
5.79
2007
4.26
2008
4.67
2009
1.00
Google
1.61
1.59
1.78
1.67
12.13
10.56
2006
4.28
4.43
4.44
4.70
14.63
12.44
4.45
2007
6.64
6.94
6.39
7.46
3.24
17.24
14.28
29.03
2008
9.08
9.56
8.40
10.05
5.64
19.94
15.76
2009
9.15
9.73
7.51
9.95
5.70
20.12
14.65
2.96
2.48
4.52
2005
5.61
2.73
5.99
16.84
12.56
2.34
8.78
6.72
3.78
0.10
18.53
7.36
10.78
3.88
2.21
3.13
7.87
12.05
3.93
32.40
Adobe
Apple
Google
Intel
2005
5.83
0.97
1.89
3.43
2006
2.05
4.03
12.09
1.29
2007
5.83
9.57
7.59
5.47
6.76
2008
5.18
4.33
25.03
2.56
8.81
2009
1.46
2.26
6.45
3.09
10.53
Leamer Figure 20 and 23 regressions estimated
Undercompensation calculated
using residuals
data after 2005 are included
predicted
3.07
1.64
2006
4.29
4.96
7.23
3.06
14.77
10.47
2007
6.48
7.79
9.36
3.38
3.41
18.08
10.61
16.70
2008
8.80
10.64
11.20
4.76
5.21
20.44
11.87
23.03
2009
8.44
10.51
9.00
4.19
4.96
20.54
9.62
2005
24.15
1.52
6.44
1.86
1.90
the conduct
period
Class
1.90
11.66
period
Pixar
1.56
Year
non conduct
RD
Lucasfilm
Google
Pixar
data
Adobe
Intuit
Apple
Lucasfilm
6.07
for
Intel
Technical Creative and
3.05
Intuit
using
Adobe
Class
Year
22 and 24
Apple
Year
RD
46
AllSalaried Employee Class
Pixar
Technical Creative and
of
Figures
Lucasfilm
Intuit
Page43
24
Employee Class
Adobe
Pixar revenue
Filed1007 13
Undercompensation Estimates in Dr Leamer’s
Non
Leamer’s Regression
Year
Source
3
Intel
Intuit
Lucasfilm
10.80
Pixar
9.28
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page44
of
46
Case5 11cv 02509LHK
Document518
3
Filed1007 13
Page45
of
46
Case5 11cv 02509LHK
Document518
Exhibit
Undercompensation Estimates Including Change
SP 500
in
Dr Leamer’s
AllSalaried
Adobe
Apple
Google
2005
0.11
0.06
0.17
0.23
0.27
0.43
0.39
0.44
0.68
1.70
2008
0.55
0.62
1.01
2009
0.66
0.66
1.01
Figures
Google
1.61
1.59
1.78
1.67
12.13
10.56
4.28
4.43
4.44
4.70
14.63
12.44
2007
6.64
6.94
6.39
7.46
3.24
17.24
14.28
2.25
2008
9.08
9.56
8.40
10.05
5.64
19.94
15.76
2.14
2009
9.15
9.73
7.51
9.95
5.70
20.12
14.65
1.64
2005
1.83
1.83
2006
0.22
1.96
2.23
2.22
0.55
2.28
2.32
0.61
2.31
Adobe
Apple
Google
2005
0.48
0.19
0.84
2006
1.20
0.69
1.82
2007
2008
1.93
2.64
1.00
1.32
1.87
1.74
4.26
5.59
2009
2.81
1.40
1.15
5.76
Lucasfilm
Intuit
RD
Class
1.90
3.07
1.64
10.80
9.28
4.29
4.96
7.23
3.06
14.77
10.47
6.48
8.80
7.79
2008
10.64
9.36
11.20
3.38
4.76
3.41
5.21
18.08
20.44
10.61
11.87
2009
8.44
10.51
9.00
4.19
4.96
20.54
9.62
3.49
1.29
2005
3.17
1.43
2006
0.71
1.59
3.38
4.37
2.21
1.86
2007
1.74
4.57
1.65
P 500
Pixar
1.56
0.41
S
RD
Lucasfilm
Google
2.12
Adobe
Intuit
Apple
Year
in
Intel
Technical Creative and
Pixar
Intuit
Leamer Figure 20 and 23 regressions including change
Adobe
Class
Lucasfilm
Intel
22 and 24
Apple
1.90
Year
Bloomberg
vs
Year
Technical Creative and
46
AllSalaried Employee Class
0.84
2007
of
26
Pixar
Intel
Page46
Undercompensation Estimates in Dr Leamer’s
in
Regression
0.17
2006
Source
Filed1007 13
Employee Class
Year
Net Total Return Index
3
Intel
Intuit
Lucasfilm
Pixar
Case5 11 cv 02509 LHK
Document518
4
Filed10 07 13
Page1
of
73
CurriculumVitae
Kevin
M Murphy
October 2012
Home
Address
Business
University
Address
of Chicago
1810 Pennington Court
Booth School of Business
New
5807 South Woodlawn Avenue
Chicago
Illinois
60637
murphy chicagobooth
email kevin
Lenox
Illinois
60451
Phone 815463 4756
Fax 815463 4758
edu
Current Positions
July
2005 Present George J
Department
Faculty
of Economics
Research
Stigler
Distinguished
Service Professor
of
and Booth School of Business University
Bureau of Economic
Associate National
Economics
of Chicago
Research
Education
University
of California Los Angeles
University
of
Thesis
Previous
Topic
Chicago
Ph
Specialization
D
and
Research and Academic
2002 2005 George J
Stigler
A B Economics
1981
1986
Human
Capital
Positions
Professor
of
Economics Department
of
Economics and
Booth School of Business University of Chicago
1993
2002 George
Relations
1989
University
1993
Pratt Shultz
Professor
of Business
Economics
and Industrial
of Chicago
Professor
of Business
Associate
Professor
Economics and
Industrial Relations
University
Chicago
1988
1989
University
of Chicago
of Business
Economics
and Industrial Relations
of
Case5 11 cv 02509 LHK
1986
1988
University
Document518
Assistant Professor
of Business
Filed10 07 13
4
Economics and
Page2
of
73
Relations
Industrial
of Chicago
1983
1986 Lecturer Booth School
1982
1983 Teaching
Associate Department of Economics University of Chicago
1979
1981 Research
Assistant
of Business University
Unicon
of Chicago
Corporation Santa Monica
Research
California
Honors and Awards
2008 John von Neumann
Lecture
2007 Kenneth J Arrow Award
1998
Rajk
with Robert
October 2005 Garfield Research
September
Award
Prize
College Corvinus University
Budapest
H Topel
with Robert
H Topel
2005 MacArthur Foundation Fellow
Elected
Academy
to the American
1997 John Bates
of Arts
Sciences
Clark Medalist
1993 Fellow of The Econometric
Society
1989
1991
Sloan
1983
1984
Earhart
1981
1983 Fellowship Friedman Fund
1980
1981 Phi Beta Kappa
1980
1981
1979
1981 Department Scholar Department
Foundation Fellowship University of Chicago
Earhart
Foundation Fellowship University
University
University
of Chicago
of California
Los
Foundation Fellowship University
of
of Chicago
Angeles
of California
Economics
Los
University
Angeles
of California
Los Angeles
Publications
Books
Social
Economics Market Behavior
Cambridge
MA Harvard
University
in a Social Environment
Press
Measuring the Gains from Medical Research
with Robert
H Topel
Chicago
University
with Gary
S Becker
2000
An
Economic Approach
of Chicago Press
edited
volume
2003
2
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
4
Page3
73
of
Articles
Government
of
Journal
and Inference
Estimation
Journal
of Cigarette Health
Regulation
Schneider 24
Lynne
Law
of
Two Step
in
Unemployment
Risk and Earnings
with Benjamin
Klein and
1981
Models
Econometric
with Robert
H Topel
Testing
Wage
for Equalizing
Differences
H
Market with Robert
Topel in Unemployment and the Structure of
Markets pp 103139 ed Kevin Lang and Jonathan S Leonard London Basil
Labor
3
1985
and Economic Statistics 370
Business
Information
and Economics 575
the
in
Labor
Blackwell
1987
The
Evolution
Topel
MA MIT
Unemployment
of
NBER
in
in the United
States
pp 11 58
Annual
Macroeconomics
1968 1985
ed
H
with Robert
Stanley Fischer
Cambridge
1987
Press
Cohort Size and Earnings in the United States with Mark Plant and Finis Welch
Economics
Ronald
Age
of Changing
D Lee W
The Family and
A Theory
Distributions
in
Brian Arthur and Gerry
State with Gary
the
Addiction
of Rational
Developed
Countries
pp
39 58 ed
1988
Rodgers Oxford Clarendon Press
S Becker
with Gary
31
Journal
S Becker
of
96
Law
Journal
Political
1988
1
and Economics
of
in
Economy 675
1988
Vertical Restraints and Contract
and Economics 265
Income
Robert
Wage
Distribution
W Vishny 104
Premiums for
Welch 18
with Finis
of
Building
Robert
Political
Quarterly
of
Economics
537
Researcher
1003
Push with
1989
17
W Vishny
of
Law
in
NBER
Andrei
Macroeconomic
the Theory and Measurement
Weiss and Gideon Fishelson
with Andrei Shleifer
and
1989
Shleifer
Annual
MA MIT
London
of
and Robert
Models
Macmillan
ed
Olivier Jean
1989
with Robert
Unemployment
W Vishny 97
with Andrei Shleifer and
pp 247 87
Press
Wages Reconsidered Theory and Evidence
in
Journal
1989
Blocks of Market Clearing Business Cycle
Efficiency
Klein 31
Graduates Recent Growth and Possible Explanations
College
Educational
Economy
Journal
Blanchard and Stanley Fischer Cambridge
Advances
with Benjamin
Market Size and Industrialization
Industrialization and the Big
Journal
Enforcement
1988
H Topel
pp 204 240 ed
in
Yoram
1990
3
Case5 11 cv 02509 LHK
Age Earnings
Empirical
Document518
Filed10 07 13
4
Profiles with Finis Welch
8
Journal
of
Page4
Labor
of
73
Economics
202
1990
Human
Capital
Tamura 98
Accounting
Political
of
Economy
Slowdown
for the
The Role
Changing
Has
Rational
Michael
Wages
106
Enterprise
Quarterly
Journal
for
Grossman 81
American
Economic
Graduates
of College
121 40 ed
Publishers
E Becker
William
Increased
in
over
Activity
Growth with
and the Effect of Price on
Addiction
F
with Chinhui Juhn
the United
in
Enterprise
Institute
Welch
Finis
States
in
Workers
Marvin Kosters
1991
Institute
Time
H
with Robert
1991
75
Andrei Shleifer and Robert
1991
Economics 503
of
and Robert
pp 39 69 ed
States
Papers on Economic
Brookings
Patterns
with
Differentials
the United
of Talent Implications
Allocation
W Vishny
in
Rate of Unemployment
the Natural
Topel and Chinhui Juhn 2
The
Patterns
DC American
Washington
Why
Trade in Wage
of International
Convergence
DC American
Marvin Kosters Washington
Wages
and Their
Black White Wage
in
S Becker
Gary
1990
S12
and Their Wages Changing
and Brooks Pierce in Workers
pp 107143 ed
1991
Growth with
and Economic
Fertility
Journal
Consumption
S Becker
with Gary
and
1991
Review 237
pp
The Economics of American Higher Education
Lewis Boston Kluwer Academic
R
1992
Changes
Katz 107
in
The
The
Transition
Journal
Wages
Structure of
Shleifer
The
to
and Robert
Division
Quarterly
of
Journal
Industrial
Wages 1963 1987
Relative
Quarterly
of
Wage
101
of
107
Pitfalls
Quarterly
Quarterly
of Partial
Journal
of
Journal
Economics 1137
the Rising
in
Planning
Importance
Economy
of Skill
in
410
1992
Reform with
1992
Andrei
with
pp 101 132 ed
with Finis
Gary
Welch
Peter Gottschalk
Sage Foundation Publications
and the Rise
Political
Economics 285
of
S Becker
107
1992
America
Russell
F
Economics 889
Labor Coordination Costs and Knowledge
York
Inequality
Journal
107
with Lawrence
1992
with Finis Welch
Economy
Demand Factors
Supply and
Economics 35
W Vishny
Change and
New
of
a Market
Tides Rising Inequality
Danziger
and Darrell
in
Uneven
and Sheldon
1993
Returns to Skill with Chinhui Juhn and Brooks Pierce
1993
4
Case5 11 cv 02509 LHK
Change and
Occupational
Document518
for Skill
1940
1990
Page5
with Finis
of
73
Welch 83
1993
Wages
and Relative
Inequality
Demand
the
Economic Review 122
American
Filed10 07 13
4
with Finis
Welch 83
Economic Review 104
American
1993
Why
Is
Rent Seeking
Vishny 83
A Simple
Journal
of
American
So
Costly to
Growth
Theory of Advertising as
a
with Andrei Shleifer
W
and Robert
1993
Review 409
Economic
Good
or
Bad
with Gary
S Becker
108
Quarterly
1993
Economics 941
Relative Wages and Skill Demand 1940 1990 with Chinhui Juhn in Labor Markets
Employment Policy and Job Creation pp 34360 ed Lewis
Solmon and Alec
C
Levenson The Milken
Westview
Cattle
Empirical
Analysis of Cigarette
Grossman 84
in
with Chinhui
Labor Market Outcomes
Juhn 1
in
Wage
Political
of
The
and Michael
Contrasting
the 1980s and Earlier
Decades
1995
Review 26
Hike
with
Donald
R Deere
and
1995
on Minimum Wages and Employment with Donald
Effects
of the
Finis
R Deere
MinimumWage on Employment pp 26 54 ed
1996
H Kosters Washington DC The AEI Press
of
Political
Economy
Inequality
Economics 72
Quality and
Wage
Policy
Review 232
Economic
Social Status Education
Journal
Journal
1994
1990 91 Minimum Wage
the Evidence
and Finis Welch
Marvin
Economic
the
American
Examining
102
S Becker
Addiction with Gary
Economic Review 396
American
Employment and
Welch 85
A Scheinkman
Jose
1994
468
Inequality
Boulder
1994
Press
Cycles with Sherwin Rosen and
Economy
An
Economics and Education
Series in
Institute
R
CO
Growth with
Chaim Fershtman and Yoram Weissm 104
1996
and Family Labor
Supply
with Chinhui
Juhn 15
Journal
of
Labor
1997
Trade with
Inequality
Economics 72
and
108
Andrei Shleifer
and Family Labor
53
Supply
Journal
of
Development
with Chinhui
Economics
Juhn 15
Journal
of
1
1997
Labor
1997
Vertical Integration
Klein 87 American
as a SelfEnforcing Contractual
Economic
Review 415
Arrangement
with Benjamin
1997
5
Case5 11 cv 02509 LHK
Unemployment
Nonemployment
Skills and Technology
M Romen
and Paul
on
Perspectives
the United
in
Page6
73
of
87 American
Economic
Growth
145 1999
Economic Review
A Competitive Perspective
Economic Review 184
Bush School
W
with
Craig Riddell
pp 283
Growth
1998
Press
Solutions
with Finis
Welch 88
on
with Gary
S Becker
Explorer
Internet
and Edward Glaeser 89
with Steven
J
Davis 90 American
2000
Change and
Industrial
Consequences
MA MIT
Canada
and Economic
1998
and Economic
Population
States and
the Social Security Crisis and Proposed
Economic Review 142
American
H Topel
with Robert
in General Purpose Technologies
309 ed Elhanan Helpman Cambridge
American
Filed10 07 13
4
1997
Review 295
Wages
and
Document518
the
Demand
of Increasing
for Skill
Economics of
Series in the
Welch in The Causes and
Welch Volume II in the
with Finis
pp 26384 ed
Inequality
Public
Finis
Policy Chicago University
of Chicago
2001
Press
Wage
Differentials
Welch
the 1990s
Is the Glass Half Full or Half
The Causes and Consequences
in
Volume
in
II
in the
Bush School
of Chicago Press
University
on
Software
in
pp 341 64
Inequality
Economics of
the
Public
ed
Finis
Finis
Welch
Policy Chicago
2001
Economic
Perspectives
with Steven
Economy
Increasing
of
Series
Empty with
J Davis
Design PC Operating Systems and Platforms
and Jack MacCrisken
Selected Essays
in Microsoft
pp 361 420 ed
Davis
Antitrust
S Evans
and the
Boston
New
MA Kluwer
2001
Current
Unemployment
Chinhui Juhn 1
The Economics
Papers on Economic
Brookings
of Copyright Fair
and Benjamin Klein 92 American
The Economic Value
M Murphy
Use
in
Economic
of Medical
Gains from Medical Research
and Kevin
Contemplated
Historically
An
Economics 299
Entrepreneurial
Japanese
Review
of
and
2002
A Networked World with
Andres Lerner
2002
Research with Robert
Economic Approach
Chicago University
H Topel
in
pp 41 73 ed
2003
Measuring the
Robert
H Topel
of Chicago Press
Market
with Sam Peltzman
22
Journal
of
2003
ability
and market selection in an infant industry
cotton spinning
Economic
79
Review 205
School Performance and the Youth Labor
Labor
Activity
H Topel
with Robert
Dynamics
industry with Atsushi
354
Ohyama and
evidence
Serguey
from the
Braguinsky
7
2004
6
Case5 11 cv 02509 LHK
Entry Pricing
Document518
and Product Design
Davis and Robert
H Topel
112
an
in
Journal
of
Filed10 07 13
4
Initially
S188
Diminishing Returns The Costs and benefits of Increased
Topel 46
Perspectives
Persuasion
in
in
Biology
Longevity
of
73
with Steven
with Robert
J
H
2004
S108
and Medicine
Market
2004
Monopolized
Economy
Political
Page7
Politics with Andrei Shleifer
94 American
May
Economic Review 435
2004
Black White
Topel 48
The
Differences
Perspectives
Equilibrium
in
in the
Biology
Distribution of
Becker and IvÆn Werning 113
The Market
for Illegal
Grossman 114
Journal
Two
Competition in
Fees
Interchange
Journal
Economic
Goods The
of
Political
of
Economy
Sided Markets
the Market for
38
The
Drugs with
2006
Antitrust
Status
S
with Gary
2005
Economy 282
Political
Case of
H
2005
Income and
Journal
Health with Robert
Value of Improving
S176
and Medicine
Gary
S Becker
and Michael
Economics of Payment Card
with Benjamin Klein Kevin Green and Lacey Place 73
Antitrust
Law
2006
571
The Value
of Health
Economy 871
and Longevity with Robert
H Topel
2006
Social Value and the Speed of
Review 433
Innovation with Robert
Marketplace
Why
Capital
Fertility
a
1 The Journal of
Journal
in
the Whole Foods
Dealing
in
97 American
the Household
Human
Capital
with Isaac Ehrlich
Case with
Robert
Economic
1
Compared
Winter
9
The Journal
H Topel
Competition for Distribution
Intensifies
Law Journal Vol 75 October
Decline
The Market
Papers
Political
to
2007
of
Human
3
2 GCP
2008
March
Robert Tamura
Women
Need
Capital
Loss Analysis
Exclusive
Antitrust
H Topel
of
2007
1 Winter
Critical
S Becker
with Gary
Does Human
Magazine
Journal
2007
Education and Consumption The Effects of Education
the
114
2
the
Baby
The Journal
for College
with Gary
Proceedings
Boom
and Economic
Human
Capital
3
Growth with
Fall 2008
Graduates and the Worldwide
S Becker
229
of
with Benjamin
Klein
2008
and William
Boom
H J Hubbard
in
Curtis
Simon and
Higher Education of
100 American
Economic Review
May 2010
7
Case5 11 cv 02509 LHK
H J Hubbard
William
Boom
the Worldwide
Explaining
Document518
Journal
Higher Education of
in
Human
of
Filed10 07 13
4
Page8
Women
Capital University
of
with Gary
73
S Becker
of Chicago Press
vol
43 203 2010
How
Exclusivity
with Benjamin
Used
is
Klein 77
to Intensify Competition for Distribution Reply to
Antitrust
Law
Maximum Long Run Growth
Achieving
City Proceedings
of
the
Annual
Hole
Jackson
No
Journal
forthcoming
Conference
Zenger
2011
2
the Federal
in
Bank
Reserve
of
Kansas
2011
Selected Working Papers
Gauging
the
Working
War
Economic
Paper
October
NBER
Working
Estimating
the Effect
Unpublished
Working
The
Interaction
Unpublished
Persuasion
The Value
Philipson
On the
Weighing
No 12092
Growth
Costs
Its
with Steven J Davis and Robert
2006
Steve
Levitt and Roland
Fryer
and
Income
with Gary
S Becker
2006
2007
and Indoctrination with Gary Becker
of Life Near
Unpublished
2006
in Population
Paper
the
March
Epidemic with
of the Crack
End
and Terminal
Care with
Gary
S Becker
and Tomas
2007
Economics of
Working
Selected
of
Paper
Paper September
Working
S Becker
2001
In Iraq Versus Containment
H Topel
11th with Gary
Impact of September
Paper
No
Climate
234
Policy
January
with
Gary
S Becker
2010 Revised September
and Robert
H Topel
2010
Comments
Comment on
Reserve
Bank
Comment
Reform
Causes of Changing
of Kansas
City
Comment on
Answers
Social
of Investment
Feldstein
Equality by Robert
pp 175 81 ed
Andrew J
of Chicago Press
Federal
Debate Medicare
and Thomas R
Rettenmaier
2000
Security and Demographic
Uncertainty
ed
2001
Based Social Security Reform
Chicago University
Z Lawrence
1998
Asking the Right Questions in the Medicare Reform
Issues and
Saving Chicago University
Aspects
Earnings
of Chicago Press
John
by Henning
Y Campbell
Bohn
in Risk
and Martin
8
Case5 11 cv 02509 LHK
Comment on High
Filed10 07 13
4
Page9
of
73
R Varian
by Hal
Technology Industries and Market Structure
2001
Reserve Bank of Kansas City
Federal
Popular
Document518
Press Articles
Rap
The Education Gap
The American
Antitrust with Gary
Rethinking
MarchApril 1990 pp 62
Enterprise
S Becker
Wall
26 2001 pp
February
Journal
Street
pA22
Out
Prosperity Will Rise
S Becker
Gary
Wall
Street
Journal
29 2001 pp pA22
October
The Economics
the request
Articles
Ashes with
of the
NFL Team
of
of the National
Ownership
Football
with Robert
League Players
H Topel
Association
report prepared
at
2009
January
About Murphy
Higher Learning
Clearly
March 12 1989 Jobs
the Middle Class Is
Extensive
reference
Unequal
Business
One
pp 1
Section
Long
article
Kleiman Chicago
Carol
The
about
Tribune
Wages
Structure of
with
Murphy
picture of
Why
Means Higher Earning by
to
Pay Widespread
Day
section
in
pp 1
17 1992
by Louis
S Richman
to
Is
Business
on income
piece
York Times August
14 1990
inequality
Rut of Poverty by Sylvia Nasar New York
Anothers
pp 1
Section
May 21 1990 pp 106
Fortune
education
US by Louis Uchitelle New
Long
Rags to Riches
Studys
Times June
Anxious
Murphy’s work on returns
Long
piece
on
the income
inequality
research
Nobels
Pile
November
Up
for Chicago
4 1993
Business
but
Is
Section
the Glory
pp 1
Featured
a photo of five of the
Murphy
and a paragraph about Murphys
This Sin Tax
is
Win Win
Commentary section
brightest stars
by Christopher
refers to
Murphy
Gone
Long
by Sylvia Nasar New York Times
on Chicago School
piece
on
the economics
faculty
of economics
including
research
Farrell Business
Becker
Week
April
11 1994 pp 30
and Grossmans work on
rational
addiction
Growing
and the economics
inequality
Sunday Globe August
details
21 1994 pp
about Murphy and
replaced
the old in
his
of fragmentation
A1 Two page
research
article
part of a series
by David Warsh Boston
with picture and biographical
about
how
the
new
generation
economics
9
Case5 11cv 02509LHK
A Pay
Impact by
Raises
Section
pp 1
Articles
featuring
January
Filed1007 13
4
Page10
of
73
New York Times January 12 1995 Business
minimum wage
of proposed increase in the
Murphy’s comments on the minimum wage appeared in numerous
including the Chicago Tribune
Murphy was
in addition
on
interviewed
26 1995
American Wall
The Undereducated
in
Louis Uchitelle
about consequences
Article
other publications
CNN
Document518
Street
Journal August
19 1996 pp A12
Changes
the rate of returns to education
M Murphy Winner
In Honor of Kevin
Welch 14
Journal
Testimony Reports and Depositions
Canadian
Vehicles
District
15 16
January
2008
M Murphy
Inc
Appliances
v
February
1 2008
Tyco Healthcare
M Murphy
of Kevin
Declaration
Corporation
v
Eastern
Inc
v
the Central District
v
Inc
Inc
Electronics
its
Inc
Trustee
States District
v
Inc
et
al
v
Electronics
its
Inc
Trustee
States District
Initial
Inc
Jaco
Hynix
Hynix
22 2008
States District
Inc
Inc
et
v
Court for the Northern
al
NA
in
District
Court for
the Matter of Sun Microsystems
Inc
Semiconductor
Inc
Hynix
District
v
Inc
et
et
al
Hynix Semiconductor
al
v
Inc
et
All American
DRAM Claims Liquidation
Semiconductor
et
al The United
of California San Francisco
Division
24 2008 in the Matter of Sun Microsystems
Inc et al Consolidated Unisys Corporation v
Jaco
Inc
Electronics
Semiconductor
Bank
Court for the Northern
Submission of Kevin
United States
al Consolidated Unisys Corporation
Semiconductor
Fargo
in the Matter of Allied Orthopedic
April
Hynix
Hynix
Wells
et
7 2008
Electronics
Bank
M Murphy
v
United States District
in the Matter of Novelis
LP The
Semiconductor
Fargo
Hynix Semiconductor
Semiconductor
Trust by
al
Court for the Northern
Hynix Semiconductor
al Edge
et
v
Wells
Deposition of Kevin
LP The
District
M Murphy March
Hynix Semiconductor
Semiconductor
Group
of California Western
Hynix Semiconductor
Trust by
Court for the
District
28 2008
February
Tyco Healthcare
Expert Report of Kevin
al Edge
States District
Motor
Division
M Murphy
Deposition of Kevin
Appliances
February
Anheuser Busch Inc The United
Ohio
of
al
New
the Matter of Allied
in
Group
Court for the Central District of California Western
et
the Matter of
The United
Export Antitrust Litigation
Orthopedic
Inc
in
Maine
of
Expert Report of Kevin
District
by Finis
2000
Last 4 Years
M Murphy
Deposition of Kevin
Medal
of the John Bates Clark
193
Economic Perspectives
of
NA
Inc
Hynix
District
v
Inc
et
al
et
Hynix Semiconductor
al
DRAM Claims Liquidation
Semiconductor
et
al The United
of California San Francisco
M Murphy October
Inc
All American
6 2008 in the
2006
Division
MSA
Adjustment
Proceeding
10
et
Case5 11cv 02509LHK
Filed1007 13
4
M
Murphy October 29 2008
myFICO Consumer Services Inc vs
Expert Report of Kevin
Corporation and
Document518
LLC Experian Information
Solutions LLC and Does I through
Information Services
VantageScore
District
Systems Inc
v News
Court for the
District
v News
FSI Inc
America
America Marketing
In Store
LLC The United
No 07 706645
Services
Deposition of Kevin
Corporation and
VantageScore
21 2008
Equifax
United States
LLC
Court
District
ak
a
News
Inc
Services
in the Matter of Valassis
a k a News
America Incorporated
America Marketing
America Marketing FSI
aa
a
News
LLC and
M Murphy December 12 2008
Services
News
In Store
American Marketing
States Third Circuit Court of Michigan Detroit
myFICO Consumer
Information Services
Division
in the Matter of Fair Issac
Inc vs Equifax Inc Equifax
LLC Experian Information
LLC and Does I through
Solutions
Solutions
X The
Inc TransUnion
United States
LLC
District
Court
of Minnesota
Deposition of Kevin
v News
Rebuttal
X The
Inc
Inc TransUnion
21 2008 in the Matter of Insignia
In Store Inc The United States District
M Murphy November
Group News
District
Solutions
of Minnesota
Communications Inc
Inc
in the Matter of Fair Issac
Equifax
M Murphy November
America Marketing
Expert Report of Kevin
District
73
of
of Minnesota
Expert Report of Kevin
Case
Page11
M Murphy December 15 2008
America Marketing
In Store Inc The
in the Matter of Insignia Systems
United States
District
Court for the
of Minnesota
Expert Report of Kevin
Valassis Communications
Marketing
Group News
Inc
M Murphy December
v News
26 2008
America Incorporated
America FSI Inc
ak
a
News
aa
Division Case
a
the Matter of
News
America Marketing
News America Marketing In Store Services Inc
In Store Services LLC The United States Third Circuit
and
in
ak
a
News
America
FSI LLC
American Marketing
Court of Michigan Detroit
No 07 706645
Final Submission of Kevin
M Murphy
January
16 2009
in
the 2006
MSA Adjustment
Proceeding
Expert Report of Kevin
v
of
M Murphy
January
Corp et al The United
New York Report submitted on behalf
Amerada Hess
of Kevin
Declaration
Inc
v News
District
M Murphy
America Marketing
January
23 2009
in
States District
the Matter of City of
Court for the Southern
of Citgo Petroleum
29 2009
In Store Inc The
in
New
York
District
Corporation
the Matter of Insignia Systems
United States
District
Court for the
of Minnesota
11
Case5 11cv 02509LHK
Document518
M Murphy
Deposition of Kevin
Communications Inc
v News
Group News
FSI Inc
America
America Marketing
Case
In Store
LLC The
Services
v
Services
News
a
a k a News
Inc
America Marketing
LLC and
America Marketing FSI
aa
News
a
News
In Store
American Marketing
Division
M Murphy
Citgos
Center on behalf
States District
all
others
Court for the Eastern
Center on behalf
supply
M Murphy March
of itself and
Deposition of Kevin
States District
RFG
share of total
Expert Report of Kevin
all
Case
Declaration
of Kevin
Declaration
of Kevin
M Murphy
v
District
17 2009
April
v
6 2009
v
April
16 2009
in
April
23 2009
in
of California San Francisco
Expert Report of Kevin
M Murphy May 11 2009
Judicial
District
v
Inc
Meritor
LLC
States District
Meritor
LLC and
States District
v
Division
Inc
Sun Microsystems
a Korean
al The
corporation
United States
Hynix
District
Division
in the Matter of Jim
Corporation
Microsoft
Sun Microsystems
al The United States
et
the Matter of
a California corporation et
State of Mississippi
Expert Report of Professor
the Western
United
the Matter of
Hynix Semiconductor
District
Petroleum Corporation
ZF
The United
of California San Francisco
Court Northern
First
ZF
The United
in the Matter of
Eaton Corporation
Inc
rel
Medical
Inc The
Division
the Matter of
in
United
Division
CR Bard
Eaton Corporation
Semiconductor America
Hinds County
vs
Medical
Inc The
in the Matter of St Francis
situated
Hynix Semiconductor
M Murphy
a California corporation
General ex
CR Bard
No 06 CV 623
a California corporation
Court Northern
vs
No 06 CV 623
M Murphy
Case
the Matter of St Francis
situated
of Missouri Southeastern
District
v
in
York Harbor
of Missouri Southeastern
similarly
M Murphy March
TransmissionCorporation
Court of Delaware
District
New
the Matter of City of
in
New
3 2009
District
and Meritor TransmissionCorporation
Deposition of Kevin
the
similarly
others
Court for the Eastern
Court of Delaware
at
M Murphy March 6 2009
of itself and
Expert Report of Kevin
Meritor
13 2009
February
Amerada Hess Corp et al The United States District Court for the Southern
of New York Report submitted
on behalf of Citgo Petroleum Corporation
regarding
Inc
in the Matter of Valassis
America Incorporated
ak
73
of
No 07 706645
District
Inc
10 2009
February
Page12
United States Third Circuit Court of Michigan Detroit
Expert Report of Kevin
York
Filed1007 13
4
Hood
The Chancery
Attorney
Court of
District
Kevin
M Murphy
Ranger Enterprises
June
12 2009
Inc The
in the Matter of
CITGO
United States District Court for
of Wisconsin
12
Case5 11cv 02509LHK
Expert Report of Kevin
v
Incorporated
District
Trial
Filed1007 13
4
M Murphy June 24 2009
Page13
of
73
in the Matter of Novell
The United
Corporation
Microsoft
Court Northern
States District
of Maryland
M Murphy
Testimony of Kevin
Communications Inc
Group News
v News
July
16 2009
the Matter of Valassis
in
a k a News
America Incorporated
ak
America Marketing
FSI Inc
a News America Marketing FSI LLC and News
In Store Services Inc
a a News American Marketing In Store
America
America Marketing
LLC The United
No 07 706645
Services
Case
Document518
a
States Third Circuit Court of Michigan Detroit
M Murphy August 14 2009
Declaration
of Kevin
Litigation
The United
Declaration
submitted
of
EBay
Antitrust
Seller
District of California
Ebay Inc s motion
of defendant
support
in
in the Matter
Court for the Northern
States District
Division
for
summary
judgment
Expert Report of Kevin
Inc
and
S Jerrold
M Murphy August 21 2009
Kaplan
v
in the Matter of
The
Corporation
Microsoft
Superior
Go Computer
Court for the
State
of California for the City and County of San Francisco
Deposition of Kevin
v
Incorporated
District
M Murphy
Microsoft
16 2009 in the Matter of Novell
The United States District Court Northern
September
Corporation
of Maryland
Deposition of Kevin
M Murphy
September
21 2009
The United
States District
California Deposition in support
of defendant
Antitrust Litigation
Ebay
in the Matter of
Court for the Northern
Ebay Inc s motion
for
Seller
of
District
summary
judgment
Expert Report of Kevin
M Murphy September
Temperature Sales Litigation The United States
Trial
Testimony of Kevin
M Murphy October
and Meritor TransmissionCorporation
Court of Delaware
Declaration
of Kevin
California Declaration
District
1 2009
the Matter of
Motor
Fuel
Court of Kansas
ZF
in the Matter of
Eaton Corporation
The United
LLC
Meritor
States District
No 06 CV 623
Case
Antitrust Litigation
v
29 2009 in
M Murphy October
The United
in
16 2009
States District
further support
in the Matter of
Ebay
Court for the Northern
of defendant
Seller
District
Ebay Inc s motion
for
of
summary
judgment
Expert Report of Kevin
Devices
Intel
Inc
and
Kabushiki
AMD
M Murphy October
International
Kaisha The United
Sales
20 2009
Service
States District
in the Matter of
LTD v
Intel
Court for the
Advanced Micro
Corporation and
District
of
Delaware
13
Case5 11cv 02509LHK
M Murphy
Deposition of Kevin
S Jerrold
and
Document518
v
Kaplan
Filed1007 13
4
24 2009
October
Court
Superior
73
of
Go Computer Inc
the Matter of
in
The
Corporation
Microsoft
Page14
the State of
for
California for the City and County of San Francisco
M Murphy
Deposition of Kevin
Expert Report of Kevin
Discount
New
Court for the Eastern District of
Valassis Communications
Group News
Inc
Card
States District
ak
ak
America FSI Inc
a
News
a
in
the Matter of
News
a
News
America
FSI LLC
America Marketing
American Marketing
Court of Michigan Detroit
No 07 706645
Testimony of Kevin
and
21 2009
America Incorporated
aa
Inc
The United
M Murphy December
v News
News America Marketing In Store Services Inc
In Store Services LLC The United States Third Circuit
Trial
Fuel
the Matter of Payment
in
Antitrust Litigation
and
Division Case
Motor
Court of Kansas
York
Supplemental Expert Report of Kevin
Marketing
the Matter of
in
States District
M Murphy December 14 2009
Fee and Merchant
Interchange
26 2009
October
The United
Temperature Sales Litigation
S Jerrold
M Murphy
Kaplan
v
11 2010
January
The
Corporation
Microsoft
in
the Matter of
Go
Computer
Court for the
Superior
State
of California for the City and County of San Francisco
Supplemental Rebuttal
America Marketing
LLC
News
and
Marketing
Detroit
Group News
v
Services
Division Case
LLC
Communications Inc
v News
Group News
FSI Inc
America
America Marketing
Services
Case
In Store
LLC The
a k a News
Inc
Services
The United
aa
a
the
in
a
News
America Marketing
News
States Third Circuit
FSI
American
Court of Michigan
26 2010
January
in
the Matter of Valassis
America Incorporated
ak
News
a
Services
Inc
a k a News
America Marketing
America Marketing FSI
aa
a
News
LLC and
American Marketing
United States Third Circuit Court of Michigan Detroit
News
In Store
Division
No 07 706645
Declaration
of Kevin
Antitrust Cases I and
the
In Store
ak
No 07 706645
M Murphy
Deposition of Kevin
14 2010
January
News America Incorporated
America FSI Inc
America Marketing
In Store
M Murphy
Expert Report of Kevin
Matter of Valassis Communications Inc
County
Declaration
M Murphy
II The
28 2010
in
the Matter of Automobile
Court of the State of California for
of San Francisco
of Kevin
M Murphy
Determination of Interim
al The United
Court for the Eastern
Court Southern
M Murphy
Fee and Merchant
District
2 2010
April
April
Discount
of
in
the Matter of the Application
Fees for The Cromwell
License
States District
Deposition of Kevin
Interchange
January
United States Superior
New
13 14
District
2010
in
of
Group Inc
New
et
York
the Matter of Payment
Antitrust Litigation
for the
and Affiliates
The United
Card
States District
York
14
Case5 11cv 02509LHK
Document518
M Murphy June 1 2010
Supplemental Expert Report of Kevin
Inc
Insignia Systems
2010
The United
v
News America Marketing
M Murphy June 21 2010
Comcast Corporation General
to
In Store
Assign Licenses or Transfer
Page15
73
of
in the Matter of
Inc
8
June
corrected
Court for the District of Minnesota
States District
Expert Report of Kevin
Filed1007 13
4
in the Matter of Applications
NBC Universal
Company and
Electric
of Licensees
Control
Inc
of
for Consent
Communications
Federal
Commission
Supplement to Expert Report of Kevin
Payment
Card Interchange
M Murphy June 24 2010
Fee and Merchant
Discount
Court for the Eastern District of
States District
Inc
Insignia Systems
District
Court for the
Deposition of Kevin
v
The United
New York
M Murphy
Second Supplemental Expert Report of Kevin
in the Matter of
Antitrust Litigation
July
6 2010
in
the Matter of
News America Marketing In Store Inc The United
District
States
of Minnesota
M Murphy
8 2010
July
in the Matter of Insignia Systems
News America Marketing In Store Inc The United
Inc
Court for the
States District
v
District
of Minnesota
M Murphy 28 2010
by Thomas W Corbett Jr
Expert Report of Kevin
Pennsylvania
July
of Pennsylvania
Commonwealth
Court of Pennsylvania
Response of Kevin
19 2010
M Murphy
of Licensees
NBC Universal Inc
Federal
Louis et al
v
of
al
et
in
the
MD 2004
Israel
and Michael Katz August
Comcast Corporation General
Electric
Control
or Transfer
Commission
M Murphy September
14 2010 in
the Matter of City of
Co et al The Circuit Court of
American Tobacco
of
St
St Louis
the City of
of Missouri
State
Deposition of Kevin
et
212
Inc
Products
for Consent to Assign Licenses
Communications
Expert Report of Kevin
No
Commonwealth
General of the
as Attorney
Pharmaceutical
Reply Report of Mark
to
in the Matter of Applications
Company and
the Matter of
his capacity
in
v TAP
Commonwealth
in
al
v
M Murphy
American Tobacco
Co
et
24 2010
September
al The
in
St Louis
the Matter of City of
Court of the City of St Louis State of
Circuit
Missouri
Supplemental Expert Report of Kevin
Commonwealth
al
in the
Commonwealth
Expert Report of Kevin
Hampshire
v
by Thomas
of Pennsylvania
General of the Commonwealth
M Murphy September
W Corbett
of Pennsylvania
v TAP
Court of Pennsylvania
M Murphy October
Hess Corporation
et
al The
No
1 2010
State of
30 2010
Jr in his
capacity
Pharmaceutical
212
in the Matter of
Inc
MD 2004
in the Matter of State
New
as Attorney
Products
of
Hampshire Superior
New
Court
15
et
Case5 11cv 02509LHK
Expert Report of Kevin
between
Cordis
Conflict
Prevention
M Murphy October
Corporation and Abbott
Cordis
Conflict
M Murphy
Prevention
of Kevin
M Murphy November
v BP
Court for the Northern
CPR
Expert Report of Kevin
8 2010
v BP
Inc
Analysis of Kevin
Dodd Frank
Comments
of Kevin
L Katz
and Michael
of
International
RWJ
States
Division
in the Matter of
RWJ
America Inc The United
19 2010
States
Division
al
in the Matter of Craft et
v
a corporation and Philip Morris Incorporated a
M Murphy
Bank
Judicial
District City of
St Louis
to Guide Interpretation
of Debit Interchange
or Transfer
on
the
November 10 2010 Report
November 24 2010
Control
v NFL
of Provisions of the
Fees November 23 2010
of America Corporation
M Murphy
Expert Report of Kevin
in
of Licensees
the Matter of Applications
NBC Universal Inc
Federal
Lockout
29 2010
Loans The
Drs Mark
of
Israel
Comcast
for Consent to Assign
Communications
M Murphy November
Lockout Insurance
of
Commission
in the Matter of Reggie
United States
District
White
Court
of Minnesota
Deposition of Kevin
v NFL
District
CPR
the Matter of
of Illinois Eastern
District
Corporation General Electric Company and
al
in
15 2010
Products North
Act Regarding Regulation
submission on behalf
al
the Matter of the
00200406 02
Economic
District
for
Institute
America Inc The United
corporation Missouri Circuit Court Twenty Second
et
in
Vascular
12 2010
M Murphy November
Philip Morris Companies
Licenses
International
of Illinois Eastern
District
Court for the Northern
No
for
Institute
the Matter of the Arbitration
in
Products North
M Murphy November
Management Company Inc
Case
International
Resolution
Prevention
Expert Report of Kevin
District
7 2010
Vascular
M Murphy November
Management Company Inc
District
73
of
in the Matter of the Arbitration
CPR
Cordis Corporation and Abbott
for Conflict
Declaration
Page16
Resolution
Testimony of Kevin
Institute
4 2010
Vascular
October
Corporation and Abbott
Arbitration between
Filed1007 13
4
Resolution
Deposition of Kevin
between
Trial
Document518
Lockout
M Murphy December 3 2010
Insurance
Lockout
in the Matter of Reggie
Loans The
United States
District
White
et
Court
of Minnesota
Deposition of Kevin
Company Inc
the Northern
v BP
District
M Murphy December 13 2010
Products North
in the Matter of
America Inc The United
of Illinois Eastern
RWJ
Management
States District
Court for
Division
16
Case5 11cv 02509LHK
Deposition of Kevin
Document518
M Murphy
Inc
Philip Morris Companies
Filed1007 13
4
17 18
January
2011
No
73
of
v
al
the Matter of Craft et
a corporation and Philip Morris Incorporated a
corporation Missouri Circuit Court Twenty Second
Case
in
Page17
St Louis
District City of
Judicial
00200406 02
Report of Kevin
M Murphy
16
Corporation on February
Consumer
15 2011
February
submitted
by
TCF Financial
2011 to the Subcommittee on Financial
Credit of the Committee
on
Financial Services of the
and
Institutions
US House
of
Representatives
Declaration
Ben
of Kevin
S Bernanke
M Murphy March
Janet
L Yellen
Kevin
2 2011
in
M Warsh
Elizabeth
and Sarah Bloom Raskin the Board of Governors
capacities and John
official
Expert Report of Kevin
LTD
M Murphy
Declaration
Court Northern
of Kevin
Board on behalf
Entertainment Consoles
Judicial
June
District
System in
Reserve
in his official
the Matter of Datel
Microsoft
v
Corporation
their
capacity
Holdings
The United
14 2011
v
filed
with
the National
Labor Relations
Players Association
in
the Matter of Datel
Corporation
Microsoft
LTD
Holdings
The United
States
of California
M Murphy
July
1 2011
in
the Matter of Certain
Gaming and
and Components Thereof The United States
Related Software
Trade Commission
Expert Report of Kevin
Inc
Basketball
M Murphy
Expert Report of Kevin
Airlines
v
M Murphy May 26 2011
Court Northern
International
of the Currency
in
Bank
National
A Duke Daniel K Tarullo
of the Federal
11 2011
TCF
of California
Development Inc
and Datel Design
District
District
of the National
Deposition of Kevin
April
Development Inc
and Datel Design
States District
Walsh Comptroller
the Matter of
v
Sabre
M Murphy August 17 2011
Inc
et
al The
Judicial
District
in the Matter of American
of Tarrant
County Texas
67th
District
Expert Report of Kevin
M Murphy August 19 2011
Temperature Sales Litigation
The United
in the Matter of
States District
Court for the
Motor
Fuel
of
District
Kansas
Deposition of Kevin
M Murphy
September
6 2011
and Entertainment Consoles Related Software
States International
v
Intel
the Matter of Certain
Gaming
Trade Commission
Expert Report of Kevin
York
in
and Components Thereof The United
M Murphy September
Corporation
The United
9 2011
States District
in the Matter of State of
Court for the
District
of
New
Delaware
17
Case5 11cv 02509LHK
Deposition of Kevin
Document518
M Murphy
Temperature Sales Litigation
14 2011
September
The United
Filed1007 13
4
Page18
in the Matter of
Court for the
States District
73
of
Motor
Fuel
District
of
Kansas
Direct Testimony
M Murphy
of Kevin
Gaming and Entertainment Consoles
York
v
Report of Kevin
NRLC
and
M Murphy
M Murphy October
railroad
810
October
The United
Corporation
Intel
2011
States District
10 2011
employees National
Hearing Testimony of Kevin
NRLC
and
Thereof The
in connection
railroad
before
M Murphy October
employees National
of
District
A13569 A13570
Emergency Board
13 2011
in
No
243
connection with dispute
Mediation Board Case
A13570 A13572 A13573 A13574 A13575 A13592
Delaware
between
with dispute
Nos
New
of
in the Matter of State
Court for the
Mediation Board Case
A13572 A13573 A13574 A13575 A13592
between
in the Matter of Certain
Software and Components
Trade Commission
United States International
Deposition of Kevin
27 2011
September
Related
before
Nos
Emergency
A13569
No
Board
243
Expert Report of Kevin
Hampshire
Declaration
v
M Murphy October
Hess Corporation
of Kevin
et
al The
17 2011
M Murphy December 1 2011
Temperature Sales Litigation
The United
Hampshire Superior
the Matter of
States District
New
in the Matter of State of
New
State of
Court
Motor Fuel
Court for the
District
of
Kansas
Expert Report of Kevin
M Murphy December 5 2011
v
the Matter of Retractable
in
Technologies
Inc and Thomas Shaw
States District
Court for the Eastern District of Texas Marshall
Trial
Testimony of Kevin
Incorporated
District
Trial
v
Microsoft
Corporation
The United
in
Company The
United
Division
the Matter of Novell
States District
Court Northern
of Maryland
Testimony of Kevin
M Murphy December
v BP
Court for the Northern
M Murphy
Technologies
Inc and Thomas Shaw
The United
States District
Court for the Eastern
M Murphy
January
and Entertainment Consoles Related Software
States International
the Matter of
RWJ
America Inc The United
of Illinois Eastern
District
Retractable
Testimony of Kevin
29 2011 in
Products North
Supplemental Expert Report of Kevin
Trial
and
M Murphy December 7 8 2011
Management Company Inc
District
Becton Dickinson
January
v
States
Division
15 2012
in
the Matter of
Becton Dickinson and Company
District
18 2012
of Texas Marshall
in
Division
the Matter of Certain
Gaming
and Components Thereof The United
Trade Commission
18
Case5 11cv 02509LHK
Document518
M Murphy February
Supplemental Expert Report of Kevin
State
New
of
v
Hampshire
Filed1007 13
4
Page19
23 2012
the Matter of
in
New
Hess Corporation et al The State of
73
of
Hampshire
Court
Superior
M Murphy March
Affidavit of Kevin
12 2012
Fruth Individually and on Behalf
Michael
of Sharon
in the Matter
of Others Similarly
Incorporated The United States Circuit Court Third
Situated
Price and
vs
Philip Morris
Court Madison
Judicial
County
Illinois
of Kevin
Declaration
M Murphy May 3 2012
v
in
the Matter
Inc and Thomas Shaw
States District
Court for the Eastern District of Texas Marshall
Comments
of Kevin
M Murphy
of the Commissions
Revision
DIRECTV Group Inc
Authority
Assignees
Subsidiaries
Sabre
Judicial
Inc
Sabre Holdings
of Kevin
Capital Partners
LLC
July
Corp and
July
The United
20 2012
67th
21 2012
July
v
Motorola Inc The United
Washington
at
the Matter of American Airlines
Ltd The
International
Court
23 2012
Corporation
Commission
in the Matter of Kirk Dahl
in
v
United States
in
District
v
District
Bain
of Massachusetts
the Matter of Kirk Dahl
Court
24 2012
July
and or
Corporation and
District
Judicial
States District
M Murphy
in
Sabre Travel
States District
M Murphy
The United
Expert Report of Kevin
Corporation and the
Time Warner Cable Inc
to
Communications
Federal
M Murphy
LLC
Expert Report of Kevin
Capital Partners
al
County Texas
Tarrant
District
Declaration
et
M Murphy
Expert Report of Kevin
the Matter of
for Consent to the Assignment
Communications
Debtorsin Possession Assignors
Subsidiaries
in
Transferors and Liberty Media Corporation Transferee for
of Control of Licenses Adelphia
Transfer
United
Division
2012
News
Program Access Rules
Company The
and
LLC June 22
DirecTV
Control Applications
Transfer
to
of
Becton Dickinson
of Retractable
Technologies
v
Bain
of Massachusetts
the Matter of Microsoft
Seattle
Deposition of Kevin
M Murphy
August
States District Court Western
22 2012
Corporation
v
Motorola Inc The United
Washington
at
in
District of
the Matter of Microsoft
Seattle
Economic
Evidence
Analysis of the Impact
from San
Diego
Commissions Program
Inc
Transferors
August
Control Applications
Federal
31 2012
News
submitted
in
District of
of Carrying an
the Matter of Revision
Corporation and the
and or
RSN
of the
DIRECTV Group
Corporation Transferee for Authority
Transfer
to Transfer
of Control
of
Corporation and Subsidiaries Debtorsin
Assignors to Time Warner
Communications
Court Western
Subscribership
for Consent to the Assignment
Licenses Adelphia Communications
Possession
on DIRECTVs
Access Rules
and Liberty Media
States District
Cable Inc
Subsidiaries
Assignees
et
al
Commission
19
Case5 11cv 02509LHK
Expert Report of Kevin
Brown
et al v
Document518
M Murphy September
The American Tobacco
California for the County of San
Deposition of Kevin
et
of San
Deposition of Kevin
Inc
Inc
v Sabre
State
of Texas
Court for the
Sabre Holdings
et
73
of
in the Matter of Willard
al Superior
in
al Superior
September
Corp and
District
24 2012
Court for the
R
State
of
the Matter of Willard
R Brown
Court for the State of California
Properties and
in
the Matter of American Airlines
Sabre Travel International
of Tarrant
M Murphy October
Innovative
District
et
14 2012
September
Co Inc
M Murphy
for the Judicial
v 3M
7 2102
Page20
Diego
Expert Report of Kevin
Corporation
Co Inc
Filed1007 13
Diego
M Murphy
al v The American Tobacco
for the County
4
LTD
for the
County
10 2102
in the Matter of Avery Dennison
3M Company
The United
States District
of Minnesota
20
Case5 11cv 02509LHK
Appendix
B Materials Relied
Document518
Filed1007 13
4
Page21
of
73
Upon
Court Documents
Motion and Motion for Class Certification
Plaintiffs Notice
of
Support
Consolidated Amended Complaint
in
Re High Tech
and
Memorandum
Law
1 2012
October
Employee Antitrust
Litigation
of
in
September
2 2011
E Leamer PhD
Expert Report of Edward
1 2012
October
Leamer Backup
Plaintiffs First Set of Requests
M
Declaration
of Tina
Declaration
of
of
Documents
October
3 2011
of Chris Galy
Declaration
for Production
in Support of Opposition to Class Certification
Evangelista
Danny McKell
in Support of Defendants
Opposition
to
Plaintiffs
Motion
for
Class Certification
Declaration
Plaintiffs
Declaration
of
Donna Morris
of
Adobe Systems Inc
in Support of
Defendants Opposition
to
Motion for Class Certification
of Frank
Wagner
in
Support of Defendants Opposition to Plaintiffs Motion for
Class Certification
Declaration
Plaintiffs
Declaration
of Jeff Vijungco
of
Adobe Systems Inc
in
Support of Defendants Opposition to
Motion for Class Certification
of Lori
McAdams
Defendants Opposition
in Support of
to Plaintiffs
Motion
for
Class Certification
Mason
Declaration
of
Stubblefield
Declaration
of Michelle
Maupin
in
Support of Defendants Opposition to Plaintiffs Motion for
Class Certification
Declaration
of Steven
Burmeister
in
Support of Defendants Opposition to Plaintiffs Motion for
Class Certification
Declaration
of
Rosemary
Opposition to Plaintiffs Motion
Deposition of Lori
Keiper of Adobe Systems Inc in Support of Defendants
Arriada
for Class Certification
McAdams and
Deposition of Arnnon Geshuri and Exhibits
Deposition of Donna Morris and Exhibits
Deposition of Jeffrey Vijungco
3 2012
October 5 2012
August
Deposition of Daniel Stover and Exhibits
Deposition of Mark Fichtner and Exhibits
Hariharan
and
24 2012
October
October
October
Exhibits
Deposition of Edward Leamer and Exhibits
23 2012
October
Exhibits
Deposition of Brandon Marshall and Exhibits
Deposition of Siddharth
2 2012
August
and Exhibits
Devine and
October
August 21 ,2012
Deposition of Mark Bentley and Exhibits
Deposition of Michael
17 2012
August
Deposition of Danielle Lambert and Exhibits
Deposition of James Morris and Exhibits
2 2012
August
Exhibits
22 2012
29 2012
15 2012
October
October
12 2012
26 2012
Case5 11cv 02509LHK
Document518
Filed1007 13
4
Deposition of Jack Gilmore and Exhibits
June
June
of
73
28 2012
Deposition of Denise Miller and Exhibits
Page22
28 2012
June
Deposition of Shawna
Dougherty
June
Deposition of John Schirm and Exhibits
Yu and
Deposition of Matthew
Deposition of Shiloh
Exhibits
Howard and
Kuz and
26 2012
29 2012
June
July
17 2012
Exhibits
26 2012
June
Exhibits
Deposition of
and
AmberGay Remaley
Exhibits
Plaintiff
Michael
March
21 2012
June 21 2012
June
and Exhibits
Cruzat and Exhibits
26 2012
June
Deposition of Mary Kathleen Galle and Exhibits
Deposition of Eleterio
26 2012
June
Deposition of Robert DeMartini and Exhibits
del Torro
17 2012
July
Deposition of Michelle Deneau and Exhibits
Deposition of Rebecca
12 2012
and Exhibits
Deposition of Mai Tran and Exhibits
Deposition of Jaime
27 2012
July
Deposition of Steven Burmeister and Exhibits
June
June
21 2012
22 2012
Answers and Objections
27 2012
Plaintiff
March
Plaintiff
Mark
Devine’s
Answers and Objections
Fichtner
to
to Defendants
Defendants
Set of Interrogatories
First
First
Set of Interrogatories
28 2012
Siddharth Hariharan’s
Plaintiff
Brandon
March
to
Defendants
First
Set of
27 2012
March
Interrogatories
Answers and Objections
27 2012
Plaintiff
March
Final
Answers and Objections
Marshall’s
to
Defendants
Daniel Stover’s Answers and Objections to Defendants
Interviews
in United
Final
States
of
Judgment in United
States
v
Adobe Systems Inc
of
America
America
v
23 2012
Jeff
August
23 2012
Donna Morris Adobe
27 2012
30 2012
August
31 2012
August
30 2012
July
25 2012
19 2012
26 2012
September
Vijungco Adobe
Interview
August
June
Set of Interrogatories
Interview
Interview
Interview
6 2012
30 2012
August
16 2012
with Steve Burmeister Apple
with Seth Williams Google
Interview
Interview
August
with Mark Bentley Apple
Interview
with Frank
with Christina
Wagner Google
Dickenson
with Danny McKell
Interview
Interview
Interview
with Chris Galy
with Michelle
with Laurie
Intel
Intel
with Mason Stubbenfeld
Intuit
Intuit
Maupin
McAdams
Lucasfilm
Pixar
et al
Lucasfilm Ltd
Conducted by Kevin Murphy
August
July
First
Set of Interrogatories
28 2012
Judgment
Proposed
July
First
March
17 2011
May 9 2011
Case5 11cv 02509LHK
Academic
Albert
Document518
Filed1007 13
4
Page23
of
73
Papers
Rees The
Role of Fairness
in
Wage Determination 11
Economics 243
of Labor
Journal
1993
Mas Pay
Alexandre
D
Angrist Joshua
Jersey Princeton
and Jrn
Vol
M
P
Ann
Science
Working Paper
J
of
Journal
of
New
Chapter 8.2
No
Bargaining with
Game Theory Aumann Robert
BV Chapter
J Borjas
and George
Borjas George
Quarterly
Mostly Harmless Econometrics
Cramton and Raymond J Deneckere
Peter
Information Handbook
Consequences
NBER
Pischke
Steffen
3 Amsterdam Elsevier
Bartel
121
Press 2009
University
Ausubel Lawrence
Incomplete
Performance
Reference Points and Police
2006
Economics 783
50
NBER
Sergiu Hart
eds
2002
Middle Age Job Mobility
161
J and
Its
Determinants
and
Working Paper Series January 1977
Job Mobility and Earnings Over the Life
Cycle Working
paper
No
233
Working Paper Series February 1978
Davidson
University
Press Inc
E Leamer
Edward
G MacKinnon Econometric
and James
Russell
Theory and Methods
Oxford
2004
Take
Lets
the
Con Out
of
Econometrics 73 The American Economic
Review 1 1983
Freeman Richard
Gary Becker
B and
Nobel Lecture
Economy 385
Political
James
Greene William
June
L Medoff
What
The Economic
Do
Way
of
Unions
Do
Looking
at
New York
Basic
Behavior 101
Books 1984
Journal
of
1993
H Econometric
Analysis
6th Edition Chapter 9.3.3
New
Jersey Pearson
Hall 2008
Prentice
Grossman
Academic
Sanford
Press revised
Hirsch Barry
Competition
Honoree
Individual
J and
T
Motty
Perry
February
Sequential
Bargaining under AsymmetricInformation
2 1986
Sluggish Institutions in a Dynamic
Coexist
Journal
Andre I and David
of
Economic
E Terpstra
and Procedural Equity
to
World Can
Perspectives vol
The
Relative
Unions and Industrial
22 1
Importance
Winter
2008
of External
Internal
Benefits
Review
Pay Satisfaction Compensation
November December 2003
Joseph Stiglitz
Economic
Robert
H Topel
Quarterly
Information and the Change in the Paradigm in
Review 460
Journal
and Michael
of
William Samuelson
Economics 92 American
2002
P Ward
Job Mobility and the Careers of
Young
Men
107 The
Economics 2 1992
Bargaining Under AsymmetricInformation Econometrica
52 1984
Websites
http online wsj comarticle SB10001424052970203750404577173031991814896
http online wsj comarticle SB124269038041932531
http techcrunch com2007 1121 facebook
html
html
stealinggooglers at an alarming
rate
http www aeaweb org honors awards clark medal php
http www dailytechcomGoogle FindsThat PerksCant KeepSomeEmployees FromLe
aving article11794 htm
Case5 11cv 02509LHK
Document518
Bates Documents
76550DOC000014
231APPLE04166
76583DOC001487
Other
Pixar Data
Pixar revenues
2005
2011 xlsx
4
Filed1007 13
Page24
of
73
Case5 11cv 02509 LHK Document5184
Filed1007 13
Page25
of
73
Appendix1A
Analysis of Hires from Other Defendants
AllSalaried Employee Class
Panel
A 2001 2012
Last Previous
Company
Company
within 1 year
Adobe
Apple
Lucasfilm
0
6
0
2
0
Pixar
3
8
6
1
2
222
218
54
293
98
37
Last Previous
Company
Percentage
12
Hiring
Google
Intel
Intuit
Lucasfilm
Pixar
Other
Total
of
Row
Adobe
Apple
Google
Intel
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
All
Defendants
Panel
35
1,369
0.00
0.44
0.00
0.15
0.00
1,367
0.22
0.59
0.44
0.07
0.15
0.88
91,971
0.24
0.24
0.06
0.32
0.11
0.04
91,014
0.73
0.04
B 2001 2004
Company
within 1 year
Adobe
Apple
Lucasfilm
0
1
0
1
0
Pixar
0
4
0
0
1
34
45
0
34
15
6
Last Previous
Company
Adobe
Apple
0
5
Percentage
3
Hiring
1,351
1,335
10
Google
Intel
Intuit
Lucasfilm
of
Row
Total
Other
Total
Adobe
Apple
Google
Intel
402
407
0.00
0.25
0.00
0.25
0.00
431
Pixar
439
0.00
0.91
0.00
0.00
0.23
0.68
23,181
0.15
0.19
0.00
0.15
0.06
0.03
Other
Total
Adobe
Apple
Google
Intel
788
799
0.00
0.63
0.00
0.13
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
All
Defendants
Panel
Hiring
3
5
23,042
0.74
0.02
C 2005 2009
Company
Google
Intel
Intuit
within 1 year
Lucasfilm
Percentage
Pixar
of
Row
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
1
0
0
5
1
3
5
1
1
6
104
97
27
167
44
17
Last Previous
Company
Adobe
Apple
Google
0
0
Pixar
All
Defendants
Panel
Hiring
0.63
674
0.15
0.45
0.74
0.15
0.15
0.89
44,069
0.24
0.22
0.06
0.38
0.10
0.04
Other
Total
Adobe
Apple
Google
Intel
161
163
0.00
0.00
0.00
0.00
657
18
0.00
43,595
0.04
D 2010 2012
Company
Intel
Intuit
within 1 year
Lucasfilm
Percentage
Pixar
of
Row
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
Defendants
Note
Source
0
2
2
Pixar
All
0
0
This analysis
Dr
1
1
0
0
3
84
76
27
92
39
14
excludes hires indicated as acquisitions
Leamer’s employee
data
and
hires
showing the same defendant
247
12
company
24,377
as their
0.00
1.23
254
0.79
0.39
0.39
0.00
0.00
1.18
24,721
0.34
0.31
0.11
0.37
0.16
0.06
immediate previous employer within one year of the hiring
0.05
Case5 11cv 02509 LHK Document5184
Filed1007 13
Page26
of
73
Appendix 1B
Analysis of Separations Going to Other Defendants
AllSalaried Employee Class
Panel
A 2001 2012
Next
Separation
Company
Adobe
Apple
Google
Company
Intel
within 1 year
Intuit
Percentage of
Lucasfilm
Pixar
Other
Total
Row
Total
Adobe
Apple
Google
Intel
0.00
0.59
0.98
0.07
0.00
0.00
1.46
0.80
0.27
0.00
0.93
0.17
0.45
0.46
0.05
0.10
0.02
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
9
15
1
0
Pixar
0
11
6
2
0
7
122
326
336
35
74
15
All
Defendants
Panel
12
1,490
1,527
726
31
72,287
752
73,226
0.79
0.04
B 2001 2004
Next
Separation
Company
Adobe
Apple
Google
Company
Intel
within 1 year
Intuit
Percentage of
Row
Total
Pixar
Other
Total
Adobe
Apple
Google
Intel
4
Lucasfilm
580
589
0.00
0.51
0.34
0.00
0.00
229
235
0.00
0.85
0.43
0.00
0.00
1.28
0.11
0.22
0.09
0.01
0.09
0.02
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
3
2
0
0
Pixar
0
2
1
0
0
3
28
55
24
3
22
5
Adobe
Apple
Google
All
Defendants
Panel
0.68
9
25,399
Pixar
Other
Total
Adobe
Apple
Google
Intel
5
655
669
0.00
0.45
0.75
0.15
0.00
340
0.00
1.18
0.88
0.59
0.00
35,375
0.20
0.42
0.51
0.05
0.11
0.02
Pixar
Other
Total
Adobe
Apple
Google
Intel
3
255
269
0.00
1.12
2.97
0.00
0.00
177
0.00
2.82
1.13
0.00
0.00
1.13
6
11,513
0.20
1.01
1.10
0.13
0.11
0.02
0.04
0.59
16
25,545
C 2005 2009
Next
Separation
Company
Company
Intel
within 1 year
Intuit
Percentage of
Lucasfilm
Row
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
1
0
0
4
3
2
0
2
151
182
17
39
8
Adobe
Apple
Google
Defendants
Panel
5
70
Pixar
All
3
329
35,858
0.75
0.04
D 2010 2012
Next
Separation
Company
Company
Intel
within 1 year
Intuit
Percentage of
Lucasfilm
Row
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
Source
0
0
0
Defendants
Note
8
This analysis
5
2
0
0
2
24
Pixar
All
3
120
130
15
13
2
excludes separations
Dr Leamer’s
employee
data
that appear as immediately
rehired by the
same defendant
168
company
11,823
within one year
1.12
0.05
Case5 11 cv 02509 LHK
Document518
4
Filed10 07 13
Page27
of
73
Appendix 1C
Analysis of Hires from Other
DNCC Defendants
AllSalaried Employee Class
Panel
A 2001 2012
Last Previous
Company
Hiring
DNCC Defendant
Company
within 1 year
Non DNCC Defendant
Percentage
of
Row
Total
Non
DNCCDefendant
Total
DNCC Defendant
1,369
1.17
98.83
1,367
1.54
98.46
91,971
0.79
99.21
Adobe
Apple
Google
Intel
Intuit
16
Lucasfilm
21
Pixar
All
1,353
Defendants
Panel
91,246
B 2001 2004
Last Previous
Company
within 1 year
Percentage
of
Row
Total
DNCC Defendant
Non DNCC Defendant
Total
DNCC Defendant
4
Company
Hiring
1,346
725
403
407
0.98
99.02
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
7
All
Defendants
Panel
432
439
1.59
98.41
110
Pixar
23,071
23,181
0.47
99.53
C 2005 2009
Last Previous
Company
within 1 year
Percentage
of
Row
Total
DNCC Defendant
Non DNCC Defendant
Total
DNCC Defendant
Lucasfilm
10
789
799
1.25
98.75
Pixar
10
664
674
1.48
98.52
346
43,723
44,069
0.79
99.21
Company
Hiring
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
All
Defendants
Panel
D 2010 2012
Last Previous
Company
within 1 year
Percentage
of
Row
Total
DNCC Defendant
Non DNCC Defendant
Total
DNCC Defendant
Lucasfilm
2
161
163
1.23
98.77
Pixar
4
250
254
1.57
98.43
269
24,452
24,721
1.09
98.91
Company
Hiring
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
All
Defendants
Notes
This analysis
excludes hires indicated as acquisitions
and
hires
showing the same defendant company as
year of the hiring
Adobe
Apple
Google
had DNCC agreements with Adobe Google
allegedly
Lucasfilm
Source
had DNCC agreements with Apple Google and Pixar
had DNCC agreements with Apple and Google
allegedly
Pixar allegedly
Dr
Intel Intuit Lucasfilm
had DNCC agreements with Apple Intel and Intuit
allegedly
Intel allegedly
Intuit
had a DNCC agreement with Apple
allegedly
allegedly
had
had DNCC agreements with Apple and Pixar
DNCC
agreements with Apple Intel and Lucasfilm
Leamer’s employee
data
and Pixar
their
immediate previous employer within one
Case5 11 cv 02509 LHK
Document518
4
Filed10 07 13
Page28
of
73
Appendix 1D
Analysis of Separations Going to Other
DNCC Defendants
AllSalaried Employee Class
Panel
A 2001 2012
Next
Separation
Company
Company
DNCC Defendant
Non
within 1 year
Percentage of
DNCCDefendant
Total
DNCC Defendant
Row
Total
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
21
Pixar
20
All
Defendants
Panel
1,506
712
1.38
72,514
73,226
97.34
0.97
752
98.62
2.66
1,527
732
99.03
B 2001 2004
Next
Separation
Company
Company
Non
within 1 year
Percentage of
Row
Total
DNCCDefendant
Total
DNCC Defendant
582
DNCC Defendant
589
1.19
98.81
235
2.13
97.87
0.45
99.55
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
7
Lucasfilm
5
Pixar
All
Defendants
Panel
230
116
25,429
25,545
C 2005 2009
Next
Separation
Company
Company
DNCC Defendant
Non
within 1 year
Percentage of
DNCCDefendant
Total
DNCC Defendant
Row
Total
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
8
661
669
1.20
98.80
Pixar
8
332
340
2.35
97.65
0.98
99.02
All
Defendants
Panel
350
35,508
35,858
D 2010 2012
Next
Separation
Company
Company
Non
within 1 year
Percentage of
Row
Total
DNCCDefendant
Total
DNCC Defendant
263
DNCC Defendant
269
2.23
97.77
177
3.95
96.05
2.08
97.92
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
6
Lucasfilm
7
Pixar
All
Defendants
170
246
11,577
11,823
Notes
This analysis
Adobe
Apple
excludes separations
had DNCC agreements with Apple Intel and Intuit
allegedly
Intel allegedly
Intuit allegedly
had DNCC agreements with Apple Google and Pixar
had DNCC agreements with Apple
Lucasfilm allegedly
Pixar
allegedly
Source
Dr
rehired by the
same
had DNCC agreements with Adobe Google Intel Intuit Lucasfilm
allegedly
Google
that appear as immediately
defendant
had a DNCC agreement with Apple
allegedly
had
and Google
had DNCC agreements with Apple
DNCC
agreements with Apple
Leamer’s employee
data
and Pixar
Intel
and Lucasfilm
and Pixar
company
within one year
Case5 11cv 02509 LHK Document5184
Filed1007 13
Page29
of
73
Appendix2A
Analysis of Hires from Other Defendants
Technical Creative and
Panel
Class
A 2001 2012
Last Previous
Company
Company
within 1 year
Adobe
Apple
Lucasfilm
0
5
0
0
0
Pixar
2
7
3
1
2
159
150
29
191
59
24
Last Previous
Company
Percentage
8
Hiring
RD
Google
Intel
Intuit
Lucasfilm
of
Row
Total
Other
Total
Adobe
Apple
Google
Intel
532
543
0.00
0.92
0.00
0.00
0.00
762
Pixar
785
0.25
0.89
0.38
0.13
0.25
1.02
53,747
0.30
0.28
0.05
0.36
0.11
0.04
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
All
Defendants
Panel
6
25
0.05
B 2001 2004
Company
within 1 year
Adobe
Apple
Lucasfilm
0
0
0
0
0
Pixar
0
3
0
0
1
17
32
0
17
7
3
Last Previous
Company
Adobe
Apple
0
5
Percentage
1
Hiring
53,110
1.10
Google
Intel
Intuit
Lucasfilm
of
Row
Total
Total
Adobe
Apple
Google
Intel
56
57
0.00
0.00
0.00
0.00
0.00
234
Pixar
239
0.00
1.26
0.00
0.00
0.42
0.42
12,349
0.14
0.26
0.00
0.14
0.06
0.02
Other
Total
Adobe
Apple
Google
Intel
387
397
0.00
1.26
0.00
0.00
Other
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
All
Defendants
Panel
Hiring
1
2
12,271
1.75
0.02
C 2005 2009
Company
Google
Intel
Intuit
within 1 year
Lucasfilm
Percentage
Pixar
of
Row
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
0
0
5
0
3
3
1
1
4
81
65
15
99
29
10
Last Previous
Company
Adobe
Apple
Google
0
0
Pixar
All
Defendants
Panel
Hiring
1.26
406
0.00
0.74
0.74
0.25
0.25
0.99
26,035
0.31
0.25
0.06
0.38
0.11
0.04
Other
Total
Adobe
Apple
Google
Intel
89
89
0.00
0.00
0.00
0.00
394
18
0.00
25,718
0.07
D 2010 2012
Company
Intel
Intuit
within 1 year
Lucasfilm
Percentage
Pixar
of
Row
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
Defendants
Note
Source
0
0
2
Pixar
All
0
0
This analysis
Dr
1
0
0
0
3
61
53
14
75
23
11
excludes hires indicated as acquisitions
Leamer’s employee
data
and
hires
showing the same defendant
134
5
15,121
company
as their
0.00
0.00
140
1.43
0.71
0.00
0.00
0.00
2.14
15,363
0.40
0.34
0.09
0.49
0.15
0.07
immediate previous employer within one year of the hiring
0.03
Case5 11cv 02509 LHK Document5184
Filed1007 13
Page30
of
73
Appendix 2B
Analysis of Separations Going to Other Defendants
Technical Creative and
Panel
RD
Class
A 2001 2012
Next
Separation
Company
Adobe
Apple
Google
Company
Intel
within 1 year
Intuit
Percentage of
Row
Total
Pixar
Other
Total
Adobe
Apple
Google
Intel
5
Lucasfilm
333
349
0.00
0.86
2.01
0.29
0.00
378
397
0.00
1.76
1.26
0.50
0.00
1.26
0.20
0.60
0.70
0.06
0.10
0.02
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
3
7
1
0
Pixar
0
7
5
2
0
5
74
223
259
23
37
9
All
Defendants
Panel
18
36,356
36,999
1.43
0.05
B 2001 2004
Next
Separation
Company
Adobe
Apple
Google
Company
Intel
within 1 year
Intuit
Percentage of
Lucasfilm
Total
Total
Adobe
Apple
Google
Intel
7
7
0.00
0.00
0.00
0.00
0.00
106
Pixar
Other
Row
111
0.00
0.90
0.90
0.00
0.00
2.70
0.19
0.23
0.11
0.01
0.10
0.03
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
0
0
0
0
Pixar
0
1
1
0
0
3
21
25
12
1
11
3
Adobe
Apple
Google
All
Defendants
Panel
0
0.00
3
11,001
Pixar
Other
Total
Adobe
Apple
Google
Intel
2
197
201
0.00
0.00
0.50
0.50
0.00
186
0.00
2.15
1.61
1.08
0.00
18,863
0.21
0.53
0.74
0.06
0.10
0.03
Pixar
Other
Total
Adobe
Apple
Google
Intel
3
129
141
0.00
2.13
4.26
0.00
0.00
100
0.00
2.00
1.00
0.00
0.00
0.00
0.18
1.43
1.55
0.15
0.09
0.00
0.03
1.08
9
11,077
C 2005 2009
Next
Separation
Company
Company
Intel
within 1 year
Intuit
Percentage of
Lucasfilm
Row
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
1
0
0
4
3
2
0
2
102
143
12
20
6
Adobe
Apple
Google
Defendants
Panel
1
41
Pixar
All
0
175
19,196
1.00
0.05
D 2010 2012
Next
Separation
Company
Company
Intel
within 1 year
Intuit
Percentage of
Lucasfilm
Row
Total
Intuit
Lucasfilm
Pixar
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
0
Source
0
0
0
Defendants
Note
6
This analysis
2
1
0
0
0
12
Pixar
All
3
96
104
10
6
0
excludes separations
Dr Leamer’s
employee
data
that appear as immediately
rehired by the
same defendant
97
6
company
6,492
within one year
6,726
2.13
0.09
Case5 11 cv 02509 LHK
Document518
4
Filed10 07 13
Page31
of
73
Appendix 2C
Analysis of Hires from Other DNCC Defendants
Technical Creative and
Panel
Class
A 2001 2012
Company
Last Previous
DNCC
Company
Hiring
RD
within 1 year
Percentage
of
Row
Total
DNCC Defendant
Total
DNCC Defendant
532
Non
Defendant
543
2.03
97.97
785
2.04
97.96
0.90
99.10
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
11
Lucasfilm
16
All
Defendants
Panel
769
482
Pixar
53,265
B 2001 2004
Company
Last Previous
DNCC
Company
Hiring
53,747
Non
Defendant
within 1 year
DNCC Defendant
Percentage
Total
of
DNCC Defendant
Row
Total
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
1
56
57
1.75
98.25
Pixar
4
235
239
1.67
98.33
61
12,288
0.49
99.51
All
Defendants
Panel
C 2005 2009
Company
Last Previous
DNCC
Company
Hiring
12,349
Non
Defendant
within 1 year
DNCC Defendant
Percentage
Total
of
DNCC Defendant
Row
Total
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
10
387
397
2.52
97.48
8
398
406
1.97
98.03
228
Lucasfilm
25,807
0.88
99.12
Pixar
All
Defendants
Panel
D 2010 2012
Company
Last Previous
DNCC
Company
Hiring
26,035
Defendant
Non
within 1 year
DNCC Defendant
Percentage
of
Row
Total
Non
DNCCDefendant
Total
DNCC Defendant
89
0.00
140
2.86
97.14
1.26
98.74
Adobe
Apple
Google
Intel
Intuit
0
Lucasfilm
4
136
193
15,170
Pixar
All
89
Defendants
15,363
100.00
Notes
This analysis
excludes
hires
indicated as acquisitions and hires showing the
same defendant company
one year of the hiring
Adobe
had a
allegedly
Intuit
allegedly
had
allegedly had
Source
Adobe Google Intel Intuit
DNCC agreements
with Apple
Google
and
Pixar
had DNCC agreements with Apple and Google
Lucasfilm allegedly
Pixar
with
had DNCC agreements with Apple Intel and Intuit
allegedly
Intel allegedly
agreement with Apple
DNCC agreements
Apple allegedly had
Google
DNCC
had DNCC agreements with Apple and Pixar
DNCC agreements
Dr Leamer’s
employee
data
with Apple
Intel and Lucasfilm
Lucasfilm
and Pixar
as their immediate previous
employer within
Case5 11 cv 02509 LHK
Document518
4
Filed10 07 13
Page32
of
73
Appendix 2D
Analysis of Separations Going to Other
Technical Creative and
Panel
RD
DNCC Defendants
Class
A 2001 2012
Next
Company
within 1 year
Percentage of
Row
Total
Total
DNCC Defendant
8
Company
DNCCDefendant
341
349
2.29
97.71
14
Separation
383
397
3.53
96.47
1.35
98.65
DNCC Defendant
Non
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
Pixar
All
Defendants
Panel
498
36,501
36,999
B 2001 2004
Next
Separation
Company
Company
Non
within 1 year
Percentage of
Row
Total
DNCCDefendant
Total
DNCC Defendant
7
DNCC Defendant
7
0.00
100.00
111
3.60
96.40
0.55
99.45
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
0
Lucasfilm
4
Pixar
All
Defendants
Panel
107
61
11,016
11,077
C 2005 2009
Next
Separation
Company
Company
DNCC Defendant
Non
within 1 year
Percentage of
DNCCDefendant
Total
DNCC Defendant
Row
Total
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
2
199
201
1.00
99.00
Pixar
8
178
186
4.30
95.70
1.29
98.71
All
Defendants
Panel
248
18,948
19,196
D 2010 2012
Next
Separation
Company
Company
Non
within 1 year
Percentage of
Row
Total
DNCCDefendant
Total
DNCC Defendant
135
DNCC Defendant
141
4.26
95.74
100
2.00
98.00
2.81
97.19
Non
DNCCDefendant
Adobe
Apple
Google
Intel
Intuit
6
Lucasfilm
2
Pixar
All
Defendants
98
189
6,537
6,726
Notes
This analysis
Adobe
Apple
excludes separations
had DNCC agreements with Apple Intel and Intuit
allegedly
Intel allegedly
Intuit allegedly
had DNCC agreements with Apple Google and Pixar
had DNCC agreements with Apple
Lucasfilm allegedly
Pixar
allegedly
Source
Dr
rehired by the
same
had DNCC agreements with Adobe Google Intel Intuit Lucasfilm
allegedly
Google
that appear as immediately
defendant
had a DNCC agreement with Apple
allegedly
had
and Google
had DNCC agreements with Apple
DNCC
agreements with Apple
Leamer’s employee
data
and Pixar
Intel
and Lucasfilm
and Pixar
company
within one year
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page33
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page34
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page35
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page36
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page37
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page38
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page39
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page40
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page41
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page42
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page43
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page44
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page45
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page46
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page47
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page48
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page49
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page50
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page51
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page52
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page53
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page54
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page55
of
73
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page56
of
73
Appendix 9A
Dr
Leamer’s Figure 20 RegressionIncluding Defendant
Defendant
Specific
Specific
Conduct Variables and Other
Interactive Effects
AllSalaried Employee Class
Dependant
Variable Log Total Annual Compensation CPI
Variable
ADOBE
APPLE
Conduct
Conduct
GOOGLE
INTEL
PIXAR
Age
Conduct
LUCASFILM
APPLE
ADOBE
APPLE
GOOGLE
INTEL
Conduct
INTUIT
PIXAR
ADOBE
Log Number
Log Number
Conduct
New
of
of
New
New
New
Log Number
of
Hires in the
New
Log Age
ADOBE
APPLE
Log Age
GOOGLE
INTEL
INTUIT
PIXAR
Log Age
Log Age
Log Age
Log Age
3.52
0.0001
0.19
22.24
0.0250
12.57
of
0.3453
0.0061
56.20
0.0323
0.0020
16.45
0.0213
0.0127
1.67
0.1142
0.0342
3.34
0.0664
0.0169
3.92
0.0976
19.15
0.0549
13.46
0.0380
6.84
0.0132
1.81
0.0576
2.46
0.1164
0.24
0.1636
1.48
0.0056
125.95
0.0027
272.85
0.0017
294.66
0.0023
286.66
0.7202
0.0059
121.40
0.6619
0.0056
117.60
0.8067
0.0360
22.42
0.2868
0.0055
52.13
0.2828
0.0028
102.17
0.3466
0.0017
207.40
0.2964
0.0023
129.91
0.2541
0.0057
44.21
0.1743
2
0.0376
0.3141
0.6721
1
0.8370
0.5121
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
0.0001
0.7265
Log Total Annual Compensation CPI
LUCASFILM
0.78
0.7079
Log Total Annual Compensation CPI
2
2
GOOGLE Log Total Annual Compensation CPI 2
INTEL Log Total Annual Compensation CPI
2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
APPLE
6.83
0.0000
0.2427
Log Total Annual Compensation CPI
ADOBE
0.0000
0.0277
Conduct
Log Total Annual Compensation CPI
3.44
0.1416
Conduct
Conduct
LUCASFILM
0.0000
0.0240
1
1
GOOGLE Log Total Annual Compensation CPI 1
INTEL Log Total Annual Compensation CPI
1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
APPLE
5.58
0.2602
Hires in the
Conduct
ADOBE
1.26
0.0000
0.7391
of
FirmNumber
FirmNumber
Conduct
LUCASFILM
0.0000
Employees
0.0053
32.60
0.1922
0.0365
5.26
0.4727
0.2194
2.15
1.0913
0.1256
8.69
1.0010
0.1547
6.47
0.2981
0.0485
6.15
0.8571
0.1696
5.05
0.0441
0.4413
0.10
of
of
FirmNumber
FirmNumber
Hires in the
Hires in the
1
1
Employees 1
Employees
0.37
1.8691
Hires in the
Conduct
INTUIT
PIXAR
of
of
Firm
Number
Firm
Number
Conduct
GOOGLE
INTEL
Log Number
Conduct
Conduct
LUCASFILM
APPLE
Log Number
Conduct
Hires in the
Hires in the
0.0074
0.0000
New
3.59
0.0002
New
0.0042
0.0000
of
0.75
0.0000
of
5.78
0.0024
0.0001
Log Number
0.0006
0.0001
Log Number
Conduct
Conduct
3.38
0.0000
Conduct
0.0020
0.0027
Conduct
5.34
0.0152
Age
1.79
0.0015
0.0018
Age
0.0026
0.0032
Age
Age 2
Age 2
GOOGLE Conduct
Age 2
INTEL Conduct
Age 2
INTUIT
Conduct
Age 2
PIXAR
Conduct
Age 2
LUCASFILM
Conduct
Age 2
ADOBE
T Value
0.0067
Age
Conduct
Error
0.0079
Age
Conduct
St
0.0047
Age
Conduct
Conduct
INTUIT
Estimate
1
Employees 1
Employees 1
Employees
of
of
Firm
Number
of
Employees
1
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page57
of
73
Appendix 9A
Log Age
LUCASFILM
0.0240
0.8306
0.03
0.0695
0.0297
2.34
0.1235
0.0170
7.24
0.1483
0.0214
6.92
0.0348
0.0066
5.30
0.1010
0.0229
4.41
0.0166
0.0605
0.27
0.0085
0.1115
0.08
Log Company Tenure Months
0.0167
0.0050
3.36
Log Company Tenure
0.0017
0.0005
3.14
0.0025
0.0005
4.62
in San Jose
1.5574
0.0183
85.30
Among Defendants
0.0770
0.0018
42.53
0.0025
0.0003
7.90
0.0441
0.0095
4.63
0.0461
0.0066
6.94
0.2261
0.0026
0.0049
0.0013
0.0808
0.0046
0.1603
0.0308
0.0217
0.0154
0.2292
0.0026
89.66
0.0915
0.0043
21.15
0.1646
0.0033
50.39
APPLE
3.3227
0.4646
7.15
GOOGLE
0.0066
0.4898
0.01
INTEL
1.6772
0.4130
4.06
INTUIT
2.9576
0.5094
5.81
PIXAR
1.3942
0.9009
1.55
LUCASFILM
0.9044
1.5907
0.57
Log Age
ADOBE
Log Age
APPLE
2
2
Log Age 2
GOOGLE
Log Age 2
INTEL
Log Age 2
INTUIT
Log Age 2
PIXAR
Log Age
LUCASFILM
2
2
Male
DLogInformation
Log Total Number
Year
Sector Employment
of Transfers
trend
ADOBE
Log Number
Log Number
APPLE
GOOGLE
INTEL
PIXAR
of
Log Number
Log Number
of
New
New
Hires in the
New
of
Log Firm Revenue
Per EmployeeCPI
DLogFirm
Revenue
State
Location
Per
of
of
FirmNumber
FirmNumber
Hires in the
Log Total Number
of
Firm
Number
FirmNumber
Hires in the
Hires in the
of
Firm
Number
Firm
Number
Hires in the
Hires in the
New
Log Number
LUCASFILM
Hires in the
New
New
of
of
New
New
of
Log Number
Log Number
INTUIT
of
1
Employees 1
Employees
of
1
1
1
Employees 1
Employees
of
of
Employees
Employees
Firm
Number
of
Employees
1
Hires
EmployeeCPI
1
1
YES
Indicators
YES
Constant
R Square
0.928
508,969
Observations
Note
Source
Significant
Dr
at
1
level
Significant
at
Leamer’s backup data and materials
5
level
Pixar
Significant
revenue
at
10 level
data after 2005 are included
86.41
3.77
17.61
5.20
1.41
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page58
of
73
Appendix 9B
Dr
Leamer’s Figure 23 RegressionIncluding Defendant
Defendant
Specific
RD
APPLE
Conduct
Conduct
GOOGLE
INTEL
PIXAR
Age
LUCASFILM
APPLE
APPLE
GOOGLE
INTEL
Conduct
INTUIT
PIXAR
ADOBE
Log Number
Log Number
Conduct
New
of
of
New
New
New
Log Number
of
Hires in the
New
Log Age
ADOBE
APPLE
Log Age
GOOGLE
INTEL
INTUIT
PIXAR
Log Age
Log Age
Log Age
Log Age
1.92
0.0002
0.41
0.0345
of
0.3276
0.0088
37.18
0.0388
0.0026
14.83
0.0750
0.0194
3.87
0.0642
0.0440
1.46
0.0820
0.0276
2.97
0.1241
0.0747
5.79
0.0494
4.21
0.0185
2.97
0.0875
2.14
0.1508
1.37
0.3662
0.56
0.0075
89.78
0.0037
192.60
0.0022
207.91
0.0029
219.78
0.6772
0.0088
76.81
0.6202
0.0084
73.65
0.7676
0.0695
11.04
0.3112
0.0074
42.05
0.2864
0.0038
74.62
0.3478
0.0021
162.51
0.3113
0.0028
109.66
0.2930
0.0085
34.49
0.0956
2
0.0482
0.1272
0.6429
1
0.9854
0.4607
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
0.0001
0.7040
Log Total Annual Compensation CPI
LUCASFILM
0.17
0.6754
Log Total Annual Compensation CPI
2
2
GOOGLE Log Total Annual Compensation CPI 2
INTEL Log Total Annual Compensation CPI
2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
APPLE
5.07
0.0000
0.2062
Log Total Annual Compensation CPI
ADOBE
0.0000
0.2066
Conduct
Log Total Annual Compensation CPI
3.01
0.1868
Conduct
Conduct
LUCASFILM
0.0000
0.0548
1
1
GOOGLE Log Total Annual Compensation CPI 1
INTEL Log Total Annual Compensation CPI
1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
APPLE
4.65
0.2078
Hires in the
Conduct
ADOBE
1.37
0.0000
0.4323
of
FirmNumber
FirmNumber
Conduct
LUCASFILM
0.0000
Employees
0.0076
12.61
0.2340
0.0702
3.34
0.3557
0.2812
1.26
1.2304
0.1670
7.37
0.1880
0.1917
0.98
0.3725
0.0699
5.33
1.0874
0.2520
4.31
0.6246
0.5776
1.08
of
of
FirmNumber
FirmNumber
Hires in the
Hires in the
1
1
Employees 1
Employees
0.20
2.2161
Hires in the
Conduct
INTUIT
PIXAR
of
of
Firm
Number
Firm
Number
Conduct
GOOGLE
INTEL
Log Number
Conduct
Conduct
LUCASFILM
APPLE
Log Number
Conduct
Hires in the
0.0182
0.0001
Hires in the
1.83
0.0001
New
0.0056
0.0000
New
0.29
0.0000
of
4.42
0.0037
0.0001
of
0.0008
0.0001
Log Number
2.93
0.0001
Log Number
Conduct
Conduct
0.0025
0.0036
Conduct
4.54
0.0102
Age
1.85
0.0020
0.0011
Age
0.0033
0.0035
Age
Conduct
ADOBE
T Value
0.0074
Age 2
Age 2
GOOGLE Conduct
Age 2
INTEL Conduct
Age 2
INTUIT
Conduct
Age 2
PIXAR
Conduct
Age 2
LUCASFILM
Conduct
Age 2
ADOBE
Error
0.0090
Age
Conduct
St
0.0062
Age
Conduct
Conduct
Estimate
Age
Conduct
Conduct
INTUIT
Class
Variable Log Total Annual Compensation CPI
Variable
ADOBE
Conduct Variables and Other
Interactive Effects
Technical Creative and
Dependant
Specific
1
Employees 1
Employees 1
Employees
of
of
Firm
Number
of
Employees
1
20.45
3.68
17.85
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page59
of
73
Appendix 9B
Log Age
LUCASFILM
0.4933
1.5449
0.32
0.0547
0.0381
1.43
0.1382
0.0228
6.07
0.0387
0.0265
1.46
0.0449
0.0095
4.73
0.1305
0.0342
3.82
0.0667
0.0793
0.84
0.0634
0.2101
0.30
Log Company Tenure Months
0.0021
0.0067
0.31
Log Company Tenure
0.0003
0.0007
0.47
0.0058
0.0008
7.21
in San Jose
1.6830
0.0250
67.20
Among Defendants
0.0854
0.0024
35.18
0.0004
0.0004
0.99
0.0497
0.0122
4.06
0.0349
0.0092
3.81
0.2318
0.0037
0.0041
0.0018
0.1109
0.0069
0.0495
0.0394
0.0296
0.0227
0.2643
0.0035
0.0435
0.0058
0.1532
0.0044
APPLE
3.4399
0.5998
5.73
GOOGLE
1.5131
0.6217
2.43
INTEL
1.6323
0.5322
3.07
INTUIT
3.2415
0.6919
4.68
PIXAR
0.8473
1.1715
0.72
LUCASFILM
1.4582
2.8740
0.51
Log Age
ADOBE
Log Age
APPLE
2
2
Log Age 2
GOOGLE
Log Age 2
INTEL
Log Age 2
INTUIT
Log Age 2
PIXAR
Log Age
LUCASFILM
2
2
Male
DLogInformation
Log Total Number
Year
Sector Employment
of Transfers
trend
ADOBE
Log Number
Log Number
APPLE
GOOGLE
INTEL
PIXAR
of
Log Number
Log Number
of
New
New
Hires in the
New
of
Log Firm Revenue
Per EmployeeCPI
DLogFirm
Revenue
State
Location
Per
of
of
FirmNumber
FirmNumber
Hires in the
Log Total Number
of
Firm
Number
FirmNumber
Hires in the
Hires in the
of
Firm
Number
Firm
Number
Hires in the
Hires in the
New
Log Number
LUCASFILM
Hires in the
New
New
of
of
New
New
of
Log Number
Log Number
INTUIT
of
1
Employees 1
Employees
of
1
1
1
Employees 1
Employees
of
of
Employees
Employees
Firm
Number
of
Employees
1
Hires
EmployeeCPI
1
1
YES
Indicators
YES
Constant
R Square
0.879
295,136
Observations
Note
Source
Significant
Dr
at
1
level
Significant
at
Leamer’s backup data and materials
5
level
Pixar
Significant
revenue
at
10 level
data after 2005 are included
63.00
2.34
16.17
1.26
1.31
76.33
7.45
35.02
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page60
of
73
Appendix 10A
Dr
Leamer’s Figure 20 RegressionUsing a Single Conduct Variable
AllSalaried Employee Class
Dependant
Log Total Annual Compensation CPI
Variable
Variable
Estimate
Conduct
St
Error
T Value
0.0344
0.0008
41.98
0.6978
0.0054
129.27
0.7416
0.0026
279.85
0.4943
0.0017
293.50
0.6687
0.0024
282.48
0.7117
0.0057
124.33
0.6961
0.0069
100.42
0.8118
0.0363
22.36
0.2934
0.0053
55.74
0.2595
0.0027
95.36
0.3734
0.0016
229.06
0.3005
0.0023
130.49
0.2522
0.0055
45.49
0.1992
0.0067
29.64
0.1798
0.0367
4.90
Log Age Years
0.0105
0.0328
0.32
Log Age
0.0076
0.0044
1.72
1
APPLE
Log Total Annual CompensationCPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
Log Total Annual Compensation CPI 1
INTUIT
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual CompensationCPI
LogTotal
LUCASFILM
1
Annual CompensationCPI
2
APPLE
Log Total Annual CompensationCPI 2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual CompensationCPI
LogTotal
LUCASFILM
Annual CompensationCPI
2
2
Log Company Tenure
Months
0.0083
0.0050
1.66
Log Company Tenure
2
0.0009
0.0006
1.66
0.0027
0.0005
5.02
San Jose
1.4135
0.0136
103.90
Among Defendants
0.0959
0.0015
63.66
0.0039
0.0003
14.53
0.0169
0.0008
21.61
0.2478
0.0021
116.78
0.1027
0.0034
30.20
0.2162
0.0033
66.49
APPLE
0.0607
0.0162
GOOGLE
1.0320
0.0174
59.42
INTEL
0.1516
0.0146
10.40
INTUIT
0.1473
0.0193
PIXAR
0.7075
0.0422
0.1256
0.0480
Male
DLog
Information Sector Employment
Log Total Number
Year
of
Transfers
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog
in
Firm Revenue
Firm
Number
Employees
1
Hires
EmployeeCPI
Per
of
EmployeeCPI
1
1
LUCASFILM
Location
State Indicators
YES
Constant
YES
RSquare
0.926
504,897
Observations
Note
Source
Significant
Dr Leamer’s
at
1
level
Significant
at
backup data and materials
5
level
Significant
at
10 level
3.75
7.64
16.77
2.61
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page61
of
73
Appendix 10B
Dr
Leamer’s Figure 23 RegressionUsing a Single Conduct Variable
Technical Creative and
Dependant
RD
Class
Log Total Annual Compensation CPI
Variable
Variable
Estimate
Conduct
St
Error
T Value
0.0234
0.0011
20.94
0.6643
0.0072
91.76
0.7212
0.0037
197.36
0.4403
0.0022
203.78
0.6407
0.0030
215.53
0.6578
0.0084
78.28
0.6523
0.0106
61.69
0.8457
0.0692
12.21
0.3158
0.0071
44.58
0.2581
0.0038
68.54
0.3629
0.0021
173.68
0.3171
0.0029
110.18
0.2967
0.0081
36.48
0.1054
0.0097
10.89
0.1456
0.0694
2.10
Log Age Years
0.1807
0.0463
3.90
Log Age
0.0146
0.0063
2.32
1
APPLE
Log Total Annual CompensationCPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
Log Total Annual Compensation CPI 1
INTUIT
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual CompensationCPI
LogTotal
LUCASFILM
1
Annual CompensationCPI
2
APPLE
Log Total Annual CompensationCPI 2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual CompensationCPI
LogTotal
LUCASFILM
Annual CompensationCPI
2
2
Log Company Tenure
Months
0.0326
0.0068
4.78
Log Company Tenure
2
0.0028
0.0008
3.78
0.0065
0.0008
7.89
San Jose
1.5271
0.0189
80.81
Among Defendants
0.0983
0.0020
48.08
0.0009
0.0004
0.0154
0.0011
14.31
0.2724
0.0029
93.07
0.0811
0.0047
17.17
0.2127
0.0044
48.43
APPLE
0.1244
0.0245
GOOGLE
1.3816
0.0259
INTEL
0.1573
0.0219
INTUIT
0.1486
0.0315
PIXAR
1.5543
0.0771
0.0296
0.1038
Male
DLog
Information Sector Employment
Log Total Number
Year
of
Transfers
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog
in
Firm Revenue
Firm
Number
Employees
1
Hires
EmployeeCPI
Per
of
EmployeeCPI
1
1
LUCASFILM
Location
State Indicators
YES
Constant
YES
RSquare
0.874
292,489
Observations
Note
Source
Significant
Dr Leamer’s
at
1
level
Significant
at
backup data and materials
5
level
Significant
at
10 level
2.52
5.08
53.33
7.19
4.71
20.17
0.29
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page62
of
73
Appendix 10C
Undercompensation Estimates Using a
Conduct Variable
in
Dr Leamer’s
AllSalaried
Year
Adobe
Apple
Google
2005
1.72
1.72
1.72
2006
4.63
4.71
2007
7.17
2008
9.80
2009
9.80
10.28
vs
Regression
Figures
Employee Class
Apple
Google
1.61
1.59
1.78
1.67
12.13
10.56
2006
4.28
4.43
4.44
4.70
14.63
12.44
14.00
2007
6.64
6.94
6.39
7.46
3.24
17.24
14.28
15.61
2008
9.08
9.56
8.40
10.05
5.64
19.94
15.76
14.52
2009
9.15
9.73
7.51
9.95
5.70
20.12
14.65
Pixar
Year
1.72
11.95
10.29
2005
4.28
4.58
14.77
12.23
7.37
6.19
7.02
3.44
17.58
10.13
8.10
9.51
5.88
20.36
7.17
9.32
5.91
20.55
Intuit
RD
Adobe
Apple
Google
2005
1.17
1.17
1.17
1.17
2006
3.12
3.19
2.86
3.09
2007
4.78
4.94
4.03
4.69
2.34
2008
6.50
6.73
5.15
6.33
3.88
2009
6.42
6.71
4.31
6.13
3.83
Intel
with age and hiring rate
Intel
Technical Creative and
Lucasfilm
Class
1.90
3.07
1.64
2006
4.29
4.96
7.23
3.06
14.77
10.47
2007
6.48
7.79
9.36
3.38
3.41
18.08
10.61
7.92
2008
8.80
10.64
11.20
4.76
5.21
20.44
11.87
6.54
2009
8.44
10.51
9.00
4.19
4.96
20.54
9.62
2005
6.85
12.27
7.45
14.22
14.40
interactions
Pixar
1.56
6.08
8.33
RD
Lucasfilm
Google
Year
Adobe
Intuit
Apple
Pixar
10.31
Intuit
Leamer Figure 20 and 23 regressions excluding conduct
Adobe
Class
Year
22 and 24
AllSalaried Employee Class
Lucasfilm
Intel
Technical Creative and
Source
Undercompensation Estimates in Dr Leamer’s
Single
Intel
Intuit
Lucasfilm
10.80
Pixar
9.28
Case5 11 cv 02509 LHK
Document518
4
Filed10 07 13
Page63
of
73
Appendix 11A
Dr
Leamer’s Figure 20 RegressionIncluding Defendant
Specific
Conduct
Variables
AllSalaried Employee Class
Variable Log Total Annual Compensation CPI
Dependant
Variable
ADOBE
Estimate
Conduct
St
Error
T Value
0.0053
0.0028
1.89
0.0139
0.0019
7.37
0.0969
0.0021
45.25
0.0304
0.0009
33.37
0.0600
0.0026
23.17
0.0396
0.0048
8.34
0.0000
0.0075
0.00
0.6855
0.0056
122.85
0.7361
0.0027
276.84
0.4858
0.0017
283.31
0.6721
0.0024
283.28
0.7173
0.0058
122.92
0.6857
0.0055
124.10
0.7984
0.0364
21.92
Log Total Annual Compensation CPI
0.3056
0.0055
56.03
Log Total Annual Compensation CPI
0.2645
0.0027
96.26
0.3741
0.0016
228.53
0.2976
0.0023
128.96
0.2466
0.0056
43.72
0.1758
0.0053
33.30
0.2003
0.0369
5.43
Log Age Years
0.0244
0.0327
0.75
Log Age
APPLE
Conduct
GOOGLE
Conduct
Conduct
INTEL
Conduct
INTUIT
PIXAR
Conduct
LUCASFILM
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
1
2
2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
APPLE
Log Total Annual Compensation CPI
LUCASFILM
2
0.0057
0.0044
1.28
Log Company Tenure Months
0.0128
0.0050
2.55
Log Company Tenure 2
0.0013
0.0006
2.42
0.0032
0.0005
1.4228
0.0136
104.42
0.0800
0.0015
53.90
0.0032
0.0003
12.13
0.0128
0.0008
16.20
0.2273
0.0021
108.21
0.0677
0.0033
20.55
0.1461
0.0029
50.95
APPLE
0.0492
0.0163
GOOGLE
1.0950
0.0176
62.24
INTEL
0.1587
0.0147
10.82
INTUIT
0.1818
0.0193
PIXAR
0.7905
0.0264
LUCASFILM
0.0271
0.0503
2
Male
DLog
Log Total Number
Year
of Transfers
Among Defendants
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog
in San Jose
Information Sector Employment
Firm Revenue
State
Location
Firm
Number
Employees
1
Hires
EmployeeCPI
Per
of
EmployeeCPI
1
1
YES
Indicators
Constant
YES
R Square
0.926
Observations
Note
Source
Significant
Dr Leamer’s
508,969
at
1
level
Significant
backup data and materials
at
5
Pixar
level
Significant
revenue
at
10 level
data after 2005 are included
5.82
3.02
9.40
29.96
0.54
Case5 11 cv 02509 LHK
Document518
4
Filed10 07 13
Page64
of
73
Appendix 11B
Dr
Leamer’s Figure 23 RegressionIncluding Defendant
Specific
Conduct
Variables
Technical Creative and
RD
Variable Log Total Annual Compensation CPI
Dependant
Variable
ADOBE
Class
Estimate
Conduct
St
Error
T Value
0.0175
0.0036
4.80
0.0227
0.0026
8.71
0.1219
0.0029
42.51
0.0124
0.0012
10.12
0.0512
0.0040
12.96
0.0800
0.0061
13.10
0.0204
0.0130
0.6517
0.0075
86.93
0.7204
0.0036
197.54
0.4279
0.0022
195.45
0.6449
0.0030
217.17
0.6682
0.0086
77.99
0.6623
0.0081
81.28
0.7861
0.0701
11.21
Log Total Annual Compensation CPI
0.3285
0.0074
44.62
Log Total Annual Compensation CPI
0.2566
0.0038
67.66
0.3684
0.0021
175.48
0.3140
0.0029
109.24
0.2870
0.0083
34.76
0.1014
0.0075
13.58
0.2148
0.0707
3.04
Log Age Years
0.2111
0.0461
4.58
Log Age
APPLE
Conduct
GOOGLE
Conduct
Conduct
INTEL
Conduct
INTUIT
PIXAR
Conduct
LUCASFILM
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
1
2
2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
APPLE
Log Total Annual Compensation CPI
LUCASFILM
2
1.57
0.0187
0.0063
2.99
Log Company Tenure Months
0.0011
0.0068
0.16
Log Company Tenure 2
0.0005
0.0008
0.73
0.0067
0.0008
1.5258
0.0189
80.88
0.0805
0.0020
40.21
0.0000
0.0004
0.0145
0.0011
13.40
0.2548
0.0029
88.38
0.0402
0.0045
0.1324
0.0038
APPLE
0.1309
0.0246
GOOGLE
1.4469
0.0261
INTEL
0.1653
0.0220
INTUIT
0.1840
0.0315
PIXAR
1.3668
0.0455
LUCASFILM
0.0872
0.1064
2
Male
DLog
Log Total Number
Year
of Transfers
Among Defendants
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog
in San Jose
Information Sector Employment
Firm Revenue
State
Location
Firm
Number
Employees
1
Hires
EmployeeCPI
Per
of
EmployeeCPI
1
1
YES
Indicators
Constant
YES
R Square
0.874
Observations
Note
Source
Significant
Dr Leamer’s
295,136
at
1
level
Significant
backup data and materials
at
5
Pixar
level
Significant
revenue
at
10 level
data after 2005 are included
8.24
0.08
8.91
34.60
5.32
55.52
7.53
5.83
30.03
0.82
Case5 11cv 02509LHK
Document518
4
Filed1007 13
Page65
of
73
Appendix 11C
Undercompensation
Estimates Using
Conduct Variables
in
AllSalaried
Year
Adobe
Apple
Google
2005
0.26
0.69
4.85
2006
0.71
1.90
2007
1.09
2008
1.49
2009
1.49
Dr
Defendant
vs
Leamer’s Regression
Figures
Employee Class
Apple
Google
1.61
1.59
1.78
1.67
12.13
10.56
2006
4.28
4.43
4.44
4.70
14.63
12.44
15.21
2007
6.64
6.94
6.39
7.46
3.24
17.24
14.28
16.76
2008
9.08
9.56
8.40
10.05
5.64
19.94
15.76
2009
9.15
9.73
7.51
9.95
5.70
20.12
14.65
Pixar
Year
1.52
0.01
11.48
2005
12.04
4.06
0.01
13.46
2.97
17.35
6.23
6.00
0.01
4.08
22.63
8.44
10.30
0.02
4.13
19.91
8.28
10.36
0.02
15.16
Intuit
RD
Adobe
Apple
Google
2005
0.87
1.13
6.09
0.62
2006
2.32
3.08
14.79
1.64
2007
3.55
4.78
20.76
2.50
5.12
2008
4.82
6.50
26.52
3.37
8.55
2009
4.74
6.47
22.04
3.27
8.46
Intel
hiring
rate and including company conduct
data after 2005 are included
3.07
1.64
2006
4.29
4.96
7.23
3.06
14.77
10.47
2007
6.48
7.79
9.36
3.38
3.41
18.08
10.61
27.50
2008
8.80
10.64
11.20
4.76
5.21
20.44
11.87
22.83
2009
8.44
10.51
9.00
4.19
4.96
20.54
9.62
8.71
age and
interactions
Class
1.90
2005
23.69
10.39
25.82
12.08
12.24
with
Pixar
1.56
21.01
interactions
RD
Lucasfilm
Google
Year
Adobe
Intuit
Apple
Pixar
7.02
Leamer Figure 20 and 23 regressions excluding conduct
Intel
Technical Creative and
Lucasfilm
Intuit
Source
Adobe
Class
Year
22 and 24
AllSalaried Employee Class
Lucasfilm
Intel
Technical Creative and
Pixar revenue
Undercompensation Estimates in Dr Leamer’s
Specific
Intel
Intuit
Lucasfilm
10.80
Pixar
9.28
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
4
Page66
of
73
Appendix 12A
Dr
Leamer’s Figure 20 RegressionUsing
Employee Class
AllSalaried
Dependant
Variable
PreConduct Periodas Benchmark
Log Total Annual Compensation CPI
Variable
Estimate
St
Error
T Value
Conduct
Age
0.0056
0.0005
10.83
Conduct
Age 2
0.0001
0.0000
11.78
Conduct
Log Number
0.0391
0.0010
40.01
0.2432
0.0111
21.97
0.7667
0.0062
122.75
0.7374
0.0033
223.86
0.5619
0.0023
245.29
0.6743
0.0026
263.51
0.7086
0.0062
114.53
0.6957
0.0056
123.46
0.7392
0.0390
18.95
0.2167
0.0061
35.43
0.2637
0.0034
77.79
0.3504
0.0020
178.13
0.2932
0.0025
118.61
0.2459
0.0059
41.50
0.1477
0.0054
27.16
0.2434
0.0395
6.16
Log Age Years
0.4166
0.0537
7.75
Log Age 2
0.0498
0.0073
6.79
Log Company Tenure Months
0.0684
0.0057
12.04
Log Company Tenure
0.0068
0.0006
10.87
0.0030
0.0006
4.83
San Jose
1.2592
0.0166
75.70
Among Defendants
0.0789
0.0018
42.98
0.0105
0.0003
29.97
0.0197
0.0010
19.03
0.2174
0.0030
71.92
0.0928
0.0045
20.50
of
New
Hires In the
Firm
Number
of
Employees
1
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
DLog
Information Sector Employment
Log Total Number
Year
of
Transfers
in
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog Firm Revenue
Firm
Number
of
Employees
1
Hires
1
EmployeeCPI 1
EmployeeCPI
0.1286
0.0033
38.95
APPLE
0.1111
0.0194
5.71
GOOGLE
0.6086
0.0217
28.00
INTEL
0.1019
0.0173
5.89
INTUIT
0.2270
0.0223
10.17
PIXAR
0.9625
0.0302
31.82
LUCASFILM
0.1298
0.0626
2.07
State
Location
Per
YES
Indicators
Constant
YES
RSquare
0.924
Observations
Note
Source
Significant
Dr Leamer’s
381,288
at
1
level
Significant
at
backup data and materials
5
level
Pixar
Significant
revenue
at
10 level
data after 2005 are included
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
4
Page67
of
73
Appendix 12B
Dr
Leamer’s Figure 23 RegressionUsing
PreConduct Periodas Benchmark
Technical Creative and
Dependant
Variable
RD
Class
Log Total Annual Compensation CPI
Variable
Estimate
St
Error
T Value
Conduct
Age
0.0061
0.0008
8.05
Conduct
Age 2
0.0001
0.0000
8.90
Conduct
Log Number
0.0546
0.0013
40.90
0.2967
0.0159
18.61
0.7426
0.0083
89.58
0.7137
0.0047
151.39
0.4868
0.0031
157.85
0.6285
0.0032
195.11
0.6641
0.0093
71.55
0.6794
0.0084
81.00
0.6826
0.0827
8.25
0.2307
0.0081
28.45
0.2675
0.0049
54.82
0.3341
0.0026
129.27
0.3232
0.0031
104.05
0.2842
0.0088
32.11
0.0644
0.0078
8.27
0.2566
0.0822
3.12
Log Age Years
0.5769
0.0798
7.23
Log Age 2
0.0720
0.0109
6.59
Log Company Tenure Months
0.0994
0.0079
12.64
Log Company Tenure
0.0093
0.0009
10.65
0.0065
0.0009
6.89
San Jose
1.1685
0.0234
49.89
Among Defendants
0.0782
0.0025
30.91
0.0042
0.0005
8.83
0.0239
0.0014
16.49
0.2084
0.0043
48.83
0.1131
0.0062
18.39
of
New
Hires In the
Firm
Number
of
Employees
1
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
DLog
Information Sector Employment
Log Total Number
Year
of
Transfers
in
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog Firm Revenue
Firm
Number
of
Employees
1
Hires
1
EmployeeCPI 1
EmployeeCPI
0.1164
0.0044
26.21
APPLE
0.0573
0.0292
1.96
GOOGLE
1.1501
0.0330
34.87
INTEL
0.1375
0.0256
5.38
INTUIT
0.2064
0.0364
5.67
PIXAR
1.5840
0.0521
30.41
LUCASFILM
0.0853
0.1652
0.52
State
Location
Per
YES
Indicators
Constant
YES
RSquare
0.866
Observations
Note
Source
Significant
Dr Leamer’s
216,253
at
1
level
Significant
at
backup data and materials
5
level
Pixar
Significant
revenue
at
10 level
data after 2005 are included
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
4
Page68
of
73
Appendix 12C
Dr
Leamer’s Figure 20 RegressionUsing Post Conduct Period as Benchmark
Employee Class
AllSalaried
Dependant
Variable
Log Total Annual Compensation CPI
Variable
Estimate
St
Error
T Value
Conduct
Age
0.0078
0.0006
13.85
Conduct
Age 2
0.0001
0.0000
13.31
Conduct
Log Number
0.0114
0.0009
12.67
0.0973
0.0121
0.7630
0.0069
110.30
0.7349
0.0029
250.23
0.5002
0.0018
277.95
0.6763
0.0034
200.70
0.8207
0.0103
79.39
0.7036
0.0058
122.35
0.8750
0.0378
23.12
0.2528
0.0070
36.11
0.2602
0.0031
85.08
0.3684
0.0017
213.20
0.3235
0.0034
95.84
0.1548
0.0104
14.95
0.1769
0.0055
32.24
0.1143
0.0382
2.99
Log Age Years
0.6760
0.0560
12.08
Log Age 2
0.0797
0.0076
10.55
Log Company Tenure Months
0.0254
0.0058
4.39
Log Company Tenure
0.0020
0.0006
3.21
0.0021
0.0006
3.34
San Jose
0.8493
0.0541
15.70
Among Defendants
0.0287
0.0019
15.14
0.0113
0.0005
23.30
0.0325
0.0012
26.15
0.0683
0.0059
11.64
0.0268
0.0040
6.61
of
New
Hires In the
Firm
Number
of
Employees
1
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
DLog
Information Sector Employment
Log Total Number
Year
of
Transfers
in
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog Firm Revenue
Firm
Number
of
Employees
1
Hires
1
EmployeeCPI 1
EmployeeCPI
8.06
0.1248
0.0032
39.43
APPLE
0.2203
0.0187
11.80
GOOGLE
1.1437
0.0196
58.31
INTEL
0.0757
0.0169
4.47
INTUIT
0.2278
0.0247
9.23
PIXAR
0.8522
0.0283
30.13
LUCASFILM
0.1705
0.0507
3.36
State
Location
Per
YES
Indicators
Constant
YES
RSquare
0.922
Observations
Note
Source
Significant
Dr Leamer’s
399,299
at
1
level
Significant
at
backup data and materials
5
level
Pixar
Significant
revenue
at
10 level
data after 2005 are included
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
4
Page69
of
73
Appendix 12D
Dr
Leamer’s Figure 23 RegressionUsing Post Conduct Period as Benchmark
Technical Creative and
Dependant
Variable
RD
Class
Log Total Annual Compensation CPI
Variable
Estimate
St
Error
T Value
Conduct
Age
0.0096
0.0008
12.31
Conduct
Age 2
0.0001
0.0000
11.96
Conduct
Log Number
0.0008
0.0012
0.70
0.1544
0.0165
9.37
0.7523
0.0092
81.89
0.7161
0.0039
181.32
0.4438
0.0023
193.37
0.6464
0.0041
156.05
0.7732
0.0151
51.22
0.7071
0.0085
83.39
0.9511
0.0719
13.24
0.2530
0.0094
26.98
0.2581
0.0041
62.57
0.3655
0.0022
165.61
0.3478
0.0041
84.01
0.1837
0.0151
12.18
0.1052
0.0078
13.57
0.0413
0.0720
0.57
Log Age Years
0.9447
0.0755
12.51
Log Age 2
0.1145
0.0102
11.21
Log Company Tenure Months
0.0094
0.0078
1.21
Log Company Tenure
0.0008
0.0009
0.98
0.0065
0.0009
6.91
San Jose
0.9430
0.0718
Among Defendants
0.0088
0.0026
3.41
0.0148
0.0006
22.84
0.0367
0.0017
21.93
0.0834
0.0078
10.64
0.0112
0.0054
2.05
of
New
Hires In the
Firm
Number
of
Employees
1
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
DLog
Information Sector Employment
Log Total Number
Year
of
Transfers
in
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog Firm Revenue
Firm
Number
of
Employees
1
Hires
1
EmployeeCPI 1
EmployeeCPI
13.14
0.1110
0.0042
26.40
APPLE
0.2949
0.0283
10.42
GOOGLE
1.4735
0.0292
50.43
INTEL
0.0390
0.0255
1.53
INTUIT
0.2932
0.0406
7.21
PIXAR
1.2492
0.0487
25.67
LUCASFILM
0.0692
0.1083
0.64
State
Location
Per
YES
Indicators
Constant
YES
RSquare
0.869
Observations
Note
Source
Significant
Dr Leamer’s
236,748
at
1
level
Significant
at
backup data and materials
5
level
Pixar
Significant
revenue
at
10 level
data after 2005 are included
Case5 11cv 02509LHK
Document518
Filed1007 13
4
Page70
of
73
Appendix 13A
Dr
Leamer’s Figure 20 RegressionEstimated Using
Variable
Log Total Annual Compensation CPI
Variable
ADOBE
Period Data
Employee Class
AllSalaried
Dependant
Non Conduct
Estimate
1
1
CPI
1
Log Total Annual Compensation CPI
St
Error
T Value
0.6108
0.0072
84.47
0.7408
0.0036
205.55
0.4578
0.0026
175.14
0.6685
0.0034
196.94
0.7266
0.0063
115.16
0.8377
0.0219
38.18
0.9990
0.0845
11.82
0.3441
0.0067
51.72
0.2708
0.0036
74.65
0.3957
0.0028
141.55
0.2620
0.0032
81.66
0.2413
0.0060
40.26
0.1329
0.0201
6.60
0.0161
0.0856
0.19
Log Age Years
0.0292
0.0436
0.67
Log Age 2
0.0122
0.0059
2.07
Log Company Tenure Months
0.0613
0.0071
8.59
Log Company Tenure
0.0064
0.0008
8.21
Log Total Annual Compensation CPI
APPLE
GOOGLE
Log Total Annual Compensation
Log Total Annual Compensation
CPI
INTEL
Log Total Annual Compensation CPI
INTUIT
Log Total Annual CompensationCPI
PIXAR
1
1
1
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation
CPI
INTEL
Log Total Annual Compensation CPI
INTUIT
Log Total Annual CompensationCPI
PIXAR
2
2
2
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
0.0041
0.0007
5.58
San Jose
1.3739
0.0252
54.58
Among Defendants
0.0610
0.0027
22.79
0.0028
0.0007
3.93
0.0365
0.0013
27.33
0.2303
0.0053
43.47
0.0961
0.0048
19.94
0.0715
0.0062
11.50
APPLE
0.2454
0.0216
11.37
GOOGLE
0.8453
0.0233
36.31
INTEL
0.1981
0.0195
10.18
INTUIT
0.0736
0.0242
3.04
PIXAR
0.0559
0.0473
1.18
0.2748
0.0708
3.88
DLogInformation
Log Total Number
Year
Sector Employment
of Transfers
in
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLogFirm
Revenue
Firm
Number
EmployeeCPI
Per
of
Employees
1
Hires
EmployeeCPI
1
1
LUCASFILM
State Indicators
Location
YES
YES
Constant
R Square
0.937
237,351
Observations
Note
Source
Significant
Dr
at
1
level
Significant
Leamer’s backup data and materials
at
5
level
Pixar
Significant
revenue data
at
after
10 level
2005 are included
Case5 11cv 02509LHK
Document518
Filed1007 13
4
Page71
of
73
Appendix 13B
Dr
Leamer’s Figure 23 RegressionEstimated Using
Technical Creative and
Dependant
Variable
RD
Non Conduct
Class
Log Total Annual Compensation CPI
Variable
ADOBE
Period Data
Estimate
1
1
CPI
1
Log Total Annual Compensation CPI
St
Error
T Value
0.5929
0.0100
59.23
0.7428
0.0049
151.07
0.4205
0.0033
129.36
0.6526
0.0043
153.41
0.7101
0.0092
76.79
0.9381
0.0359
26.12
0.9713
0.1224
7.94
0.3475
0.0092
37.69
0.2392
0.0050
48.28
0.3895
0.0036
108.96
0.2660
0.0040
66.55
0.2593
0.0087
29.69
0.0343
0.0307
1.12
0.0629
0.1247
0.50
Log Age Years
0.2740
0.0614
4.46
Log Age 2
0.0282
0.0083
3.38
Log Company Tenure Months
0.0758
0.0096
7.89
Log Company Tenure
0.0086
0.0011
8.09
Log Total Annual Compensation CPI
APPLE
GOOGLE
Log Total Annual Compensation
Log Total Annual Compensation
CPI
INTEL
Log Total Annual Compensation CPI
INTUIT
Log Total Annual CompensationCPI
PIXAR
1
1
1
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation
CPI
INTEL
Log Total Annual Compensation CPI
INTUIT
Log Total Annual CompensationCPI
PIXAR
2
2
2
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
0.0071
0.0011
6.43
San Jose
1.3635
0.0362
37.70
Among Defendants
0.0650
0.0038
17.33
0.0034
0.0011
3.16
0.0495
0.0018
26.92
0.2480
0.0078
31.98
0.0458
0.0067
6.82
0.0388
0.0086
4.51
APPLE
0.1750
0.0326
5.37
GOOGLE
0.9977
0.0343
29.13
INTEL
0.2041
0.0293
6.96
INTUIT
0.1603
0.0388
4.13
PIXAR
0.1585
0.0893
1.77
0.5484
0.1265
4.34
DLogInformation
Log Total Number
Year
Sector Employment
of Transfers
in
trend
Log Number
of
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLogFirm
Revenue
Firm
Number
EmployeeCPI
Per
of
Employees
1
Hires
EmployeeCPI
1
1
LUCASFILM
State Indicators
Location
YES
YES
Constant
R Square
0.895
137,271
Observations
Note
Source
Significant
Dr
at
1
level
Significant
Leamer’s backup data and materials
at
5
level
Pixar
Significant
revenue data
at
after
10 level
2005 are included
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
4
Page72
of
73
Appendix 14A
Dr
Leamer’s Figure 20 RegressionIncluding Change
Variable
SP 500
Employee Class
AllSalaried
Dependant
in
Log Total Annual Compensation CPI
Variable
Estimate
St
Error
T Value
Conduct
Age
0.0066
0.0005
13.98
Conduct
Age 2
0.0001
0.0000
13.83
Conduct
Log Number
0.0043
0.0008
5.54
0.1309
0.0100
13.04
0.6894
0.0054
126.98
0.7449
0.0027
280.12
0.4988
0.0017
293.05
0.6678
0.0024
282.12
0.7070
0.0058
122.77
0.6943
0.0069
100.22
0.8204
0.0363
22.62
0.3023
0.0053
57.04
0.2581
0.0027
94.33
0.3694
0.0016
225.49
0.3012
0.0023
130.80
0.2567
0.0056
46.04
0.1985
0.0067
29.56
0.1737
0.0366
4.74
Log Age Years
0.3495
0.0415
8.42
Log Age 2
0.0380
0.0056
6.74
Log Company Tenure Months
0.0039
0.0050
0.78
Log Company Tenure
0.0005
0.0006
0.92
0.0027
0.0005
4.93
San Jose
1.5373
0.0151
101.59
Among Defendants
0.0566
0.0020
27.69
0.0656
0.0023
28.72
0.0026
0.0003
7.45
0.0135
0.0009
14.55
0.2182
0.0024
92.01
0.1319
0.0037
36.14
0.2371
0.0033
70.97
APPLE
0.0747
0.0162
4.62
GOOGLE
1.0592
0.0174
60.95
INTEL
0.1542
0.0146
10.59
INTUIT
0.1485
0.0193
7.71
PIXAR
0.7001
0.0422
16.60
0.1483
0.0480
3.09
of
New
Hires In the
Firm
Number
of
Employees
1
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
DLog
Information Sector Employment
Log Total Number
DLog
SP 500 Net
Year
of
Transfers
in
trend
Log Number
of
Total Return Index CPI
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog Firm Revenue
Firm
Number
Employees
1
Hires
EmployeeCPI
Per
of
EmployeeCPI
1
1
LUCASFILM
State
Location
YES
Indicators
Constant
YES
RSquare
0.926
Observations
Note
Source
Significant
Dr Leamer’s
504,897
at
1
level
Significant
at
backup data and materials
5
level
Significant
at
10 level
Case5 11 cv 02509 LHK
Document518
Filed10 07 13
4
Page73
of
73
Appendix 14B
Dr
Leamer’s Figure 23 RegressionIncluding Change
Technical Creative and
Dependant
Variable
RD
in
SP 500
Class
Log Total Annual Compensation CPI
Variable
Estimate
St
Error
T Value
Conduct
Age
0.0077
0.0007
11.44
Conduct
Age 2
0.0001
0.0000
11.18
Conduct
Log Number
0.0099
0.0010
9.44
0.1717
0.0141
12.16
0.6662
0.0073
91.42
0.7299
0.0037
199.33
0.4425
0.0022
202.73
0.6405
0.0030
215.77
0.6672
0.0085
78.91
0.6508
0.0106
61.63
0.8548
0.0691
12.37
0.3141
0.0071
44.00
0.2505
0.0038
66.22
0.3607
0.0021
171.44
0.3177
0.0029
110.53
0.2888
0.0082
35.32
0.1053
0.0097
10.90
0.1398
0.0692
2.02
Log Age Years
0.5757
0.0587
9.80
Log Age 2
0.0676
0.0080
8.46
Log Company Tenure Months
0.0204
0.0068
3.00
Log Company Tenure
0.0016
0.0008
2.14
0.0064
0.0008
7.86
San Jose
1.5716
0.0209
75.07
Among Defendants
0.0443
0.0028
16.05
0.0881
0.0031
28.55
0.0078
0.0005
16.67
0.0213
0.0013
16.62
0.2308
0.0033
70.79
0.1028
0.0051
20.31
0.2359
0.0045
52.12
APPLE
0.1328
0.0244
5.44
GOOGLE
1.4013
0.0259
54.09
INTEL
0.1574
0.0218
7.20
INTUIT
0.1378
0.0315
4.38
PIXAR
1.5355
0.0770
19.94
0.0399
0.1036
0.38
of
New
Hires In the
Firm
Number
of
Employees
1
Conduct
1
APPLE
Log Total Annual Compensation CPI 1
GOOGLE
Log Total Annual Compensation CPI 1
INTEL
Log Total Annual Compensation CPI 1
INTUIT
Log Total Annual Compensation CPI 1
PIXAR
Log Total Annual Compensation CPI 1
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
1
2
APPLE
Log Total Annual Compensation CPI 2
GOOGLE
Log Total Annual Compensation CPI 2
INTEL
Log Total Annual Compensation CPI 2
INTUIT
Log Total Annual Compensation CPI 2
PIXAR
Log Total Annual Compensation CPI 2
ADOBE
Log Total Annual Compensation CPI
Log Total Annual Compensation CPI
LUCASFILM
2
2
Male
DLog
Information Sector Employment
Log Total Number
DLog
SP 500 Net
Year
of
Transfers
in
trend
Log Number
of
Total Return Index CPI
New
Hires In the
New
Log Total Number
of
Log Firm Revenue
Per
DLog Firm Revenue
Firm
Number
Employees
1
Hires
EmployeeCPI
Per
of
EmployeeCPI
1
1
LUCASFILM
State
Location
YES
Indicators
Constant
YES
RSquare
0.875
Observations
Note
Source
Significant
Dr Leamer’s
292,489
at
1
level
Significant
at
backup data and materials
5
level
Significant
at
10 level
4
Case5 11 cv 02509 LHK
Document424
2
Filed05 17 13
Page1
of
62
IN THE UNITED STATES DISTRICT COURT
FOR THE NORTHERN DISTRICT OF CALIFORNIA
SAN JOSE DIVISION
CONFIDENTIAL
TO BE FILED
SUBJECT TO PROTECTIVE
IN RE HIGH TECH EMPLOYEES ANTITRUST
LITIGATION
THIS
UNDER SEAL
ORDER
No 11 CV2509LHK
DOCUMENT RELATES TO
ALL ACTIONS
SUPPLEMENTAL EXPERT REPORT OF EDWARD
May
E LEAMER PH D
10 2013
REDACTED
Case5 11 cv 02509 LHK
Document424
Filed05 17 13
2
Page2
of
62
TABLE OF CONTENTS
I
Introduction Assignment and
II
Defendants Use
III
Empirical Methodologies for Exploring the
of
Summary
of
Conclusions
1
4
Compensation Structures
Somewhat
Rigid
Salary Structure
A
B
C
IV
6
6
VI
of
Regression Analysis
of Compensation
byTitle
Correlation Analysis
Title
byTitle
7
Structure
7
Multiple Regressions
of Compensation
10
Structure
14
Based Correlations and Multiple Regressions
18
Decile
Based
Correlation Analysis
18
Decile
Based
Multiple Regression Results
20
Additional Exploration of
1
2
3
4
VII
Structure
Based Correlations and Multiple Regressions 10
Title
Decile
A
B
Compensation
Correlation Analysis
Results of Title
A
B
V
Choice of Aggregation Level
Adobe
Adobe Correlations
22
22
Correlation Results
Headcount Matters
for Interpreting
24
Correlations
Correlations
25
Outliers
26
29
Internal Versus External Forces
i
Supplemental Expert Report
of
Edward
E Leamer
Ph D
Case5 11 cv 02509 LHK
Document424
Filed05 17 13
2
Page3
62
of
5 10 2013
CONFIDENTIAL
I
Summary
Introduction Assignment and
1
been asked by counsel
I have
my
RD
and
2
my
1
job
somewhat
titles
structure
would
be
to
require
p283
the parallel
movementsof
orderly and prices
move
in parallel
compensation
rigid
that
of
For
for so
for
these firms
studies
vast
also evidence
compensation
I
found
same group from two
This again
is
zinc
I
Class
forces
are
that the
not
that
this
dictated
so
do not normally
with a
somewhat
compensation
me
and found
I have
titles
and the
the
the internal forces
compared
also
empirical
to separate
that
in all
average
non technical
employees
compensation curves of these two
firms move in a
compensation curves
for
much more disparate
are evident but the external
to detect
present correlations
To
allow
parallel while the
different
that
movementsof
role in determining
Class of
highly
average compensation of each
Technical
parallel
I
typically are not
copper and
silver
saying that the internal forces
more difficult
In this Report
forces
for the Technical
groups within each firm are
4
have added
I should
Markets
titles
models
estimated regression
for all the defendants
forces are
what
specifically in favor of the hypothesis
played an important
are evident but the external
way
to
In this report I confirm this opinion with two additional
I have
of a
the existence
Could a nonrigid wage
not only to be consistent
contributions of internal and external
the
Class employees in
market conditions which
labor
numbers of
titles
Creative
report
lines I responded
reason I regard the
many
my initial
could
it
example gold
that
wage structure but
internal equity
Yes
with
unusual external
highly
been asked to focus
I have
Defendant
asked in the deposition
a hypothetical
in
to the
can be
analysis that
over time further confirming
each
at
respond
proposed Technical
to the
structure as you’ve defined it lead to parallel
thought
further
compensation of Technical
total
move together
pay
Answer When
the
and
in this case
data
Classidentified
Technical
Does
rigid
available
prior analysis
the employees belonging
Class
Question
specific
3
on
response
Conclusions
for Class Plaintiffs in this matter to
following questions regarding
conducted based on the
of
title
accommodate
that
compare
the
movement
over
time of the
with the average compensation of the firms
titles
that
cannot be accessed
on a
by
title
Page 1
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11 cv 02509 LHK
Document424
Filed05 17 13
2
Page4
62
of
5 10 2013
CONFIDENTIAL
title
due to
basis
insufficient
titles but representing just
analyzed correlations
of
data
approximately 63 percent of Technical
6 percent
relatively
of Class Period
narrow groups
computed
are
of employees
each
Class employees of that
approximately a tenth of the Technical
correlations
employee years
for all titles not just
20
They
reveal
Class
I also
comprising
firm These
that
there
is
topdown
amount
large
co movement
of
among most
of compensation
Class titles of each defendant These correlations
method in which
budgeting
receive
a
common compensation
and
possibly
by
The
evident
substantial
what
membersof
title
I previously
the firm in any given year
called
a
is
depending
common firmwide
with a
are consistent
increment which
within the
individual
what
creates
all
of the Technical
adjusted
on
by
title
circumstances
specific
component
somewhat
somewhat
of compensation
is
rigid salary structure which
allows the effects of the anti cold calling conspiracy
to spread broadly across
each firm
5
2 Do
Question
partly drove
as internal equity
opposed to only
6
Answer
I
the data show additional
external
have analyzed a model of sharing
bytitle
title
relatively
compensation of the
year byyear
I report
increases
explanatory variables
the previous years
average
title
Technical
below estimated
in
of Technical
other Technical
that
titles
title
by
Class
cannot be accessed
Class titles but representing
1 increases
in
multiple
the
regression models that explain
title
average Technical
ratio of average Technical
3 the previous
title
against
Class employees
average compensation at the
revenue divided by the average
in
movementsof
narrow groups of employees
overall
compensation
software jobs
of compensation effects
of Class Period employee years I also analyzed the
compensation of
Specifically
compensation structures as
Again to accommodate
approximately 70 percent
just 8.4 percent
such
market forces
employees compensation
7
the Defendants
within Defendant firms relative to
title
evidence that internal factors
level in terms of four
Class compensation
2
Class compensation divided by the
years
compensation
ratio of
firmwide average
4 the
percent change in
the San Jose Sunnyvale Santa Clara Metropolitan
Area hereafter San Jose
the
Statistical
MSA
Page 2
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11 cv 02509 LHK
Document424
Filed05 17 13
2
Page5
of
62
5 10 2013
CONFIDENTIAL
8
the vast majority of individuals
I find that
1 positive
within
fall
contemporaneous sharing of compensation
across time that would spread gains in compensation
This
is
with
consistent
my
previous opinion that
employees would have been impacted
Furthermore the
an
9
3 Do
No
Answer
do not
agreements
the existence
of
member compensation
of large groups
would not have been harmed by
necessarily
I
have performed the above mentioned
find persuasive
evidence
compensation structure The
this regard
roles
a
The
show
ripple
analysis
shows
case bycase basis
the collection
it is
of other
they differ in the job
4
Question
job
titles
Answer
based
No
correlations
on
there are sizeable groups whose
be candidates
statistically
titles
among
positive
An
each
of these positive
vast
rigid
performed in
across employees in very different
title
or group
Technical
statistical
is
separately on
studied
how much
titles
variability
statistical
many
strong positive
exhibit
One
Class there certainly are exceptions
with the overall
correlations
of
Class
relationships
from the class However
the thousands
corrected estimated of
correlative
majority of
titles
Technical
some
even
the true correlations
if
all
negative
for this kind of data
when
this
is
done the
of these negatives
is
Class
this is not justified
can cause
model
pooling of evidence across titles and
that
analysis I
to identify and exclude from the Technical
with negative
appropriate
and regression
somewhat
I
they contain are found to be tied closely together
for exclusion
because
estimates
when
from Defendants
or groups All these groups no matter
with the overall
titles
analyses
statistical
found that compensation almost always moves with
titles
a lack
on
to suggest that
spillover effects
possible
it
a restriction
grouped by compensation level
correlation
that
Although the
might consider
to
Is
and
of class
of employees
compensation might have been disconnected
12
Defendants
all
indicates
of class
the data show the existence
for distinct subgroups
separately
11
titles
coldcalling
memberswho
10
2 sharing
market forces
than only external
Question
or almost
non compete
the structure
and
effects
across other job
of gains over time strongly
sharing
internal sharing force driving
rather
by the
all
show
or groups that
titles
correlation
allows
are
some
analysis indicates
positive
In other words
Page 3
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11 cv 02509 LHK
Document424
Filed05 17 13
2
Page6
of
62
5 10 2013
CONFIDENTIAL
matters for interpreting the evidence about each
it
estimated correlations
13
In
sum
the
statistical
and econometric
of a somewhat
are positive
analysis I
evidence
conduct
herein
my original
in
rigid pay structure
at
II
Most
provide them information about
finding
each Defendant that would have
broadly including throughout
subscribe to services
market
prices
the imminent threat
opportunities with or without
However
these external
relevance and
the
offer
of loss of
is
jobs Such
an employee
suggest
with the information conveyed
That can ring off a loud alarm that
to
averages with limited
Regardless of what these services
compare
are intended
in line with the external
sources provide broad industry
reliability
information cannot
that
for various
information helps them keep compensation packages
outside
my original
Compensation Structures
of
not all of these defendants
if
with the economic
Class
Defendants Use
14
conjunction
reports supports
transmitted the effects of the agreements
Technical
the vast majority of
that
title
heard
all
their
by an
actual
way up
the
to the
CEO
15
The
information by an outside offer or even a cold
by management
A
affected
that
chain of
individual
threat
move
against
that
threat
similar employees
compensation
action
aware
by an
compensation
to salary increases
for a
First when
one
for
also of the implicit competitive
may
increase
Though
directly
opportunity
outside
the
feel
in
wise to
it
make a
compensation
market
does
for these
not require a
until they actually
receive
an
can minimizethe disruption to employee
might occur when an employee discovers
unfairly undercompensated
analogous
individual
bump in
for these similar individuals
offer preemptive
loyalty that
attractive
and management
to similar individuals
bump in
the specific
can transmit a
may make management
newlythreatened
outside
similarities
becomes aware of an
this
preemptive
much beyond
broadly across a firm for two reasons
single individual
management
can go
a response
can stimulate
call
A broad
preemptive
that
he or she had been
response
is
completely
that are tied to information provided
by
Page 4
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11 cv 02509 LHK
Document424
Filed05 17 13
2
Page7
62
of
5 10 2013
CONFIDENTIAL
employment
These responses
16
Similarity in
broad and not
are
worth
is
a
bump in
compensation
individual
jobs
can be broad
productivity
employees
who are
any
to treat
employees
individual
until
are likely
to have
an
perceive
that
compared
Fairness
some
with
is
feels
their
is
salary
employees
equivalent
As
doesnt
actually
increases
own views
in
their productivity
it
is
essential for
an
require
increase
in
but the force
materializes
compensation
adversely
above a
In addition
affected
similarity
if
they
unfairly high compensation
are receiving
reflect
and there
really
is
reason
no sure way to know
who got
care The
title
that
why companies
bump in
and grade
managements views of what
is fair
and
employee
of
structure
it
may
fairness beliefs
tend to follow guidelines laid out in terms of
ranges so employees can be assured
reasonable
high
contented
of job and performance
the perception of similarity that determines
the
come from
discussed in the paragraph
threat
to the employee
who doesnt
compensation may
This
encourage
that
them
compensation and
influence
outside
a matter of personal opinion
who
policies
and commitment
fairly
preemptive
and these employees can have
exactly
HR
salaried
range of achieved
levels of productivity
the contentment
fairness can necessitate
employees
17
yet the
market view of employee compensation
salary of
of
highest
can be
worker in most
of each
is
committed to the mission of the enterprise In order to
maintain or to increase
management
with accuracy
Fairness
marketbased
problem with
Firms need to use
The
levels of productivity
A critical
tied together
for a single individual
that the productivity
to determine
is difficult
strictly
is
can be
salaries
compensation
transmitted broadly across a firm
based
necessarily individual
one reason why
why
the second reason
market
regarding the compensation offered by the
services
that their
compensation
falls
within
range of their colleagues
Page 5
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11 cv 02509 LHK
Document424
Filed05 17 13
2
Page8
of
62
5 10 2013
CONFIDENTIAL
III
Somewhat
Empirical Methodologies for Exploring the
Rigid
Salary Structure
A Choice of Aggregation
18
The
data set
explore
I
on
individuals
composed
title
Based on
of compensation records of salaried
of the Defendants
the payrolls
the Defendants by
by grade
is
Level
and
some
for
These
of the Defendants
19
the
the
including
have chosen
I
data
individual
the individual
at the individual
studied
more aggregated groups
because
is likely
titles
are
grouped
the
title
in the
been
have
that
level
work
to
at
first
level or
with the
title
to be dominated by forces
which can make
level
titles
grouped by
Class titles 1
as Technical
These data could be
the
are
from counsel regarding the employees
instructions
Class except for Lucasfilm I limit the inquiry to the
identified
individuals
it
difficult
averages
that operate
at
the firm wide effects
to detect
spread of the anti cold calling agreements
some
broadly across the
titlelevel
firms
Averaging
across individuals
in a title
more transparent
thus making the firmwide effects
effects
documentary evidence shows Defendants
20
I have
internal equity
discovered
there are
some
that
titles
individual
estimate
adequate
The
title
and
which
behavior
still
for the statistical
Because Lucasfilm
inquiry
is
expanded
meaning
Technical Class for
did
the
data
analysis
works
briefly
and
well for
effect
many
but not plentiful
not provide
title
statisticians
the
idiosyncratic
that
Titles that
call
for
Apple which had
I
am
seeking to
statistically
data prior to 2006 there are insufficient
all
Defendants expect Lucasfilm
for
whom
it
applies
all
in
This
is
2005 and
of data unless the
presented below
to
annual
insignificant
estimates
years
is
have fewer
a title restructuring
Lucasfilm employees Hence the analysis
cover
but
titles
are other titles that
there
sets are too small to yield accurate
to
all
many employees
employees
masks the firmwide
work
the
structures
data set contains only eleven annual observations which
particularly troublesome
1
In addition a
used to manage their
seem much influenced by
that
observations tend to produce what
results
their
were used only
that
are sparsely populated
among
bytitle
the
individual
perspective on the compensation
analysis provides a clearer
and maintain
can average out the
is
limited
to
employees
Page 6
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11 cv 02509 LHK
Document424
Filed05 17 13
2
Page9
62
of
5 10 2013
CONFIDENTIAL
which did not provide
for Lucasfilm
include
few
a
just
and furthermore
individuals
for sparsely populated
21
To
of
the
much from
data the
the individual
with highly
titles
title
across
compensation
come and
go
I give
and
headcounts
variable
that
titles
averaging
the
of the titlebytitle data I also include
limitations
work but applied
statistical
In addition
median ages
variable
deal with
2006
can vary wildly as individuals
Adobe
of
prior to
not benefit
unlike
titles
some examples below
highly
may
individuals
titles
the ten groups of
titles
to ten groups
by ordering the
of
titles
titles
the same type
firm
in each
I have
formed
by average base compensation and
employeeyears
then
the
splitting
into
titles
ten deciles
based
on
the
number of
2
B
22
Correlation Analysis of
Economists often look
closely
different
The
relationship
question
move
compensation
23
There
are
to correlation
move
variables
from 0
absolute value
to
sign
in the
Compensation Structure
together
1 One indicates
on
the
correlation
same direction
structures
two types of
and
correlations
Correlation coefficients
perfect
indicates
I begin
how
range in
correlation zero indicates
whether or not the
my analysis
no
series in
Defendant
of
correlations
relevant
for determining
if
the
of compensation
series are similar correlation
of compensation changes
The
correlations
the levels of compensation emphasize longer run movements
correlations
statistically
with compensation correlations
compensation movementsof two
levels
to measure
coefficients
of the log of
and the
of the change in the log of the levels focus on year byyear
movements
C Regression
24
Analysis of
compensation and
For several
number
of
title
sharing
2
Correlation of
but could
effects
Defendants
certain
large
Compensation Structure
titles
also
made
class
come from
splits into
ten
compensation could come from
third variables
that
groups impractical
In
operate
on both
those cases a smaller
groups was used
Page 7
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page10
62
of
5 10 2013
CONFIDENTIAL
title
and
class
compensation
of a somewhat
confirm the existence
my
correlation
analysis
I
model which forces the
an explanation
of title
This regression
25
model
compensation
the
class
equation
increases
two
26
The
regression
class
effect
one
the
variable
includes
title
a
they
and
for the other
sense the
second
move apart
sharing
variable which
in the previous year
the
in
closely
If
the
the ratio of
is
While the
two compensation
coefficient
which the
is
on
positive
is
levels
action
first
move
taken at
this
class average
abnormally high compared
tends to get a special increase
in
the coefficient
two would be
the extent to which corrective
is
If
variables
effect
compensation
company
setting
the extent to which
market forces
in that
that following periods in
at the
regression
after controlling
for
increase
encompasses the
In the multiple
measures the extent to which
means
it
compensation
title
model
company when
as
then the employee would inherit 100 percent
sharing
first
the second measures
together
variables
By including the
at a particular defendant
after controlling
the
compensation to
sharing
regression
average real inflation adjusted
regression
variables
of the class compensation changes
is
a multiple
in class average real total
move together
were equal to
This
title
variables 3
equation the
in the
in particular
tied together
in
one of four explanatory
compensation
of this variable
revealed by
structure
compete with other
to
the increase
includes
allows us to determine
this variable
and
compensation
class
analysis of these
correlation
compensation
rigid
market forces To
example
examine company by company
explains
compensation
in class
title
as
for
compensation
compensation and
total
same time
at the
with the
compensation to bring it back
title
the
in line with
the class
27
The
regression
against
model
is
divided by the average
For each
title
title
both
two other determinants of
these other variables
3
requires
regression
I
of these sharing
title
compensation
the previous years ratio of
title
compensation
exclude from the class average
variables to
at
company
firm
wide
This variable
real total
the
compete
One
of
average revenue
allows us to determine
compensation the compensation
of
the
itself
Page 8
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page11
62
of
5 10 2013
CONFIDENTIAL
which
that
titles
on
28
share increases
and
firms
fourth variable
the percent
is
workers
there are general
2
other Technical
increase
in
coefficients
with this
the firm and
ones
are the
may
It
who would
title
increases
a title that
Figure
in software jobs in the
which
1 as
for the
is
intended
effect
two sharing
in the
compensation
received
variables
a
move
how hot
a relatively small percent
company
years This indicates
last
year
sharing
in a fashion that helps align workers
meaning
increase
if
1
Class titles
increase
and
at
relative to
tend to receive
will
a positive
In this
are positive
of other Technical
or
MSA
title 4
compensation
MSA
San Jose
to reflect
estimated for one Intel
can expect to receive
Class titles at the
in subsequent
forces
have revenue
have an accentuated
effect Both the contemporaneous and lagged coefficients
equity
might be expected
job market generally not just in the San Jose
this regression
example the two
that
growth
job market variable
cold was the technical
I illustrate
workers
overall
success
This the external
29
creative
in firm revenue
with their firms since they
relationships
the
The
any
technical
critical
sharing
if
a larger
internal equity
suggest
that internal
compensation together
with that of employees in other roles at the firm
4
As mentioned
afford
before this regression
a sufficient
Estimated
and
are
number
excluded
of
is
estimated separately
observations
from the
6 observations
coefficient
for
or
distribution
each
title
7 consecutive
calculations
and
company
Titles that
years are treated
as
do not
Not
presented in this report
Page 9
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page12
of
62
5 10 2013
CONFIDENTIAL
Figure 1
Example of Compensation
Illustrative
Intel Named
Plaintiff
Title
Sharing
Variable
RD
Total
Average Annual
0.784
0.064
12.238
0.000
1
0.251
0.098
2.562
0.051
1
0.032
0.094
0.346
0.743
Employment
0.092
0.126
0.731
0.498
0.541
0.411
0.698
Compensation
Total
Compensation
1
Avg Annual Total Comp
Avg Annual
Title
Total Compensation
Forces Variables
Log
Firm Revenue
Avg Annual
Title
DLog
San
Information
Jose
1
Per Employee
Total Compensation
Sector
Constant
Observations
10
Rsquared
1
2
3
0.986
Significant
Title
RD
1
at
All
Compensation
Defendants
A Titleby
is
is
computed
employee
are
5
level
Significant
computed as the average of
over
all
Technical
title
Creative and
10
level
RD
total
compensation
employees other than the
tilte
itself
Adjusted
Inflation
compensation
data
Title Correlation Analysis of
for all Defendants
Exhibit 2
other Defendants
but here
it
is
3 which
indicate
enough
to
at the
Below
between
of
titles
title
Compensation Structure
are reported in Exhibit
I will
summarize the
the fractions
correlations
compensation
at
employee’s annual
Based Correlations and Multiple Regressions
correlations
positive
at
Significant
Variables
Results of Title
The
level
Average Compensation
Avg Total Comp
Source
30
5
Variable
Effect
Log
IV
4
Variable
Effect
RD
DLog
Note
Pvalue
value
3
0.223
Average Annual
Title
Contemporaneous
External
T
2
Variable
DLog
Lagged
7
Std Error
Coefficient
1
Dependant
Model
Regression
SOFTWARE ENGINEER
discuss
overall
the
1
Adobe
and
results in detail
results with Figure
weighted
2 and Figure
by employee years with
compensation and Technical
same firm restricted
Adobe
to titles with
six
Class
or more annual
Page 10
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page13
of
62
5 10 2013
CONFIDENTIAL
The
observations
Although there
that
among
32
are
error
which
thousands
the
Moreover
statistical
that
of
titles
enough
produce some
to
included even
if
all
of cases
the true correlations
for each
Once
title
toward
this shrinkage
is
is
strong support
Share
are formal
on
will
all
the
titles
the results indicate
that
which
for
be positive strengthening the
movementswith
the class overall
Change
Correlations
Change
of
Compensation
mean across
the
done
the corrected results
that all titles in the class share
2 Large
were positive
based
titles
with
estimates
There
are positive
mean
probably have similar correlations These methods
estimates
of these negatives
negative
are positive
allow pooling of results across
titles
computed
are
true correlations
that
of course positive
Figure
large
is
all
the
would shrink the
conclusion
These estimates
are negative
that
methods
assumption
many
estimated correlations that does not
negative
the fact that the vast majority
for the conclusion
is
some
any true correlations
statistical
more extreme
of correlations
distribution
31
with five or fewer tend to produce a
titles
Correlation by Titles
are Positive
100
80
60
Share
40
20
0
ADOBE
GOOGLE
APPLE
INTEL
Negative
Source
Note
Distribution
of
Defendant
growth
in
Employee
Positive
Compensation
avg compensation
Weighted by class
PIXAR
INTUIT
correlation
period
Data
over
employee
Correlation
titles
with
Analysis
six
or more
years
of
data
years
Page 11
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page14
62
of
5 10 2013
CONFIDENTIAL
Figure
3 Large
Share
of Level
Compensation
Correlations
Correlation
by
are Positive
Titles
100
80
60
Share
40
20
0
ADOBE
GOOGLE
APPLE
INTEL
INTUIT
Negative
Source
Note
Distribution
Defendant
of log
avg
Weighted
33
It
is
not just
statistical
Positive
Employee Compensation
compensation
variability
correlation
by class
period
that
Data
over
with the class overall
examination of
correlations
some
of the
with
six
Adobe
Analysis
or
more
years
of
data
years
can explain
the
negative
or small
of employees within a
employees come and go can cause changes
close
Correlation
titles
employee
correlations Changes in the composition
normal correlation
PIXAR
in title
that
as
compensation and mask the
I will illustrate
titles
title
this point
below with
a
have low or negative
with the class
Page 12
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page15
of
62
5 10 2013
CONFIDENTIAL
Figure 4
Summary of Compensation Change
Positive
Employer
Sign
Not
Significant
Percent
Negative
Significant
Sign
Not
Significant
Percent
Correlation
Percent
Significant
Percent
Total
Percent
ADOBE
67
32
0
0
100
APPLE
54
35
1
10
100
GOOGLE
76
22
0
2
100
INTEL
94
6
0
1
100
INTUIT
81
14
0
5
100
PIXAR
86
13
0
1
100
Source Defendants
Note
Distribution of growth
in
employee
compensation
data Correlation Analysis
compensation
over
correlation
titles
by class period employee
Weighted
with
six
or
more years of data
years
Figure 5
Summary of Compensation Level
Positive
Employer
Sign
Not
Significant
Percent
Correlation
Negative
Significant
Not
Significant
Percent
Sign
Significant
Percent
Percent
Total
Percent
ADOBE
92
5
0
3
APPLE
78
16
1
5
100
GOOGLE
83
16
0
1
100
INTEL
85
14
0
1
100
INTUIT
45
40
2
12
100
PIXAR
84
15
0
0
100
Source Defendants
Note
Distribution of log avg
employee
compensation
Weighted
compensation
correlation
100
data Correlation Analysis
over
titles
by class period employee
with
six
or
more yearsof data
years
Page 13
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page16
of
62
5 10 2013
CONFIDENTIAL
B
34
Titleby Title Multiple Regressions
As
above
described
compensation
that
I also analyzed a multiple
the year byyear increases
explains
Class compensation
average Technical
1 increases
jobs in the San Jose
The
data set
many
is
title
words
36
seven
least
6 and
observations
Those
results
positive
show
sharing
relationship
significant
for statistical
Figure
Second
of titles
If
This model is completely appropriate
on the other hand
title
A
if
only to
would
yield a small
largely
determined by variability in compensation
estimating
makes
it
and probably insignificant
less likely
effect
of
force
title
all
of
majority
these are
To
not more likely to find a sharing
titles
by sharing
body of
relationships
significant
across
class overall
attempt to link
a
are statistically
in the context of
statistically
came from the
B
Class
or the lagged
relationship
all
In
sum
Class titles
equally
A to
across
all
at
a
titles
the class overall
unless the variability in compensation
title
with at
have
titles
a small fraction
supported
my
Technical
wins
employee years
results occur
B then
In other
titles
of
The
5
effects
class wide results for
is
for internal relationships
the sharing
were connected
results are particularly
it
contemporaneous
Many of
statistically
that
First the vast
relationship
a heavy
is
sharing
their Class Period
a general
is
2011 and
significant coefficients
of those that are negative
this analysis provides support
5
in software
revenue sharing
approximately 30 percent
Third even these negative
majority
the
significance
7 below
in either the
regression
insignificant
statistically
the following
effect
evidence that there
for the vast
statistically
more
and more than 91 percent of
37
change
number of
in the
market effects and
have
in the competition
I present in Figure
title
average revenue
percent
the
four variable
reflected
is
The
prevalent for the external
variables
A
observations
coefficients
two sharing
firm
wide
observations from 2001 to
limited to eleven annual
burden with such data which
insignificant
4
compensation
ratio of
years
firmdivided by the average
the
at
average
in
MSA
have fewer
titles
2 the previous
previous years ratio of
divided by the average
35
firm
Class compensation
3 The
compensation
the
at
compensation
in average
at the title level in terms of four explanatory variables
Technical
model of
regression
of
the class
were
put this in simple terms the model that I
am
effect
Page 14
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page17
of
62
5 10 2013
CONFIDENTIAL
firm that would tend to
make
impact
agreements
of the
common to
all
Class
members
38
Thus
the vast majority
relationship
is
6 Large
my
to support
non compete
common across
Figure
titles
with other Technical
of these results
alleged
of these
Share
of
a positive
Class titles
at
internal equity
the same firm
previous conclusion
agreements
the Technical
have
would be
The
sharing
implication
that the impact
common across
of the
the class and
Class employees in particular
Contemporaneous
Contemporaneous
Coefficients
Coefficient
by
are Positive
Titles
100
80
60
Share
40
20
0
ADOBE
GOOGLE
APPLE
INTEL
Negative
Source
Note
Distribution
of
Defendant
estimated
Employee
PIXAR
INTUIT
Positive
Compensation
Data
Regression
Analysis
contemporaneous
coefficient over
titles
with seven
Weighted by class period
employee years
or
more
years
of
data
Page 15
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page18
of
62
5 10 2013
CONFIDENTIAL
Figure
7 Large
Share
of
Lagged
Lagged
Coefficients are Positive
by
Coefficient
Titles
100
80
60
Share
40
20
0
ADOBE
GOOGLE
APPLE
INTEL
Negative
Source
Note
Distribution
Employee
Defendant
of estimated
lagged
PIXAR
INTUIT
Positive
Compensation
coefficient
Weighted by class
over
period
Data
titles
Regression
with
employee
seven
Analysis
or
more
years
of
data
years
Figure 8
Summary of Contemporaneous and Lagged Net
Positive
Employer
Sign
Not
Significant
Percent
Negative
Significant
Percent
Sign
Not
Significant
Effect
Percent
Total
Significant
Percent
Percent
ADOBE
22
75
0
3
100
APPLE
23
62
0
14
100
GOOGLE
12
69
2
17
100
INTEL
88
11
0
1
100
INTUIT
PIXAR
73
23
0
4
100
60
39
0
0
100
Source
Note
Distribution
of
Defendants
the
employee
compensation
sum of estimated contemporaneous
Weighted
data Regression Analysis
and lagged
coefficients
by class period employee
over
titles
with
six or
more years of data
years
Page 16
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page19
62
of
5 10 2013
CONFIDENTIAL
39
It
may be
important
relationships
Adobe
many
that
parameters
with
separately
40
The
regression
all
much
results for
1
Adobe
The
but instead
titles
relevant
here because
regression
coefficients
first
a
titles
t statistic
in
data
it
is
18
in excess of
excess of
2
has a
tvalue
above
this
would
yield a small
the
R
Among the
Column 15 The
has one negative
title
A
2 only 4 of 41
and no
information about
of
correlations
Column
9 which
compensation
is
is
together in Section
said to
4
is
produce
for a sharing
were connected
force
fit
sq in the
and probably insignificant
effect
of
titles
B
Column
market variable
variable
tstats The
significant
B then
unless
tstatistics more
and the revenue
results are
of observations diminishes
comes from the
title
reason
with eleven years of
titles
external
number
title
that
my
class overall
equally
A to
attempt to link
the variability in
The model that
I
across
all
the class
compensation
of the class
am
makes
estimating
it
effect
of the model associated
last
that
only to
For
than the class wide
positive
into the table as the
not more likely to find a sharing
The increment in the
comparing
7
by conventional standards
were largely determined by variability in compensation
7
not
of the estimated regression 7
part
in absolute value
Column 16
variable
in excess of
model looks
on the other hand
less likely
title
wide variables6
by the
are sorted
in average real
is
each
is far
regressions
are reported to their right in Section
are highlighted
effect
more mixed deeper
noted
2
corrective
Column 17
overall
descriptive
of the four variables are collected
significant estimate
contemporaneous
If
The
Adobe
the two sharing variables that jump out with high
often the
As I
which
connecting
Column
this correlation
require
tstatistics
statistically
titles
would
or more years of data are
two Sections give
tstatistics
3 and the corresponding
6
this
with the
with seven
titles
between the percent change
the correlation
Roughly
with other
For example with 101
titles
simpler structure
the log levels of average real compensation
41
by the conspiracy
more observations
the data and the two correlations These
The
a matrix of sharing
to estimate with only eleven years of data
of the other
reported in Exhibit
more
is
the estimation of a 101 by 101 matrix of connections
have estimated have a
I
there
in principle
directly affected
titles
in the class with six or
titles
potentially
too
connect
that
that
are tied together with these affected
that
titles
to understand
with the
column with the squared
last
three
explanatory
variables
can be found
by
of the correlation
Page 17
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page20
62
of
5 10 2013
CONFIDENTIAL
42
This confirms the
titles
The
summaryabove
almost always positive
to the lagged ratio of
indicates
that
the
if
title
V
Decile
43
equal
variable
compensation
in the
title
from its normal relationship
with
taken to either raise or lower compensation
in
Based Correlations and Multiple Regressions
titlebased
groups of
titles
regressions
form the
ten
deciles based
A Decile
The
To
on which
I split each
groups
these
to conduct
the
Defendants
I ranked
correlation
correlation
on the
study
analysis
I have
formed
and the multiple
To
Class titles into ten groups
Technical
basis of average
lifetime of the title
analysis of the ten groups
Class employees
ten groups that had positive
for the levels correlation
inflation adjusted
and then divided these up
shows both immediate
structures
yields strong evidence
for each
into
correlations
its
firms
and longrun
with and supports
of both short
subgroup of the Defendants
Figure 9 and Figure 10 indicate
with the Technical
and 10 out of 10
every group shares in
consistent
in this
that are
Correlation Analysis
and long run compensation
Thus
titles
titles
on employee years 8
Based
Technical
titles
excludes
necessity
include
compensation over the
total
by
study just described
populated
infrequently
is
departs
corrective
title
The
44
is
the
relative to title
compensation
action
on
coefficients
compensation
the class then corrective
the
providing direct evidence of sharing across
for the percent
the numbers of the
Class 10 out of 10
change
correlations
compensation structure Every group
correlation
my conclusion
structure
that
for every
group
This
the Defendants
compensation was semirigid
8
Since Lucasfilm
Although
I
did
not provide
title
data individuals
were ranked
attempted to break the firms up into 10 equal
sized
in
groups
a similar fashion
equal
for
Lucasfilm
based on employee
years some
groups end up being larger than others because of some big titles
Page 18
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page21
of
62
5 10 2013
CONFIDENTIAL
Figure
9 Large
Share
of
Change
Compensation Change
Correlations
Correlation
are Positive
by Deciles
100
80
60
Share
40
20
0
ADOBE
GOOGLE
APPLE
INTEL
INTUIT
Negative
Source
Note
Distribution
Defendant
of growth
in
Employee
LUCASFILM
PIXAR
Positive
Compensation
avg compensation
correlation
Data
Correlation
by class
weighted
Analysis
period
employee
years
Figure 10 Large Share of Level Correlations are Positive
Compensation
by Deciles
Correlation
100
80
60
Share
40
20
0
ADOBE
GOOGLE
APPLE
INTEL
INTUIT
Negative
Source
Note
Distribution
Defendant
of log
avg
Employee
compensation
LUCASFILM
PIXAR
Positive
Compensation
correlation
Data
weighted
Correlation
by
Analysis
class period
employee
years
Page 19
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page22
45
62
5 10 2013
CONFIDENTIAL
B
of
Decile
Based
Multiple Regression Results
have
Multiple regressions
summarized
in Figure
contemporaneous and
also
been estimated with these
11 and Figure
laggedare
12 below
the
positive
decile
data As
sharing
effects
both
rule
Figure 11 Large Share of Contemporaneous Coefficients are Positive
Contemporaneous
Coefficient
by Deciles
100
80
60
Share
40
20
0
ADOBE
APPLE
GOOGLE
INTEL
INTUIT
Negative
Source
Note
Distribution
Defendant
of estimated
Employee
LUCASFILM
PIXAR
Positive
Compensation
contemporaneous
coefficient
Data
Regression
weighted
by class
Analysis
period
employee
years
Page 20
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page23
of
62
5 10 2013
CONFIDENTIAL
Figure 12 Large Share of Lagged Coefficients are Positive
Lagged
by Deciles
Coefficient
100
80
60
Share
40
20
0
ADOBE
GOOGLE
APPLE
INTEL
Negative
Source
Note
46
The
Employee
Defendant
Distribution
of
almost always positive
indicate
that
the
if
title
Compensation
Data
Regression
by class
weighted
on the
coefficients
compensation of
Analysis
period
employee
years
corrective
a decile departs
with the class then corrective
relationship
PIXAR
Positive
lagged coefficient
estimated
LUCASFILM
INTUIT
action
is
variable
from
in Figure
12
normal
its
taken to either
or lower
raise
classwide
compensation
impact
in the decile
suppressing wages in
The cold
some
calling conspiracy
titles
would have some
effect
averages which in turn would suppress compensation
the
47
titles
are several
important
in either the
Second those
things to note
contemporaneous
group Many
of these are
of positive
statistically
support for internal relationships
impact
common to
of negative
First every group
relationship
are not statistically
that are negative
in the context of evidence
make
in all
direct
on the
or almost
all
of
in the class
Figure 11 and Figure 12 contain a few instances
effect
would have
that
sharing
across
all
has a positive
There
sharing
or the lagged relationship
significant
relationships
significant
estimates
In
Third these occur
for almost every
sum this
analysis provides
these groups that would tend to
each
Page 21
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page24
62
of
5 10 2013
CONFIDENTIAL
48
Here
I
want
to issue another
coefficients
It is important
the changing
composition
number of
adding a
compensation
workers
impact
To
into account
test this opinion
Adobe
dataset
as set
examined the
any of the
51
Adobe
The
have
forth
for each
closely
conclusion
I previously
statistical
which
the
reported
10
in their
headcounts
in
Class
and
outputs for the
correlation
my view
confirm
that the title
is
consistent
title
I
find nothing
have
similarly
in that
data to
that
the
movement
Class with the
compensation
of a
in the rest of the Technical
and the
class have
with sharing
Common Factors
characteristics
compare
movement
of the
Analysis
moves
in
a way
A
that
Class thus supporting the
coordinated compensation
that
compensation
individual
characteristics
title
title
title
levels
a
of gains and broad impact of the anti cold
on measured
within a
of real
Class overall but excluding the selected
means
but also depends
in
the Technical
confirmation that at least some individual
changes in the individual
the
correlation
demonstrated with the
on
examined the
below They
title in
compensation
similar to
any year depends
for titles with just a
find no compelling reason
from the Technical
of the other defendants
numerical correlations
high positive
9
of individual
large changes10
I
the
Adobe Correlations
compensation of the Technical
fact
average
company and
variability
of these data
titles
For instance
titles
be a bigger problem
to
9
Correlation Results
compensation
is
title
by
to the question of whether
Idiosyncratic
experience
the limitations
I
can be affected
this conclusion
contradict
1
data
going
title is
Additional Exploration of
50
within each
workers might bring down the
for titles that
this analysis to exclude
VI
of the workforce
of the unlawful agreements
within a
of negative
to realize that these coefficients
or vice versa for reasons unrelated
few employees and
Taking
about misinterpretation
share broadly in things such as the gains of the
characteristics
49
junior
warning
matter
can cause changes
at
characteristics
in
the individual
including
level in
age This
is
and this raises the possibility that
title
compensation
that
can mask
firmwide common component
Though
a stable
headcount
can come
from equal numbers of departures and new
arrivals
Page 22
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page25
of
62
5 10 2013
CONFIDENTIAL
whether
calling conspiracy
the Technical
52
directly affects
it
on 5
correlations
based
The
sorted
to
6 and
is
first
shaded
Technical
53
The
This
first
is
in
green
they are populated
title
Titles with the weakest
since
6 or more years
shaded
statistically
in
insignificant
Class overall
with the Technical
statistical
The
was populated from 11
with the Technical
correlation
statistical
are
in
of years the title
of the
Class at
correlation
Adobe
with the
yellow
in Exhibit
1 has the
first
year of data for each
title
the early years from 2001 to 2003 had a sharp decline
Class compensation for
It
under study or the rest of
observations are often
by the number
column of numbers
early years thus are
together
or fewer
Adobe
Class at
important
Technical
if
then by the correlation
Titles with the strongest
are
title
Class
Titles are included in the table
table
the
an important
would not be
Adobe
test
surprising
in Figure
as illustrated
bed for
identifying
which
weaker
to find statistically
in
13 and these
titles
moved
results if these
years are not included
Figure 13
Class Average Total Compensation
Technical
Compensation
Adobe
Total
Average
Source
Note
Inflation
Defendant
adjusted
average
Employee
Compensation
compensation
with
Data
2011 as base
year
Page 23
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page26
62
of
5 10 2013
CONFIDENTIAL
54
The second column
populated This
the
reports
number of
also important
is
since the
during which the
years
truncated this table
the number
at
For
that
6 or more
of years equal to
reason I have
since
with 5 or fewer years populated
55
are estimated with greater statistical
The
number
2
56
It
third
column measures
Headcount Matters
is
my view
that
title
structure
and personal
characteristics
consequently
some
changes
go
variability
titles
in average
may have
as
which masks
is
and
comes from
employees come and
unusual employee
substantially
a close
fully
tenure
on compensation
of employees
lose or gain a large fraction
that
company
at the title level
of employee characteristics
that
structure but not
age experience
compensation
in
compensation
of these characteristics
like
title
have an impact
have just a few employees
and
characteristics
Variables
to
error
Correlations
influenced by the
are likely
of the change
in the distribution
Titles that
is
the cases
of employee years
for Interpreting
compensation
by the
determined
the
was
accuracy of the estimate of
statistical
depends on the number of observations
correlation
title
connection
influenced by
may have
variability
with the Technical
Class overall
57
The
Technical
Figure
may
14
Class overall
are losing
gaining
may have
a rising
headcounts
employee
headcount
as illustrated
similar to the Technical
characteristics
while
in
Class
titles
that
workers much more rapidly than the Technical
average compensation histories different
Class not because
employees
in
similar movementsin
workers or
Class overall
Technical
movement
Titles with
experience
has experienced
there
in the title is changing
is
from the
no sharing but because the group of
enough to mask the sharing
Page 24
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page27
62
of
5 10 2013
CONFIDENTIAL
Figure 14
Adobe
Class Average Headcount
Technical
per
Title
40
30
Title
20
Employees
10
0
2001
2002
2003
2004
2005
Source
3
58
2006
2007
Employee
Defendant
Compensation
2009
2010
2011
Data
Correlations
As
described
determining
correlations
above
if
the
there are two types of correlations
movementsof
Section
compensation
2
the
in Exhibit
in the title
two
third
change in the logarithm of average real total
Technical
The
for
are relevant
The
first
column
1 compares the logarithm of average
The
Class
which
series are similar
and the logarithm of average
of the rest of the Technical
59
2008
real total
of
total real
compensation
column of Section 2 compares the
compensation
of the title with the
Class excluding the title
tstatistics
corresponding
for these correlations
are reported immediately
tstatistics
following each correlation
greater than
years in
which the
two
title is
and the
are
statistically
shaded
populated
The
significant correlations
table is sorted first
and second by the
by
correlation
the
with
number of
between the
log levels
60
The
statistically
from the
longest
most
significant
time series with
correlations
all
with the shaded
tstatistics come
eleven years of data populated That
is
a
Page 25
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page28
of
62
5 10 2013
CONFIDENTIAL
of any
feature
significant
statistically
61
There
are
no
negative
populated These
statistically
4
62
exercise
statistical
is
time series the more
the
are the findings
correlations
positive
significant
the longer
in
for the
correlations
41 titles
with
all
eleven years
larger than zero
are statistically
39 out of the 41 cases
Outliers
To
fully
these correlations and the significance
understand
anomalies it may be
16 have the average
employees
helpful
real
to look at
some
Class overall
Figure 15
eleven years of data that are most highly correlated
overall and Figure 16 has the least correlated
together
which
is
The
title
with the lowest correlation
different but not dramatically
of the
data displays Figure 15 and Figure
compensation for ten Adobe
in the Technical
not
or
titles
is
titles
and
illustrates
for the
the five
with the Technical
All these titles
Adobe
titles
with
Class
move
TECHNICALWRITER
2
so
Page 26
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page29
of
62
5 10 2013
CONFIDENTIAL
Figure 15 Selected Adobe Titles with a Full 11 years of Data
Most Correlated
Average Total Compensation
Compensation
Titles
Total
Average
Source
Note
Titles
Defendant
with
Inflation
highest
adjusted
Employee Compensation Data Correlation Analysis
log compensation
correlation
among fully populated
average total compensation
with 2011 as base year
titles
Figure 16
Compensation
Correlated Titles Average Total Compensation
Compensation
Least
Total
Total
Average
Average
Source
Note
Titles
Defendant
with highest
Inflation
adjusted
Employee
log
Compensation
compensation
average
total
Data
correlation
compensation
Correlation
among
with
fully
Analysis
populated
2011 as base
titles
year
Page 27
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page30
62
of
5 10 2013
CONFIDENTIAL
63
However
as noted
above when
may change
characteristics
with the lowest correlation
change
headcounts
too The
substantially
are illustrated
is
for the two
headcounts
in Figure
17 The
headcount
very volatile with a standard
the percent
change equal to 72 percent compared
benchmark
of 11 percent
away
compared
with the Technical
with an average annual percent
Class benchmark
17 Headcounts
Least
with the Technical
of
for
of
Class
basically
12 percent
of
increase
titles
deviation
title is
withering
Figure
employee
substantially
5 percent
Least Correlated Titles
Correlated
Titles
Headcount
12
10
8
Headcount
6
Title
4
2
0
2001
2002
2003
2004
Source
Note
64
The
variability
problem
It
Titles
2005
Defendant
with
lowest
2006
2008
Employee
Compensation
Data Correlation
Analysis
compensation
correlation
among fully populated
for these
substantially
two
to the
smooth
elevation
of
age of
compensation
titles
not just a hypothetical
Class overall
jump upward
is
disconnect
in the Technical
2011
titles
which
in Figure
18
are
In
of the median age of the class the median age
has a big
contribute to the apparent
titles is
2010
the median ages for these
contrasted with the median age of the Technical
contrast
2009
log
in the headcounts
has affected
2007
in
2006 and
highly volatile
These
between compensation
Class overall
And
in any
the median
facts surely
in these titles
event these
and
results
Page 28
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page31
of
62
5 10 2013
CONFIDENTIAL
offer
no reason
pay structure
my conclusion
to question
that
applied to
all
these titles I offer these
two
presence of a few outlier
titles
how
conclusions about
its
let
been harmed
examples simply to
in the analyses
rigid
in
the point that the
illustrate
our
does not challenge
basic
these companies pay their employees which are also
alone convincing
evidence
18 Median ages
Least
Correlated
I
have not seen any
any of these
that
by the anti competitive behavior
Figure
a somewhat
exhibits
employees including those
salaried
by economic theory and the evidentiary
supported
evidence
of
Adobe
that
I
titles
would not have
have studied
Least Correlated Titles
Titles
Median
Age
45
40
Age
Median
35
30
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Technical
Source
Note
VII
Titles
Defendant
with
Employee
Compensation
Data Correlation
Analysis
compensation
correlation
among fully populated
lowest
log
2011
Class
titles
Internal Versus External Forces
65
The
regression
are generally
forces
I
analysis reported
more
detectable
expand on
above
indicates
generally
this finding
in this section
correlation
or the external
Class employees and the
within firms between
effects
market
with an examination of the
Class employees of each of the defendants
more
the internal sharing
than either revenue sharing
average real compensation for the Technical
Technical
that
I
show
non
here that there
is
these two groups than between
Page 29
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page32
62
of
5 10 2013
CONFIDENTIAL
firms for either group
Thus again
very evident while the external
66
Figure 19 below
for the Technical
compensation
average
that
For most defendants
closely
tracks
compensation
total
than between
RD and
is
firms In other
to detect
non Technical
for the
these two subgroups
should
It
much more
generally
forces are
the average total compensation
one another
words
more difficult
are
defendant
Class employees
the internal sharing
that
market forces
for each
illustrates
NRD
employees
I observe
also
Class
have total
be evident
similar within each
that
firm
the internal sharing forces dominate and
keep the compensation of the Technical
Class employees and the
non Technical
Class employees closely aligned
67
This visual observation
numerically by the computation of the
confirmed
is
over time of the change
correlations
in logarithms
of the average total
compensation between these fourteen groups of employees reported
The boxes down
Correlations in excess of 0.9 are shaded
RD and NRD
the within firm correlations
between
boxes refer to comparisons
between firms Four out of
boxes and
are in these
correlations
correlation
of 0.86
correlation
in
every
row and column
making
by very
contaminated
it
hard to estimate
large
bonuses
in
five of the
Google has an
the
percent
and
1
these
shaded
internal
is
the largest
Lucasfilm has a very
change
The
correlation
for producers
Table
Correlations outside
except for Lucasfilm
little variability
in
the diagonal contain
Furthermore the within firm correlation
short time series with very
compensation
in addition
real
in
Pixar
directors
data are
in
2002 and
2006
68
Table 2 has the levels correlations
that
capture the longer term
of the compensation series These confirm the importance
forces compared
with the external
forces
sense that the within firm correlation
column except
because
the
for Lucasfilm
Lucasfilm data
compensation
at
is
is
forces for
all
of the internal
but Lucasfilm
the largest correlation
Lucasfilm and Intel appear to
confined
co movements
in every
move
in the
row and
together only
to a brief period of stable growth of
both firms
Page 30
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page33
of
62
5 10 2013
CONFIDENTIAL
Figure 19 Defendant
RD vs NRD Average
Total
Compensation
Page 31
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page34
of
62
5 10 2013
CONFIDENTIAL
Table
Correlations of
Changes
in
1
Defendants Average
Total
Compensation
2001 2011
Adobe
Apple
Google
Intel
Intuit
Lucasfilm
Pixar
NRD RD
NRD RD
NRD RD
NRD RD
NRD RD
NRD RD
NRD RD
NRD
1.00
0.94
0.66
0.56
0.17
0.16
0.47
0.60
0.63
0.60
0.19
0.62
0.53
0.53
RD
0.94
1.00
0.64
0.65
0.13
0.24
0.34
0.45
0.53
0.51
0.12
0.67
0.51
0.37
NRD
0.66
0.64
1.00
0.93
0.48
0.17
0.02
0.16
0.85
0.73
0.08
0.87
0.56
0.16
RD
0.56
0.65
0.93
1.00
0.42
0.07
0.12
0.00
0.77
0.63
0.11
0.83
0.45
0.05
NRD
0.17
0.13
0.48
0.42
1.00
0.86
0.51
0.39
0.20
0.17
0.49
0.89
0.62
0.21
RD
0.16
0.24
0.17
0.07
0.86
1.00
0.53
0.50
0.09
0.06
0.68
0.83
0.50
0.19
NRD
0.47
0.34
0.02
0.12
0.51
0.53
1.00
0.97
0.31
0.30
0.01
0.92
0.00
0.89
RD
0.60
0.45
0.16
0.00
0.39
0.50
0.97
1.00
0.38
0.33
0.23
0.70
0.03
0.89
NRD
0.63
0.53
0.85
0.77
0.20
0.09
0.31
0.38
1.00
0.91
0.15
0.17
0.43
0.28
RD
0.60
0.51
0.73
0.63
0.17
0.06
0.30
0.33
0.91
1.00
0.51
0.55
0.63
0.34
NRD
0.19
0.12
0.08
0.11
0.49
0.68
0.01
0.23
0.15
0.51
1.00
0.24
0.03
0.38
RD
0.62
0.67
0.87
0.83
0.89
0.83
0.92
0.70
0.17
0.55
0.24
1.00
0.58
0.29
NRD
Adobe
0.53
0.51
0.56
0.45
0.62
0.50
0.00
0.03
0.43
0.63
0.03
0.58
1.00
0.29
RD
0.53
0.37
0.16
0.05
0.21
0.19
0.89
0.89
0.28
0.34
0.38
0.29
0.29
1.00
Apple
Google
Intel
Intuit
Lucasfilm
Pixar
Note
Values
above
0.9
shaded
Source Defendants employee compensation data
Page 32
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page35
of
62
5 10 2013
CONFIDENTIAL
Table
Correlations of
2
Defendants Average
Total
Compensation
2001 2011
Adobe
NRD RD
Apple
NRD RD
Google
Intel
NRD RD
NRD RD
Lucasfilm
Intuit
NRD RD
NRD RD
Pixar
NRD RD
NRD
1.00
0.88
0.17
0.17
0.43
0.73
0.18
0.58
0.50
0.41
0.15
0.04
0.33
0.38
RD
0.88
1.00
0.24
0.27
0.05
0.63
0.47
0.72
0.69
0.61
0.40
0.32
0.48
0.51
NRD
0.17
0.24
1.00
0.99
0.91
0.38
0.65
0.33
0.64
0.68
0.74
0.58
0.48
0.39
RD
0.17
0.27
0.99
1.00
0.90
0.33
0.69
0.37
0.64
0.66
0.83
0.72
0.46
0.40
NRD
0.43
0.05
0.91
0.90
1.00
0.67
0.53
0.13
0.36
0.44
0.81
0.59
0.46
0.28
RD
0.73
0.63
0.38
0.33
0.67
1.00
0.05
0.44
0.20
0.08
0.47
0.04
0.22
0.12
NRD
0.18
0.47
0.65
0.69
0.53
0.05
1.00
0.87
0.64
0.66
0.93
0.98
0.54
0.86
RD
0.58
0.72
0.33
0.37
0.13
0.44
0.87
1.00
0.65
0.62
0.91
0.96
0.48
0.90
NRD
0.50
0.69
0.64
0.64
0.36
0.20
0.64
0.65
1.00
0.94
0.63
0.54
0.55
0.54
RD
0.41
0.61
0.68
0.66
0.44
0.08
0.66
0.62
0.94
1.00
0.78
0.91
0.72
0.62
NRD
0.15
0.40
0.74
0.83
0.81
0.47
0.93
0.91
0.63
0.78
1.00
0.88
0.63
0.83
RD
0.04
0.32
0.58
0.72
0.59
0.04
0.98
0.96
0.54
0.91
0.88
1.00
0.62
0.86
NRD
Adobe
0.33
0.48
0.48
0.46
0.46
0.22
0.54
0.48
0.55
0.72
0.63
0.62
1.00
0.65
RD
0.38
0.51
0.39
0.40
0.28
0.12
0.86
0.90
0.54
0.62
0.83
0.86
0.65
1.00
Apple
Google
Intel
Intuit
Lucasfilm
Pixar
Note
Values
above
0.9
shaded
Source Defendants employee compensation data
Page 33
Supplemental Expert Report
of
Edward
E Leamer PhD
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page36
of
62
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Exhibit 1
Page37
of
62
Case5 11cv 02509LHK
Document424
Exhibit
2
Filed0517 13
Page38
62
of
1
Adobe
Section
First
Job
Title
Year
1
Years
of
1
Section
Total
Emp
Data
2
Level
Years
3
Avg
Emp
4
dlog
Avg
5
dlog
Std
Dev
6
Correlation
Coeff
7
T
Stat
8
2
Section
Change
Correlation
Regression
9
T Stat
Contemp
10
Coeff
11
Lagged
12
3
Section
Coefficients
Revenue
Regression
SJ
13
Emp
14
Contemp
15
Lagged
16
4
T
Section
Net
Stats
Revenue
SJ
17
Emp
18
C
L
19
5
Section
6
Effect
T Stat
Obs
r2
20
21
22
2001
11
170
15
0.27
0.34
0.90
6.07
0.89
5.55
1.18
1.04
0.12
0.02
5.15
6.71
1.77
0.07
2.22
8.15
10
0.98
2001
11
311
28
0.05
0.19
0.89
5.89
0.78
3.55
1.07
1.18
0.09
0.31
0.67
1.38
0.25
0.25
2.25
1.66
10
0.74
2001
11
371
34
0.11
0.16
0.89
5.73
0.79
3.59
0.67
1.33
0.12
0.34
0.66
1.95
0.45
0.36
2.01
1.99
10
0.81
2001
11
29
3
0.16
0.65
0.87
5.37
0.78
3.56
2.67
1.08
0.33
0.48
1.49
1.80
0.80
0.32
3.75
2.24
10
0.79
2001
11
82
7
0.10
0.25
0.85
4.87
0.72
2.97
0.89
1.09
0.46
0.58
0.65
1.99
1.23
0.39
1.97
1.39
10
0.77
2001
11
108
10
0.03
0.40
0.84
4.73
0.82
4.08
0.93
0.88
0.04
0.51
2.43
3.32
0.37
1.38
1.81
3.34
10
0.94
2001
11
96
9
0.12
0.37
0.84
4.65
0.85
4.56
0.80
0.59
0.05
0.84
1.93
2.68
0.45
1.89
1.38
2.66
10
0.95
2001
11
250
23
0.04
0.16
0.84
4.60
0.85
4.47
1.28
0.97
0.08
0.19
2.60
3.59
0.47
0.37
2.25
3.83
10
0.93
2001
11
559
51
0.11
0.20
0.83
4.53
0.88
5.31
0.94
0.80
0.21
0.04
2.27
2.28
1.45
0.08
1.74
3.24
10
0.92
2001
11
93
8
0.11
0.26
0.81
4.19
0.67
2.54
3.21
0.89
0.24
1.55
1.03
0.75
0.30
0.62
4.10
1.49
10
0.63
2001
11
14
1
0.00
0.45
0.80
3.97
0.63
2.29
2.50
0.06
0.51
0.17
0.50
0.04
0.40
0.04
2.57
0.56
10
0.57
2001
11
152
14
0.28
0.15
0.78
3.74
0.72
2.96
0.54
0.65
0.13
0.54
0.98
1.60
0.89
1.07
1.18
1.43
10
0.81
2001
11
202
18
0.06
0.25
0.78
3.74
0.70
2.78
0.68
1.24
0.21
0.34
1.30
4.27
1.40
0.67
1.91
3.24
10
0.92
2001
11
550
50
0.06
0.18
0.78
3.70
0.95
8.29
0.99
0.15
0.06
0.43
2.87
0.54
0.47
0.94
1.14
2.66
10
0.94
2001
11
234
21
0.07
0.22
0.78
3.68
0.73
2.98
0.97
1.14
0.12
0.29
1.56
2.19
0.43
0.48
2.11
2.22
10
0.82
2001
11
273
25
0.17
0.19
0.77
3.60
0.74
3.11
0.34
1.32
0.23
0.33
0.60
2.67
1.59
0.66
1.66
2.77
10
0.86
2001
11
327
30
0.11
0.14
0.74
3.34
0.82
4.00
0.66
0.40
0.11
0.19
1.39
1.12
0.74
0.38
1.06
1.67
10
0.78
2001
11
434
39
0.07
0.18
0.74
3.29
0.65
2.39
0.72
1.09
0.21
0.30
1.29
2.84
1.33
0.56
1.82
2.39
10
0.84
2001
11
196
18
0.13
0.24
0.74
3.27
0.82
4.06
1.23
0.57
0.09
0.02
1.48
1.38
0.29
0.02
1.80
1.87
10
0.78
2001
11
353
32
0.06
0.19
0.73
3.23
0.56
1.91
0.81
1.43
0.17
0.44
1.59
4.09
1.21
0.94
2.23
3.21
10
0.87
2001
11
309
28
0.08
0.23
0.71
3.03
0.61
2.20
0.96
1.13
0.06
0.24
1.27
2.23
0.24
0.34
2.09
1.95
10
0.73
2001
11
94
9
0.08
0.27
0.71
3.03
0.62
2.25
0.65
1.02
0.11
0.58
0.89
2.65
0.49
0.79
1.68
1.74
10
0.83
2001
11
2095
190
0.05
0.13
0.70
2.91
0.69
2.68
0.26
0.49
0.12
0.35
0.60
1.35
0.88
0.79
0.75
1.25
10
0.72
2001
11
514
47
0.08
0.22
0.70
2.90
0.63
2.27
0.71
0.97
0.08
0.45
0.91
2.30
0.29
0.57
1.68
1.66
10
0.77
2001
11
35
3
0.00
0.32
0.69
2.90
0.53
1.75
0.58
1.09
0.15
0.15
0.45
2.12
0.47
0.09
1.67
1.05
10
0.81
2001
11
215
20
0.07
0.53
0.69
2.88
0.46
1.48
0.35
1.26
0.07
0.47
0.51
3.49
0.39
0.69
1.61
1.88
10
0.82
2001
11
496
45
0.05
0.20
0.67
2.74
0.75
3.18
0.08
0.47
0.14
0.56
0.17
1.29
0.89
0.91
0.56
0.87
10
0.83
2001
11
466
42
0.06
0.11
0.67
2.74
0.69
2.71
0.27
0.62
0.10
0.27
0.49
1.62
0.59
0.48
0.89
1.33
10
0.71
2001
11
234
21
0.09
0.33
0.67
2.71
0.77
3.39
0.10
0.27
0.17
1.23
0.21
1.12
1.01
2.21
0.38
0.63
10
0.87
2001
11
1441
131
0.06
0.19
0.65
2.55
0.48
1.56
0.24
0.71
0.11
0.54
0.35
1.51
0.58
0.89
0.94
0.98
10
0.61
2001
11
302
27
0.00
0.21
0.64
2.49
0.91
6.03
0.62
0.10
0.17
0.94
2.20
0.67
1.72
2.57
0.72
2.18
10
0.95
2001
11
222
20
0.09
0.15
0.63
2.44
0.62
2.22
0.05
0.45
0.11
0.75
0.07
1.04
0.51
0.95
0.50
0.52
10
0.70
2001
11
975
89
0.12
0.23
0.63
2.42
0.48
1.55
0.24
0.49
0.00
0.40
0.39
1.05
0.01
0.71
0.73
0.86
10
0.42
2001
11
2041
186
0.05
0.20
0.61
2.33
0.57
1.94
0.07
0.43
0.14
0.55
0.14
1.04
0.80
1.04
0.50
0.67
10
0.62
2001
11
56
0.03
0.54
0.61
2.32
0.52
1.70
0.27
1.04
0.08
1.06
0.36
2.96
0.39
1.55
1.30
1.43
10
0.83
2001
11
2064
0.05
0.08
0.61
2.29
0.52
1.71
0.07
0.44
0.13
0.65
0.14
1.13
0.82
1.29
0.37
0.52
10
0.66
2001
11
100
9
0.09
0.31
0.60
2.27
0.61
2.20
1.92
0.91
0.00
3.12
1.44
1.96
0.00
2.95
2.83
2.36
10
0.86
2001
11
1008
92
0.06
0.27
0.59
2.17
0.56
1.91
0.36
0.56
0.26
0.29
0.57
1.18
1.41
0.48
0.91
1.09
10
0.62
2001
11
41
0.59
0.58
2.11
0.34
1.02
0.41
1.61
0.19
0.56
0.42
2.35
0.55
0.42
2.01
1.37
10
0.71
11
66
4
6
0.00
2001
0.06
0.72
0.51
1.77
0.37
1.13
1.62
0.86
0.57
1.57
4.28
3.06
4.84
5.82
2.48
3.98
10
0.91
5
188
Case5 11cv 02509LHK
Document424
Exhibit
2
Filed0517 13
Page39
62
of
1
Adobe
Section
First
Job
Title
Year
1
Years
of
1
Section
Total
Emp
Data
2
Level
Years
Avg
3
Emp
4
dlog
Avg
5
dlog
Std
Dev
6
Correlation
Coeff
7
T
Stat
8
2
Section
Change
Correlation
Regression
9
T Stat
Contemp
10
Coeff
11
Lagged
12
3
Section
Coefficients
Revenue
Regression
SJ
13
Emp
14
Contemp
15
Lagged
16
4
T
Section
Net
Stats
Revenue
SJ
17
Emp
18
C
L
19
5
Section
6
Effect
T Stat
Obs
r2
20
21
22
2005
7
22
3
0.18
0.41
0.76
2.64
0.15
0.31
0.14
0.93
0.38
0.36
0.11
1.48
0.68
0.21
1.07
0.60
6
0.91
2001
7
42
6
0.27
0.76
0.57
1.56
0.39
0.84
3.13
2.20
0.57
3.68
2.63
2.79
1.65
2.92
0.93
1.11
6
0.93
2001
7
88
13
0.41
0.33
0.53
1.38
0.38
0.82
3.36
5.49
1.61
7.47
4.12
6.77
4.51
5.53
2.13
6
1.00
2001
7
17
2
0.00
0.36
0.48
1.21
0.93
4.88
0.58
0.42
0.13
0.77
0.54
0.84
0.54
0.89
1.00
0.71
6
0.95
2005
7
93
13
0.00
0.27
0.40
0.98
0.97
7.56
1.30
0.10
0.07
0.02
2.06
0.28
0.24
0.03
1.40
1.76
6
0.94
2005
7
59
8
0.05
0.36
0.08
0.18
0.52
1.21
0.49
0.70
0.24
0.26
0.34
0.76
0.40
0.13
1.19
0.61
6
0.73
2001
6
46
8
0.14
0.21
0.98
0.90
3.49
2001
6
25
4
0.36
0.95
0.97
8.18
0.86
2.98
2001
6
19
3
0.06
0.45
0.96
7.28
0.93
4.41
2001
6
87
15
0.03
0.12
0.96
6.72
0.83
2.55
2001
6
13
2
0.28
1.05
0.94
5.50
0.94
4.92
2001
6
89
15
0.11
0.43
0.94
5.29
0.82
2.47
2001
6
108
18
0.01
0.23
0.93
5.23
0.74
1.90
2001
6
20
3
0.00
0.20
0.93
5.11
0.78
2.17
2001
6
16
3
0.06
0.70
0.92
4.77
0.58
1.23
2001
6
33
6
0.08
0.33
0.92
4.62
0.66
1.52
2001
6
22
4
0.03
0.74
0.89
3.99
0.94
4.80
2001
6
23
4
0.22
0.49
0.89
3.90
0.67
1.54
2001
6
35
6
0.09
0.26
0.89
3.87
0.91
3.90
2001
6
57
10
0.06
0.53
0.88
3.77
0.47
0.91
2001
6
10
2
0.22
0.32
0.88
3.74
0.50
1.00
2001
6
24
4
0.25
1.15
0.88
3.70
0.83
2.11
2001
6
21
4
0.36
0.59
0.88
3.66
0.49
0.97
2001
6
92
15
0.19
0.16
0.87
3.60
0.78
2.16
2001
6
68
11
0.00
0.21
0.86
3.44
0.66
1.51
2001
6
13
2
0.00
0.29
0.86
3.43
0.59
1.28
2001
6
27
5
0.42
0.63
0.86
3.38
0.74
1.92
2001
6
8
1
0.00
0.49
0.85
3.28
0.93
4.31
2001
6
15
3
0.08
0.34
0.85
3.18
0.27
0.49
2001
6
26
4
0.04
0.41
0.82
2.84
0.76
2.03
2006
6
7
1
0.14
0.31
0.81
2.81
0.85
2.85
2001
6
18
3
0.00
0.51
0.67
1.79
0.43
0.82
2001
6
105
18
0.04
0.36
0.66
1.74
0.68
1.59
2006
6
27
5
0.14
0.46
0.62
1.57
0.61
1.34
2006
6
19
3
0.08
0.52
0.61
1.55
0.54
1.11
2001
6
15
3
0.14
0.90
0.61
1.54
0.14
0.24
2001
6
12
2
0.22
0.32
0.57
1.39
0.76
2.05
2001
6
15
3
0.22
0.32
0.57
1.38
0.56
1.17
2006
6
19
0.53
0.34
0.72
0.21
0.38
6
6
3
1
0.28
2004
0.00
0.00
0.13
0.26
0.28
0.50
10.31
10.60
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Exhibit 2
Page40
of
62
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page41
62
of
Exhibit 2
Apple
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
Coeff
2
Section
Change
Correlation
Correlation
T
Coeff
Stat
3
Section
Regression
Stat
Contemp
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
Section
T Stats
Net
Revenue
SJ Emp
C
5
Section
T Stat
L
6
Effect
r2
11
294
0.98
13.53
0.74
3.11
0.80
0.04
0.34
0.06
1.64
0.05
0.81
0.13
0.84
0.76
0.71
11
501
0.98
13.42
0.87
4.91
2.46
1.09
0.70
0.67
5.33
1.71
1.82
1.18
3.56
4.85
0.92
11
229
0.98
13.33
0.65
2.41
1.15
0.97
0.09
0.08
2.58
1.52
0.26
0.19
2.12
2.15
0.73
11
169
0.97
12.72
0.70
2.79
1.29
1.49
0.57
0.28
2.17
1.67
1.00
0.46
2.78
2.20
0.72
11
352
0.95
9.16
0.71
2.82
0.92
0.22
0.76
0.16
1.56
0.39
1.55
0.26
0.71
0.72
0.78
11
189
0.93
7.38
0.84
4.39
1.68
0.36
0.20
0.87
1.81
0.38
0.26
0.87
2.04
1.39
0.82
11
428
0.91
6.72
0.65
2.45
0.51
4.63
2.48
1.62
0.53
2.82
2.28
1.57
5.14
2.93
0.82
11
156
0.88
5.54
0.39
1.21
0.71
0.25
0.28
0.62
0.95
0.38
0.40
0.67
0.96
0.77
0.29
11
118
0.68
2.82
0.36
1.09
0.58
0.17
0.11
0.23
0.86
0.31
0.16
0.24
0.75
0.70
0.16
11
686
0.49
1.69
0.43
1.33
0.66
0.47
0.15
0.49
0.68
0.60
0.18
0.40
1.13
0.73
0.52
11
58
0.50
1.71
0.07
0.20
0.03
0.11
0.21
0.27
0.05
0.28
0.49
0.47
0.09
0.11
0.10
10
82
0.67
2.52
0.03
0.08
0.38
0.08
0.18
0.01
0.39
0.10
0.22
0.01
0.30
0.19
0.34
10
184
0.81
3.84
0.25
0.68
0.17
0.08
0.18
0.91
0.20
0.11
0.24
0.81
0.09
0.07
0.40
10
110
0.81
3.93
0.71
2.64
0.69
0.07
0.04
0.53
2.98
0.36
0.18
1.86
0.76
2.06
0.75
10
66
0.89
5.57
0.04
0.11
0.14
0.06
0.06
0.20
1.03
0.53
0.47
1.12
0.20
0.92
0.36
0.85
4.33
0.55
1.59
0.43
0.03
0.14
0.95
1.37
0.14
0.54
1.36
0.39
0.79
0.83
11.69
0.59
1.27
1.84
3.27
2.40
1.69
9
116
8
44
0.98
8
35
0.97
9.97
0.78
2.48
0.30
0.21
1.02
0.21
1.13
0.37
3.49
0.93
0.50
0.73
0.99
8
19
0.76
2.89
0.62
1.78
0.16
0.16
0.02
0.78
0.78
0.97
0.13
1.91
0.00
0.01
0.86
8
52
0.82
3.57
0.02
0.05
0.14
0.08
0.13
0.07
0.50
0.28
0.51
0.36
0.22
0.40
0.57
8
13
0.96
7.90
0.24
0.55
0.09
0.05
0.03
0.22
0.84
0.50
0.27
0.69
0.14
0.78
0.51
7
71
0.99
22.21
0.95
5.95
0.54
0.46
0.07
0.06
1.39
0.22
0.15
0.04
0.08
0.03
0.94
7
193
0.99
20.45
0.95
6.20
1.49
1.49
0.41
0.82
12.36
3.86
2.99
1.89
2.98
6.80
1.00
7
626
0.99
16.77
0.94
5.77
1.41
1.40
0.29
0.07
30.92
4.57
3.71
0.27
2.82
8.34
1.00
7
184
0.99
16.70
0.96
6.91
1.16
1.48
0.31
0.23
3.69
0.99
0.69
0.27
2.64
1.81
0.97
0.99
14.96
0.92
4.55
0.88
0.60
0.16
0.65
10.23
3.85
1.64
3.23
1.48
7.27
0.99
0.99
13.76
0.81
2.81
0.24
0.38
0.08
0.22
0.48
0.29
0.14
0.12
0.14
0.08
0.80
7
7
2566
29
7
253
0.98
12.12
0.92
4.72
0.76
1.16
0.20
0.64
1.85
1.01
0.73
0.66
1.92
1.84
0.95
7
130
0.98
10.75
0.89
3.94
0.47
5.06
1.65
5.63
0.64
1.93
1.97
1.78
4.59
2.36
0.97
7
447
0.98
10.68
0.95
6.15
1.48
0.65
0.02
0.45
2.89
0.47
0.04
0.35
2.12
1.64
0.96
7
244
0.98
10.66
0.88
3.63
0.18
4.02
1.70
0.93
0.73
3.21
3.80
7.34
4.20
2.81
1.00
7
125
0.98
9.93
0.86
3.39
0.99
1.14
0.05
0.09
4.26
3.10
0.20
0.19
2.14
5.47
0.98
0.98
9.91
0.93
4.96
0.85
0.41
0.34
1.08
5.64
1.91
2.09
2.89
1.26
4.61
0.99
0.97
9.77
0.81
2.81
1.59
2.35
1.09
2.20
5.11
4.37
4.08
2.80
3.94
6.55
0.98
7
7
1364
54
7
236
0.97
9.58
0.97
7.42
0.99
0.57
0.28
0.18
2.55
1.16
0.76
0.24
1.56
3.63
0.97
7
475
0.97
9.33
0.84
3.04
0.55
0.80
0.42
1.16
2.01
1.71
1.67
1.34
1.35
2.55
0.95
0.97
9.17
0.81
2.81
0.66
0.37
0.03
0.87
9.39
3.50
0.50
5.68
1.03
6.50
0.99
4.22
3.00
79.73
1.00
7.66
1.52
14.05
1.00
7
1304
7
110
0.97
8.72
0.95
6.06
1.93
1.07
0.23
0.24
108.02
31.38
7
902
0.97
8.62
0.82
2.84
0.83
0.68
0.49
1.09
13.99
9.36
14.63
7.98
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page42
62
of
Exhibit 2
Apple
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
7
17
T
Coeff
2
Section
Change
Correlation
Correlation
T
Coeff
Stat
3
Section
Regression
Stat
Contemp
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
Section
T Stats
Net
Revenue
0.95
7.08
0.71
2.01
1.88
6.66
3.36
7.09
7.10
9.22
8.26
SJ Emp
6.61
16.05
C
5
Section
T Stat
L
6
Effect
r2
10.88
1.00
0.75
7.07
1.00
8.54
7
127
0.95
6.94
0.52
1.21
0.56
0.19
0.28
1.66
15.61
2.26
7.87
7
142
0.95
6.80
0.83
2.99
0.30
3.49
0.40
0.56
0.28
2.08
0.91
0.41
3.19
2.94
0.95
1.09
2.55
0.84
2.00
2.49
4.60
2.54
2.18
3.64
5.17
0.98
2.37
0.57
0.11
0.28
3.89
0.73
0.43
0.38
1.80
3.15
0.98
7
63
0.95
6.73
0.69
1.92
7
45
0.95
6.73
0.99
12.42
7
98
0.95
6.52
0.84
3.11
0.42
0.03
0.15
0.86
2.29
0.16
0.60
1.47
0.39
1.18
0.93
7
70
0.94
6.46
0.88
3.72
1.03
3.36
0.26
1.34
1.02
0.65
0.33
0.29
4.39
1.02
0.95
0.94
6.42
0.96
7.04
1.85
0.66
0.02
0.43
20.57
4.80
0.28
1.92
2.51
19.20
1.00
0.94
6.33
0.60
1.52
0.75
0.73
0.18
0.36
3.05
2.18
0.96
0.70
1.48
2.83
0.92
7
7
182
2915
7
134
0.94
6.30
0.66
1.76
0.94
1.02
0.16
0.07
8.01
7.07
1.52
0.25
1.97
9.04
0.99
7
143
0.94
6.27
0.48
1.10
0.38
0.26
0.73
1.64
0.87
0.46
1.94
1.39
0.64
0.68
0.84
7
476
0.94
6.23
0.91
4.31
3.20
2.66
1.18
5.55
2.00
1.31
1.16
1.44
0.53
0.75
0.96
0.94
6.18
0.79
2.54
1.14
0.91
0.12
0.64
3.07
1.95
0.41
0.78
2.05
3.12
0.98
7
53
7
275
0.94
6.09
0.70
1.97
0.82
0.80
0.45
1.06
2.39
1.55
1.68
1.39
1.62
2.24
0.97
7
255
0.93
5.78
0.74
2.21
0.07
2.18
0.57
1.09
0.15
4.59
2.06
1.39
2.11
4.69
0.98
7
300
0.93
5.69
0.38
0.82
0.33
0.33
0.09
0.42
1.51
1.22
0.67
1.12
0.66
1.43
0.82
7
125
0.93
5.69
0.79
2.56
0.64
1.88
0.06
0.58
5.01
16.56
0.79
2.97
2.52
18.16
1.00
7
262
0.93
5.65
0.51
1.18
0.99
1.54
0.46
0.24
4.29
4.47
2.58
0.49
2.53
5.03
0.97
0.93
5.63
0.72
2.10
1.20
1.08
0.14
0.10
2.30
1.36
0.24
0.08
2.28
2.10
0.97
0.93
5.58
0.27
0.57
0.71
0.94
0.29
1.78
0.41
0.33
0.52
1.46
1.65
0.36
0.76
6.30
4.69
2.75
12.46
1.00
7
7
16
115
7
33
0.93
5.56
0.55
1.31
1.06
1.69
0.48
0.89
11.73
10.86
7
16
0.93
5.55
0.47
1.06
2.57
3.07
1.01
2.89
2.51
2.15
1.27
1.18
5.64
2.42
0.92
7
35
0.93
5.46
0.68
1.85
0.43
0.40
0.43
1.40
0.92
0.30
0.85
1.10
0.83
0.53
0.92
0.92
5.42
0.84
3.04
0.57
1.74
0.21
0.65
0.73
2.15
0.46
0.55
2.30
2.76
0.95
7
297
7
57
0.92
5.39
0.72
2.05
0.69
0.70
0.36
0.74
2.04
2.46
0.95
0.86
1.39
2.85
0.94
7
58
0.92
5.35
0.78
2.48
0.81
0.46
0.29
0.50
3.21
2.06
0.77
0.78
1.28
3.10
0.94
7
26
0.92
5.30
0.67
1.80
2.23
2.43
1.17
0.57
5.76
2.33
1.86
0.37
4.66
3.32
1.00
58.99
1.00
7
115
0.92
5.30
0.64
1.68
0.86
0.53
0.05
1.73
81.85
34.93
6.57
83.66
7
103
0.92
5.23
0.35
0.74
0.71
2.91
1.10
0.68
1.67
3.08
2.22
0.72
3.62
3.03
0.94
1.39
7
35
0.92
5.21
0.59
1.45
0.67
4.66
1.96
0.59
1.56
5.68
4.15
0.64
5.33
5.82
0.99
7
49
0.92
5.14
0.67
1.79
1.20
0.72
0.03
2.50
2.41
0.57
0.03
1.91
1.92
1.15
0.98
7
23
0.92
5.12
0.89
3.94
1.50
0.38
0.73
0.15
3.16
0.60
1.79
0.15
1.12
1.44
0.98
0.91
5.03
0.24
0.50
0.05
0.05
0.05
0.41
0.10
0.09
0.19
0.45
0.01
0.01
0.23
0.31
2.38
6.99
3.91
0.96
9.56
1.00
35.05
1.00
0.84
0.79
24.05
1.00
225.62
1.00
7
431
7
21
0.91
4.94
0.54
1.30
3.18
3.81
0.09
4.43
4.28
3.52
7
64
0.91
4.93
0.33
0.71
0.14
0.85
0.65
1.56
2.39
11.13
11.42
6.65
0.99
7
56
0.91
4.86
0.93
4.90
3.28
0.05
0.48
3.16
26.16
0.30
13.49
7.06
3.23
7
14
0.91
4.86
0.40
0.86
0.07
0.01
0.16
0.43
1.14
0.23
1.66
0.08
7
59
0.91
4.83
0.88
3.68
1.77
1.31
0.18
0.90
13.53
9.61
7
48
0.90
4.69
0.20
0.42
0.20
0.71
0.09
0.37
102.47
285.17
1.50
1.45
64.33
2.78
73.80
3.09
0.91
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page43
62
of
Exhibit 2
Apple
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
Coeff
2
Section
Change
Correlation
Correlation
T
Coeff
Stat
3
Section
Regression
Stat
Contemp
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
Section
T Stats
Net
Revenue
SJ Emp
C
5
Section
T Stat
L
6
Effect
r2
7
37
0.88
4.25
0.04
0.07
0.57
0.88
0.53
0.36
1.77
2.44
2.49
0.57
1.45
2.29
0.89
7
34
0.88
4.25
0.15
0.30
1.13
2.90
1.26
0.78
2.87
5.65
3.68
0.93
4.03
5.04
0.98
7
8
0.88
4.20
0.89
3.94
1.47
0.70
0.91
1.65
9.23
2.78
5.87
4.83
0.78
2.62
1.00
0.88
4.17
0.40
0.87
0.34
0.62
0.43
0.72
1.28
1.65
1.75
1.23
0.96
1.71
0.99
7
103
7
7
0.88
4.15
0.72
2.05
0.53
0.22
0.32
0.27
3.42
1.74
1.18
0.56
0.75
3.01
0.94
7
8
0.88
4.11
0.04
0.09
0.44
1.15
0.14
0.78
0.71
1.55
0.31
0.53
1.58
1.29
0.81
7
28
0.88
4.08
0.45
1.02
0.07
3.01
0.73
2.35
0.56
12.67
6.31
6.82
3.09
11.64
1.00
7
61
0.88
4.08
0.26
0.55
1.31
2.69
1.24
1.51
3.08
3.00
2.32
2.70
4.00
3.06
0.99
7
25
0.87
4.01
0.59
1.45
0.28
3.71
1.43
0.39
8.63
82.49
54.77
6.26
3.99
78.44
1.00
7
7
0.87
3.98
0.26
0.53
1.98
2.62
1.42
5.06
1.68
2.14
1.30
1.91
4.61
2.14
0.98
0.87
3.94
0.85
3.21
3.43
3.62
0.07
5.76
2.07
1.57
0.14
1.48
0.19
0.18
0.94
0.87
3.94
0.61
1.53
0.61
1.04
0.29
1.97
1.40
2.07
0.84
1.88
1.64
2.22
0.93
0.87
3.93
0.50
1.16
0.27
0.05
0.31
0.87
0.23
0.04
0.66
0.51
0.22
0.10
0.51
0.87
3.91
0.49
1.14
0.28
2.39
0.62
0.66
0.16
0.99
0.48
0.16
2.11
0.87
0.81
2.27
0.94
7
7
7
7
501
74
192
11
7
116
0.87
3.89
0.21
0.43
6.50
7.89
2.48
6.52
2.32
2.22
1.95
1.50
14.39
7
239
0.87
3.89
0.89
3.90
0.95
0.13
0.56
0.89
1.43
0.16
1.08
0.59
0.82
0.81
0.90
7
10
0.86
3.83
0.54
1.30
4.35
6.24
1.52
7.36
0.67
0.87
0.63
0.57
1.89
0.66
0.80
7
44
0.86
3.78
0.52
1.22
0.32
0.27
0.00
0.96
0.20
0.17
0.00
0.74
0.59
0.21
0.54
7
21
0.86
3.69
0.69
1.91
0.77
0.40
0.84
1.36
0.94
0.44
1.35
0.74
0.37
0.30
0.95
7
17
0.85
3.65
0.68
1.84
1.99
1.43
0.04
0.81
2.93
1.69
0.07
0.45
3.42
2.63
0.97
0.85
3.60
0.92
4.56
1.94
0.26
0.17
0.60
0.89
0.17
0.12
0.17
1.68
1.12
0.84
7
563
7
12
0.85
3.58
0.06
0.12
0.12
0.13
0.26
0.46
0.51
0.68
1.37
0.74
0.25
0.65
0.79
7
57
0.85
3.58
0.46
1.03
0.26
1.45
0.06
1.52
0.14
1.18
0.06
0.40
1.19
0.60
0.89
0.85
3.57
0.90
4.16
1.96
0.40
0.23
2.66
15.41
5.76
2.47
9.44
1.55
13.27
1.00
0.85
3.55
0.04
0.07
0.55
0.93
0.28
2.78
0.76
0.91
0.66
3.50
1.48
0.86
0.95
7
7
145
33
7
131
0.85
3.55
0.76
2.36
0.54
0.17
0.73
1.81
1.90
0.72
2.37
2.35
0.71
1.63
0.96
7
267
0.84
3.52
0.16
0.32
0.22
0.30
1.27
2.14
0.14
0.19
0.35
0.20
0.51
0.17
0.51
7
47
0.84
3.43
0.29
0.60
0.83
1.09
0.45
1.22
1.10
1.76
0.48
0.69
1.91
1.62
0.85
7
60
0.84
3.42
0.52
1.21
0.83
0.25
0.30
0.36
0.54
0.17
0.26
0.29
1.09
0.41
0.36
7
8
0.84
3.40
0.06
0.12
0.13
3.20
1.30
2.42
0.36
3.70
2.62
2.29
3.33
3.00
0.97
7
50
0.83
3.35
0.61
1.56
0.65
0.05
0.93
1.56
4.31
0.32
7.83
4.65
0.70
2.62
1.00
7
57
0.83
3.34
0.11
0.22
0.25
0.75
0.33
0.60
0.87
2.96
1.13
0.67
1.00
2.18
0.95
7
20
0.83
3.33
0.35
0.75
0.24
0.46
0.59
1.46
0.59
1.17
1.77
1.65
0.70
1.04
0.99
7
20
0.83
3.32
0.38
0.83
0.34
1.47
0.20
0.34
2.79
7.80
1.91
1.02
1.14
3.94
1.00
7
40
0.82
3.24
0.94
5.74
1.96
0.82
0.43
0.51
3.60
1.74
1.46
0.51
1.14
2.01
0.98
0.82
3.24
0.91
4.27
1.43
0.33
0.57
0.59
1.18
0.30
0.81
0.26
1.11
0.79
0.89
1.37
5.78
2.74
18.75
0.55
1.16
1.17
1.69
7.16
0.96
0.99
7
144
7
23
0.82
3.21
0.55
1.31
7
72
0.82
3.17
0.01
0.02
0.59
0.65
1.04
2.39
0.45
0.50
0.46
0.44
1.24
0.49
0.22
7
47
0.81
3.07
0.71
2.01
1.22
0.50
0.87
1.01
2.88
1.31
2.57
1.08
1.72
2.53
0.98
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page44
62
of
Exhibit 2
Apple
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
Coeff
2
Section
Change
Correlation
Correlation
T
Coeff
Stat
3
Section
Regression
Stat
Contemp
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
Section
T Stats
Net
Revenue
SJ Emp
7
26
0.70
2.22
0.23
0.48
0.23
0.43
0.86
0.85
15.24
28.90
49.37
7
25
0.70
2.20
0.68
1.86
0.94
0.69
0.33
3.50
0.89
0.45
0.39
1.53
7
38
0.70
2.20
0.79
2.56
9.17
2.35
7.19
19.15
1.03
2.58
0.73
0.53
7
18
0.66
1.95
0.11
0.22
2.32
2.16
0.46
7.71
1.39
1.67
0.35
1.99
7
58
0.66
1.95
0.07
0.15
0.76
0.61
1.32
2.62
1.44
1.90
2.40
7
26
0.65
1.90
0.43
0.95
1.80
1.36
0.95
0.78
1.91
1.85
7
13
0.65
1.90
0.51
1.18
1.56
2.39
0.40
6.21
0.64
7
51
0.64
1.88
0.23
0.47
1.80
1.79
0.28
0.82
7
14
0.64
1.87
0.38
0.82
0.56
0.52
0.89
7
57
0.64
1.86
0.03
0.05
0.09
0.08
7
11
0.63
1.82
0.45
1.01
1.68
7
24
0.63
1.80
0.57
1.40
0.62
1.79
0.04
7
127
12.29
C
5
Section
T Stat
L
6
Effect
r2
25.59
1.00
0.82
0.88
1.20
0.97
4.48
1.69
0.93
1.36
0.14
0.20
0.97
0.84
0.32
3.16
2.12
0.83
1.99
0.29
1.79
0.83
0.26
0.97
1.11
1.29
0.21
0.21
3.59
1.32
0.74
4.00
1.07
1.02
2.18
3.39
1.08
1.20
0.97
1.16
3.51
2.94
2.71
47.78
0.01
0.26
1.00
1.26
0.17
1.18
3.40
3.03
0.42
1.01
2.93
3.65
0.97
0.12
7.51
4.87
14.39
0.13
2.03
2.22
3.54
7.63
1.69
0.99
0.08
2.05
1.96
4.08
9.17
7.13
7.37
9.16
8.99
4.01
7.51
0.99
40.82
0.66
1.62
11.52
7
45
0.62
1.79
0.82
2.90
1.18
0.46
0.62
0.77
1.08
0.92
0.50
0.58
1.64
1.07
0.97
7
36
0.58
1.58
0.86
3.38
3.09
0.55
1.14
3.47
0.92
0.56
0.39
0.63
3.64
0.88
0.87
7
52
0.57
1.57
0.56
1.34
0.91
0.24
2.01
5.19
0.41
0.17
1.13
1.29
0.67
0.21
0.91
0.56
1.51
0.25
0.51
0.93
0.88
0.89
1.03
2.28
2.86
1.87
1.16
1.81
2.82
0.94
7
137
7
18
0.55
1.49
0.33
0.69
0.11
0.48
2.73
0.70
0.25
1.33
3.40
0.55
0.59
0.78
0.98
7
13
0.55
1.48
0.52
1.23
0.42
1.07
2.09
2.76
0.47
1.46
3.04
1.39
0.65
0.46
0.97
7
59
0.55
1.46
0.06
0.12
0.37
0.17
0.75
5.12
0.25
0.10
0.87
2.95
0.54
0.18
0.93
7
16
0.54
1.45
0.47
1.07
3.59
2.10
0.38
3.17
2.25
1.44
0.44
0.55
5.69
1.95
0.93
7
34
0.54
1.42
0.41
0.90
0.50
0.48
1.73
2.69
1.25
1.52
5.22
2.97
0.01
0.02
0.98
7
35
0.53
1.39
0.50
1.17
0.35
0.64
1.85
0.64
0.37
0.92
2.26
0.30
0.30
0.20
0.94
7
41
0.53
1.38
0.52
1.21
0.82
0.14
0.66
1.97
1.45
0.32
0.93
1.55
0.96
1.08
0.86
7
46
0.52
1.36
0.33
0.69
1.08
1.05
0.12
0.61
6.60
8.54
0.68
1.70
2.13
8.30
1.00
7
15
0.52
1.35
0.73
2.16
0.40
0.56
0.89
2.38
0.20
0.51
0.62
0.68
0.96
0.38
0.84
0.52
1.35
0.00
0.00
0.17
0.16
0.08
0.05
2.19
2.67
0.66
0.17
0.33
2.68
0.95
7
646
7
14
0.51
1.33
0.20
0.41
0.55
0.31
1.05
0.73
0.34
0.25
0.34
0.15
0.86
0.37
0.55
7
47
0.51
1.31
0.96
6.64
1.90
0.37
0.00
0.63
3.25
1.24
0.00
0.50
1.53
2.14
0.97
7
27
0.50
1.30
0.11
0.23
1.00
1.69
0.29
2.35
5.14
10.67
1.89
4.56
0.69
2.25
1.00
7
17
0.49
1.25
0.19
0.38
0.61
0.30
1.50
1.86
1.24
0.79
3.44
1.64
0.91
1.17
0.98
7
13
0.49
1.24
0.72
2.07
2.54
2.26
1.08
0.56
0.86
0.91
0.51
0.10
0.28
0.06
0.95
7
63
0.47
1.20
0.14
0.29
0.42
0.43
1.07
1.20
0.49
0.68
1.03
0.38
0.01
0.01
0.91
7
85
0.47
1.18
0.43
0.96
0.01
0.28
1.09
0.67
0.01
0.41
0.72
0.15
0.27
0.14
0.90
7
60
0.45
1.11
0.74
2.18
0.54
0.10
0.20
0.50
1.72
0.44
0.46
0.35
0.64
1.31
0.88
7
19
0.44
1.10
0.46
1.03
0.96
1.03
1.13
5.03
2.16
3.27
4.56
5.89
1.98
2.79
1.00
7
10
0.44
1.08
0.78
2.50
5.10
0.41
0.38
7.95
0.71
0.11
0.12
0.81
4.69
0.46
0.79
7
69
0.42
1.04
0.10
0.20
0.58
2.48
1.20
12.14
0.30
0.89
1.11
0.64
3.06
1.33
0.87
7
36
0.42
1.03
0.34
0.73
0.26
0.23
0.76
0.15
1.39
1.93
2.82
0.32
0.03
0.10
0.94
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page45
62
of
Exhibit 2
Apple
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
7
7
T
Coeff
29
117
2
Section
Change
Correlation
Correlation
T
Coeff
Stat
3
Section
Regression
Stat
Contemp
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
Section
T Stats
Net
Revenue
SJ Emp
C
5
Section
T Stat
L
6
Effect
r2
0.08
0.18
0.33
0.70
0.04
0.09
1.00
2.10
0.04
0.12
0.78
0.64
0.13
0.08
0.73
0.04
0.08
0.26
0.55
0.56
1.36
6.15
1.05
1.15
3.45
3.63
0.97
0.80
1.44
0.96
7
26
0.04
0.08
0.21
0.43
0.76
0.43
1.14
2.09
0.93
0.73
1.69
1.20
0.34
0.26
0.99
7
22
0.04
0.10
0.17
0.34
4.02
1.91
5.35
23.44
2.41
1.95
2.23
2.84
5.94
2.26
0.97
7
31
0.07
0.16
0.29
0.62
0.47
1.28
2.00
1.97
0.26
0.93
1.01
0.32
1.74
0.61
0.64
7
11
0.27
0.63
0.23
0.48
0.75
0.14
0.01
0.87
0.25
0.05
0.00
0.37
0.89
0.17
0.21
7
46
0.28
0.66
0.02
0.03
2.17
1.69
6.68
6.27
1.26
1.10
1.77
1.60
0.48
0.21
0.82
7
52
0.36
0.87
0.37
0.79
1.19
0.84
0.81
2.05
2.75
2.07
1.17
2.10
2.04
2.78
0.95
7
50
0.43
1.06
0.96
6.86
0.30
0.06
0.07
0.09
6.12
1.64
0.89
0.54
0.24
3.24
0.99
7
49
0.48
1.23
0.27
0.57
0.03
0.11
0.46
1.13
0.06
0.26
0.43
0.48
0.14
0.18
0.55
0.49
1.25
0.44
0.97
0.12
0.34
0.76
0.70
1.22
3.94
3.94
2.00
0.22
1.43
0.96
7
166
7
36
0.50
1.29
0.05
0.10
1.28
3.22
5.96
8.31
0.99
1.06
0.97
1.00
4.50
1.10
0.61
7
21
0.54
1.42
0.80
2.66
1.42
0.36
0.68
1.28
6.57
1.85
1.97
2.37
1.77
4.71
0.99
7
59
0.62
1.79
0.31
0.65
0.43
0.52
0.51
0.18
0.46
0.58
0.70
0.24
0.94
0.59
0.48
7
40
0.65
1.92
0.35
0.74
0.75
0.85
0.63
0.30
0.43
0.46
0.41
0.22
1.61
0.50
0.32
6
16
0.98
9.32
0.93
4.31
6
19
0.96
7.34
0.85
2.85
6
54
0.96
7.16
0.89
3.46
6
48
0.93
4.91
0.94
4.62
6
44
0.87
3.58
0.64
1.18
6
20
0.87
3.48
0.45
0.72
6
73
0.85
3.24
0.41
0.78
6
19
0.77
2.41
0.51
1.03
6
6
0.76
2.35
0.46
0.91
6
15
0.76
2.31
0.90
3.49
6
24
0.75
2.27
0.08
0.12
6
6
0.75
2.26
0.53
1.07
6
57
0.73
2.13
0.47
0.92
6
8
0.72
2.05
0.36
0.55
6
10
0.71
2.04
0.55
1.14
6
6
0.67
1.81
0.59
1.26
6
6
0.63
1.61
0.81
1.95
6
8
0.63
1.61
0.82
2.00
6
11
0.60
1.49
0.83
2.59
6
19
0.59
1.45
0.05
0.08
6
12
0.48
1.08
0.06
0.09
6
19
0.47
1.07
0.04
0.07
18
0.42
0.93
0.61
1.09
0.42
0.92
0.55
1.14
6
6
166
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page46
62
of
Exhibit 2
Apple
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
Coeff
2
Section
Change
Correlation
Correlation
T
Coeff
Stat
Stat
6
17
0.16
0.32
0.07
0.13
6
16
0.29
0.60
0.78
2.16
6
6
0.30
0.62
0.55
1.13
6
40
0.31
0.65
0.11
0.19
6
6
0.45
1.02
0.84
2.67
0.65
1.70
0.32
0.59
6
1398
3
Regression
6
15
0.76
2.36
0.93
4.48
6
19
0.85
3.22
0.43
0.83
Contemp
Lagged
Revenue
Section
Coefficients
Regression
SJ
Emp
Contemp
Lagged
4
Section
T Stats
Revenue
Net
SJ Emp
C
L
5
Section
Effect
T Stat
r2
6
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page47
of
62
Exhibit 2
Google
Section
Years
Job
Title
of
Data
1
Section
Total
Emp
Years
Level
Correlation
T
Coeff
2
Change
Section
T
Coeff
Stat
Regression
Correlation
Contemp
Stat
Lagged
3
Section
Regression
Coefficients
Revenue
SJ
Emp
Contemp
Lagged
4
T
Section
5
Section
6
Net Effect
Stats
Revenue
SJ
Emp
C
T
L
Stat
r2
0.94
8.15
0.89
5.63
0.08
0.07
1.36
2.10
0.45
0.26
3.49
3.85
0.15
0.37
0.96
0.91
6.58
0.88
5.21
0.26
0.10
0.73
0.87
1.01
0.27
1.53
1.28
0.36
0.62
0.91
0.91
6.51
0.83
4.27
0.80
0.26
0.48
1.30
0.87
0.13
0.35
0.64
1.06
0.37
0.88
0.86
5.00
0.76
3.30
0.16
0.08
0.70
1.49
0.40
0.14
0.89
1.13
0.24
0.26
0.75
0.82
4.29
0.82
4.05
0.08
1.78
2.60
0.26
0.11
1.70
2.30
0.15
1.86
1.10
0.89
0.79
3.89
0.78
3.55
0.21
1.42
2.46
2.14
0.56
2.52
4.01
2.41
1.63
1.80
0.94
0.79
3.86
0.75
3.22
0.45
0.57
0.45
2.87
0.99
0.55
0.79
1.95
1.02
0.69
0.77
0.79
3.83
0.61
2.21
0.27
0.71
2.24
3.07
0.83
1.34
4.09
3.87
0.98
1.19
0.95
0.79
3.82
0.84
4.31
0.61
0.50
0.12
1.31
1.49
0.56
0.20
1.16
1.11
0.87
0.79
0.78
3.75
0.82
4.01
0.38
0.24
0.53
2.31
1.00
0.27
0.99
1.54
0.62
0.50
0.80
0.74
3.33
0.75
3.24
0.64
0.88
0.45
0.85
2.62
1.79
1.17
0.82
1.52
2.14
0.74
0.71
3.05
0.72
2.91
0.30
2.66
3.51
1.03
0.32
1.73
2.31
0.42
2.97
1.23
0.86
0.71
3.01
0.83
4.25
0.68
0.53
0.03
1.25
1.35
0.47
0.04
0.83
1.21
0.75
0.75
0.70
2.90
0.70
2.78
0.29
1.04
1.65
1.88
0.93
2.14
2.97
1.92
1.33
1.73
0.84
0.67
2.68
0.50
1.64
0.72
1.63
2.36
3.79
2.59
3.56
4.96
5.62
2.35
3.28
0.91
0.62
2.39
0.47
1.52
0.27
0.41
0.37
1.40
0.48
0.50
0.37
0.72
0.68
0.51
0.59
0.59
2.20
0.55
1.84
1.63
4.50
5.16
4.24
1.47
2.51
2.86
1.61
6.13
2.16
0.82
0.56
2.05
0.53
1.77
2.49
7.13
7.79
5.04
2.28
3.94
4.41
1.94
9.62
3.40
0.91
0.51
1.78
0.23
0.66
1.01
1.63
2.56
2.55
1.52
1.63
2.14
1.56
2.64
1.62
0.68
0.48
1.63
0.39
1.21
0.98
2.45
3.07
5.23
0.85
1.26
1.94
2.93
3.43
1.12
0.83
0.27
0.84
0.02
0.05
0.15
0.67
0.31
4.53
0.32
0.91
0.40
3.20
0.82
0.70
0.75
0.81
3.90
0.77
3.21
0.35
0.43
0.23
2.19
1.13
0.64
0.53
1.75
0.78
0.81
0.77
0.80
3.75
0.72
2.51
0.11
0.45
1.71
3.16
0.14
0.24
1.71
2.76
0.56
0.21
0.90
0.75
3.16
0.85
4.29
1.58
2.53
1.92
2.75
3.14
2.44
2.19
1.43
4.11
2.77
0.92
0.71
2.82
0.47
1.42
1.78
3.60
2.30
0.40
2.18
2.42
1.61
0.12
5.38
2.41
0.86
0.66
2.47
0.50
1.53
1.25
1.78
1.19
1.94
3.31
3.15
1.67
1.23
3.03
3.39
0.89
0.52
1.74
0.62
2.09
0.46
0.10
0.22
1.96
0.71
0.09
0.15
1.13
0.56
0.33
0.63
0.32
0.95
0.68
2.45
1.20
1.43
0.38
3.13
1.21
0.71
0.24
1.47
2.62
0.89
0.77
0.84
4.08
0.82
3.45
1.37
2.09
0.38
0.78
4.96
3.34
0.84
0.51
3.46
4.07
0.97
0.78
3.27
0.77
2.94
0.96
1.43
0.46
1.25
5.78
3.93
1.70
1.37
2.40
4.80
0.96
0.73
2.80
0.80
3.23
1.06
1.36
0.75
0.45
2.63
1.44
1.12
0.23
2.42
1.86
0.82
0.71
2.63
0.70
2.43
1.73
2.75
2.01
1.05
7.82
6.48
5.33
0.90
4.48
7.35
0.97
0.67
2.38
0.71
2.45
0.80
0.83
0.13
0.74
2.41
1.03
0.21
0.54
1.62
1.47
0.93
0.64
2.18
0.60
1.84
0.28
0.10
0.34
0.24
0.63
0.10
0.55
0.18
0.38
0.27
0.80
0.56
1.79
0.83
3.70
0.12
0.02
1.64
0.59
0.18
0.03
1.22
0.27
0.14
0.11
0.92
0.44
1.28
0.63
2.00
2.00
0.63
0.47
0.85
0.89
0.16
0.13
0.07
2.63
0.45
0.77
0.34
0.95
0.18
0.46
1.05
1.92
0.72
0.01
1.31
1.32
0.55
0.00
2.97
1.39
0.63
0.31
0.86
0.54
1.58
0.17
0.39
2.01
1.80
0.23
0.39
1.39
0.70
0.56
0.34
0.85
0.26
0.72
0.45
1.12
0.44
0.25
0.04
1.69
0.59
0.24
0.03
0.85
0.69
0.39
0.60
0.22
0.59
0.30
0.77
0.23
1.16
2.30
0.22
0.78
2.06
4.60
0.12
1.39
1.72
0.97
0.09
0.23
0.11
0.27
0.35
0.55
0.79
2.64
1.22
1.12
0.93
1.48
0.91
1.23
0.74
0.06
0.17
0.01
0.02
0.56
1.41
0.72
1.11
1.04
1.55
0.68
0.37
1.96
1.43
0.74
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page48
of
62
Exhibit 2
Google
Section
Years
Job
Title
of
Data
1
Section
Total
Emp
Years
Level
Correlation
T
Coeff
2
Section
Change
T
Coeff
Stat
Regression
Correlation
Contemp
Stat
Lagged
3
Section
Regression
Coefficients
Revenue
SJ
Emp
Contemp
Lagged
4
T
Section
5
Section
6
Net Effect
Stats
Revenue
SJ
Emp
C
T
L
Stat
r2
0.11
0.26
0.05
0.12
1.78
4.82
3.95
8.75
0.53
0.77
0.59
0.38
6.61
0.72
0.69
0.10
0.25
0.40
0.98
0.64
1.19
2.95
1.74
0.96
1.03
2.17
0.98
1.83
1.01
0.98
0.09
0.22
0.47
1.20
0.22
0.67
2.13
1.85
0.22
0.39
1.04
0.85
0.89
0.33
0.96
0.08
0.19
0.61
1.74
0.11
0.73
1.64
0.18
0.35
1.26
2.37
0.16
0.84
0.97
0.92
0.00
0.00
0.54
1.44
0.19
1.04
2.39
4.19
0.27
0.75
1.17
1.85
1.24
0.59
0.95
0.19
0.47
0.36
0.87
0.44
1.21
2.37
2.43
0.85
1.21
2.05
1.36
1.66
1.10
0.94
0.94
6.31
0.98
0.92
0.44
0.15
1.14
1.60
0.34
0.13
0.94
1.36
0.76
0.99
0.88
4.22
0.98
9.66
1.71
1.08
1.17
1.74
2.76
0.95
0.95
1.42
2.78
1.63
0.99
0.81
3.05
0.93
5.04
2.09
1.73
1.40
4.09
5.52
4.20
3.82
7.88
1.00
0.80
2.97
0.89
3.87
1.89
2.59
2.38
0.19
1.24
0.96
0.73
0.07
4.48
1.07
0.91
0.78
2.79
0.92
4.85
0.04
1.56
2.30
0.05
0.07
1.45
2.12
0.04
1.60
0.99
0.99
0.77
2.68
0.87
3.50
0.01
0.93
1.40
1.72
0.03
1.46
2.31
2.49
0.94
1.01
0.99
0.76
2.60
0.79
2.55
2.08
3.14
6.08
2.19
1.36
1.38
1.95
0.97
5.22
1.38
0.98
0.73
2.36
0.77
2.38
0.48
1.11
2.62
0.84
6.23
8.70
7.53
1.59
7.81
1.00
0.72
2.31
0.73
2.15
2.48
6.19
6.26
2.27
3.18
3.57
2.61
8.67
3.46
1.00
0.70
2.22
0.77
2.40
0.78
1.84
3.07
1.89
9.88
0.69
2.14
0.75
2.28
0.69
2.40
3.41
7.95
0.25
0.42
0.61
1.33
3.09
0.37
0.93
0.67
2.00
0.86
3.38
1.48
1.36
0.94
2.69
0.97
0.51
0.33
0.73
2.85
0.69
0.94
0.64
1.87
0.87
3.48
0.04
0.79
1.30
0.83
0.15
1.63
2.67
1.56
0.83
1.15
0.99
0.63
1.80
0.55
1.14
0.39
0.10
2.24
0.62
1.76
0.63
1.61
0.92
2.25
3.15
0.31
4.54
5.33
8.35
0.79
3.17
5.10
1.00
0.61
1.74
0.68
1.83
0.01
0.21
1.26
0.28
0.02
0.15
0.74
0.18
0.20
0.09
0.89
0.60
1.68
0.64
1.66
0.89
1.99
3.14
0.82
5.88
6.81
2.59
2.88
6.54
1.00
0.60
1.67
0.75
2.29
0.41
0.22
0.58
1.15
0.85
0.25
0.60
1.23
0.64
0.47
0.99
0.57
1.56
0.90
4.02
0.15
0.71
1.44
1.90
0.22
0.49
1.27
1.40
0.56
0.26
0.97
0.56
1.52
0.76
2.33
0.78
0.82
0.11
0.71
1.67
0.94
0.12
0.79
1.60
1.20
0.99
0.50
1.29
0.39
0.83
4.23
8.54
8.63
7.90
1.16
1.18
1.07
1.13
0.49
1.26
0.67
1.78
1.37
4.14
4.70
0.11
0.20
0.22
0.91
0.47
1.20
0.38
0.82
0.80
1.63
2.83
2.19
3.13
3.15
6.13
0.44
1.11
0.37
0.81
1.66
2.94
4.48
6.60
0.97
0.89
1.31
0.44
1.09
0.42
0.92
0.82
1.60
2.92
2.97
0.73
0.68
0.43
1.06
0.45
0.99
0.65
1.18
2.15
1.97
0.59
0.57
0.41
1.02
0.49
0.79
1.37
2.80
2.02
0.00
0.40
0.97
0.54
1.30
5.72
5.70
1.24
0.23
0.53
0.45
1.01
0.28
0.43
0.82
0.22
0.38
0.22
0.51
0.16
0.22
2.68
4.65
1.97
0.00
0.21
0.49
0.41
0.90
0.83
3.92
4.02
7.39
2.91
0.18
0.41
0.31
0.66
0.20
0.67
2.19
2.29
0.92
0.13
0.29
0.00
0.01
0.36
0.84
1.88
1.39
0.30
0.69
0.11
0.22
3.76
6.86
6.03
0.30
0.69
0.60
1.51
1.75
2.91
2.70
0.94
5.52
0.96
5.86
10.15
13.34
10.00
11.51
12.40
18.00
4.53
19.74
10.69
11.61
2.62
11.63
1.00
12.58
10.24
1.17
0.85
2.77
0.08
0.84
4.50
2.43
3.16
0.99
1.73
4.59
0.92
0.93
1.34
1.06
2.42
0.70
0.88
0.99
0.91
1.83
0.58
0.98
1.29
1.52
1.11
1.27
0.94
0.26
0.38
0.10
0.71
0.30
1.00
5.34
6.73
2.49
4.76
4.89
0.99
1.76
4.10
1.40
0.87
1.55
0.98
1.32
1.38
3.83
0.58
1.20
1.43
0.99
2.52
6.36
5.97
5.30
2.11
1.26
2.35
2.34
2.92
1.03
24.13
12.77
19.06
10.62
4.65
6.14
1.00
2.36
0.94
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page49
of
62
Exhibit 2
Google
Section
Years
Job
Title
of
Data
1
Section
Total
Emp
Years
Level
Correlation
T
Coeff
2
Section
Change
T
Coeff
Stat
Regression
Correlation
Contemp
Stat
0.44
0.99
0.60
1.32
0.42
0.93
0.50
0.99
0.38
0.83
0.42
0.81
0.35
0.74
0.27
0.49
0.34
0.72
0.64
1.45
0.30
0.63
0.95
3.20
0.30
0.63
0.18
0.32
0.29
0.61
0.17
0.30
0.25
0.51
0.18
0.32
0.22
0.45
0.08
0.14
0.19
0.39
0.55
1.13
0.15
0.31
0.30
0.45
0.14
0.29
0.37
0.69
0.12
0.23
0.15
0.27
0.10
0.20
0.58
1.24
0.09
0.18
0.01
0.01
0.07
0.13
0.07
0.12
0.04
0.09
0.37
0.69
0.05
0.11
0.28
0.51
0.24
0.48
0.60
1.31
Lagged
3
Section
Regression
Coefficients
Revenue
SJ
Emp
Contemp
Lagged
4
T
Section
Revenue
5
Section
Net Effect
Stats
SJ
Emp
C
L
T
Stat
r2
6
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page50
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
11
432
0.96
10.82
0.95
8.41
2.03
0.51
0.64
0.34
6.11
0.78
1.25
0.76
1.52
1.78
0.95
11
1501
0.96
9.78
0.94
7.56
1.56
0.30
0.32
0.54
6.76
0.36
0.73
1.63
1.86
2.07
0.96
11
233
0.94
8.46
0.91
6.14
1.47
1.33
0.23
0.09
4.71
0.74
0.25
0.15
2.80
1.46
0.92
11
3042
0.94
8.03
0.89
5.67
0.61
0.39
0.20
0.31
7.76
2.09
1.33
1.93
1.00
4.39
0.95
11
5042
0.92
7.30
0.91
6.21
0.81
2.22
0.06
0.63
3.59
2.93
0.23
2.53
3.03
4.40
0.96
11
293
0.91
6.73
0.89
5.46
2.30
0.95
0.19
0.45
4.05
0.63
0.18
0.54
3.25
1.88
0.88
11
724
0.88
5.65
0.94
8.07
1.43
0.58
0.19
0.55
1.48
0.38
0.39
1.04
2.00
2.26
0.91
11
59
0.88
5.56
0.72
2.91
1.12
0.73
0.22
0.33
2.35
0.84
0.37
0.49
1.85
1.54
0.81
11
394
0.88
5.52
0.88
5.34
0.63
0.35
0.13
0.06
4.97
1.77
0.54
0.30
0.98
3.77
0.87
11
3991
0.88
5.51
0.96
9.32
1.21
0.07
0.45
0.45
5.45
0.12
2.00
1.73
1.28
2.52
0.97
11
715
0.86
4.96
0.96
9.29
1.41
0.28
0.49
0.32
4.26
0.51
1.60
0.87
1.13
2.18
0.95
11
437
0.85
4.85
0.84
4.41
0.76
0.75
0.30
0.49
4.90
1.85
1.46
2.05
1.51
3.13
0.95
11
6082
0.85
4.85
0.94
7.51
0.81
0.45
0.34
0.48
6.95
1.58
2.34
2.61
1.27
4.17
0.97
11
912
0.85
4.76
0.94
7.60
0.95
0.69
0.20
0.59
3.95
1.52
0.76
1.49
1.64
3.53
0.94
11
31
0.84
4.74
0.82
4.00
0.59
0.35
0.44
0.13
3.17
0.95
1.78
0.52
0.94
2.06
0.91
11
216
0.83
4.50
0.83
4.23
0.66
0.62
0.09
0.03
4.10
2.02
0.34
0.08
1.28
3.57
0.93
11
1681
0.83
4.45
0.92
6.69
0.78
0.39
0.30
0.37
5.05
1.16
1.60
1.35
1.17
3.20
0.96
11
103
0.81
4.17
0.87
4.91
0.76
0.70
0.09
0.30
4.60
2.74
0.41
1.11
1.46
4.40
0.93
11
2903
0.81
4.12
0.95
8.50
0.92
0.20
0.32
0.30
8.74
0.80
2.51
1.67
1.12
4.24
0.98
11
413
0.81
4.11
0.95
8.85
0.88
0.38
0.07
0.09
5.34
1.23
0.34
0.31
1.26
3.91
0.95
11
1438
0.81
4.08
0.93
7.04
0.96
0.63
0.02
0.19
3.97
1.40
0.08
0.43
1.58
3.38
0.92
11
2235
0.80
4.01
0.89
5.55
0.73
0.22
0.42
0.36
7.48
1.12
3.34
2.29
0.95
4.04
0.98
11
4821
0.80
4.00
0.96
9.45
0.80
0.19
0.27
0.26
1.28
3.21
2.23
1.00
5.90
0.99
11
638
0.80
3.98
0.91
6.09
0.77
0.53
0.13
0.22
4.39
1.74
0.59
0.74
1.31
3.66
0.94
11
760
0.80
3.97
0.93
7.45
0.94
0.34
0.23
0.29
5.66
1.03
1.16
1.11
1.28
3.47
0.96
11
501
0.79
3.91
0.88
5.24
0.75
0.24
0.46
0.50
4.67
0.68
2.22
1.90
0.99
2.42
0.96
11
1538
0.79
3.90
0.91
6.15
0.78
0.20
0.22
0.05
3.77
0.59
0.79
0.17
0.98
2.32
0.90
11
292
0.79
3.89
0.82
4.10
0.70
0.83
0.05
0.23
3.30
2.23
0.16
0.52
1.53
3.43
0.85
11
528
0.79
3.81
0.75
3.23
0.84
1.07
0.36
0.95
4.51
2.41
1.37
3.86
1.91
3.58
0.96
11
75
0.78
3.80
0.81
3.88
2.04
0.36
0.21
0.24
3.00
0.25
0.19
0.23
2.40
1.22
0.83
11
244
0.78
3.78
0.90
5.76
0.68
0.61
0.06
0.23
9.04
4.38
0.55
1.62
1.29
7.24
0.97
11
5735
0.78
3.75
0.91
6.32
0.76
0.29
0.30
0.31
6.40
1.23
2.00
1.53
1.06
3.83
0.97
11
2120
0.78
3.72
0.95
9.08
0.74
0.29
0.11
0.08
2.62
1.25
0.67
1.03
7.72
0.99
11
328
0.77
3.66
0.77
3.41
0.75
0.71
0.38
0.88
4.32
2.20
1.67
3.46
1.46
3.53
0.93
11
1011
0.77
3.64
0.91
6.37
0.74
0.36
0.06
0.16
6.31
1.72
0.35
0.66
1.09
4.25
0.95
11
811
0.77
3.62
0.84
4.31
0.67
0.44
0.10
0.20
3.33
1.31
0.35
0.63
1.11
2.49
0.81
11
262
0.77
3.61
0.91
6.02
0.75
0.54
0.02
0.17
4.38
2.21
0.07
0.64
1.28
4.18
0.92
11
1332
0.77
3.61
0.92
6.65
0.79
0.51
0.18
0.35
4.64
1.60
0.85
1.17
1.30
3.57
0.94
11
104
0.77
3.57
0.84
4.35
0.53
0.19
0.54
0.50
4.55
0.98
3.37
2.61
0.72
2.80
0.96
11
91
0.76
3.52
0.89
5.55
1.09
0.23
0.37
0.29
3.84
0.37
0.82
0.50
1.32
2.15
0.83
11
127
0.75
3.44
0.90
6.00
0.35
0.00
0.00
0.08
3.84
0.02
0.00
0.63
0.35
1.75
0.86
12.44
11.59
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page51
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
11
1906
0.71
3.04
0.97
0.85
0.22
0.13
0.14
9.03
1.52
1.05
0.83
1.07
6.07
0.98
11
3531
0.71
3.03
0.89
5.61
0.72
0.21
0.33
0.26
7.95
1.32
2.79
1.66
0.93
4.71
0.98
11
934
0.71
3.03
0.92
6.73
0.72
0.36
0.04
0.02
7.71
2.74
0.33
0.10
1.08
6.22
0.98
11
1873
0.71
3.02
0.96
9.25
0.85
0.36
0.21
0.43
9.91
2.60
2.01
2.80
1.21
7.24
0.99
11
130
0.71
2.99
0.90
5.77
0.86
0.03
0.42
0.38
3.38
0.07
1.28
0.96
0.89
1.59
0.89
11
2037
0.70
2.98
0.92
6.42
0.63
0.23
0.18
0.11
8.43
1.93
1.78
0.85
0.86
5.50
0.98
11
88
0.70
2.98
0.91
6.08
0.69
0.06
0.27
0.13
3.97
0.21
1.20
0.54
0.75
2.06
0.91
11
366
0.70
2.95
0.95
8.65
0.67
0.18
0.16
0.11
2.16
2.20
1.16
0.85
8.03
0.99
11
137
0.70
2.94
0.67
2.53
0.71
0.37
0.76
0.67
3.37
1.01
2.72
2.35
1.08
2.16
0.96
11
828
0.70
2.92
0.93
7.12
0.63
0.27
0.10
0.06
5.89
1.78
0.60
0.37
0.89
4.25
0.92
11
969
0.70
2.91
0.91
6.08
0.66
0.35
0.18
0.26
6.39
2.57
1.14
1.47
1.01
5.16
0.94
11
87
0.69
2.89
0.75
3.25
0.92
1.57
0.16
0.84
4.41
3.84
0.58
2.69
2.49
5.00
0.93
11
179
0.69
2.87
0.87
5.06
0.64
0.05
0.57
0.77
5.02
0.28
3.39
4.03
0.69
2.74
0.96
11
8983
0.69
2.87
0.96
9.77
0.78
0.25
0.03
0.09
12.24
2.56
0.35
0.73
1.03
8.27
0.99
11
934
0.69
2.86
0.96
0.83
0.15
0.12
0.04
12.79
1.57
1.42
0.31
0.98
8.03
0.99
11
1049
0.69
2.85
0.89
5.67
0.68
0.28
0.40
0.60
4.91
1.15
2.29
2.62
0.96
3.24
0.96
11
146
0.69
2.84
0.65
2.41
0.39
0.43
0.29
0.16
1.82
1.34
1.00
0.52
0.82
1.81
0.84
11
509
0.69
2.84
0.89
5.51
0.70
0.18
0.30
0.17
4.88
0.78
1.58
0.74
0.88
2.97
0.95
11
1402
0.69
2.83
0.94
7.53
0.77
0.19
0.26
0.34
4.41
0.81
1.16
1.27
0.96
3.11
0.94
11
2097
0.68
2.81
0.97
0.78
0.15
0.07
0.02
1.65
0.91
0.15
0.93
8.04
0.99
11
268
0.68
2.77
0.95
8.82
0.83
0.00
0.24
0.10
0.01
1.66
0.55
0.83
3.85
0.97
11
546
0.68
2.76
0.94
7.55
0.72
0.29
0.07
0.04
10.66
2.99
0.76
0.29
1.01
7.77
0.99
11
12004
0.68
2.75
0.95
8.95
0.76
0.28
0.02
0.07
16.18
4.10
0.24
0.81
1.04
11
577
0.67
2.74
0.96
9.51
0.82
0.18
0.02
0.02
6.42
1.06
0.12
0.08
1.00
4.25
0.95
11
50
0.67
2.72
0.45
1.42
1.17
0.66
0.28
0.65
1.34
0.35
0.19
0.47
1.83
0.73
0.66
11
358
0.67
2.72
0.85
4.50
0.58
0.40
0.23
0.30
4.89
2.47
1.27
1.51
0.98
4.17
0.90
11
753
0.67
2.70
0.97
0.91
0.20
0.12
0.25
2.81
1.70
2.54
1.11
11
517
0.67
2.69
0.84
4.39
0.49
0.28
0.06
0.02
3.39
1.48
0.26
0.07
0.77
2.76
0.87
11
547
0.67
2.68
0.95
9.06
0.78
0.29
0.08
0.16
9.05
2.41
0.68
1.08
1.07
6.51
0.98
11
834
0.66
2.67
0.94
7.57
0.81
0.02
0.36
0.27
8.80
0.16
3.17
1.75
0.83
4.89
0.99
11
556
0.66
2.66
0.89
5.49
0.73
0.28
0.05
0.02
3.34
1.08
0.15
0.07
1.00
2.64
0.84
11
361
0.66
2.65
0.55
1.88
1.08
1.40
0.82
0.92
4.01
3.26
2.79
2.83
2.48
3.84
0.79
11
955
0.66
2.65
0.95
8.72
0.67
0.22
0.09
0.12
6.90
1.62
0.63
0.72
0.89
4.72
0.95
11
188
0.66
2.64
0.88
5.23
0.67
0.43
0.06
0.11
7.43
3.15
0.46
0.60
1.09
5.97
0.97
11
169
0.66
2.63
0.92
6.63
0.78
0.01
0.43
0.34
4.71
0.04
2.09
1.30
0.77
2.49
0.96
11
91
0.66
2.62
0.84
4.34
1.85
0.51
0.49
0.64
3.26
0.37
0.49
0.66
2.36
1.55
0.91
11
94
0.66
2.60
0.84
4.32
0.61
0.00
0.79
0.89
1.50
0.00
1.68
1.45
0.61
0.89
0.87
11
59
0.65
2.59
0.81
3.93
0.97
0.78
0.29
0.10
2.52
1.75
0.57
0.15
1.75
2.67
0.82
11
537
0.65
2.59
0.97
0.81
0.20
0.12
0.17
2.35
1.33
1.54
1.01
8.64
0.99
11
249
0.65
2.59
0.78
3.47
0.69
0.54
0.26
0.23
3.24
1.77
0.82
0.62
1.23
2.81
0.77
11
557
0.65
2.58
0.90
5.76
0.61
0.06
0.35
0.26
4.22
0.27
1.84
1.16
0.67
2.34
0.94
10.58
10.05
11.50
11.28
11.61
13.19
13.52
7.42
18.00
12.42
11.58
11.93
0.99
0.99
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page52
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
11
283
0.62
2.37
0.94
7.52
0.65
0.05
0.30
0.25
7.48
0.42
2.54
1.81
0.71
4.13
0.98
11
142
0.62
2.37
0.83
4.28
0.72
0.31
0.18
0.06
8.01
2.32
1.48
0.33
1.03
5.73
0.99
11
2959
0.62
2.36
0.92
6.75
0.72
0.20
0.13
0.05
8.49
1.69
1.18
0.35
0.92
5.64
0.98
11
880
0.62
2.36
0.93
7.34
0.70
0.32
0.11
0.18
5.74
1.75
2.20
1.03
11
202
0.61
2.34
0.85
4.49
0.77
0.28
0.22
0.61
5.44
1.53
1.16
2.85
1.05
3.95
0.94
11
1662
0.61
2.32
0.91
6.38
0.61
0.23
0.05
0.04
6.83
1.91
0.38
0.23
0.85
4.89
0.97
11
731
0.61
2.29
0.94
7.46
0.95
0.31
0.11
0.18
5.47
1.49
0.46
0.60
1.26
4.15
0.94
11
2205
0.61
2.29
0.68
2.62
0.68
0.79
0.41
0.54
4.01
2.68
2.34
2.66
1.47
3.36
0.81
11
2086
0.61
2.29
0.95
8.39
0.76
0.23
0.10
0.10
2.61
1.13
0.81
0.99
8.19
0.99
11
1156
0.61
2.29
0.71
2.86
0.74
0.86
0.51
0.62
6.12
4.08
4.06
4.03
1.60
5.14
0.91
11
91
0.61
2.29
0.64
2.33
0.90
1.18
0.26
1.23
2.02
1.38
0.41
2.07
2.07
1.92
0.87
11
1393
0.60
2.26
0.84
4.44
0.72
0.19
0.02
0.05
2.62
0.50
0.06
0.13
0.90
1.70
0.74
11
96
0.60
2.26
0.84
4.38
0.60
0.15
0.30
0.21
2.16
0.44
0.84
0.51
0.75
1.50
0.84
11
281
0.60
2.25
0.80
3.74
0.73
0.65
0.11
0.15
3.57
2.18
0.38
0.43
1.37
3.32
0.84
11
128
0.60
2.24
0.94
7.89
0.75
0.26
0.11
0.04
6.10
1.57
0.60
0.20
1.01
4.27
0.93
11
601
0.60
2.23
0.91
6.20
0.57
0.17
0.08
0.06
7.66
1.70
0.72
0.46
0.74
5.09
0.97
11
303
0.60
2.23
0.55
1.87
0.48
0.48
0.59
0.82
1.84
1.14
1.74
2.12
0.96
1.62
0.90
11
147
0.59
2.21
0.85
4.64
0.47
0.12
0.26
0.16
5.01
0.93
2.02
1.06
0.59
3.12
0.96
11
261
0.59
2.20
0.68
2.60
0.63
0.93
0.49
1.91
1.52
1.85
1.07
3.78
1.56
2.22
0.93
11
282
0.59
2.19
0.68
2.59
0.54
0.63
0.41
0.42
4.24
3.97
1.95
1.85
1.18
4.79
0.89
11
223
0.59
2.18
0.88
5.13
0.59
0.14
0.15
0.09
2.98
0.53
0.52
0.31
0.72
1.93
0.83
11
5107
0.59
2.18
0.95
8.21
0.84
0.24
0.15
0.28
5.52
1.16
0.75
1.03
1.08
3.87
0.95
11
213
0.59
2.18
0.82
4.07
0.45
0.03
0.45
0.36
2.95
0.14
2.19
1.50
0.48
1.57
0.92
11
347
0.58
2.15
0.93
6.90
0.76
0.02
0.21
0.24
3.83
0.08
0.78
0.66
0.79
1.98
0.88
11
135
0.58
2.15
0.76
3.34
0.38
0.13
0.15
0.03
2.29
0.56
0.61
0.13
0.50
1.49
0.80
11
1471
0.58
2.13
0.93
7.06
0.65
0.32
0.21
0.23
8.65
3.35
1.84
1.76
0.97
6.81
0.96
11
2090
0.58
2.13
0.95
9.03
0.60
0.18
0.03
0.01
8.79
1.97
0.27
0.07
0.79
5.91
0.97
11
197
0.58
2.13
0.91
6.18
0.77
0.16
0.05
0.05
3.57
0.62
0.15
0.13
0.94
2.36
0.86
11
35
0.58
2.12
0.76
3.32
0.76
0.57
0.11
0.36
1.09
0.86
0.14
0.34
1.33
1.21
0.72
11
159
0.57
2.11
0.85
4.48
0.98
0.74
0.47
0.30
4.51
2.97
1.57
0.79
1.72
4.57
0.90
11
126
0.57
2.10
0.69
2.71
1.14
1.07
0.90
0.54
4.38
3.53
2.14
1.07
2.21
4.76
0.86
11
223
0.57
2.09
0.95
8.28
0.68
0.18
0.13
0.18
6.80
1.40
0.96
1.01
0.86
4.66
0.97
11
934
0.57
2.08
0.91
6.20
0.82
0.33
0.01
0.06
6.86
2.24
0.05
0.28
1.15
5.30
0.97
11
403
0.57
2.07
0.87
4.91
0.55
0.13
0.29
0.30
3.25
0.58
1.23
1.05
0.68
2.07
0.89
11
1801
0.57
2.06
0.96
9.45
0.70
0.22
0.06
0.09
3.09
0.78
0.93
0.91
9.09
0.99
11
400
0.57
2.06
0.85
4.49
0.67
0.45
0.41
0.40
5.67
2.89
2.06
1.68
1.11
4.91
0.90
11
390
0.57
2.06
0.88
5.26
0.57
0.16
0.17
0.11
3.80
0.78
0.81
0.43
0.73
2.48
0.91
11
115
0.56
2.04
0.57
1.97
0.29
0.20
0.31
0.21
1.24
0.60
0.89
0.54
0.49
1.00
0.64
11
556
0.56
2.03
0.95
8.49
0.65
0.18
0.00
0.02
6.90
1.45
0.01
0.13
0.84
4.56
0.95
11
120
0.56
2.03
0.62
2.25
0.48
0.36
0.00
0.07
1.78
0.90
0.00
0.15
0.83
1.46
0.50
11
5274
0.56
2.02
0.92
6.52
0.60
0.23
0.29
0.32
6.74
1.85
2.08
2.13
0.83
4.58
0.93
16.35
11.64
13.13
12.77
0.99
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page53
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
11
147
0.51
1.78
0.54
1.81
1.24
1.41
0.21
0.81
1.90
1.26
0.23
0.75
2.66
1.76
0.77
11
86
0.51
1.77
0.79
3.65
1.01
0.67
0.58
0.49
3.20
1.85
1.27
1.00
1.68
2.97
0.78
11
102
0.50
1.75
0.81
3.91
0.54
0.33
0.22
0.52
3.49
1.69
0.97
1.88
0.87
2.97
0.90
11
4667
0.50
1.75
0.98
0.61
0.16
0.13
0.18
23.02
4.67
3.16
3.88
0.77
11
1283
0.50
1.74
0.96
9.47
0.92
0.32
0.18
0.20
11.04
3.34
1.57
1.40
1.24
8.41
0.98
11
54
0.50
1.74
0.57
1.94
0.57
0.03
0.42
0.12
0.99
0.03
0.60
0.17
0.54
0.38
0.54
11
222
0.49
1.67
0.70
2.76
0.62
0.56
0.36
0.29
2.67
1.82
0.99
0.69
1.18
2.61
0.70
11
43
0.48
1.66
0.60
2.11
0.79
1.05
0.64
0.46
2.16
2.61
1.17
0.66
1.84
2.77
0.79
11
56
0.47
1.62
0.76
3.30
0.53
0.16
0.41
0.70
1.48
0.37
0.91
1.23
0.70
1.05
0.81
11
536
0.46
1.56
0.88
5.16
0.70
0.04
0.16
0.24
3.19
0.13
0.51
0.82
0.66
1.60
0.81
11
7841
0.46
1.55
0.94
7.67
0.82
0.32
0.37
0.32
9.49
2.99
2.82
2.15
1.14
7.10
0.96
11
325
0.46
1.55
0.68
2.65
0.21
0.18
0.74
0.69
1.37
0.86
3.29
2.77
0.04
0.12
0.89
11
249
0.46
1.54
0.53
1.79
1.23
1.07
0.31
0.93
1.94
0.98
0.36
0.84
2.29
1.50
0.62
11
666
0.46
1.54
0.96
9.70
0.68
0.13
0.01
0.03
6.56
1.02
0.06
0.15
0.81
4.14
0.94
11
150
0.46
1.54
0.91
6.38
0.52
0.03
0.28
0.35
6.29
0.24
2.44
2.80
0.55
3.42
0.96
11
106
0.44
1.49
0.78
3.50
0.66
0.53
0.14
0.07
2.86
2.01
0.44
0.16
1.19
2.82
0.87
11
101
0.44
1.46
0.72
2.94
0.57
0.04
0.50
0.56
1.39
0.07
0.93
0.89
0.62
0.76
0.76
11
1976
0.44
1.46
0.83
4.16
0.68
0.48
0.47
0.38
6.73
3.82
2.95
2.20
1.16
5.99
0.92
11
353
0.43
1.43
0.82
4.00
0.71
0.28
0.25
0.20
2.97
0.92
0.68
0.53
0.99
2.16
0.72
11
56
0.43
1.42
0.49
1.57
1.04
1.39
0.40
0.48
1.87
1.86
0.52
0.57
2.43
2.24
0.67
11
137
0.43
1.42
0.87
4.89
0.81
0.36
0.30
0.35
3.47
1.33
0.87
0.85
1.18
2.78
0.83
11
105
0.42
1.38
0.86
4.75
0.84
0.39
0.31
0.05
6.05
2.44
1.50
0.24
1.23
4.84
0.92
11
125
0.41
1.34
0.58
2.03
0.57
0.70
0.34
0.12
2.36
2.39
0.99
0.34
1.27
2.77
0.77
11
117
0.41
1.33
0.58
2.03
0.53
0.23
0.87
1.07
0.83
0.25
1.07
1.28
0.30
0.24
0.67
11
65
0.40
1.32
0.02
0.07
0.48
1.30
0.35
0.08
1.01
2.07
0.47
0.10
1.78
1.85
0.59
11
156
0.38
1.22
0.74
3.13
0.60
0.32
0.49
0.61
3.02
1.23
1.54
1.64
0.92
2.34
0.73
11
35
0.35
1.14
0.59
2.08
0.13
0.31
0.80
0.34
0.31
0.61
1.51
0.55
0.18
0.23
0.82
11
98
0.35
1.12
0.57
1.97
0.63
0.55
0.53
0.51
1.92
1.28
1.03
0.93
1.18
1.83
0.50
11
225
0.34
1.10
0.71
2.82
0.58
0.08
0.58
0.82
1.30
0.14
0.92
1.07
0.50
0.59
0.67
11
171
0.34
1.08
0.80
3.76
0.70
0.12
0.43
0.34
3.96
0.49
1.54
1.13
0.82
2.35
0.78
11
45
0.34
1.08
0.50
1.62
0.09
0.43
1.15
1.06
0.44
1.56
3.87
3.50
0.34
0.82
0.87
11
533
0.34
1.07
0.41
1.28
1.15
1.12
0.12
1.23
1.70
1.00
0.13
1.01
2.27
1.42
0.66
11
243
0.33
1.05
0.86
4.84
0.61
0.24
0.31
0.42
4.09
1.26
1.28
1.53
0.85
2.92
0.85
11
774
0.33
1.04
0.83
4.27
0.45
0.16
0.02
0.16
3.29
0.89
0.08
0.75
0.60
2.26
0.86
11
47
0.29
0.92
0.73
3.05
0.47
0.13
0.47
0.46
1.38
0.30
1.06
0.98
0.34
0.53
0.69
11
199
0.27
0.84
0.60
2.10
0.44
0.37
0.19
0.36
1.43
0.96
0.38
0.55
0.81
1.32
0.68
11
111
0.25
0.76
0.48
1.56
0.31
0.18
0.21
0.29
1.00
0.46
0.51
0.68
0.49
0.81
0.53
11
30
0.21
0.64
0.09
0.25
0.14
0.54
0.12
0.12
0.33
0.99
0.19
0.19
0.68
0.80
0.43
11
31
0.17
0.52
0.66
2.46
0.23
0.65
0.88
0.73
0.98
1.97
2.75
2.07
0.42
0.85
0.79
11
361
0.12
0.38
0.79
3.70
0.59
0.11
0.24
0.14
3.26
0.46
0.90
0.50
0.70
1.96
0.71
11
734
0.03
0.08
0.47
1.51
0.65
0.02
0.22
0.63
3.20
0.07
0.77
2.03
0.63
1.49
0.84
12.47
14.83
0.99
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page54
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
10
149
0.69
2.70
0.84
3.75
0.24
1.11
0.15
0.83
0.32
2.13
0.30
1.34
1.35
1.70
0.92
10
1401
0.68
2.61
0.96
9.53
0.72
0.27
0.06
0.09
9.38
2.52
0.50
0.62
0.99
6.88
0.98
10
81
0.68
2.61
0.75
2.96
1.20
1.12
0.39
0.27
2.19
1.58
0.57
0.35
2.31
2.29
0.76
10
1872
0.63
2.29
0.95
8.08
0.69
0.29
0.05
0.06
8.10
2.53
0.42
0.42
0.98
6.08
0.97
10
53
0.62
2.26
0.46
1.25
0.68
0.75
0.23
0.32
5.31
4.85
1.71
1.26
1.42
5.73
0.97
10
31
0.61
2.20
0.94
7.06
1.28
0.42
0.82
1.13
5.65
1.31
2.71
3.26
0.86
2.03
0.98
10
40
0.60
2.09
0.89
5.10
1.03
0.47
0.29
0.81
2.39
0.97
0.62
1.30
1.50
2.17
0.90
10
951
0.59
2.06
0.93
6.70
0.62
0.30
0.24
0.25
3.72
2.57
2.31
0.92
7.78
0.97
10
20
0.58
2.04
0.56
1.66
0.30
0.27
0.47
0.38
1.05
0.68
1.35
0.78
0.03
0.04
0.87
10
37
0.58
2.04
0.89
4.84
1.29
0.23
0.09
0.05
2.39
0.41
0.19
0.08
1.51
2.05
0.90
10
113
0.57
1.98
0.73
2.61
0.21
0.27
0.20
0.09
0.34
0.52
0.37
0.15
0.48
0.52
0.81
10
464
0.57
1.97
0.82
3.86
0.84
0.93
0.34
0.18
1.89
1.44
0.72
0.27
1.77
2.36
0.95
10
86
0.55
1.88
0.56
1.64
1.30
2.76
0.18
0.29
2.28
1.89
0.22
0.34
4.05
2.12
0.73
10
29
0.48
1.55
0.90
5.35
0.63
0.27
0.06
0.22
4.22
1.40
0.28
0.76
0.90
3.16
0.90
10
107
0.48
1.54
0.78
3.31
0.67
0.81
0.22
0.33
3.00
3.39
0.89
0.77
1.48
4.16
0.98
10
878
0.47
1.52
0.92
6.26
0.96
0.40
0.12
0.15
4.86
1.53
0.43
0.32
1.37
3.79
0.93
10
42
0.46
1.45
0.87
4.28
0.72
0.53
0.35
0.76
0.50
0.79
0.45
1.06
1.24
0.73
0.95
10
281
0.45
1.42
0.66
2.34
0.30
0.20
0.23
0.09
1.79
0.88
0.85
0.28
0.50
1.48
0.78
10
49
0.37
1.13
0.94
7.27
0.64
0.15
0.13
0.28
5.60
0.83
0.83
1.38
0.49
2.03
0.93
10
340
0.34
1.02
0.92
6.08
0.52
0.16
0.11
0.21
6.64
1.52
0.96
1.61
0.68
4.43
0.96
10
44
0.26
0.78
0.91
5.82
1.04
0.03
0.32
0.06
3.33
0.08
0.84
0.10
1.01
1.83
0.91
10
42
0.26
0.76
0.79
3.13
3.52
1.68
0.54
1.64
6.75
2.12
0.79
1.94
5.21
4.68
0.97
10
157
0.23
0.68
0.40
1.17
0.28
0.30
0.16
0.07
0.52
0.43
0.22
0.09
0.58
0.54
0.43
10
20
0.28
0.83
0.32
0.88
0.07
0.37
0.13
1.18
0.16
0.68
0.21
1.75
0.44
0.52
0.68
10
40
0.34
1.02
0.48
1.45
0.16
0.33
0.16
1.00
0.30
0.52
0.25
1.38
0.17
0.16
0.68
9
72
0.84
4.12
0.73
2.59
2.09
0.76
0.09
1.59
1.57
0.30
0.04
0.56
2.86
0.82
0.81
9
46
0.78
3.34
0.77
2.94
1.06
0.67
0.54
0.76
1.37
0.24
0.30
0.43
1.73
0.56
0.81
105
0.78
3.31
0.79
3.13
1.15
0.86
0.01
0.49
9.29
0.16
3.20
2.02
9
18
0.77
3.16
0.75
2.57
0.57
0.15
0.76
0.64
0.99
0.19
1.43
0.64
0.72
0.80
0.89
9
50
0.75
3.01
0.85
3.89
0.77
0.92
0.37
1.82
0.50
0.82
0.35
0.91
1.69
0.90
0.87
9
64
0.75
2.98
0.92
4.79
3.72
0.33
1.05
1.80
1.60
0.23
0.69
0.79
4.05
1.75
0.92
172
0.72
2.73
0.85
3.92
0.82
0.28
0.19
0.33
1.36
0.33
0.19
0.26
1.10
0.91
0.75
9
50
0.61
2.03
0.70
2.19
0.92
0.94
0.21
0.16
3.01
1.38
0.49
0.26
1.86
2.00
0.97
9
67
0.43
1.26
0.21
0.49
0.05
0.30
0.88
0.96
0.13
0.54
1.61
1.59
0.26
0.31
0.71
9
17
0.36
1.01
0.55
1.31
5.91
3.81
2.42
0.48
2.49
2.36
2.09
0.41
9.72
3.51
0.96
9
13
0.17
0.46
0.58
1.41
0.10
0.15
0.52
0.29
0.10
0.12
0.49
0.29
0.05
0.02
0.79
9
52
0.08
0.22
0.60
1.81
1.09
0.34
0.38
0.65
3.50
1.05
1.09
0.99
1.43
2.58
0.95
8
283
0.99
17.90
0.97
9.74
0.86
0.01
0.14
0.01
6.72
0.02
1.05
0.05
0.85
1.38
0.97
8
864
0.98
12.28
0.98
9.96
0.75
0.36
0.18
0.24
1.90
2.63
1.88
1.12
5.69
0.99
8
1526
0.98
11.20
0.96
7.28
0.74
0.02
0.19
0.29
4.74
0.04
1.16
0.93
0.72
1.51
0.95
8
50
0.97
10.69
0.96
7.81
0.91
0.17
0.09
0.12
4.85
0.29
0.41
0.31
1.08
1.69
0.94
9
9
10.29
16.00
12.01
14.45
0.99
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page55
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
8
359
0.91
5.42
0.84
3.53
0.59
0.10
0.50
0.47
4.46
0.38
3.44
1.76
0.49
1.50
0.97
8
152
0.91
5.32
0.67
2.01
0.43
0.75
0.12
0.29
3.08
3.03
0.66
1.03
1.18
3.69
0.92
98
0.90
5.00
0.94
6.09
1.44
0.21
0.34
0.88
2.61
0.25
0.71
0.98
1.65
2.00
0.94
8
374
0.90
4.94
0.70
2.19
0.58
0.04
0.53
0.61
1.66
0.04
1.49
1.07
0.62
0.51
0.92
8
689
0.90
4.92
0.94
6.39
0.87
1.50
0.11
0.83
1.67
1.34
0.30
1.87
2.37
3.38
0.99
8
203
0.88
4.53
0.95
6.80
2.06
0.25
0.42
0.73
2.43
7.04
9.31
2.32
8
291
0.87
4.41
0.94
6.05
0.64
0.46
0.13
0.17
7.62
3.81
1.42
1.07
1.10
7.90
0.99
65
0.86
4.07
0.65
1.90
1.50
0.76
1.51
1.81
3.06
1.28
1.46
3.42
2.26
5.65
0.95
318
0.84
3.77
0.91
4.83
0.65
0.35
0.12
0.06
4.53
1.09
0.46
0.15
1.00
2.51
0.98
24
0.83
3.68
0.74
2.49
0.68
2.29
0.04
3.04
1.06
2.31
0.04
1.88
2.97
2.25
0.94
8
217
0.82
3.50
0.96
7.70
0.63
0.17
0.21
0.27
6.94
8.81
6.75
0.80
8
201
0.82
3.49
0.84
3.43
0.54
0.30
0.04
0.29
3.36
1.40
0.22
0.89
0.84
2.75
0.88
8
214
0.81
3.36
0.94
6.24
0.62
0.24
0.01
0.06
5.00
1.47
0.07
0.25
0.86
3.82
0.96
8
304
0.81
3.36
0.52
1.37
0.23
0.07
0.37
0.43
0.84
0.18
1.44
0.91
0.16
0.25
0.72
8
266
0.80
3.32
0.91
4.94
0.48
0.43
0.41
0.06
4.05
1.36
2.10
0.27
0.05
0.12
0.98
8
116
0.80
3.30
0.91
4.95
0.62
0.34
0.34
0.76
6.72
2.40
3.96
4.87
0.95
6.30
0.99
8
180
0.78
3.03
0.88
4.13
0.40
0.16
0.13
0.04
2.75
0.51
0.48
0.12
0.56
1.37
0.94
8
1077
0.77
2.97
0.92
5.43
0.57
0.26
0.09
0.20
5.81
2.02
0.83
1.08
0.83
4.79
0.98
8
155
0.77
2.92
0.95
6.98
0.93
0.43
0.15
0.20
9.92
4.50
1.94
1.44
1.36
8
57
0.76
2.91
0.54
1.45
1.12
1.05
0.14
1.02
2.78
1.86
0.43
1.65
2.17
2.40
0.80
8
48
0.76
2.90
0.40
0.99
0.70
0.73
0.12
0.95
4.75
3.51
0.65
3.27
1.44
4.66
0.95
8
64
0.76
2.90
0.56
1.52
0.81
1.22
0.14
0.05
1.51
1.36
0.12
0.04
2.02
1.58
0.70
8
246
0.76
2.87
0.93
5.66
0.99
0.13
0.37
0.25
6.76
0.56
2.17
0.84
0.86
3.16
0.97
8
157
0.75
2.81
0.88
4.13
0.60
0.45
0.59
0.51
1.01
0.65
1.45
0.67
1.06
1.60
0.94
8
33
0.75
2.81
0.83
3.26
2.33
0.41
1.63
0.25
2.80
0.74
2.32
0.76
2.74
6.78
0.98
8
41
0.75
2.81
0.39
0.95
0.84
0.76
0.40
0.84
2.52
1.52
1.13
1.30
1.60
2.23
0.90
8
87
0.75
2.79
0.11
0.24
0.77
1.02
0.04
0.44
1.60
1.67
0.10
0.68
1.79
1.75
0.62
8
62
0.75
2.77
0.94
6.08
0.71
0.47
0.37
0.27
1.17
129.23
1.00
8
72
0.75
2.75
0.37
0.89
0.42
0.15
0.43
0.35
0.63
0.13
0.81
0.32
0.57
0.33
0.67
8
69
0.72
2.56
0.20
0.46
0.04
0.06
0.24
0.32
0.34
0.35
1.65
1.25
0.10
0.41
0.76
8
10
0.72
2.51
0.15
0.34
0.79
1.77
0.84
0.75
0.58
1.22
0.48
0.26
2.56
1.06
0.64
460
0.71
2.50
0.91
5.04
0.53
0.31
0.03
0.20
8.28
3.96
0.37
1.66
0.84
7.81
0.99
8
29
0.71
2.44
0.29
0.67
0.18
0.87
0.37
1.00
0.59
1.92
1.14
1.66
1.05
1.65
0.93
8
53
0.70
2.38
0.41
1.00
0.47
0.78
0.27
0.65
0.89
1.19
0.45
0.62
1.25
1.31
0.74
102
0.69
2.34
0.66
1.76
1.06
2.22
0.21
0.04
3.56
2.14
0.18
0.03
3.28
2.68
0.95
33
0.69
2.32
0.74
2.47
1.60
0.84
1.27
1.94
8.75
5.92
5.09
5.55
2.44
324
0.67
2.23
0.58
1.60
0.29
0.18
0.23
0.05
2.30
1.15
1.57
0.20
0.47
2.02
0.90
8
8
8
8
8
8
8
8
43.99
32.69
85.90
58.69
59.28
22.97
23.84
22.98
12.17
10.41
1.00
1.00
1.00
0.99
14
0.67
2.20
0.55
1.48
1.25
0.61
0.13
1.89
6.47
2.56
0.58
4.51
1.86
5.42
0.98
132
0.65
2.11
0.94
5.98
0.89
0.41
0.13
0.37
5.32
2.31
0.80
1.43
1.30
5.86
0.98
8
34
0.65
2.10
0.52
1.38
0.59
0.38
0.09
0.89
5.92
1.33
7.48
0.96
8
79
0.65
2.08
0.85
3.63
0.62
0.05
0.59
1.07
0.11
1.98
1.94
0.57
8
8
10.55
1.95
10.05
1.11
0.99
0.95
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page56
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
17
0.09
0.23
0.04
0.10
0.56
0.38
1.10
1.61
0.55
0.23
0.59
0.80
0.94
0.37
0.70
7
104
0.99
14.44
0.82
2.85
1.14
1.18
0.09
0.01
1.48
1.03
0.25
0.01
2.32
1.59
0.85
7
163
0.99
13.23
0.85
3.20
0.57
0.15
0.26
0.17
0.71
117.74
1.00
7
283
0.98
10.30
0.90
4.19
0.89
0.35
0.06
0.14
2.34
1.14
0.36
0.51
1.24
1.91
0.97
7
245
0.97
9.67
0.79
2.57
1.32
0.61
0.15
0.49
3.19
2.10
0.84
1.67
1.94
2.94
0.95
7
236
0.97
8.77
0.68
1.87
1.14
0.88
0.17
0.10
6.89
2.33
2.02
7
18
0.96
8.21
0.38
0.82
0.14
0.87
0.01
0.31
0.12
1.14
0.02
0.36
0.73
0.49
0.87
7
43
0.95
7.10
0.23
0.47
0.31
0.93
0.23
0.25
0.21
0.95
0.33
0.22
1.24
0.59
0.60
7
26
0.95
6.90
0.26
0.53
0.70
0.15
0.73
1.01
0.49
0.12
1.14
1.09
0.55
0.21
0.84
116
0.95
6.82
0.67
1.83
0.38
0.04
0.24
0.10
2.45
0.44
4.00
1.02
0.42
1.74
0.99
38
0.95
6.61
0.71
2.03
0.38
0.23
0.56
0.87
0.16
0.16
0.55
0.48
0.61
0.19
0.77
7
118
0.94
6.35
0.25
0.52
0.97
1.19
0.03
0.07
3.96
6.27
0.33
0.46
2.16
5.16
0.99
7
331
0.94
6.31
0.74
2.23
0.48
1.24
0.22
0.07
0.32
0.76
0.30
0.05
1.72
0.69
0.87
7
23
0.94
6.28
0.30
0.64
1.69
1.01
0.35
1.06
8.89
8.16
3.65
7.14
2.70
9.52
0.99
7
47
0.94
6.16
0.04
0.08
0.69
0.12
0.28
1.02
4.52
0.77
3.15
7.38
0.81
3.08
0.99
7
58
0.94
6.02
0.84
3.08
0.65
0.16
0.42
0.70
0.37
0.12
0.54
0.54
0.81
0.32
0.84
7
389
0.93
5.80
0.58
1.41
0.81
0.37
0.06
0.32
1.48
1.15
0.25
0.87
1.19
1.41
0.83
7
8
7
7
170.84
17.95
47.89
15.99
194.68
80.66
18.15
1.00
114
0.92
5.11
0.86
3.44
0.98
1.26
0.50
0.28
9.71
4.95
2.37
1.04
2.24
7.23
0.99
7
78
0.91
4.78
0.84
3.06
0.64
0.46
0.57
0.63
3.30
1.03
1.65
1.35
1.10
1.96
0.98
7
11
0.91
4.77
0.56
1.36
0.52
0.56
0.59
0.99
0.18
0.33
0.46
0.38
1.09
0.28
0.74
7
10
0.90
4.64
0.21
0.43
0.52
0.03
0.29
0.55
0.13
0.02
0.15
0.17
0.48
0.09
0.38
154
0.90
4.59
0.89
3.86
0.70
0.35
0.38
0.27
0.72
0.53
0.94
0.44
0.34
0.22
0.92
7
38
0.89
4.28
0.91
4.34
2.43
1.12
0.12
0.31
1.55
1.02
0.19
0.28
3.56
1.80
0.95
7
57
0.88
4.22
0.01
0.03
0.30
1.33
0.01
0.44
0.14
0.80
0.01
0.23
1.63
0.45
0.68
7
14
0.88
4.19
0.79
2.58
1.41
1.63
0.08
0.17
3.38
1.26
0.13
0.20
3.04
1.94
0.96
7
93
0.88
4.10
0.51
1.17
0.39
0.53
0.14
0.10
0.19
0.43
0.15
0.07
0.91
0.31
0.54
7
12
0.87
3.95
0.28
0.59
1.73
1.98
0.00
0.02
0.38
3.81
3.72
205.65
1.00
7
61
0.86
3.80
0.51
1.18
2.12
1.62
1.89
2.71
2.11
1.70
3.49
4.25
3.75
1.93
0.99
7
40
0.86
3.79
0.46
1.03
0.91
0.00
0.55
0.22
0.56
0.00
0.99
0.23
0.91
0.33
0.75
7
70
0.86
3.74
0.39
0.84
0.20
0.03
0.30
1.19
0.15
0.03
0.65
1.40
0.23
0.11
0.92
7
81
0.86
3.72
0.78
2.53
1.55
1.09
1.49
0.68
2.23
2.74
5.13
1.23
0.46
0.53
1.00
7
45
0.86
3.70
0.69
1.91
1.92
1.03
0.04
0.46
1.64
1.63
0.07
0.53
2.95
1.79
0.93
7
35
0.85
3.68
0.64
1.66
0.36
0.35
0.80
1.47
0.46
0.60
2.23
2.45
0.71
0.54
0.97
7
8
0.85
3.62
0.43
0.95
3.96
4.09
3.34
7.26
2.93
3.93
4.56
5.07
8.05
3.88
0.98
7
90
0.85
3.55
0.67
1.79
1.34
0.61
0.07
0.26
6.07
4.89
0.77
1.70
1.94
5.87
0.99
7
82
0.84
3.43
0.15
0.31
2.16
1.13
0.81
1.24
2.08
1.99
1.91
1.82
3.29
2.12
0.83
7
31
0.84
3.42
0.72
2.06
1.23
1.76
0.59
0.20
0.98
0.50
0.27
0.13
2.99
0.65
0.75
7
155.52
275.61
569
0.83
3.32
0.32
0.67
0.93
0.64
0.26
0.11
1.39
1.68
0.95
0.23
1.57
1.55
0.89
7
15
0.82
3.24
0.74
2.23
2.26
1.27
0.47
0.09
0.83
0.48
0.55
0.07
3.53
0.67
0.78
7
17
0.82
3.23
0.32
0.69
0.87
2.52
0.26
1.72
0.39
1.54
0.33
0.81
3.38
0.89
0.95
7
39
0.82
3.22
0.14
0.27
5.32
2.53
3.26
4.08
1.61
1.16
1.98
1.77
7.85
1.51
0.91
7
Case5 11cv 02509LHK
Document424
Filed0517 13
2
Page57
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
7
12
0.60
1.69
0.69
1.90
1.46
4.18
1.01
6.31
1.14
1.33
0.55
2.46
5.64
1.53
0.99
7
16
0.60
1.67
0.68
1.85
0.81
0.54
0.75
0.35
0.11
0.15
0.24
0.07
1.35
0.14
0.82
224
0.59
1.64
0.33
0.70
1.58
0.63
0.86
0.91
1.44
1.03
2.22
1.42
2.21
1.32
0.87
7
27
0.59
1.62
0.59
1.47
2.44
0.08
0.44
1.28
0.37
0.02
0.24
0.25
2.36
0.24
0.75
7
52
0.58
1.59
0.79
2.62
0.26
0.23
0.06
0.07
0.49
0.88
0.33
0.20
0.03
0.04
0.92
7
31
0.54
1.45
0.67
1.83
2.56
0.74
0.13
0.92
0.54
0.30
0.06
0.26
3.30
0.50
0.76
878
0.50
1.31
0.59
1.48
1.85
0.75
0.70
0.51
7
88
0.49
1.27
0.79
2.57
0.24
0.32
0.37
0.19
7
9
0.49
1.26
0.61
1.34
2.99
4.04
3.39
3.51
7
14
0.42
1.04
0.60
1.48
5.60
3.82
2.25
7
15
0.39
0.95
0.62
1.59
7.30
2.86
2.56
7
68
0.38
0.91
0.51
1.17
4.24
1.64
7
34
0.36
0.85
0.62
1.57
0.33
7
11
0.34
0.81
0.14
0.27
3.17
7
12
0.31
0.74
0.60
1.29
8.69
7
47
0.24
0.55
0.29
0.61
2.29
1.15
0.73
0.52
7
24
0.12
0.26
0.09
0.17
4.06
2.08
1.49
2.58
7
7
46.85
35.03
2.49
6.95
1.36
2.90
3.35
5.09
1.33
1.91
0.65
0.02
2.67
1.80
11.14
12.08
45.73
19.62
11.05
3.14
2.25
2.54
0.88
4.58
3.84
3.05
0.76
0.62
0.88
0.79
0.24
0.27
1.18
0.05
1.00
0.34
0.63
0.44
2.60
43.87
1.00
0.08
0.54
1.00
9.42
3.07
0.96
5.44
0.97
5.88
0.72
0.64
0.27
0.32
0.19
0.93
0.15
5.84
0.45
0.55
0.29
3.44
0.65
0.46
10.16
6.51
0.66
18.07
0.61
20.25
0.66
14.21
15.65
6.14
19.43
1.00
14
0.08
0.17
0.24
0.50
0.71
0.45
0.92
0.81
0.43
0.54
1.27
0.70
0.27
0.11
0.95
187
0.08
0.17
0.37
0.78
0.10
0.08
0.12
0.82
0.08
0.12
0.24
0.97
0.18
0.10
0.77
7
10
0.18
0.42
0.29
0.62
5.13
4.95
4.84
5.17
5.01
0.98
7
15
0.22
0.50
0.53
1.26
1.02
0.23
0.62
0.63
0.20
0.09
0.27
0.16
0.79
0.10
0.56
7
17
0.43
1.07
0.48
1.10
5.55
2.37
2.37
1.63
3.35
2.98
3.45
1.37
7.92
3.33
0.96
201
0.97
7.68
0.90
3.51
6
98
0.96
7.13
0.97
6.67
6
8
0.96
6.83
0.92
4.03
6
222
0.95
5.98
0.92
4.09
6
8
0.95
5.93
0.72
1.48
6
28
0.93
5.17
0.09
0.15
6
72
0.92
4.79
0.48
0.95
6
17
0.92
4.72
0.83
2.13
6
25
0.91
4.36
0.24
0.35
131
0.91
4.26
0.91
3.08
6
12
0.90
4.06
0.78
1.78
6
18
0.90
4.03
0.86
2.35
402
0.89
3.99
0.79
2.26
6
41
0.89
3.96
0.90
2.05
6
77
0.89
3.95
0.77
2.12
6
12
0.88
3.76
0.76
1.68
6
36
0.88
3.74
0.03
0.05
6
8
0.87
3.57
0.13
0.22
6
93
0.87
3.55
0.56
1.16
7
7
6
6
6
15.96
30.79
17.07
35.69
46.74
Case5 11cv 02509LHK
Document424
2
Filed0517 13
Page58
of
62
Exhibit 2
Intel
Section
Years
Job
Title
of
1
Section
Total
Level
Emp Years
Data
T
2
Section
Change
Correlation
Coeff
Stat
Correlation
T
Coeff
149
0.71
2.03
0.98
7.11
6
22
0.68
1.86
0.36
0.66
6
10
0.61
1.56
0.98
8.07
6
8
0.55
1.32
0.12
0.18
6
14
0.52
1.20
0.93
2.58
6
10
0.51
1.19
0.65
1.47
6
34
0.51
1.18
0.76
1.67
6
15
0.49
1.14
0.50
0.99
6
9
0.42
0.93
0.35
0.52
6
31
0.41
0.90
0.16
0.16
6
12
0.27
0.55
0.80
2.34
6
8
0.24
0.49
0.33
0.61
6
13
0.23
0.47
0.89
2.81
6
10
0.21
0.42
0.67
1.28
6
40
0.18
0.37
0.60
1.29
6
24
0.09
0.18
0.42
0.65
6
11
0.02
0.04
0.58
1.23
10
0.41
0.90
0.20
0.20
170
0.74
2.21
0.06
0.10
6
6
6
3
Regression
Contemp
Stat
Revenue
Section
Coefficients
Lagged
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
Effect
T
Stat
r2
6
Case5 11cv 02509LHK
Document424
Exhibit
Filed0517 13
2
Page59
62
of
2
Intuit
Section
Job
Title
of
1
Section
Total
Years
Emp
Data
Level
Years
T
Coeff
11
2981
11
2
Section
Change
Correlation
T
Coeff
Stat
0.60
2.26
0.97
597
0.59
2.18
0.95
11
293
0.54
1.91
0.97
11
150
0.40
1.29
11
140
0.26
10
170
10
1571
10
69
10
194
9
9
57
1073
Correlation
3
Section
Regression
Contemp
Stat
Coefficients
Lagged
Revenue
Regression
SJ
Emp
12.05
1.50
1.01
0.26
0.34
8.57
1.13
1.33
0.48
0.04
11.05
1.50
1.17
0.49
0.76
3.31
2.01
1.70
0.81
0.05
0.13
0.69
0.78
3.55
0.98
10.93
0.55
1.85
0.79
0.49
1.60
0.40
Contemp
10.44
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
5
T
L
Section
6
Effect
Stat
r2
2.21
1.05
1.42
2.51
4.97
0.99
8.97
3.99
3.14
0.29
2.46
5.57
0.98
0.08
8.38
2.13
1.64
0.29
2.67
3.97
0.97
0.80
0.27
4.41
1.72
1.21
0.33
3.71
2.77
0.87
1.28
0.43
1.77
1.41
2.27
0.74
2.01
1.97
2.14
0.71
1.08
0.18
0.15
0.12
4.91
0.37
0.47
0.23
0.89
1.50
0.97
3.16
1.34
1.01
0.36
0.02
6.15
3.76
0.14
2.35
0.30
0.78
0.19
0.68
0.18
0.15
0.28
1.47
0.42
0.17
0.50
0.57
0.52
1.25
0.76
2.86
1.39
1.36
0.33
0.44
1.89
0.78
0.27
0.43
2.75
1.12
0.94
0.67
2.39
0.08
0.21
0.62
0.82
0.05
0.38
0.53
0.91
0.07
0.24
1.44
0.92
0.40
0.64
2.22
0.69
2.34
1.15
0.25
0.30
0.41
3.94
0.68
1.77
0.85
1.40
2.74
0.89
13.75
11.01
0.99
9
94
0.59
1.94
0.57
1.56
1.10
0.36
0.01
1.56
2.52
0.28
0.01
2.86
1.47
1.11
0.90
9
81
0.54
1.70
0.77
2.94
1.63
1.09
0.15
0.23
4.23
1.86
0.49
0.46
2.71
4.12
0.92
0.53
1.67
0.68
2.05
0.34
0.90
0.56
0.09
0.33
0.28
0.37
0.02
0.56
0.14
0.51
0.17
0.46
0.74
2.70
2.01
0.71
0.11
0.23
2.20
0.66
0.16
0.18
2.73
2.07
0.75
9
9
758
46
9
486
0.01
0.02
0.46
1.28
1.34
1.60
0.55
0.31
4.91
3.62
2.13
0.90
2.94
4.97
0.94
8
113
0.80
3.25
0.91
4.90
0.44
0.22
1.21
2.04
1.78
0.33
2.25
5.15
0.66
0.73
1.00
8
24
0.68
2.25
0.72
2.32
1.52
2.13
0.81
0.39
0.39
0.24
0.10
0.06
3.65
0.29
0.83
8
29
0.61
1.87
0.76
2.62
2.07
2.81
1.72
0.60
1.19
0.79
0.53
0.22
4.88
0.93
0.83
0.46
1.25
0.81
3.08
1.40
1.62
1.07
0.50
0.84
0.48
0.34
0.17
3.01
0.61
0.74
0.33
0.87
0.04
0.10
0.37
0.68
0.51
1.04
0.95
0.84
0.41
0.53
1.05
0.99
0.97
8
8
114
22
8
177
0.33
0.85
0.94
5.94
2.15
2.42
2.11
1.22
2.70
1.46
1.39
0.96
4.57
1.88
0.95
8
206
0.63
2.00
0.13
0.30
1.48
5.60
4.14
2.16
1.84
1.56
1.55
1.36
7.08
1.74
0.93
7
48
0.82
3.26
0.65
1.73
2.10
0.32
0.98
3.09
6.73
1.26
4.45
4.93
2.42
8.37
0.99
7
22
0.74
2.48
0.87
3.60
2.05
1.38
0.10
0.31
1.40
1.17
0.13
0.20
3.43
2.57
0.93
7
7
0.72
2.33
0.86
3.41
3.15
0.40
0.59
0.24
1.69
0.08
0.14
0.08
3.54
0.77
0.95
7
43
0.70
2.17
0.54
1.28
0.89
1.50
0.51
0.15
2.01
1.58
1.42
0.52
2.39
1.78
0.82
0.65
1.93
0.79
2.61
1.31
2.39
0.84
0.14
6.24
3.53
3.28
0.44
3.70
5.12
0.98
0.62
1.75
0.71
2.01
0.76
3.57
1.38
2.21
0.73
1.19
0.94
1.30
4.33
1.71
0.87
7
7
354
58
7
110
0.31
0.72
0.45
1.01
0.86
1.35
0.69
2.45
2.20
2.04
1.87
3.56
0.49
0.54
0.99
7
143
0.21
0.48
0.90
4.19
1.05
0.28
0.30
0.40
5.44
0.61
1.27
1.51
0.77
1.34
0.98
0.04
0.10
0.21
0.43
1.11
1.49
0.29
2.38
0.35
0.28
0.06
0.55
2.60
0.31
0.83
0.10
0.23
0.09
0.18
1.45
2.96
1.25
0.62
2.56
3.71
3.39
1.23
4.41
3.45
0.93
0.33
0.78
0.12
0.25
0.39
1.05
1.03
0.99
0.83
1.32
2.01
2.15
0.66
0.61
0.96
0.55
1.49
0.73
2.11
1.15
4.61
3.16
0.29
1.93
1.31
1.27
0.30
5.76
1.56
0.86
7
7
7
7
26
136
16
378
7
25
0.73
2.36
0.14
0.28
0.19
0.70
0.18
0.23
0.04
0.23
0.07
0.03
0.52
0.08
0.62
7
15
0.83
3.37
0.60
1.52
0.27
0.93
0.52
2.08
0.62
1.93
1.36
4.48
1.20
1.59
0.98
6
16
0.95
6.25
0.98
8.84
0.93
5.09
0.93
4.44
6
180
Case5 11cv 02509LHK
Document424
Exhibit
Filed0517 13
2
Page60
62
of
2
Intuit
Section
Job
Title
of
1
Section
Total
Years
Emp
Data
Level
Years
T
Coeff
2
Section
Change
Correlation
T
Coeff
Stat
Correlation
Contemp
Stat
6
96
0.71
2.02
0.95
5.47
6
39
0.71
2.01
0.74
1.93
6
91
0.71
2.00
0.49
0.97
6
8
0.69
1.92
0.68
1.62
6
26
0.67
1.81
0.19
0.33
6
26
0.58
1.41
0.28
0.51
6
31
0.57
1.39
0.77
2.08
6
9
0.54
1.27
0.38
0.71
6
8
0.52
1.22
0.78
2.14
6
405
0.46
1.02
0.60
1.30
6
230
0.43
0.96
0.69
1.63
6
14
0.42
0.93
0.36
0.67
6
23
0.41
0.91
0.09
0.15
6
15
0.40
0.88
0.17
0.30
6
8
0.38
0.82
0.03
0.06
6
12
0.38
0.81
0.44
0.85
6
18
0.35
0.75
0.27
0.49
6
78
0.33
0.70
0.38
0.70
6
38
0.33
0.69
0.85
2.82
0.29
0.60
0.09
0.15
0.28
0.58
0.59
1.27
0.23
0.48
0.66
1.51
6
6
6
115
37
102
6
74
0.07
0.14
0.05
0.09
6
24
0.05
0.10
0.48
0.94
0.01
0.01
0.43
0.82
6
338
6
17
0.00
0.01
0.30
0.55
6
6
0.05
0.09
0.13
0.23
6
16
0.09
0.17
0.15
0.26
6
54
0.12
0.25
0.93
4.33
6
98
0.13
0.27
0.81
2.40
0.24
0.50
0.34
0.63
6
179
3
Regression
6
23
0.26
0.54
0.09
0.16
6
19
0.29
0.61
0.07
0.13
6
35
0.36
0.78
0.83
2.61
6
18
0.38
0.83
0.22
0.40
6
15
0.40
0.87
0.53
1.08
6
16
0.46
1.02
0.80
2.29
6
10
0.47
1.06
0.69
1.36
6
38
0.85
3.22
0.92
3.98
Lagged
Revenue
Section
Coefficients
Regression
SJ
Emp
Contemp
Lagged
4
T
Section
Net
Stats
Revenue
SJ
Emp
C
L
5
Section
Effect
T
Stat
r2
6
Case5 11cv 02509LHK
Document424
Exhibit
Filed0517 13
2
Page61
62
of
2
Pixar
Section
Job
of
Title
1
Section
Total
Years
Level
Emp Years
Data
T
Coeff
2
Section
Change
Correlation
Correlation
T
Coeff
Stat
3
Section
Regression
Stat
Contemp
Coefficients
Lagged
Revenue
Regression
SJ
Emp
Contemp
Lagged
4
Section
T Stats
Net
Revenue
SJ
Emp
C
L
5
Section
6
Effect
T
Stat
r2
TECHNICAL DIRECTOR
11
1872
0.94
8.31
0.89
5.65
0.55
0.31
0.03
0.02
3.08
0.63
0.60
0.06
0.86
1.32
0.82
ARTIST SKETCH
11
141
0.91
6.64
0.82
4.06
1.29
1.53
0.12
0.18
7.17
4.44
1.77
0.40
2.82
6.78
0.94
ENGINEER
SOFTWARE
11
503
0.91
6.41
0.93
7.25
0.95
0.70
0.01
0.25
6.38
1.64
0.14
0.62
1.65
3.78
0.91
ANIMATOR
SUPERVISING
11
70
0.82
4.35
0.89
5.41
0.23
2.42
0.22
2.26
0.18
1.94
1.18
1.85
2.65
5.34
0.89
11
772
0.81
4.21
0.78
3.53
0.55
0.48
0.06
0.82
5.27
1.97
1.47
3.57
1.03
3.32
0.92
11
44
0.77
3.57
0.89
5.59
1.79
3.71
0.06
2.65
1.16
2.92
0.44
2.22
1.92
3.94
0.92
11
129
0.75
3.37
0.79
3.68
0.91
1.27
0.15
0.47
3.97
3.23
1.90
0.79
2.18
5.50
0.92
ANIMATOR
ANIMATOR
LAYOUT
DIRECTING
ARTIST
ENGINEER
SR SOFTWARE
11
53
0.74
3.31
0.79
3.59
0.70
1.61
0.00
0.79
1.75
2.89
0.03
1.11
2.32
5.27
0.89
DESIGNER
PRODUCTION
11
62
0.73
3.20
0.86
4.86
0.52
2.50
0.22
3.16
0.22
1.55
0.97
1.44
1.98
2.14
0.83
11
73
0.72
3.10
0.75
3.21
0.53
1.60
0.05
0.10
0.86
2.81
0.33
0.10
2.12
4.47
0.83
ART DIRECTOR
11
70
0.70
2.95
0.76
3.26
1.18
0.70
0.04
1.55
4.33
1.74
0.33
1.81
1.89
3.36
0.83
ENGINEER
11
54
0.58
2.16
0.82
4.06
0.72
1.11
0.24
0.86
1.07
1.77
1.00
0.75
1.83
3.79
0.80
11
91
0.56
2.04
0.81
3.97
1.07
0.56
0.12
0.70
5.49
2.03
1.65
1.48
1.63
4.81
0.90
11
247
0.55
1.98
0.46
1.48
1.27
1.09
0.01
0.41
2.96
2.26
0.07
0.43
2.36
2.98
0.70
11
11
0.51
1.79
0.81
3.89
1.08
0.42
0.01
1.19
4.76
1.69
0.09
1.88
1.50
4.24
0.86
SYSTEMS ADMINISTRATOR
11
133
0.50
1.75
0.29
0.86
0.74
1.15
0.06
0.16
1.93
2.43
0.51
0.20
1.89
2.50
0.62
SCIENTIST
11
62
0.50
1.74
0.39
1.21
1.06
1.26
0.09
0.07
2.05
2.72
0.49
0.06
2.31
2.91
0.68
ANIMATOR
FIX
QUALITY ASSURANCE
SYSTEMS ADMINISTRATOR
SR
ARTIST STORY
MGR DESKTOP
SYSTEMS
SR
TECH DIRECTOR
SUPERVISING
11
70
0.49
1.67
0.72
2.95
1.91
0.66
0.15
3.54
4.54
1.97
0.89
3.08
2.56
4.81
0.87
MGR FINANCIAL
SYSTEMS
11
11
0.43
1.41
0.84
4.41
0.91
0.34
0.00
0.90
5.48
1.95
0.03
2.06
1.24
4.99
0.88
MANAGER
11
11
0.42
1.38
0.83
4.20
0.88
0.24
0.08
0.56
4.82
1.10
1.22
1.12
1.12
3.60
0.86
11
11
0.42
1.38
0.88
5.34
0.84
0.21
0.04
0.53
5.76
1.20
0.67
1.39
1.05
4.31
0.88
11
42
0.42
1.37
0.63
2.29
1.15
0.84
0.08
1.67
3.63
2.51
0.76
1.85
1.98
3.68
0.79
11
24
0.38
1.22
0.86
4.72
0.60
0.02
0.09
0.13
4.06
0.10
1.73
0.36
0.62
2.11
0.84
SVCS
11
11
0.34
1.09
0.84
4.35
0.95
0.24
0.06
0.73
4.89
1.21
0.87
1.37
1.19
4.01
0.86
PRODUCTS
11
11
0.21
0.63
0.79
3.66
1.01
0.25
0.03
1.20
4.52
1.44
0.42
2.01
1.25
4.30
0.85
11
44
0.19
0.59
0.26
0.75
0.57
0.92
0.18
1.39
2.12
3.91
1.80
1.63
1.49
3.88
0.85
11
22
0.17
0.52
0.41
1.29
0.84
0.35
0.07
1.70
4.85
2.20
1.10
4.11
1.19
4.57
0.92
11
35
0.12
0.36
0.12
0.35
0.77
0.92
0.01
1.08
1.17
1.57
0.04
0.60
1.69
1.58
0.39
10
35
0.50
1.62
0.71
2.65
1.47
0.68
0.03
4.53
2.67
2.62
0.15
2.08
2.15
3.58
0.85
MGR SYSTEMS
10
10
0.41
1.28
0.74
2.66
1.03
0.40
0.20
2.10
3.42
1.19
0.93
1.93
1.44
2.70
0.81
ENGINEER
10
15
0.28
0.83
0.68
2.45
1.10
0.49
0.02
0.34
2.08
1.33
0.06
0.13
1.59
2.68
0.67
VP SOFTWARE ENGINEERING
10
12
0.26
0.76
0.56
1.79
3.29
0.66
0.72
9.33
2.20
1.19
1.18
2.35
3.95
2.37
0.89
USER INTERFACE DESIGNER
10
20
0.14
0.40
0.66
2.35
0.65
0.35
0.02
0.43
1.94
1.17
0.19
0.35
0.99
2.17
0.61
9
9
0.34
0.95
0.78
3.01
1.66
0.14
0.12
2.32
3.77
0.40
0.55
1.91
1.80
3.41
0.88
9
15
0.17
0.45
0.43
1.07
1.85
1.06
0.30
1.74
5.23
12.17
6.25
4.55
2.92
7.02
0.99
8
25
0.58
1.73
0.73
2.36
0.34
1.69
0.31
2.68
0.22
2.03
0.66
1.00
1.35
1.15
0.85
8
13
0.35
0.92
0.63
1.60
0.56
0.96
0.85
6.04
17.44
10.07
16.87
1.52
8
8
0.34
0.89
0.81
3.05
1.03
0.02
0.22
2.32
6.00
0.08
2.32
3.35
1.05
3.89
0.97
8
20
0.27
0.70
0.03
0.06
0.05
0.57
0.11
1.05
0.10
2.80
0.51
0.29
0.52
0.90
0.86
ENGINEERING
ENGINEER
ARTIST
ASSOCIATE
GRAPHIC
TECH DEPT
TECH DIRECTOR LEAD CRTV
ADMINISTRATOR
DEVELOPER
RENDERMAN
TECH DIRECTOR
CRTV
SVCS
SCULPTOR
ENGINEER
PROJECT
PRODUCTION
MGR STUDIO
TOOLS
OPERATIONS
RENDERMAN
SUPPORT
DIR RENDERMAN
DESIGNER
SUPPORT
PRODUCT
DEV
ENVIRONMENTAL
ARTIST AFTER EFFECTS
TECHNICAL WRITER
TECHNICAL LEAD RENDERING
ARTIST STORY DEVELOPMENT
11.18
20.27
1.00
Case5 11cv 02509LHK
Document424
Exhibit
Filed0517 13
2
Page62
62
of
2
Pixar
Section
Job
of
Title
ENGINEER
SOFTWARE
ENGINEER
IMAGE
TECHSUPPORT
1
Section
Total
Years
Level
Emp Years
Data
T
Coeff
2
Section
Change
Correlation
Correlation
T
Coeff
Stat
Stat
7
7
0.86
3.77
0.01
0.03
6
8
0.92
4.74
0.54
1.13
6
6
0.92
4.65
0.75
1.97
6
6
0.88
3.76
0.79
2.24
6
6
0.88
3.69
0.78
2.18
6
65
0.74
2.20
0.53
1.07
6
6
0.60
1.50
0.76
2.00
6
6
0.52
1.22
0.57
0.98
FINANCIAL APPS DEVELOPER
6
6
0.46
1.03
0.80
2.31
MGR SR PROJECT
6
6
0.46
1.03
0.21
0.31
6
6
0.42
0.93
0.27
0.49
6
8
0.12
0.24
0.35
0.66
TECHNICAL
ENGINEER
LEAD TELECOM
ROOM
SCREENING
MGR IMAGE
CGI
MASTERING
MASTERING
PAINTER
DESIGNER
CAMERA
ENGINEER
APPLICATIONS
STUDIO
TOOLS
LAYOUT ARTIST LEAD
MEDIA SYSTEMS COORDINATOR
3
Section
Regression
Contemp
0.51
Coefficients
Lagged
Revenue
0.02
0.01
Regression
SJ
Emp
2.20
Contemp
0.63
Lagged
0.07
4
Section
T Stats
Net
Revenue
0.03
SJ
Emp
1.07
C
L
0.49
5
Section
6
Effect
T
Stat
0.55
r2
0.58
Disclaimer: Justia Dockets & Filings provides public litigation records from the federal appellate and district courts. These filings and docket sheets should not be considered findings of fact or liability, nor do they necessarily reflect the view of Justia.
Why Is My Information Online?