Declaration of Justin McCrary - Expert Report

1
2
3
4
5
6
7
8
9
10
11
12
13
GIBSON, DUNN & CRUTCHER LLP

THEODORE J. BOUTROUS, JR., SBN 132099
[email protected]
DEBRA WONG YANG, SBN 123289
[email protected]
MARCELLUS A. MCRAE, SBN 140308
[email protected]
THEANE D. EVANGELIS, SBN 243570
[email protected]
DHANANJAY S. MANTHRIPRAGADA, SBN 254433
[email protected]
BRANDON J. STOKER, SBN 277325
[email protected]
333 South Grand Avenue
Los Angeles, CA 90071-3197
Telephone:
213.229.7000
Facsimile:
213.229.7520
JOSHUA S. LIPSHUTZ, SBN 242557
[email protected]
KEVIN J. RING-DOWELL, SBN 278289
[email protected]
555 Mission Street, Suite 3000
San Francisco, CA 94105-0921
Telephone:
415.393.8200
Facsimile:
415.393.8306
14
15
Attorneys for Defendant

UBER TECHNOLOGIES, INC.
16
UNITED STATES DISTRICT COURT
17
NORTHERN DISTRICT OF CALIFORNIA
18
19
20
DOUGLAS OCONNOR, THOMAS

COLOPY, MATTHEW MANAHAN, and
ELIE GURFINKEL, individually and on
behalf of all others similarly situated,
21
Plaintiffs,
22
23
24
25
Case No. CV 13-03826-EMC

DECLARATION OF DR. JUSTIN
MCCRARY IN SUPPORT OF UBERS
OPPOSITION TO PLAINTIFFS MOTION
FOR CLASS CERTIFICATION
v.
UBER TECHNOLOGIES, INC.,
Defendant.
26
27
28
Gibson, Dunn &
Crutcher LLP
DECLARATION OF DR. JUSTIN MCCRARY
CV 13-03826-EMC
1
2
TABLE OF CONTENTS
Page
I.
INTRODUCTION ................................................................................................ 4
A. Qualifications ..................................................................................................... 4
B.
Assignment......................................................................................................... 6
II.
SCOPE OF OPINIONS ........................................................................................ 7
III.
TECHNOLOGICAL SOLUTIONS HAVE REDUCED MARKET

FAILURES ......................................................................................................... 10
A. Internet Platforms Help Solve Traditional Coordination Problems ................ 11
9
10
B.
Internet Platforms Help Solve Traditional Information Problems .................. 14
11
C.
New Methods of Connecting People to Work Opportunities .......................... 16
12
IV.
13
LEAD GENERATION PLATFORMS LIKE UBER HAVE CHANGED

COMPETITION AND CHOICES IN DRIVING SERVICES .......................... 18
14
A. Coordination Issues in Transportation Prior to New Technologies ................ 19
15
B.
The Transportation Industry Prior to Lead Generation Platforms ................... 19
C.
New Options and Sources of Competition Emerge with the

Development of Lead Generation Platforms and TNCs .................................. 28
16
17
18
V.
THERE IS NO TYPICAL DRIVER WHO USES THE UBER APP................ 33
19
20
21
A. Some Drivers Who Use the Uber App Use Competing Platforms at the
Same Time They Use the Uber App ................................................................ 34
B.
Californians Working as Drivers Before the Existence of Uber and Lead

Generation Platforms ....................................................................................... 36
C.
24
Californians Who Were Not Working as Drivers Before the Existence of

Uber and Other Lead Generation Platforms .................................................... 47
25
D. Small Businesses That Did Not Exist Before Uber ......................................... 56
26
E.
22
23
27
28
VI.
Drivers Using the Uber App Have Different Backgrounds, Reasons for
Using the App and Outside Options ................................................................ 56
THE RELATIONSHIP BETWEEN EACH DRIVER AND THE UBER
Gibson, Dunn &

Crutcher LLP
CV 13-03826-EMC
1
2
3
4
5
APP IS UNIQUE AND INDEPENDENT ......................................................... 59

A. Drivers Who Use the Uber App Are Not Required to Drive Any
Particular Time, Amount or Place Creating Highly Individualized Work
Circumstances for Each Driver ........................................................................ 59
B.
Supervision of Drivers on Jobs is Minimal and Drivers Have Other Fulland Part-time Commitments ............................................................................ 77
6
7
8
9
VII. PLAINTIFFS TIPPING AND EXPENSES THEORY IMPLIES NO

HARM FOR THE PROPOSED CLASS ON AVERAGE AS WELL AS
CLASS CONFLICT ........................................................................................... 84
10
A. It Is Not Possible to Calculate Potential Damages from Ubers Tipping

Practices on a Class-wide Basis ....................................................................... 85
11
B.
If Uber Were to Implicitly or Explicitly Pay Drivers for Tips There Is

Meaningful Potential for Class Conflict .......................................................... 88
C.
It Is Not Possible to Calculate Potential Damages Incurred from

Expenses on a Class-wide Basis ...................................................................... 91
12
13
14
15
D. If Uber Paid for Driver Expenses There Is Meaningful Potential for

Class Conflict ................................................................................................... 93
16
17
18
19
20
21
VIII. MANY DRIVERS WOULD BE HARMED IF ACCEPTING A

PASSENGER REFERRAL FROM UBER MADE THE DRIVER AN
UBER EMPLOYEE ........................................................................................... 94
A. Many Current Drivers Would Be Harmed If Use of the Uber App
Rendered Every Driver an Uber Employee ..................................................... 95
B.
22
23
IX.
Many Current Drivers Would Be Harmed If Uber Only Allowed

Companies, and Not Individuals, to Use the Platform .................................... 99
CONCLUSION ................................................................................................ 100
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
3
CV 13-03826-EMC
I.
INTRODUCTION
2
3
A.
Qualifications
1.
I am an economist with expertise in economic modeling, statistical methods, law and
economics, labor economics, antitrust, and antitrust damages. I received my A.B. in Public Policy
from Princeton University in 1996. After working at National Economics Research Associates in
White Plains, New York, and the Federal Reserve Bank of New York from 1996-1998, I began my
Ph.D. in Economics at the University of California, Berkeley (Berkeley), completing the degree in
June 2003 with field specializations in labor economics and econometrics. After close to five years
10
as Assistant Professor at the University of Michigan, I became an Assistant Professor of Law at
11
Berkeley in January 2008 and was promoted to Professor in July 2010. While at Berkeley, I have
12
taught courses on introductory, intermediate, and advanced statistics to J.D. students, L.L.M.
13
students, and Ph.D. students; on law and economics to J.D. students as well as undergraduates; on
14
business law to J.D., L.L.M., and M.B.A. students; and on labor economics to Ph.D. students.
15
2.
In addition to my post as Professor, I am the Founding Director of D-Lab, the Social
16
Sciences Data Laboratory at Berkeley. At D-Lab I lecture on and advise graduate students and
17
faculty regarding high-performance computing, statistical software, and appropriate statistical
18
techniques.
19
3.
From September 2009 until July 2014, when I began to direct the D-Lab, I co-directed
20
the Law and Economics Program at Berkeley Law with Bob Cooter and Dan Rubinfeld (20082011)
21
and with Bob Cooter and Eric Talley (20122014).
22
4.
Since 2008 I have codirected the Economics of Crime Working Group of the National
23
Bureau of Economic Research (NBER). The NBER is the preeminent professional association of
24
economists in the world, with approximately 1,300 members worldwide. I was invited to become a
25
Faculty Research Fellow of the NBER in 2006 and remained in that position until 2012, when I was
26
invited to become a Faculty Research Associate.
27
28
5.
As noted, I previously worked at the University of Michigan. From 2003 through
2007, I was Assistant Professor of Public Policy and Assistant Professor of Economics. While at
Gibson, Dunn &

Crutcher LLP
4
CV 13-03826-EMC
Michigan, I taught introductory statistics and advanced microeconomic theory to M.P.P. students, and
advanced econometric theory to Ph.D. students.
6.
My research spans a diverse range of topics, including antitrust, crime, employment
discrimination, econometric and statistical methodology, education, fertility, financial markets,
income inequality, and monetary policy. Many of my articles have been published in leading field
journals within economics, such as the Review of Economics and Statistics and the Journal of
Econometrics. In addition, I have written or co-written three papers that were published in the top
economics journal in the world, the American Economic Review, and have co-edited a book,
Controlling Crime: Strategies and Tradeoffs. Over the years, my research has been supported by the
10
University of Michigan, the University of California, Berkeley, the MacArthur Foundation, the
11
NBER, the National Institutes of Health, the National Science Foundation, and the Robert Wood
12
Johnson Foundation.
13
7.
I am frequently asked to review articles for the leading journals within economics,
14
including Econometrica, the American Economic Review, the Quarterly Journal of Economics, the
15
Journal of Political Economy, the Review of Economic Studies, the Review of Economics and
16
Statistics, and the American Law and Economics Review. Since coming to Berkeley Law, I have also
17
been asked to comment on empirical papers submitted to law reviews and to peer-reviewed law
18
journals, including the California Law Review, the Law and Society Review, the Journal of Law and
19
Economics, and the Journal of Empirical Legal Studies.
20
8.
My consulting experience has spanned a wide range of industries and markets. For
21
example, I have previously analyzed: the extent to which alleged collusive behavior among health
22
care providers affected prices; the extent of infringing sales in a patent lawsuit pertaining to
23
pharmaceuticals; the potential anti-competitive implications of a proposed telecommunications
24
merger; damages associated with an alleged price-fixing conspiracy in the corrugated packaging
25
industry; damages associated with an alleged price-fixing conspiracy in several prominent high-
26
technology product markets; and damages associated with an alleged price-fixing conspiracy in the
27
sale of retail gasoline.
28
Gibson, Dunn &
Crutcher LLP
5
CV 13-03826-EMC
9.
Finally, I am frequently asked to give talks regarding the utilization of statistical
methodologies in empirical legal studies and for the past three summers have given day-long lectures
for the week-long Causal Inference Workshop, as well as its more advanced version, the Advanced
Causal Inference Workshop, both organized by Bernie Black (a Professor of Law and Business at
Northwestern University) and Matthew McCubbins (a Professor of Law and Political Science at
Duke University).
10.
A copy of my curriculum vitae, including a list of previous testimony and depositions,
is included as Appendix A.
11.
I am being compensated at my standard billing rate of $750 per hour. I have been
10
assisted in this matter by staff of Cornerstone Research, who worked under my direction. I receive
11
compensation from Cornerstone Research based on its collected staff billings for its support of me in
12
this matter. Neither my compensation in this matter nor my compensation from Cornerstone
13
Research is in any way contingent or based on the content of my opinion or the outcome of this or
14
any other matter.
15
16
B.
Assignment
17
12.
Uber Technologies, Inc. (Uber) is a defendant in a federal putative class action
18
lawsuit in which drivers using the Uber lead generation software application (Uber App) allege that
19
they are employees and not independent contractors. Plaintiffs also claim that Uber has violated the
20
California Labor Code by neither reimbursing expenses nor remitting tips to drivers that passengers
21
allegedly believe they are paying. I understand that the putative class is limited to individuals who
22
have worked as [UberBLACK], [UberSUV], or [uberX] drivers in California.
23
24
25
Second Amended Class Action Complaint and Jury Demand, Douglas OConnor et al. v. Uber
Technologies, Inc., 11/12/14 (SAC) 23. While I understand this to be the core of Plaintiffs
claims, I dispute that there is a clear meaning to the terms UberBLACK driver, UberSUV
driver, and uberX driver, as I clarify below.
SAC 9.
26
27
28
Gibson, Dunn &
Crutcher LLP
6
CV 13-03826-EMC
13.
I have been retained by counsel for Uber. I have been asked to serve as an
independent expert witness in connection with Ubers opposition to Plaintiffs motion for class
certification. I will provide an economic analysis of the experiences of drivers using the Uber
platforms, with a particular focus on the variety of experiences by drivers in California as this relates
to Ubers purported uniform control over drivers use of the Uber App and the other factors set forth
in the Borello case. I understand that I may be called as an expert to testify in these legal
proceedings.
14.
In developing my opinions, I have drawn on the record in this matter, as well as data
from the Defendants. A complete list of these data and materials is included in Appendix B to my
10
report. My analysis on this matter is ongoing and I reserve the right to amend my opinions as new
11
information becomes available.
12
13
II.
SCOPE OF OPINIONS
14
15
15.
Uber and other lead generation platforms provide a new, technological solution to the
16
existing regulatory approaches designed to connect providers of driving services (i.e., drivers) to
17
customers (i.e., passengers). Regulatory approaches were traditionally necessary because the market
18
for driving services confronts an important information and coordination problem: it is hard to
19
coordinate the labor supply of drivers to match passenger demand. The technological solution
20
provided by lead generation platforms offers a more efficient method of solving these problems than
21
any previous method and confers benefits upon drivers and riders alike.
22
23
24
25
26
27
28
S.G. Borello & Sons, Inc. v. Department of Industrial Relations, 48 Cal. 3d 341, 350351 (1989).
Generally, lead generation platforms are software applications that help to match prospective buyers
with sellers. In the context of driving services, lead generation platforms are software applications
that match individuals seeking rides to drivers interested in driving passengers. Because of the
short-term, mobile nature of the demand for driving services, potential riders access the described
software applications nearly exclusively using cell phones; similarly, since drivers are by
definition mobile, they also access the software applications nearly exclusively using cell phones.
Gibson, Dunn &

Crutcher LLP
7
CV 13-03826-EMC
16.
Certain types of lead generation platforms connect passengers to drivers who are using
their own personal vehicles. In California, these platforms are regulated by the California Public
Utilities Commission (PUC) as Transportation Network Companies (TNCs). TNCs do not, and
under PUC regulations are not allowed to, own cars. Rather, they provide referrals to individuals
who choose whether or not to drive those referrals. Because some drivers using the uberX platform
drive their personal vehicles (while others drive liveried cars with special licenses as discussed
below), Rasier-CA LLC, a wholly owned subsidiary of Uber is regulated as a TNC.
8
9
17.
Lead generation platforms provide a new option for existing and potential drivers
seeking to find passengers. These platform services have transformed the marketplace of driving
10
services similar to how internet platforms have transformed other marketplaces (e.g., eBay and Etsy).
11
These platform services give drivers and consumers choices they did not previously have. Each
12
existing and potential driver will experience, use, and value these choices differently.
13
18.
In my opinion, there is no typical driver who uses the Uber App. Each driver has a
14
number of individual choices that will determine his/her work circumstances. Each driver controls
15
what months to drive, what days to drive, what times of day to drive, which city to drive in, which
16
neighborhood of a given city to drive in, the route for driving a passenger from one place to another,
17
how fast to drive, and so on. Drivers choose whether, when, and how often to turn on the Uber App
18
and have the option, but not the obligation, of accepting referrals passed along to them by the Uber
19
App.
20
19.
Each drivers relationship with Uber is unique. This is not unexpected; indeed, I am
21
not aware of any evidence that Uber exercises control uniformly, reserves the right of control
22
uniformly, or has any uniform policy describing control over the physical conduct of drivers work.
23
As a consequence, each driver makes their own choices about how to go about the work of driving.
24
Some drivers work for transportation companies that contract with Uber, while others contemplate
25
accepting ride referrals using the Uber App directly. Some drivers drive liveried cars with special
26
licenses, while others drive their personal vehicles (as discussed below).
27
28
20.
Moreover, drivers have diverse backgrounds and diverse alternative work options.
Some drivers moonlight as drivers, accepting Uber App ride referrals when convenient for them,
Gibson, Dunn &

Crutcher LLP
8
CV 13-03826-EMC
and yet work a full-time job as the employee of an employer; indeed, some of these drivers have
employment contracts that disallow them from being the employee of any other employer. Other
drivers accept referrals from the Uber App when convenient for them and work one or more freelance
jobs, such as a property manager, realtor, or graphic designer. Yet other drivers have no employer and
only rarely accept referrals from the Uber App, perhaps because they care for a loved one who is sick;
they have small children; or they have a weak attachment to the labor force for other reasons.
Finally, some individuals may turn on the Uber App at the same time as they turn on the application
of a competitor for Uber, such as Lyft, SideCar, or InstantCab, for example. Some of Ubers
licensing agreements allow this; some do not.
10
21.
Just as work options vary greatly, so do tipping practices and driver expenses. Drivers
11
who have used the Uber App for at least one passenger referral would likely not be uniformly better
12
off if Uber changed its service to add tips or pay driver expenses. I am not aware that Plaintiffs have
13
articulated any notion of a but-for world that might form a basis for claims of harm. However, under
14
some notions of the but-for world, no member of the putative class would experience harm. Under
15
other notions of the but-for world, some members of the putative class benefit from the status quo
16
and others are harmed by it. Determining which member is of which type and what magnitude of
17
harm they may have experienced would involve individual inquiry and investigation.
18
22.
Underscoring these concerns, I conclude that drivers use the platform in such a large
19
variety of ways, under such a large variety of working relationships, that many current drivers who
20
have used the Uber App for ride referrals would be harmed if it were commonly found that the use of
21
the Uber App turned every driver into an employee of Uber. Uber would, under any such
22
determination, be expected to contemplate many different reorganizations of its business model, and
23
many of these would be associated with potentially substantial harms to drivers.
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
9
CV 13-03826-EMC
III.
TECHNOLOGICAL SOLUTIONS HAVE REDUCED MARKET FAILURES
2
3
23.
Innovations in market coordination over the past decade have dramatically altered
U.S. markets. Many marketplaces that were not efficient due to coordination problems have utilized
these technologies, with benefits to buyers and sellers alike.
24.
Perfect markets can achieve economic efficiency in the sense that no one can be
5
made better off without harming someone else. Economists think of perfect markets as an ideal
concept rather than a fact that we encounter in the real world; yet the ideal can help to clarify our
thinking. Many markets do not achieve economic efficiency due to market imperfections. In
10
markets where consumers struggle to connect to service providers, as in driving services, imperfect
11
information can significantly impede efficiency.
12
13
25.
Increased coordination between buyers and sellers, perhaps through intermediaries or

8
matchmakers, can help solve imperfect information problems. The dynamic of an intermediary
14
15
16
Joseph Stiglitz, Economics (New York: W. W. Norton & Company, 1993), pp. 384385. In
perfect markets actors are fully and equally informed about all facts relevant to all potential
transactions, there are no transaction costs, and firms compete on price until profits are dissipated.
The notion of efficiency I reference here is due to the 19th century Italian economist and
sociologist, Vilfredo Pareto.
Frank, Robert and Ben Bernanke, Principles of Economics (New York: McGraw-Hill/Irwin,
2001), p. 219. Three common market imperfections are market power (e.g., monopoly),
externalities (e.g., network effects), and imperfect information.
17
18
19
20
21
22
23
24
25
26
27
28
Imperfect information is often described in terms of incomplete information (when a market

participant is not fully informed and would need additional information to achieve an efficient
transaction; see Joseph Stiglitz, Economics (New York: W. W. Norton & Company, 1993), p.
493), and in terms of asymmetric information (when one party has more information than the
other; see Robert Frank and Ben Bernanke, Principles of Economics (New York: McGrawHill/Irwin, 2001), p. 305, Joseph Stiglitz, Economics (New York: W. W. Norton & Company,
1993), p. 493, or Pindyck, Robert and Daniel Rubinfeld, Microeconomics, Third Edition
(Englewood Cliffs: Prentice-Hall, 1994), p. 596).
Intermediaries can sometimes provide information and value by ensuring that goods and services
find their way to the consumer who values them most. Einav, Liran, The Economics of Peer-toPeer Internet Markets, Presentation prepared for the Federal Trade Commission Workshop on
the Sharing Economy, 6/9/15, p. 7. Frank, Robert and Ben Bernanke, Principles of Economics,
(New York: McGraw-Hill/Irwin, 2001), pp. 298, 300.
Gibson, Dunn &

Crutcher LLP
10
CV 13-03826-EMC
coordinating buyers and sellers is under some conditions described as a two-sided market. Two-
sided markets allow end-users to reach an efficient agreement that would otherwise be unlikely; value
is created by coordination, which means that as a general economic proposition, platforms facilitating
end-user transactions are best understood as products in their own right.
10
26.
In this section, I discuss recent technological innovations that have improved
coordination in a number of marketplaces. Finally, I detail how these platforms help connect
potential workers to work opportunities.
8
9
A.
Internet Platforms Help Solve Traditional Coordination Problems
10
27.
11
In markets with coordination problems, better information often unlocks substantial
12
value, leading to meaningful benefits for consumers, sellers, workers, and firms. Many recent
13
technological innovations have done just this. Many markets with coordination problems are
14
characterized by many small, independent market participants who are quite differently situated from
15
one another, have little in common, and place a premium on flexibility.
16
17
1.
Garage Sales/Flea Markets and Online Marketplaces for Used Goods
18
28.
19
20
21
22
Advertising for garage sales is typically local (e.g., word-of-mouth or flyers). Local advertising helps
used goods sellers coordinate with buyers.
Markets can be two-sided in that 1) two sets of agents interact through an intermediary or
platform, and 2) the decision of each set of agents affects the outcomes of the other set of agents.
Rysman, Marc, The Economics of Two-Sided Markets, The Journal of Economic Perspectives
23(3), 2009, pp. 125143 at p. 125. This in turn means that the structure, and not only the level
of prices charged by platforms matters. Rochet, Jean-Charles and Jean Tirole, TwoSided
Markets: A Progress Report, The RAND Journal of Economics 37(3), pp. 645667 at p. 645.
10
Rochet, Jean-Charles., and Jean Tirole, TwoSided Markets: A Progress Report, The RAND
Journal of Economics 37(3), pp. 645667 at p. 649. Therefore, by facilitating voluntary
transactions to take place, two-sided market platforms provide benefits similar to those known as
gains from trade. Joseph Stiglitz, Economics, (New York: W. W. Norton & Company, 1993),
p. 55.
23
24
25
26
27
28
Garage sales and flea markets are traditional venues for the informal sale of goods.
Gibson, Dunn &

Crutcher LLP
11
CV 13-03826-EMC
29.
A much more effective coordination method has emerged in recent years on the
Internet. A well-known example is the website run by eBay, which boasts 157 million active buyers
worldwide and currently has 800 million items listed on its platform.
register on eBay to purchase used goods from each other; buyers can browse for thousands of
products through a search engine. The existence of eBay radically expands the options for both
sellers and buyers; both groups benefit from the coordination provided by eBay.
11
Sellers and buyers can
12
13
7
8
30. Lead generation platforms such as Uber similarly coordinate transactions between drivers and
passengers.
9
10
2.
11
31.
Craft or Art Fairs/Antique Shows and Online Marketplaces for Crafts
Craft and art fairs are traditional venues for the sale of works from independent artists
12
and artisans. A fair is more enticing for a potential buyer if many sellers are in attendance and it is
13
easy to search among their wares; paradoxically, however, the more sellers, the harder it becomes to
14
search among their wares. Moreover, coordinating buyers and sellers to attend fairs is challenging.
15
Fairs happen over a day or a week and in a single location, require advertising, and require both
16
sellers and buyers to be physically present in order to coordinate transactions.
17
32.
These challenges have largely been overcome with the recent emergence of online
18
marketplaces that seek to match sellers of craft goods with potential buyers. The most famous of
19
these is Etsy, which focuses on coordinating the sale of handmade items, vintage goods and craft
20
supplies.
14
It currently has 1.4 million active sellers and 20.8 million active buyers and its presence
21
22
11
Other examples include Craigslist and other local platforms, such as the French website
leboncoin.fr. Craigslist, Factsheet, http://www.craigslist.org/about/factsheet, accessed on
7/3/15. eBay Inc., One Company, http://www.ebayinc.com/who_we_are/one_company,
accessed on 7/1/15.
12
Dinerstein, Michael et al., Consumer Price Search and Platform Design in Internet Commerce,
Working Paper, 8/14 at Figure 2.
13
Initially, eBay used only auctions as the pricing and buying mechanism but subsequently it has
introduced other forms of transaction. Einav, Liran et al., Sales Mechanisms in Online Markets:
What Happened to Internet Auctions?, Working Paper, 2013, pp. 45.
14
Etsy, About Etsy, https://www.etsy.com/about/?ref=ftr, accessed on 7/1/15.
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
12
CV 13-03826-EMC
15
has increased the options for both, in large part due to scale, but also because of its convenience.
Etsy is a good example of an online marketplace mitigating coordination problems due to scale, but it
is hardly the only such example.
33.
16
Because any barriers to entry in these online marketplaces are readily overcome, Etsy
17
has competitors, which further benefits buyers and sellers.
vintage items, eBay and Etsy compete. The existence of platforms like eBay and Etsy also allows
small sellers (e.g., individual craftsmen, antique collectors, etc.) to compete effectively with larger
companies (e.g., jewelry stores, furniture stores, etc.), which in the absence of these platforms would
have been difficult. Platforms like Etsy thus are often said to have a democratizing influence on
Furthermore, for sales of antiques and
18
10
commerce.
11
34.
Lead generation platforms such as Uber similarly have a democratizing influence on
12
the market for driving services by lowering the barriers to entry and facilitating entrepreneurism,
13
seeking to enable carpooling and shift cost structures so that fewer individuals need to buy and
14
possess their own car.
15
16
17
18
15
Etsy, About Etsy, https://www.etsy.com/about/?ref=ftr, accessed on 7/1/15.
16
Another prominent example of a successful online marketplace solving a long-standing

coordination problem is in dating. Online dating services such as Match.com and eHarmony have
changed how people find mates. See match.com, About Match.com,
http://www.match.com/help/aboutus.aspx?lid=4, accessed on 7/1/15; eHarmony, Who We Are,
http://www.eharmony.com/about/eharmony/, accessed 7/1/2015; or OKCupid, About,
https://www.okcupid.com/about, accessed on 7/1/15. It is estimated now that 35% of couples
married between 2005 and 2012 met online, and almost half of those couples met through an
online dating service. Of the 34.95% of people who met online, 45.01% met through an online
dating site. Cacioppo, et al., Marital Satisfaction and Break-ups Differ across On-line and Offline Meeting Venues, Proceedings of the National Academy of Sciences 110(25) 2013, pp.
1013510140 at p. 10136.
17
Competitors of Etsy include Zibbet and the German website DaWanda. Zibbet, About Us,
https://www.zibbet.com/about-us, accessed on 7/1/15; DaWanda Secures 4 million Euros to Go
International, TechCrunch, 11/30/11, http://techcrunch.com/2011/11/30/dawanda-secures-e4million-to-go-international/, accessed on 7/1/15.
18
For example, see Michael Wolf, The New Era of Democratized Business, Forbes, 11/5/12,
http://www.forbes.com/sites/michaelwolf/2012/11/05/the-new-era-of-democratized-business/,
accessed on 7/3/15.
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
13
CV 13-03826-EMC
B.
Internet Platforms Help Solve Traditional Information Problems
35.
A particular variety of an information problem confronting markets is the problem of

19
asymmetric information, or a problem whereby one party has better information than the other.
The one may seek to take advantage of the other, leveraging such information.
other is aware of this informational disadvantage, she may opt out of the market altogether in order to
avoid being taken advantage of. Markets under such circumstances have the tendency of being thin
and in the extreme can even collapse.
20
21
8
9
10
Moreover, when the
36.
22
A reputation presents a partial solution to asymmetric information problems.

23
reputation is a type of guarantee that can help sellers earn more by establishing trust.
Some online
platforms provide mechanisms for creating and sustaining reputations through rankings or reviews.
11
12
1.
13
14
37.
Chain Restaurants and Yelp
Reputations are particularly important in the restaurant market. Many restaurants do

24
not have enough repeat customers to develop a reputation.
Branded chain restaurants seek to build
15
16
19
The canonical example of asymmetric information is the market for lemons, where it is argued
that sellers of used cars know more about the products than the buyers and therefore have an
advantage in the transaction (see Akerlof, George, The Market for Lemons: Quality
Uncertainty and the Market Mechanism, The Quarterly Journal of Economics 84(3), 1970 pp.
488500 at pp. 488489; Frank, Robert and Ben Bernanke, Principles of Economics, (New York:
McGraw-Hill/Irwin, 2001), pp. 306307.)
20
Online transactions are prone to asymmetric information problems because the products cannot be
physically inspected prior to purchase and delivery. Frank, Robert and Ben Bernanke, Principles
of Economics, (New York: McGraw-Hill/Irwin, 2001), p. 312. Jin, Ginger Zhe, and Andrew
Kato, Price, Quality, and Reputation: Evidence from an Online Field Experiment, The RAND
Journal of Economics 37(4) 2006, pp. 9831005; Einav, Liran, et al., The Economics of Peer-toPeer Internet Markets, Presentation prepared for the Seventh Annual Federal Trade Commission
Microeconomics Conference, 10/16/14, p. 10.
21
For example, the market for shares in privately held companies may be thin due to asymmetric
information.
22
Joseph Stiglitz, Economics (New York: W. W. Norton & Company, 1993), p. 504.
23
Joseph Stiglitz, Economics,(New York: W. W. Norton & Company, 1993), p. 504.
24
Pindyck, Robert and Daniel Rubinfeld, Microeconomics, Third Edition, (Englewood Cliffs:
Prentice-Hall, 1994), p. 598.
17
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
14
CV 13-03826-EMC
a reputation across locations through standardization, and the brand may serve as a guarantee of
consistency. Independent restaurants have a much harder time establishing reputations. Examples
of traditional solutions to the reputation problem for independent restaurants include newspaper food
critics or restaurant guides such as Zagats or the Michelin guide.
25
38.
Internet services dramatically increased the ability of independent restaurants to
develop reputations. Yelp is the most famous example of online consumer review aggregation, with
over 77 million reviews of restaurants throughout the United States. Reviews can reduce the
informational advantage currently held by restaurant chains: a small independent restaurant can have
reviews available that provide consumers with information on the quality of the food in restaurants
26
10
separate from simple brand recognition. In fact, research has shown that higher Yelp penetration
11
leads to an increase in revenue for independent restaurants, but a decrease in revenue for chain
12
restaurants.
13
reputations and overcome the asymmetric information problem.
27
14
15
39.
This evidence is consistent with the ability of online reviews and rankings to establish
Lead generation platforms such as Uber and Lyft utilize rankings in an analogous
fashion to overcome problems in driving services with asymmetric information.
16
17
2.
18
19
40.
Other Examples of Online Reputation Mechanisms
Several of the online platforms that I discussed above have developed mechanisms for
sellers and buyers to establish and maintain reputations. Both eBay and Etsy encourage buyers to
20
21
22
23
24
25
Pindyck, Robert and Daniel Rubinfeld, Microeconomics, Third Edition, (Englewood Cliffs:
Prentice-Hall, 1994), p. 598. Michael Luca, Reviews, Reputation, and Revenue: The Case of
Yelp.com, Harvard Business School Working Paper, No. 12016, 2011, pp. 4, 18; Joseph
Stiglitz, Economics, (New York: W. W. Norton & Company, 1993), p. 504.
26
Yelp, About Us, http://www.yelp.com/about, accessed on 7/1/15.
27
Michael Luca, Reviews, Reputation, and Revenue: The Case of Yelp.com, Harvard Business
School Working Paper, No. 12016, 2011, p. 19.
25
26
27
28
Gibson, Dunn &
Crutcher LLP
15
CV 13-03826-EMC
28
provide feedback and ratings of sellers.
transaction and display seller ratings prominently on the platform.
41.
These sites aggregate the reviews collected from every
These ratings are viewed credibly by buyers and have concrete consequences for trade.
For example, on eBay, the growth rate of a sellers transactions drops after receiving negative
feedback.
platform and that sellers with negative feedback are more likely to exit the platform.
findings show that reputation is taken seriously by both sellers and buyers and influences
transactions. However, reputation alone may not be sufficient to maintain compliance with the
platforms policies. If a seller is found to not deliver a good enough service or to commit fraud, she
10
29
Furthermore, there is evidence that sellers invest in building a strong reputation on the
30
can be prevented from using the platform.
These
31
11
12
C.
New Methods of Connecting People to Work Opportunities
13
14
1.
15
Technological Innovations Help Solve the Coordination Problem by

Disaggregating Assets
16
42.
Another way in which coordination problems can arise is in the exact definition of the
17
product. Some products are not tradable because they are not easily divisible. Recent technological
18
innovations enable the disaggregation of physical assets in space and in time, and therefore allow
19
20
21
28
eBay, All About Feedback, http://pages.ebay.com/help/feedback/allaboutfeedback.html,

accessed on 7/1/15; Etsy, The Review System, https://www.etsy.com/help/article/102, accessed
on 7/1/15.
29
Cabral, Luis, and Ali Hortacsu, The Dynamics of Seller Reputation: Evidence from eBay, The
Journal of Industrial Economics 58(1), 2010, pp. 5478 at pp. 54, 75.
30
Cabral, Luis, and Ali Hortacsu, The Dynamics of Seller Reputation: Evidence from eBay, The
Journal of Industrial Economics 58(1), 2010, pp. 5478 at p. 56.
31
For example, Etsys policy states that refusal or inability to provide a minimum level of customer
service may ultimately result in the closure of your shop. (See Etsy, Seller Service Level
Standards, https://www.etsy.com/help/article/5690, accessed on 7/1/15.) eBay can also suspend
an account due to a rule or policy violation. (See eBay, Suspended Accounts,
http://pages.ebay.com/help/account/suspended-accounts.html, accessed on 7/1/15.)
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
16
CV 13-03826-EMC
32
for a better allocation of resources.
resources: spare rooms or apartments, idle automobiles, or the labor of underemployed individuals
such as individuals who might use spare hours to contemplate accepting ride referrals from a lead
generation platform.
empty home, a two-hour job for an idle person, etc.), online platforms have helped society to more
fully realize the value of its resources (e.g., existing cars and homes being used more).
43.
33
The FTC has described numerous examples of this better use of
By defining the product at a more granular level (i.e., a one-night rental of an
For labor markets, this has meant the ability to provide workers with options that are
disaggregated and flexible. Workers now have increased ability to choose how much time to devote
to an activity since online platforms usually do not specify a minimum commitment. For example, a
10
survey conducted by Etsy in 2012 showed that Etsy sellers invest on average 12 hours a week in their
11
shops and that 26% of U.S. Etsy sellers are full-time employees of another company.
12
below, many people who contemplate accepting ride referrals from an Uber platform are similarly
13
full-time employees of another company.
34
As I discuss
14
15
2.
16
44.
Technological Innovations Foster Entrepreneurship
Academic researchers have found that innovations in two-sided market platforms have
17
tremendous potential to expand grassroots entrepreneurship and innovation by allowing society to
18
tap into individual abilities and aspirations that would otherwise not have been realized.
19
example, according to a survey conducted by Etsy in 2012, 74% of Etsy sellers consider their Etsy
20
shop a business and 18% of sellers sell goods on Etsy full-time.
36
35
For
Etsy has also shown that the costs
21
22
32
Sundarajan, Arun, From Zipcar to the Sharing Economy, Harvard Business Review, 1/3/13.
23
33
Federal Trade Commission, The Sharing Economy: Issues Facing Platforms, Participants, and
Regulators: A Federal Trade Commission Workshop, 6/9/15, p. 3.
34
Etsy, Redefining Entrepreneurship: Etsy Sellers Economic Impact, online at

http://extfiles.etsy.com/Press/reports/Etsy_RedefiningEntrepreneurshipReport_2013.pdf.
35
Cohen, Molly and Arun Sundarajan, Self-Regulation and Innovation in the Peer-to-peer Sharing
Economy, The University of Chicago Law Review Dialogue 82, 2015, pp. 116133, at 129.
36

24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
17
CV 13-03826-EMC
of starting a business on Etsy are smaller than the costs of starting a small business as measured in the
2007 Census Survey of Business Owners.
45.
37
YouTube, the video sharing platform, allows individuals to develop careers offering
38
specialized instruction and services.
make-up tutorial channel has over 7 million subscribers.
subscription cosmetics company expected to yield $120 million in revenues in 2015 and start a line of
makeup with LOral.
played an important role in fostering entrepreneurism.
One example of this phenomenon is Michelle Phan, whose

39
40
In 2012, this allowed her to start a
As I describe below, lead generation platforms such as Uber have also
9
10
IV.
11
COMPETITION AND CHOICES IN DRIVING SERVICES
LEAD GENERATION PLATFORMS LIKE UBER HAVE CHANGED
12
13
46.
The driving services industry is one example of a market that suffers from
41
14
coordination challenges. The development of lead generation platforms, like Uber, has helped
15
lessen these coordination problems. Lead generation platforms offer a much improved, more
16
17
37

38
Einav, Liran, et al., The Economics of Peer-to-Peer Internet Markets, Presentation prepared for
the Seventh Annual Federal Trade Commission Microeconomics Conference, 10/16/14, p. 15;
Todd Spangler, New Breed of Online Stars Rewrite the Rules of Fame, Variety, 8/5/14,
http://variety.com/2014/digital/news/shane-dawson-jenna-marbles-internet-fame-1201271428/,
accessed on 7/1/15.
39
Michelle Phan, About, YouTube.com, https://www.youtube.com/user/MichellePhan/about,

accessed on 7/1/15. Ms. Phans videos have more than 1 billion views.
40
30 under 30, Art & Style, Forbes, 2015, http://www.forbes.com/30under30/#/art-style, accessed

on 7/1/15.
41
Later in this report I will refer to transportation network companies (TNCs). TNCs, which are
regulated by the California Public Utilities Commission, are a particular type of lead generation
platform that arranges rides between passengers and drivers using their own personal vehicles.
TNCs provide prearranged transportation services for compensation using an online-enabled
application or platform (such as smart phone apps) to connect drivers using their personal
vehicles with passengers. California Public Utilities Commission Transportation License
Section, Basic Information for Passenger Carriers and Applicants, p. 10.
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
18
CV 13-03826-EMC
efficient referral service to connect potential passengers and potential drivers interested in on-demand
prearranged driving services, increasing choice for both parties, much as eBay and Etsy have done for
buyers and sellers. Prearrangement of transportation previously required brand recognition,
advertising, and staffing of phone lines; against this backdrop, lead generation platforms have plainly
made prearrangement quicker and cheaper. A side effect of this efficiency is that taxi companies now
face more direct competition, but this is a difference of degree rather than kind. For example,
limousine services have long competed against taxi companies using prearrangement.
47.
In this section, I first describe in more detail the coordination issues the driving
services industry was facing prior to the development of lead generation platforms. Then, I discuss
10
the different driving services available to passengers, how the industry developed in ways to attempt
11
to overcome the coordination challenges, and what options were available to workers who wanted to
12
provide driving services. I close the section by describing the origins of Uber and how it and other
13
lead generation platforms have changed the market for driving services.
14
15
A.
Coordination Issues in Transportation Prior to New Technologies
16
48.
The coordination challenges faced by the driving services industry are similar to those
17
present in many two-sided markets. The driving services industry depends critically on the ability to
18
match passengers to available drivers in a physical location at the times that the driver wants to drive
19
and the passenger wants the ride, as well as ensuring that drivers are with passengers who will pay
20
and are safe, and passengers are with drivers who will take them safely and efficiently to where they
21
want to go. Efficiently coordinating passengers and drivers to solve these information challenges is
22
complex. Prior to the emergence of lead generation platforms, governments regulated driving
23
services with an eye towards mitigating these same problems.
24
25
B.
The Transportation Industry Prior to Lead Generation Platforms
26
49.
Historically there have been a series of regulations governing the driving services
27
industry, many of which attempted to solve the coordination problems described above. Within this
28
regulatory framework, many different driving service options were created, including taxis and black
Gibson, Dunn &

Crutcher LLP
19
CV 13-03826-EMC
cars or limousines. A person wanting to drive passengers was by regulation only allowed to provide
driving services for hire through one of these driving service options.
3
4
1.
50.
Regulation of Taxis
Passengers value the ability to access driving services without much advance notice.
How does one solve the coordination and safety issues that might arise in a system that would allow a
potential passenger to stand on the corner and be picked up by a stranger? Current taxi regulations are
an attempt to do so. One way immediate driving service can be arranged is from the side of the street
between a taxi driver and a potential passenger. This method of matching passengers to drivers
10
(hailing) is unique to taxis; in California as in most other states, other driving services are required
11
to prearrange rides.
12
that taxis were the only viable option (as described below, prearrangement was of limited use because
13
of the time lag necessary to prearrange). It follows that if a driver wanted to offer driving services to
14
passengers who were unable or unwilling to invest in a prearranged ride, driving a taxi was the only
15
way that driver could compete to do so.
42
16
51.
Passengers seeking a ride for a short or unexpected trip would typically find
In California, taxis are not regulated by the state, but are instead regulated by cities
43
17
and counties.
18
appearance of taxis (e.g., the color schemes used to paint taxis, lights on the roof, ID numbers for cars
In order to make them easily identifiable to passengers, cities commonly regulate the
19
20
21
42
California Public Utilities Commission Transportation License Section, Basic Information for
Passenger Carriers and Applicants, p. 7. The term hailing may take on different meanings in
different contexts. For example, it is often used to juxtapose the matching method used by taxis
from those used by limousine drivers, commonly referred to as prearrangement. However, lead
generation platforms provide a service that is not hailing and yet is more immediate than
traditional prearrangement. As Judge Chen noted in, for example, the Order Granting
Defendants Motion for Judgment on the Pleadings, they provide a service that might be
described as hailing on demand (Uber provides a service that gives consumers the ability to
hail a participating car service driver on demand using their mobile phone p. 2). In my report,
I refer to the services offered by lead generation platforms as prearrangement. In my view,
these platforms represent an innovation that makes prearrangement much more effective and
indeed much closer to hailing on demand.
43
Passenger Carriers and Applicants, p. 7.
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
20
CV 13-03826-EMC
44
and drivers, and so on) and the means of calculating fares (e.g., meters).
to reduce some of the imperfect information issues inherent in passengers getting into a strangers car
(i.e., finding the driver, safety of the ride). So do other potential regulations including vehicle
inspections, driver reviews, and daily hours limits.
These regulations attempt
45
52.
46
Cities also may limit how many vehicles are allowed to operate as taxis within the
city.
franchising.
these cities a driver of a taxi must either own a taxi medallion herself or rent a medallion, or taxi,
from someone who does (often a taxi company). Because cities limit the number of medallions
Two methods used to control the number of taxis operating in a city are taxi medallions and
47
Several California cities, including San Francisco, use a system of taxi medallions. In
10
available, and there is a financial opportunity to be one of a limited set of vehicles that can pick up
11
hailing passengers, the cost of buying or leasing a medallion can be quite high. For example, the cost
12
of purchasing a medallion in San Francisco was $250,000 in 2014 and in my opinion would be higher
13
if medallions could be purchased on the secondary market.
14
must rent one. Since many taxi drivers do not own their own taxi, medallion rentals are often paired
15
with a one day rental of a taxi. The cost of this rental (often referred to as the gate fee) could be
48
Taxi drivers who dont own a medallion
16
17
18
19
44
City and County of San Francisco, Transportation Code, Article 1100, Sections 1106 and 1113.
20
45
Taxicab Services Ordinance, http://ci.santa-rosa.ca.us/departments/finance/revenue/taxfees/Pages/taxi.aspx, accessed on 6/22/15; Taxicab Rules and Regulations of the Board of
Taxicab Commissioners: City of Los Angeles, Rule 790, 9/14.
46
For example, as of 5/19/15, San Francisco had 1900 taxi medallions outstanding. Medallion
Holders, https://www.sfmta.com/services/taxi-industry/medallions/medallion-holders, accessed
on 7/1/15.
47
Houston Taxi Study, Tennessee Transportation & Logistics Foundation, 1/20/14, p. 17. See
also Bruce Schaller, Entry Controls in Taxi Regulation: Implications of US and Canadian
experience for taxi regulation and deregulation, 2007, pp. 67.
48
In San Francisco, medallions are non-transferrable, so there is no open market for medallions.
The price is fixed at $250,000 with a waiting list based on driver seniority for the right to
purchase a medallion. City and County of San Francisco, Transportation Code, Article 1100,
Sections 1105 and 1116. See also A-Card Seniority List (Proposed),
https://www.sfmta.com/services/taxi-industry/medallions/seniority-list, accessed on 7/2/15.
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
21
CV 13-03826-EMC
over $100 per shift (typically 10 hours); in San Francisco, in any given week, the average gate fee
that a cab company can charge for a 10-hour shift is $106.25.
49
53.
Rather than taxi medallions, some cities in California, such as Los Angeles and Santa
50
Rosa, offer a limited number of franchise licenses.
the city in order to operate a taxi service in these cities. To be granted a franchise, a taxi company
must pay an annual fee and/or monthly fees which can be charged per vehicle.
taxi companies must pay for medallions and/or franchises in order to gain access to the restricted
ability to compete for hailing passengers.
Taxi companies must be granted a franchise by
51
Taxi owners and
9
10
2.
11
54.
Regulation of black cars and limousines
As noted above, in California, black car and limousine companies are regulated by the
12
PUC. This is in contrast to taxis and taxi companies which are regulated by cities and counties. The
13
PUC classifies a black car or limousine company as a Transportation Charter Party (TCP).
14
55.
A certificate or permit is required to operate as a TCP in California. Certificates and
15
permits are issued to individuals, general partnerships, corporations, limited liability companies
16
(LLC), and limited partnerships (LP) and are valid for three years.
17
must be met by applicants. Each vehicle to be used to provide TCP services must be registered with
18
the state as a commercial vehicle or a limousine.
52
53
54
A number of requirements
The applicant must submit a one-page statement
19
20
49
21
50
For example, see Taxicabs,

http://ladot.lacity.org/WhatWeDo/TaxicabsAmbulancesPipelines/Taxicabs/index.htm, accessed
on 7/3/15; and Taxicab Services Ordinance, http://ci.santarosa.ca.us/departments/finance/revenue/tax-fees/Pages/taxi.aspx, accessed on 7/3/15.
51
For example, see Los Angeles Municipal Code, Chapter VII, Article 1, Section 71.05.
52
53
54
22
23
24
25
26
27
28
City and County of San Francisco, Transportation Code, Article 1100, Section 1124.
Gibson, Dunn &

Crutcher LLP
22
CV 13-03826-EMC
estimating revenues and expenses over a stated initial period of operations, either 90, 120, 180 days,
or one year.
and property damage insurance on file with the PUC.
are properly completed, TCP authority is granted typically within three to six weeks.
55
Before a permit or certificate will be granted, all TCPs must have valid public liability
56
Assuming all criteria are met and all forms

57
56.
The PUC also regulates the factors used by TCPs to determine fares. With the
exception of transportation used for sight-seeing tours, TCPs are required to charge fares based on the
vehicle mileage, time of use, or a combination of both factors and are not allowed to charge fares
based on the number of passengers.
58
57.
TCPs are distinct from taxi services in several important ways according to the PUC.
10
First, taxis are regulated by cities and counties, whereas TCPs are regulated by the state, specifically
11
by the PUC. Taxis have meters and top lights, whereas TCPs do not. Most importantly, TCPs are not
12
allowed to be hailed by potential passengers. Each ride provided by a TCP must be arranged
13
beforehand, or prearranged, and the driver must maintain a waybill that contains detailed information
14
about the arrangement, including the name of passenger or person arranging the trip, the time of
15
arrangement, and the origin and destination of the trip.
59
16
17
3.
18
19
58.
Driving choices
Prior to the recent technological changes in the marketplace, a person interested in
working as a driver for passenger trips had the choice between driving a taxi or driving a black car or
20
21
55
56
57
58
59
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
23
CV 13-03826-EMC
limousine. Within each of these driving options, many work arrangements are possible. In addition,
drivers providing each service type have historically used a variety of different methods to coordinate
with potential passengers.
a.
59.
Driving a taxi
Taxi drivers in California work under many different arrangements. Some are
60
employees of a taxi company, but others are self-employed.
medallions, but others do not. Taxi drivers who do not own their own medallion the majority of
drivers must pay to rent the rights to a medallion during their shift. This rental fee is usually
included in the gate fee that is charged by taxi companies. A gate fee typically covers not only the
Some taxi drivers own their own
61
62
10
cost of renting the rights to the medallion, but also the cost of insurance and of renting the taxi.
11
Drivers usually pay for their own gas.
63
12
60.
Gate fees can be quite expensive. For example, in any given week, the average gate
64
13
fee that a cab company can charge for a 10-hour shift is $106.25.
14
order to lease a medallion and taxi for the day, drivers often find it necessary to tip employees at
15
the taxicab companies or dispatch companies.
65
Furthermore, I understand that in
This can cost the driver more than an additional $10
16
17
18
60
Bureau of Labor Statistics, U.S. Department of Labor, Occupational Outlook Handbook, 2014-15
Edition, Taxi Drivers and Chauffeurs, http://www.bls.gov/ooh/transportation-and-materialmoving/taxi-drivers-and-chauffeurs.htm, accessed on 7/3/2015
61
In San Francisco, for example, there are 1,735 full time medallions, but over 7,000 licensed taxi
drivers. Hara Associates, Best Practices Studies of Taxi Regulation: Meter Rates & Gate Fees,
8/27/13, p. 1-2.
62
Hara Associates, Best Practices Studies of Taxi Regulation: Meter Rates & Gate Fees, 8/27/13, p.
1-2.
63
Hara Associates, Best Practices Studies of Taxi Regulation: Meter Rates & Gate Fees, 8/27/13, p.
1-2.
26
64
27
65
Hara Associates, Best Practices Studies of Taxi Regulation: Taxi Driver Survey, Draft, 3/31/13,
accessed at https://www.sfmta.com/about-sfmta/reports/hara-associates-draft-taxi-user-survey, p.
3.
19
20
21
22
23
24
25
28
Gibson, Dunn &
Crutcher LLP
24
CV 13-03826-EMC
66
per shift.
standard gate fee, leading to charges from former DeSoto Cab President Hansu Kim that brokers in
possession of medallions are extorting drivers Such backroom deals are common in the cab
industry.
that day and is provided with no guarantee that fares earned during the day will cover the cost of
renting and operating the taxi. That is, some taxi drivers may work a full day and yet experience a
net loss.
67
8
9
61.
Moreover, on the underground market, renting a medallion may cost many times the
By paying the gate fee, the driver is paying an upfront fee for the opportunity to work
Taxi drivers who own their own medallions sometimes manage them on their own, but
many lease their medallion to a taxi company. The taxi company then has the right to operate a taxi
10
using the medallion and to rent the medallion to other drivers on a shift-by-shift basis. The medallion
11
owner, in return, is paid a medallion lease by the taxi company, but may then be required to pay the
12
taxi company a gate fee in order to use the medallion during his or her shifts.
13
medallion owner wants to manage his medallion, he still needs to pay a monthly fee to affiliate with a
14
Color Scheme or taxi company and use its branding and dispatch services.
68
Alternatively, if a
69
15
62.
Historically, several methods have been used by taxi drivers to find passengers,
16
including hailing, taxi stands, and dispatch services. Passengers can arrange for taxi service by
17
hailing a taxi from the side of the street. While the ability to pick up passengers in this manner is
18
unique to taxis, it has considerable limitations. For a match to occur, both partiesdriver and
19
passengermust be in the same physical location and recognize that the match is desired and
20
available. Because of this, and to increase the chance of finding their next passenger, taxi drivers
21
22
66
For example, 29% of drivers say they give more than $10 in tips on a desirable Friday or Saturday
night shift. Hara Associates, Best Practices Studies of Taxi Regulation: Taxi Driver Survey,
Draft, 3/31/13, accessed at https://www.sfmta.com/about-sfmta/reports/hara-associates-draft-taxiuser-survey, p. 5.
67
Rachel Swan, Color Schemes: The Underground Taxi Market, SF Weekly, 4/3/13,
http://www.sfweekly.com/sanfrancisco/color-schemes-the-underground-taximarket/Content?oid=2188488, accessed on 7/3/15.
68
SFMTA Taxicab Advisory Council, Medallion Sales Pilot Program Report, 3/9/12, p. 2.
23
24
25
26
27
28
69
Hara Associates, Best Practices Studies of Taxi Regulation: Meter Rates & Gate Fees, 8/27/13, pp.
1-2, 1-3; SFMTA Taxicab Advisory Council, Medallion Sales Pilot Program Report, 3/9/12, p. 2.
Gibson, Dunn &

Crutcher LLP
25
CV 13-03826-EMC
often drive around the city, looking out the windows of their vehicle for potential passengers on the
side of the street. Once the potential passenger and driver are in the same location, they both must
simultaneously see each other and jointly signal their intentions, or else the match may fail, resulting
in lost time for both parties. For passengers, signaling often involves waiving a hand or whistling.
Drivers often use lights on the top of the vehicles (top lights) to indicate their availability. This
solution to the coordination problem is effective in certain circumstances, but limited in others. For
example, this solution is less effective in neighborhoods where there are fewer taxis, and fewer riders,
on average.
63.
Taxi stands provide an alternative solution for matching potential passengers to
10
available taxis. The taxi stand provides a fixed location for potential passengers to seek out available
11
drivers. In this way, taxi stands are similar to the coordination solution offered by craft fairs and
12
garage sales, which I discussed in Section III. However, as with craft fairs and garage sales, this
13
solution to the coordination problem has its drawbacks. The passenger must physically be at the taxi
14
stand in order to arrange a ride. There is ebb and flow in passenger demand and willingness to wait
15
in the taxi stand line, which is hard for drivers to anticipate. A taxi driver consequently might have to
16
wait for a long time at a taxi stand before obtaining a customer. Conversely, a passenger might also
17
to have to wait at a taxi stand.
18
64.
A third method used to mitigate the coordination problem facing drivers and
19
passengers is dispatch services. Both hailing and taxi stands are limited by the need for passengers
20
and drivers to be in the same physical location in order to arrange a ride. Dispatch services help
21
alleviate this particular problem by centralizing and disseminating information about passenger
22
requests to a group of taxi drivers. Dispatch services can take several forms. Some taxi companies
23
own and operate their own dispatch (branded dispatch).
24
services only coordinate rides among taxi drivers who are either employed by the taxi company or
25
work with the taxi company as an independent contractor. Other dispatch services are not limited to a
26
particular taxi company, or brand. Rather, these companies provide only dispatch services
70
By construction, branded dispatch
27
28
70
For example, see Luxor Cabs in San Francisco, http://www.luxorcab.com/, accessed on 7/2/2015.
Gibson, Dunn &

Crutcher LLP
26
CV 13-03826-EMC
71
(independent dispatch); they do not own or operate any taxis.
subscribe to these independent dispatch services are responsible for providing their own taxi,
equipment, and insurance.
value of dispatch services are limited by the number of users of the service. That is, passengers
prefer dispatch services that will give them the best service (i.e., one aspect of service is quick and
reliable connection to a driver, which is aided by having many drivers on the service) and taxi drivers
get more value from a dispatch service that is used by many passengers. Passengers historically have
had to identify the dispatch service and phone number, call and communicate with a person on the
other end of the call, who then needed to relay the information accurately to the pool of available
10
72
Independent taxi drivers who
Like the two-sided platforms described in Section III, the efficacy and
73
drivers.
11
12
a)
Driving a black car or limousine
13
65.
14
There are several options for a driver who wants to drive passengers but does not want
15
to drive a taxi. Drivers can own their own black car or limousine and operate as an independent
16
business. To do so, the driver must register with the PUC and obtain a certificate or permit to operate
17
a chartered driving service (TCP). As described above, the process to obtain authority to operate a
18
TCP is nontrivial.
19
66.
74
Black cars and limousines are only allowed to offer prearranged transportation they
20
are not permitted to pick up potential customers seeking to hail them which has traditionally limited
21
the methods available to them for finding and coordinating with passengers. Drivers operating
22
71
For example, see American Taxi Dispatch,

https://www.americantaxi.com/ATOnlineOrderWeb/main.htm, accessed on 7/2/2015.
72
For example, see American Taxi Dispatch,

https://www.americantaxi.com/ATOnlineOrderWeb/main.htm, accessed on 7/2/2015.
73
For example, San Francisco has 26 taxi companies, each with their own dispatch service. Taxi
Company Directory, https://www.sfmta.com/getting-around/taxi/taxi-companies, accessed on
7/2/2015.
74
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
27
CV 13-03826-EMC
independent black car companies rely on passengers contacting them for rides. These business
owners must develop a reputation among potential passengers in order to increase the likelihood that
a passenger will choose to prearrange with their company. To do this, they can invest in advertising
or self-marketing or rely on word of mouth. Alternatively, owners of black car businesses can pay for
passenger referrals, which could involve subscribing to a dispatch service. Finally, as an owner of an
independent business, the black car driver would be responsible for paying for their business
expenses. All of this adds to the cost and risk of operating an independent black car business.
8
9
67.
Rather than operating their own independent black car service, a driver could instead
75
76
work for a black car company as either an employee or as an independent contractor.
In doing so,
10
the driver may alleviate the need to invest their own money in methods for identifying passengers and
11
may not be responsible for expenses. However, the available options for identifying and coordinating
12
with passengers are similar for the multi-car black car company rides must be prearranged and
13
passengers must somehow learn about the reputation and availability of the drivers.
14
15
C.
16
New Options and Sources of Competition Emerge with the Development of Lead
Generation Platforms and TNCs
17
68.
Historically, and as discussed, passengers could hail taxis to arrange for immediate,
18
quick, or unexpected trips, but had to prearrange transportation in order to arrange travel with any
19
non-taxi service, such as a black car or limousine. Hailing taxis may be effective in dense urban
20
areas during certain hours of the day, but may be less effective in many other areas and
21
circumstances. Prearrangement of transportation typically required coordinating over the phone well
22
in advance of the trip, which required that drivers invest in infrastructure (dispatch systems and call
23
centers) and advertising and that passengers incur search costs to identify a reliable, and available,
24
driver. These combined solutions only partially addressed the coordination challenges faced by the
25
26
27
75
For example, see Declaration of Patrick Canlas, 8/11/14, UBEOCO0000911420, 19.
28
76
For example, see Declaration of Oleg Lantsman, 5/19/15, 12.
Gibson, Dunn &

Crutcher LLP
28
CV 13-03826-EMC
driving services industry. Lead generation platform referral services have offered a new solution to
these problems and in the process expanded choices for drivers and passengers.
69.
Uber began in 2009, as a service offered to licensed black car and limousine
companies to help them find customers for prearranged trips. The platform service called UberCab
was launched in San Francisco in May 2010, and eventually became what is now UberBLACK.
The Uber App facilitates the prearrangement of driving services in a way that was not previously
possible. The Uber App provided potential passengers and black car and limousine companies with a
far more effective solution to their coordination problems. Passengers could easily signal their desire
to arrange a ride to many companies at once. They no longer needed to identify and contact a
77
10
particular company, nor spend the time to call a dispatch service or seek out a taxi stand to arrange a
11
ride. Likewise, black car and limousine companies and independent owners were able to much more
12
efficiently identify potential customers. The Uber App reduced the need to invest in the advertising
13
and infrastructure that was previously necessary to attract and identify passengers.
14
70.
In addition, the UberBLACK platform could, and does, quickly match black car
15
drivers to passengers in the same geographic location, dramatically reducing the lead time needed to
16
execute prearranged transportation. This could allow passengers who valued the service offered by
17
black car companies to rely on prearranged transportation options for more rides. This put small
18
black car businesses and black car independent drivers in a better position to compete directly with
19
taxi companies for certain customers.
20
71.
Ubers initial service, UberCab/UberBLACK, was limited to arranging rides for
21
licensed black cars and limousines. These car services provided high quality rides, but usually at a
22
higher price than a taxi fare.
78
In the summer of 2012, Uber launched uberX,
79
a lower cost
23
24
77
Ubers Founding, 12/22/10, http://newsroom.uber.com/2010/12/ubers-founding/, accessed on

7/7/15.
78
Leena Rao, UberCab Takes The Hassle Out of Booking a Car Service, TechCrunch , 7/5/10.
http://techcrunch.com/2010/07/05/ubercab-takes-the-hassle-out-of-booking-a-car-service/,
accessed 7/2/15; Michael Arrington, What if Uber Pulls an Airbnb? Taxi Business Could
(Finally) Get Some Disruption, TechCrunch, 8/31/10, http://techcrunch.com/2010/08/31/what-ifubercab-pulls-an-airbnb-taxi-business-could-finally-get-some-disruption/ accessed 7/6/15.
25
26
27
28
Gibson, Dunn &
Crutcher LLP
29
CV 13-03826-EMC
80
alternative to UberBLACK.
licenses could use the uberX platform,
vehicles, including non-hybrids, could also use the uberX platform.
a driver to have a TCP license in order to use the Uber App and the vehicles did not have to meet the
size and quality requirements of UberBLACK.
At this time only drivers driving liveried hybrid vehicles with TCP
81
but starting in February 2013, drivers using their personal

82
It was no longer necessary for
83
72.
Use of the uberX platform provided new job opportunities for people who wanted to
offer driving services. Drivers using uberX did not need to own their own medallion or taxi. Nor did
they need to pay, explicitly or implicitly for the use of someone elses taxi or black car. Because they
could use the Uber App to help find passengers, it was no longer necessary for drivers to work for, or
10
contract with, a taxi company or black car company, to pay for a dispatch service, or to find their own
11
passengers by driving around town looking for potential customers. Thus uberX and UberBLACK
12
led to increased options and lower entry costs for drivers and put increased competitive pressure on
13
the value of the ability to pick up hailing customers that was limited by regulation to taxi companies
14
with medallions or franchises. According to a 2014 to 2015 report of business travelers from expense
15
management system provider Certify, in San Francisco 71 percent of business receipts for rides went
16
to Uber with 29 percent going to taxis, and in Los Angeles 49 percent of business receipts went to
17
18
19
20
79
Alexia Tsotis, Uber Opens Up Platform to Non-limo Vehicles with UberX, Service will be 35%
Less Expensive, TechCrunch, 7/1/12, http://techcrunch.com/2012/07/01/uber-opens-upplatform-to-non-limo-vehicles-with-uber-x-service-will-be-35-less-expensive/, accessed 7/1/15.
80
Less Expensive, TechCrunch, 7/1/12, http://techcrunch.com/2012/07/01/uber-opens-upplatform-to-non-limo-vehicles-with-uber-x-service-will-be-35-less-expensive/, accessed 7/1/15;
http://www.greencarreports.com/news/1078381_uberx-service-offers-hybrid-cars-lower-fares-inny-san-francisco, accessed on 7/2/15.
81
Email from Thomas Colopy to Philip Acevedo, Fw: Introducing A New Opportunity With
Uber, 5/21/14, COLOPY00017580 at COLOPY000175.
82
Deposition of Michael Coleman (Driver Operation Specialist, Uber), 7/9/14, pp. 3536.
83
SF / Bay Area P2P Vehicle Requirements, http://www.sfuberpartners.com/acceptedvehicles.html, accessed on 6/30/15.
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
30
CV 13-03826-EMC
84
Uber with 51 percent going to taxis.
market shares is partially a result of its convenience and competitive pricing for consumers and
partially a result of its evident attractiveness to drivers.
The ability of Uber to maintain such high (and increasing)
85
73.
Uber launched uberTAXI in several different cities over the course of 2012, beginning
86
with Chicago in April, New York in September, and San Francisco in October. uberTAXI rides are
provided in a licensed taxi, rather than a personal vehicle.
utilize the Uber App to arrange taxi rides and to alleviate the traditional coordination challenges of
prearranged rides.
74.
10
87
uberTAXI allows taxi companies to
On December 20, 2012, the PUC started a rulemaking process for companies such as
Uber that have recently begun using mobile internet, social media, and location services to offer new
11
12
13
84
Sharing the Road: Uber on the Rise Nationwide, Certify, 2015,

https://www.certify.com/CertifySpendSmartReport.aspx, accessed on 7/3/15. These figures have
been trending up, with Ubers share of business travel rising from 58 percent to 71 percent in San
Francisco from the first quarter of 2014 to the first quarter of 2015, and from 23 percent to 49
percent in Los Angeles over the same period of time. Indeed, in all eight cities listed in Certifys
analysis, Uber gained ground on taxis. In New York, Ubers share of business travel expenses
rose from 9 to 21 percent; in Chicago it rose from 8 to 25 percent; in Dallas it rose from 27 to 56
percent; in Washington, D.C., it rose from 20 to 49 percent; in Miami, it rose from 0 to 23
percent; and in Atlanta it rose from 8 to 41 percent. According to the Boston Globe, Certifys
analysis is based on 28 million business expense receipts nationally. Scott Mayerowitz, More
business travelers seen going with Uber, Boston Globe, 4/7/15.
https://www.bostonglobe.com/business/2015/04/07/uber-popularity-surges-business-travelersavoiding-taxis/x9SRbGh7B7RaCqmvL7e1wM/story.html, accessed 7/3/15.
85
See Exhibit 3, below. Generally, uberX is less costly than a traditional taxi service. Certifys
analysis of national business expenses is consistent with this and indicates that the average receipt
in the first quarter of 2015 was $31.24 using Uber and $35.40 using taxis, limousines, and airport
shuttles. Sharing the Road: Uber on the Rise Nationwide, Certify, 2015,
https://www.certify.com/CertifySpendSmartReport.aspx, accessed on 7/3/15.
86
Can a Chicago Taxi be Uber? Uber, 4/18/12,

http://newsroom.uber.com/chicago/2012/04/chicago-taxi-uber/, accessed on 7/6/15; Introducing
Uber Taxi NYC + One Free Ride for Every New Yorker, Uber, 9/5/12,
http://newsroom.uber.com/nyc/2012/09/introducing-uber-taxi-nyc-one-free-ride-for-every-newyorker/, accessed on 7/6/15; Taxi is Arriving in San Francisco, Uber, 10/17/12,
http://newsroom.uber.com/sf/2012/10/taxi-is-arriving-in-san-francisco/, accessed on 7/6/15.
87
Less Expensive, TechCrunch, 7/1/12, http://techcrunch.com/2012/07/01/uber-opens-upplatform-to-non-limo-vehicles-with-uber-x-service-will-be-35-less-expensive/, accessed 7/1/15.
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
31
CV 13-03826-EMC
88
ways of arranging transportation of passengers over public highways for compensation.
completing its examination of these new companies, on September 19, 2013, the PUC ruled that it
had jurisdiction over TNCs. According to the PUC, a TNC provides prearranged transportation
services for compensation using an online-enabled application (app) or platform to connect
passengers with drivers using their personal vehicles.
After
89
75.
Drivers using uberX can use their own personal vehicles or drive liveried cars with
TCP licenses, whereas drivers using UberBLACK can only use commercially registered black cars
and limousines. The PUC ruling thus has important regulatory implications for drivers. According to
the PUC, the primary distinction between TNCs and other companies regulated by the PUC that
10
provide prearranged driving services is that a TNC connects riders to drivers who drive their
11
personal vehicle, not a vehicle such as a limousine purchased primarily for a commercial purpose. To
12
that end, a TNC is not permitted to itself own vehicles used in its operation or own fleets of
13
vehicles Thus, Rasier-CA LLC, a wholly owned subsidiary of Uber, is regulated as a TNC,
14
because it contracts with drivers who use the uberX platform and drive personal vehicles (although
15
drivers of liveried cars with TCP licenses also use the uberX platform). The licensed black car and
16
limousine companies that use UberBLACK to connect to riders are regulated by the PUC, but not as
17
TNCs.
90
91
18
19
88
California Public Utilities Commission, Order Instituting Rulemaking on Regulations Relating to

Passenger Carriers, Ridesharing, and new Online-Enabled Transportation Services, 12/27/12, p.
1.
89
California Public Utilities Commission, Decision Adopting Rules and Regulations to Protect
Public Safety While Allowing New Entrants to the Transportation Industry, 9/19/13, p. 2.
90
91
See California Public Utilities Commission, Decision Modifying Decision 13-09-045, 11/20/14, p.
23: Uber Technologies has multiple transportation offerings, however, only UberX provides
TNC services. The other transportation offerings by Uber Technologies are licensed as limo
drivers and regulated by this Commission. All Uber offerings other than UberX such as Uber or
Uber Black or Uber SUV are all and should be licensed professional drivers. I also understand
that it is under consideration by the PUC whether Uber itself should be regulated as a TCP. See
also California Public Utilities Commission, Assigned Commissioner and Assigned
Administrative Law Judges Scoping Memo and Ruling for Phase II of Proceeding, 11/26/14, pp.
23.
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
32
CV 13-03826-EMC
76.
The PUC also imposed a series of rules and regulations that apply to all TNCs. In
particular, the PUC established that TNCs needed to obtain a permit from the PUC, require criminal
background checks for each driver using the platform, establish a driver training program, implement
a zero-tolerance policy on drugs and alcohol, and require insurance coverage, as well as provide
annual reports on several issues such as safety and accessibility.
92
77.
Uber is not the only lead generation platform. Indeed, several other companies
compete with Uber to provide ride coordination services. Two of Ubers largest competitors are Lyft
and Sidecar, but many other companies compete with Uber as well, including Hailo and Kappa Phi.
93
78.
In summary, the emergence of lead generation platforms, including Uber, has led to
10
increased choice for those offering driving services. Lead generation platforms have improved the
11
efficiency of prearranged driving services such as limousines. This efficiency put prearranged
12
driving services in a better position to compete against hailing driving services such as taxis, but that
13
does not mean that taxi drivers are necessarily worse off. Taxi drivers, if they choose to, can accept
14
referrals from a lead generation platform. Moreover, if lead generation platforms are in fact growing
15
the pie, for example by persuading people that they do not need to own a car, then there may be more
16
business for prearranged and hailing driving services alike.
17
18
V.
THERE IS NO TYPICAL DRIVER WHO USES THE UBER APP
19
20
79.
In this section I discuss the many different types of drivers who use the Uber App and
21
describe the many ways in which the Uber App has changed their options for working and earning.
22
This highlights the individualized nature in which workers use the Uber App and shows that there is
23
no typical driver using the Uber App.
24
25
26
27
28
92
93
Im not sure Id be able to list all of Ubers competitors. Some that come to mind are platforms
that maybe use a similar model. Im not too familiar with exactly how they operate their
business. Sidecar, Halo, Lyft, Kappa Phi. I believe theres just many more. Deposition of
Michael Coleman (Driver Operation Specialist, Uber), 7/9/14, p. 149.
Gibson, Dunn &

Crutcher LLP
33
CV 13-03826-EMC
A.
Some Drivers Who Use the Uber App Use Competing Platforms at the Same
Time They Use the Uber App
80.
A driver who chooses to sign up to use the Uber App may simultaneously do work on
behalf of herself or third parties; Uber does not have a uniform policy preventing that person from
working full-time or part-time for themselves or third parties, and indeed does not have a uniform
policy preventing that person from multi-apping, or directly and simultaneously competing with the
Uber App by completing referrals from other lead generation platforms such as Lyft.
in this section, some drivers have additional work within the driving services industry; some even use
competing platforms at the same time as using the Uber App. Other drivers have additional work in
10
non-car service industries. For some people that non-car service industry work is part-time whereas
11
for others it is full-time.
94
12
81.
As I describe
While not on a ride, some drivers who use the Uber App may decide whether to take
13
an Uber referral or the referral from another platform such as Lyft or Sidecar. Many driver
14
declarations report actively serving another lead-generating platform.
95
Other drivers also explain
15
16
94
Uber has used many different licensing agreements. Some of these contracts specifically indicate
that drivers have the right to work at another job, including jobs obtained through competing lead
generation platforms. For example, see Rasier Transportation Provider Service Agreement,
5/15/13, p. 3 (The Company neither has nor reserves the right to restrict you from performing
other transportation services for any company, business or individual, or from being engaged in
any other occupation or business.) and Rasier Software License and Online Services Agreement,
11/10/14, 2.4 (You acknowledge and agree that you have complete discretion to provide
services or otherwise engage in other business or employment activities. For the sake of clarity,
you understand that you retain the complete right to; (i) use other software application services in
addition to the Uber Services; and (ii) engage in any other occupation or business.). On the
other hand, some licensing agreements state that drivers cannot seek ride referrals from
competing lead generation platforms while the driver is signed onto the Uber App. For example,
see Rasier Software Sublicense and Online Services Agreement, 6/21/2014, p. 3 (The Parties
recognize that both you and the Company are, or may be, engaged in similar agreements with
others. Nothing in this Agreement shall preclude the Company from doing business with other
independent transportation service providers, nor preclude you from entering into contracts
similar to this Agreement with other lead generation providers. The Company neither has nor
reserves the right to restrict you from performing other transportation services for any company,
business or individual, or from being engaged in any other occupation or business. However,
during the time you are actively signed into the Software, you shall perform transportation
services only for Requests received by you via the Software.).
95
For example, see Declaration of Sebastien Trouplin, 11/21/14, UBE-OCO0000954954, 2;

Declaration of Casey Cheung, 5/19/15, 5; Declaration of David Wong, 5/18/15, 4; Declaration
of Elmer Brigido A. Javier, 5/18/15, 8; Declaration of Euripedes Silva, 5/18/15, 3; Declaration
17
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
34
CV 13-03826-EMC
that they run their Uber and alternative platform app simultaneously to increase the chances of getting
a driving job.
from operating multiple ride sharing apps simultaneously. I always operate the Uber App and the
Lyft App simultaneously whenever I am driving. I accept a fare from whatever App sends me a lead
first.
time as Uber. That application is called Lyft. The benefit of using two applications is that I can stay
busier. There is less downtime. I can reject or accept leads. Uber has never suspended or threatened
to suspend my application use because I did not accept enough leads. Uber does not restrict my use
of other applications like Lyft.
97
96
For example, Sevuloni Ratabumusu explains: Uber has also never prevented me
Michael Banko has a similar strategy: I occasionally use another application at the same
98
Some drivers make a point of running Uber and the alternative app
10
at mutually exclusive times. Hazim Jajo explains that he would keep switching until [he] got a
11
customer.
12
82.
99
The share of each drivers trips that are arranged using the Uber App versus another
13
platform varies from driver to driver. Umar Faruk Azam, for example, uses Lyft off and on, but
14
uses Uber 95 percent of the time.
15
San Francisco fares are arranged using the Lyft platform and 30 percent are arranged using the Uber
16
App.
17
platform can change over time; for example, Sebastien Trouplin states: Currently, about 90% of my
100
101
For Sevuloni Ratabumsu, on the other hand, 70 percent of his
For some drivers, the split between the share of rides picked up from Uber versus another
18
of Christopher J. Armentrout, 5/20/15, 3; Declaration of E. Bond Francisco, 5/19/15, 4;
Declaration of Frank T. Chow, 5/19/15, 4; Declaration of Hanibal Poolisrezaeih, 5/19/15, 4;
and Declaration of Jay Townsend, 5/20/15, 5.
19
20
96
For example, see Declaration of Sebastien Trouplin, 11/21/14, UBE-OCO0000954954, 15;

Declaration of Casey Cheung, 5/19/15, 5; Declaration of David Wong, 5/18/15, 4; Declaration
of Elmer Brigido A. Javier, 5/18/15, 8; Declaration of Euripedes Silva, 5/18/15, 6; Declaration
of Jeremy Thomas Paz, 5/19/15, 4; Declaration of Ladawna Williams, 5/20/15, 6; Declaration
of Michael Banko, 5/19/15, 5; Declaration of Richard Charles Elliott, 5/18/15, 4; and
Declaration of Sevuloni Ratabumusu, 5/20/15, 5.
24
97
25
98
Declaration of Michael Banko, 5/19/15, 5.
26
99
Declaration of Hazim Jajo, 5/28/15, 9.
27
100
Declaration of Umar Faruk Azam, 5/20/15, 5.
28
101
21
22
23
Gibson, Dunn &

Crutcher LLP
35
CV 13-03826-EMC
fares are booked using the Uber app and about 10% are booked using Sidecar. I do not use the Lyft
app at this point, but at one time (up until February 2014), 30% or more of my fares came from Lyft-
generated passengers.
102
4
5
B.
Californians Working as Drivers Before the Existence of Uber and Lead

Generation Platforms
83.
Californians who were working as drivers prior to the existence of Uber and other lead
generation platforms have been impacted by the entry of lead generation platforms in different ways.
Among U.S. drivers who have driven passenger referrals using the Uber App, only 19 percent were in
103
10
driving professions directly prior to their use of the platform.
11
who have used the Uber App in the United States have never at any point in their career worked as a
12
driver prior to joining Uber.
13
of a better method for finding passengers. However, the impact of this new technology on
14
individuals work is varied. I show this below by describing a number of different driver archetypes
15
as well as real-world examples.
104
And over half (51 percent) of drivers
These drivers who have used the Uber App share the new availability
16
17
1.
18
84.
Example 1: A self-employed driver of a black car
Consider a self-employed individual who owns her own business and drives her own
19
black car as an independent contractor. She would have to obtain a P-Permit from the PUC, which
20
includes a $1,000 application fee every three years; she would have to obtain a Charter-Party Carrier
21
Permit from the Department of Motor Vehicles; and she would have to pay her own expenses and
22
23
24
102
Declaration of Sebastien Trouplin, 11/21/14, UBE-OCO0000954954, 11.
25
103
Hall, Jonathan, and Alan Krueger, An Analysis of the Labor Market for Ubers Driver-Partners
in the United States, Princeton University Industrial Relations Section, Working Paper 587,
1/22/15, p. 10.
104
1/22/15, p. 10.
26
27
28
Gibson, Dunn &
Crutcher LLP
36
CV 13-03826-EMC
105
insurance as well as invest in her own methods of finding passengers.
passengers might include word of mouth, handing out cards, or networking and advertising.
Methods for finding

106
85.
The existence of Uber and other lead generation platforms provides this limousine
driver with a meaningful option: the opportunity to use these platforms to find fares when she is not
driving her regular passengers. In order to accept a passenger referral from UberBLACK, she must
certify that the age and model of her car are consistent with it being an UberBLACK vehicle.
return for doing the job, she obtains insurance coverage for its duration
80 percent of the fare collected from the passenger, with the remainder going to Uber as a licensing
fee.
107
108
109
In
and collects approximately
She is free to continue to use word of mouth, networking, advertising or paying a larger
10
11
105
Passenger Carriers and Applicants, pp. 5, 79.
106
For example, see Declaration of Daniel Bisneto, 9/4/14, UBE-OCO0000906873, 21 (I pay a

webmaster to do some advertising on google.com. I also have business cards, and pay for the
domain fee to maintain my companys website); and Declaration of Ali Bouhelal, 5/30/14, UBEOCO0009020 5, 9 (I do not formally advertise my business, but I have built a substantial
client base through networking and word of mouth, such as recommendations from friends,
neighbors, former colleagues, and current clients).
107
The age and model requirements differ across cities. For example, see SF/Bay Area P2P Vehicle
Requirements, http://ubersupport.weebly.com/accepted-vehicles.html, accessed on 6/27/15; and
OC UberBLACK / SUV / LUX Requirements, http://ubersupport.weebly.com/av-oc-b.html,
accessed on 7/3/15.
108
When on trip with Uber, Uber provides $1 million of commercial auto coverage on every trip.;
while on platform but not on trip, Uber offers contingent liability coverage in case a drivers
personal policy does not provide coverage (Driver-jobs, https://www.uber.com/driver-jobs,
accessed on 7/5/15).
109
Uber San Francisco states: Uber BLACK partners in SF pay Uber 20% of the client payment as
a commission and receive the remaining 80% of the client payment When a client specifically
requests an SUV, Uber SUV partners in SF pay Uber 28% of the client payment as a commission
and receive the remaining 72% of the client payment UBERx partners in SF pay Uber 20% of
the client payment and receive the remaining 80% of the client payment (Email from Uber San
Francisco to Douglas OConnor, Quality Report 4.9.13, 4/9/13, UBE-OCO000002557, at
UBE-OCO000002567). The exact share of the ride that the driver pays to Uber as its licensing
fee depends on the service (UberBLACK, uberX, UberSUV, etc.), can differ across cities, and has
changed over time. For example, see Deposition of Michael Colman (Driver Operation
Specialist, Uber), 7/9/14, pp. 156158, 162164; and Email from Uber SF Team to Thomas
Colopy, Exciting New Changes to UberX!, 10/2/12, UBE-OCO0000092325 at UBEOCO00000923 (Commission for UberX will change from 80% to 90%. That means that Uber's
cut will be reduced from 20% to 10% on every trip fare until January 1st, 2013. Note: This does
not apply to Uber Black or SUV rates).
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
37
CV 13-03826-EMC
limousine company or dispatch service for referrals while using Uber and other lead generation
platforms. She can use her old methods of finding passengers as her sole source of fares when they
are sufficient to provide her with enough fares, and use Uber and other lead generation platforms to
find passenger referrals only when she is between her normal passengers or it is a slow day. Thus,
the option of Uber and other lead generation platforms leads to a potential expansion in her business.
The degree to which she, on any given day or in any given hour, chooses to use UberBLACK to
connect her to one or more passengers will depend on the value it provides to her as a referral service
in that moment the more value UberBLACK provides relative to her prior approach and other lead
generation platforms, the more she will use it.
10
86.
Consider Mohammed Ahmed, who is the sole owner of Pacific Beach Limo, and
11
drives his own black car. Before he started using Uber, Mohammed Ahmed solicited customers using
12
a variety of methods; the introduction to Uber simply allowed him to expand his business: I had a
13
number of customers who continued to use my services after I left San Diego Cab Company and a
14
few others that I either acquired through word of mouth or who or who had obtained my contact
15
information through my business cards I was looking for ways to increase my revenue. Uber
16
sounded like a great opportunity and an easy way for me to get more consistent fares.
110
17
87.
Consider Patrick Canlas, who is the owner of Black Limousines, LLC. Patrick Canlas
18
drives a black car and manages his employees; he hired these employees to drive additional cars in
19
Black Limousines fleet. Patrick Canlas initially sourced fares through traditional means: In about
20
2008, I bought a used Town Car and obtained the proper licenses At this time I was operating as a
21
sole proprietor and had my own clients that I got through word of mouth referrals from friends.
22
Clients could book my services through email, text, or by calling.
23
begin using the Uber App gave him access to a much larger client base, and leads sourced from Uber
24
now make up a substantial portion of his business: Transportation provided to non-Uber passengers
111
Patrick Canlass decision to
25
26
27
110
Declaration of Mohammed Ahmed, 7/14/14, UBE-OCO000090916, 3.
28
111
Declaration of Patrick Canlas, 8/11/4, UBE-OCO0000911420, 2, 4, 12.
Gibson, Dunn &

Crutcher LLP
38
CV 13-03826-EMC
makes up about 40% of my businesss revenue. The remaining 60% of Black Limousines business is
generated using the Uber [A]pp Since gaining access to the Uber [A]pp, my business has grown
substantially faster than it was growing before The business opportunity it presents speaks for
itself in only four years, I have gone from 1 car and 1 driver to 8 cars and 18 drivers.
112
5
6
2.
88.
Example 2: An employee of a limousine company
Consider a driver who is an employee for a limousine company. He has to perform
various regulatory requirements such as a daily written report on the state of the vehicle, and would
likely receive insurance, expense reimbursement and passenger referrals from his employer.
113
He
10
could be paid hourly for his work or paid by the ride, depending on the nature of his contract with his
11
employer. The employer may require he wear a uniform, offer soft drinks, candy and magazines in
12
the limousine and only play classical music.
13
89.
How does the existence of Uber or other lead generation platforms change this
14
employees job? With Uber or other lead generation platforms, his employer has another option:
15
using the platform to find passengers for its driver-employees. To appreciate that this is simply
16
another option that employers might or might not use depending on what suits them, consider that
17
Halim Arbouzs multi-driver business is 60% private clients and 40% from the Uber [A]pp,
18
whereas Steve Galadjians business receives more than 90% of its revenue from the Uber [A]pp.
114
115
19
90.
With the added option of Ubers platform, the employer receives a passenger referral
20
from Uber as well as insurance coverage for its employee who is driving that passenger. In return,
21
the employer pays Uber approximately 20 percent of the fare collected from the passenger as a
22
23
24
112
Declaration of Patrick Canlas, 8/11/4, UBE-OCO0000911420, 1315.
113
California Public Utilities Commission, A Guide to Filling out Form PL706-I, p. 2. See also,
Declaration of Daniel Bisneto, UBE-OCO0000906873, 9/4/14, 9, 19.
27
114
Declaration of Halim Arbouz, 5/19/15, 10.
28
115
Declaration of Steve Galadjian, 10/18/14, UBE-OCO0000903942, 8.
25
26
Gibson, Dunn &

Crutcher LLP
39
CV 13-03826-EMC
116
licensing fee.
services. The employer continues to be free to set its employees wages and hours, to choose which
passengers the driver must drive (whether found through UberBLACK platform or through other
means), to require a uniform, treats in the car, etc.
91.
This allows the employers business to compete much more directly with taxi
The employer has more choice than before; it can continue to use word of mouth,
networking, advertising or paying a referral service while using Uber and other lead generation
platforms to find passengers for its business and drivers. However, the existence of Uber and other
lead generation platforms might not change the employee-drivers job at all (other than his employer
having access to more potential passengers that it can direct its drivers to drive). On the other hand,
10
it might change the employee-drivers job in potentially significant ways, including higher pay due to
11
the employers higher profits. However, the employee could continue to be compensated in the exact
12
same way (whether per hour or per ride), at even the exact same rates as he had before, as that
13
compensation is not required to be directly connected to the fare a passenger pays the employer for
14
the business service.
117
15
16
17
116
117
For example, Eternity Limousine Service employs 9 drivers and generates about 60% of its
revenue through the Uber platform. Eternity's drivers are paid hourly regardless of how their
fares are booked, and so while Uber gives these drivers more opportunities to find fares, it does
not change their compensation structure (Declaration of Daniel Bisneto, 9/4/14, UBEOCO0000906873, 78, 10). Other companies, such as Bay Area Network Limo, pay their
drivers based on the fares they generate through the Uber app, but take a flat fee for the use of the
car (Declaration of Boris Doubnov, 5/19/15, 5).
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
40
CV 13-03826-EMC
92.
Consider drivers who are, or were in the past, employees of multi-driver businesses
that decided to use Uber to generate some of its leads. For example, Daniel Bisneto worked as an
employee for Jet Car Limousine, which started to use Uber to generate some of its leads: I first
learned about Uber about four years ago when I was working as an employee driver for a limousine
company called Jet Car Limousine During the time I was working for Jet Car, someone from Uber
contacted my boss and asked if Jet Car would be interested in partnering with Uber and using the
Uber app as a source of referrals. After Jet Car partnered with Uber, my boss told me to go to the
Uber office and sign up under Jet Cars account.
who started hiring employees because his business grew so rapidly once he started using Uber:
118
Mark Forester is a multi-driver business owner
10
After we signed up to use the Uber app, our business immediately grew so much that we hired our
11
first employee We currently employ about 34 drivers.
119
12
13
3.
14
93.
Example 3: A taxi driver who rents her cab
Consider a taxi driver who, unlike the limousine driver above, is an independent
120
15
contractor and drives for 10 hours twice a week for a taxi company.
16
reflects not only the wear and tear of using the car, but also the investment that the owner of the cab
17
has made to have the exclusive right to pick up customers who hail a ride (either by franchise or
She pays a rental fee that
18
19
20
21
118
Declaration of Daniel Bisneto, 9/4/14, UBE-OCO0000906873, 23.
22
119
Declaration of Mark Forester, 11/20/14, UBEOCO0000955560, 910.
23
120
There are a large number of different financial and work relationships any given taxi driver may
have related to the taxi he or she is driving. Some driver relationships have the characteristics of
an independent contractor relationship while others an employee relationship. Courts have
validated both relationships. The Employment Development Department of the State of
California states that taxicab drivers typically operate under three business arrangements: 1. The
taxicab company acknowledges the driver as an employee. 2. The driver owns and operates the
taxicab, independently arranges fares, and personally pays for required licenses, permits, and
insurance [Driver is an independent contractor]. 3. The driver performs services as a lease driver
on either a fixed-fee or percentage-of-receipts basis [Detailed analysis required to determine
whether the driver is an independent contractor or an employee] (State of California
Employment Development Department, Information Sheet, Taxicab Industry).
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
41
CV 13-03826-EMC
121
medallion). The driver pays a fixed amount every day to use the cab, as well as her own expenses.
She has to meet a multitude of regulatory requirements,
company or individual renting the cab to her may provide insurance while she is driving.
free to drive the cab as much as she would like in her ten-hour rental period, but she is aware that she
must use the time wisely as her investment has a fixed time period, even though this may lead to an
incentive to drive while tired.
taxi company as well as passengers found while she drives around looking on the street or waiting at
taxi stands.
94.
122
123
and so does the car she uses.
The taxi
124
125
She is
She finds passengers through any referrals she receives from the
126
How does the existence of Uber and other lead generation platforms change this
10
drivers work situation? She now has new competition on the road for finding many types of
11
passengers as Uber and other lead generation platforms have made it much easier and quicker to
12
connect passengers to drivers in prearranged rides. A company or driver using UberBLACK or
13
uberX to find prearranged passengers does not have to invest in the franchise or medallion in order to
14
15
121
Declaration of Ernest Avetisyan, 5/19/15, 7; Declaration of Hiis Ilmi, 7/9/14, UBEOCO0000907984, 2.
122
As a San Francisco cab driver, for example, she would first have to meet prerequisite
requirements including having proof of residency, meeting personal hygiene and health
requirements, having a current California drivers license, being able to drive for at least four
hours per day, having no prohibiting criminal background, being at least 21 years of age, and
speaking reading and writing the English language. She would then have to complete a driver
pre-screening, attend and complete taxi training at one of the approved schools, get fingerprinted,
obtain and pass a background check, obtain a 10 year printout of her driving record, obtain an
offer from a taxi company, pay an application fee (currently waived due the shortage of cab
drivers), take the class and test, drive with a temporary permit for 8 weeks and then exchange it
for an A-card and a badge (How to Become a Taxi Driver, SFMTA Municipal Transportation
Agency, https://www.sfmta.com/services/taxi-industry/become-taxi-driver, accessed on 7/3/15).
123
A taxicab in San Francisco, for example, needs to, among other things, have an installed
taximeter, be no older than seven model years, have a starting mileage lower than 100,000 miles,
have a current mileage lower than 375,000 miles, have rubber floormats, have rear seats
upholstered with vinyl or leather in good repair, and matching the vehicles interior colors etc.
26
124
Declaration of Kathy J. Robinson, 5/20/15, 5.
27
125
Declaration of Ernest Avetisyan, 5/19/15, 15; Declaration of Tesfaye Mekonnen, 5/18/15, 3.
28
126
Declaration of Adama Traore, 9/29/14, UBE-OCO0000901419, 22.
16
17
18
19
20
21
22
23
24
25
Gibson, Dunn &

Crutcher LLP
42
CV 13-03826-EMC
use the service. She might find this competition leads to her referrals from the taxi company coming
less often, or that there are fewer riders with their arms raised on the streets as they opt for
prearranged services instead. The earnings she can make for her daily rented investment may fall, as
she continues to pay rent for the cab (including the cost of the franchise right or medallion), but finds
it more difficult to fill her cab with passengers each day.
95.
Yet this cab driver also has new options. She can continue to pay close to $100 to rent
the cab each day (implicitly paying for the wear and tear as well as the franchise or medallion) or she
can rent a car without a medallion and find passengers using uberX.
which path is more attractive for her as a driver. Were she to continue to drive her cab, she might find
127
Her choice will depend on
10
that uberTAXI was a useful way for her to find prearranged rides and results in less idle time as she
11
can find prearranged fares when she is less successful finding potential passengers hailing.
12
taxi company from whom she rents her cab might also invest in its own prearrangement services and
13
make them available to her. Were she to use uberX to find passengers, she would need to meet the
14
regulatory requirements imposed on TNCs by the PUC,
15
Uber,
128
130
129
The
very few additional requirements by
and then she would receive a passenger referral from uberX as well as insurance coverage
16
17
127
Breeze, as one example, leases fuel-efficient cars that drivers can return at anytime after their first
month with just two weeks notice (Breeze, https://www.joinbreeze.com/, accessed on 7/3/15).
The cost to lease a 2014 or 2015 Prius through Breeze is a one-time $250 fee, $195 per week, and
$0.15 per mile for each mile over 30,000 a year (How It Works,
https://www.joinbreeze.com/how_it_works, accessed on 7/1/15).
128
Drivers have attested to the fact that uberTAXI helps them receive prearranged rides. For
example, Ernest Avetisyan contracted with Eco Taxi and used Uber to receive fares: After
signing up with Uber, I continued to contract with Eco Taxi and to provide taxi service in the
same manner as before (including paying a gate fee)By using the Uber Taxi service, I was
able to increase the number of fares I got, thereby increasing my gross revenue and profitability
(Declaration of Ernest Avetisyan, 5/19/15, 9).
129
The PUC requires drivers using their own vehicle to pass a background check, meet a zerotolerance policy on drugs and alcohol, be 21 years of age, have a valid California drivers license,
personal and commercial insurance etc. (California Public Utilities Commission, Decision
Adopting Rules and Regulations to Protect Public Safety while Allowing New Entrants to the
Transportation Industry, 9/19/13).
130
In addition to the requirements put forth by the PUC, Uber requires that the car is a four-door car
and is either from 2000 or newer, or from 2005 or newer, depending on the city. (Vehicle
Requirements, http://ubersupport.weebly.com/uberx-and-uberxl.html, accessed on 6/27/15).
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
43
CV 13-03826-EMC
for each job she takes driving a passenger, and in return she would pay 20 percent of the fare
collected from the passenger as a licensing fee.
competition from new drivers for taxi passengers and a potential reduction in the value of the taxi
medallion or franchise (i.e., its exclusive rights to pick up hailing passengers), she as a driver has the
new option to rent and drive a car that does not have the medallion or franchise fee associated with it.
131
96.
While the ease of prearrangement has led to
Consider Hiis Ilmi, who used to drive for a taxi company until his friends and other
taxi cab drivers told him about the Uber App; now he takes online classes and provides consulting
services to individuals who want to start their own business, with a focus on using the Uber App to
start a transportation business: I began driving a taxi cab for USA cab, then Yellow Cab, and finally
10
West Coast Cab... At the beginning of 2013, some of my friends and other cab drivers started using
11
the Uber [A]pp they could now drive their own cars and make more money than they had ever
12
made. He adds: The flexibility of using the Uber app has allowed me to enroll in online courses
13
through Arizona State University. In addition to going to school, I also started a second business
14
called Halabuur Enterprise (halabuur.com). My business provides consulting services to individuals
15
in the San Diego area on how to start a small business, with a focus on using the Uber [A]pp to start a
16
transportation business.
132
17
18
97.
Consider Kathy J. Robinson, who was first a taxi driver, then split her time between
driving for a taxi company and driving using Uber, and ultimately decided to drive only using Uber:
19
20
131
132
Declaration of Hiis Ilmi, 7/9/14, UBEOCO0000957984, 23, 13.
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
44
CV 13-03826-EMC
I drove with only DeSOTO for a few months, and then in mid-December I onboarded with Uber so I
could use the [U]berX platform to get fares. I would drive some days with my own car and get fares
through [U]berX, and then some days I would drive a DeSOTO cab. I drove with both for about five
months to decide which service was most profitable, but ultimately it was driving my own car and
using [U]berX because I didn't have that gate fee to pay.
133
6
7
4.
8
9
98.
Example 4: Truck driver-employee
Consider a driver who was an employee for a local food delivery company, but did not
drive passengers. His employer only does deliveries during the day, and he has been hired to do a 9
134
10
to 5 shift. He would have to fulfill various regulatory requirements,
11
many requirements on his time, including that he drives where he is told to and follows his scheduled
12
hours.
and the employer may place
135
13
99.
The existence of Uber and other lead generation platforms gives such a driver a new
14
option. A driver may value the flexibility of total freedom in his working hours or he may not. He
15
might be a single father who needed to be home in the mornings and in the evenings, but also needed
16
time off mid-day (to care for a relative, pursue his music career, etc.) and the ability to change his
17
18
133
19
134
Truck drivers need to have commercial drivers license (CDL): According to California CDL
requirements, applicants must be at least 18 years old, have a valid California driver's license,
pass a vision exam and obtain a health certificate. The U.S. Department of Transportation
(USDOT) regulates interstate transportation of cargo. If you wish to transport materials across
state lines, you must be at least 21 years old. All CDL applicants then need to pass a general
knowledge exam and a skills test, which includes a pre-trip vehicle inspection, a basic control test
and an on-road driving exam Additional knowledge and skills tests may be required depending
on the endorsements you need (California CDL License Requirements ,
http://www.besttruckingschools.com/cdl-requirements/california-cdl-license-requirements,
accessed on 7/3/15; see also CDL program information and how to apply for a license,
http://www.dmv.ca.gov/portal/dmv/?1dmy&urile=wcm:path:/dmv_content_en/dmv/dl/dl_info#C
DL, accessed on 6/27/15).
135
For example, Richard Clark Nelson states: Before I began operating my own business, I had
been employed as a commercial truck driver. My work as an employee/commercial truck driver
is not at all like my current business. For example, when I was an employee/commercial truck
driver, I reported to a dispatcher. I also had scheduled hours. I was told where to go by my former
employers. As a self-employed businessman, I no longer have any of those constraints
(Declaration of Richard Clark Nelson, 5/28/15, 3).
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
45
CV 13-03826-EMC
schedule when unexpected issues came up with his children. He might place a high value on the
ability to set daily his own hours even if it required driving his own car and paying his own expenses.
Were he to begin accepting ride referrals using the uberX platform and driving using his car, he
would need to meet the above described regulatory requirements imposed on TNCs by the PUC, and
very few additional requirements by Uber,
imposed on truck drivers.
insurance coverage for each job he takes driving a passenger, and in return he would pay to Uber
approximately 20 percent of the fare collected from the passenger as a licensing fee.
now choose to drive all day, or not at all, any day of the week.
10
100.
137
136
all of which are generally much simpler than those
He would then receive a passenger referral from uberX as well as
138
He could
However, a driver who did not value this flexibility would not experience the tradeoffs
11
in the same way. The driver who would choose to move from his existing employment to driving his
12
own car and using uberX to find passengers would be someone who valued the combination of
13
features that these types of driving jobs offered relative to his prior position as an employee driving a
14
delivery truck.
15
16
136
In addition to the requirements put forth by the PUC, Uber requires that the car is a 4-door car and
is either from 2000 or newer, or from 2005 or newer, depending on the city. (Vehicle
Requirements, http://ubersupport.weebly.com/uberx-and-uberxl.html, accessed on 6/27/15.)
137
As discussed above, truck drives have to, among other things, pass a vision exam and obtain a
health certificate (California CDL License Requirements ,
http://www.besttruckingschools.com/cdl-requirements/california-cdl-license-requirements,
accessed on 6/27/15; see also CDL program information and how to apply for a license,
http://www.dmv.ca.gov/portal/dmv/?1dmy&urile=wcm:path:/dmv_content_en/dmv/dl/dl_info#C
DL, accessed on 6/27/15) .
138
17
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
46
CV 13-03826-EMC
101.
Consider Sean Aldeguer, who worked as a commercial truck driver but switched over
to driving using Uber in order to have more flexibility and spend more time with his kids: At the
time, I was a commercial truck driver for an independent trucking company and a mobile paper
shredding company When I would take long hauls for the trucking company, I would have to travel
a few days a week and I wouldn't be home for my kids. I wanted to be more independent, wanted
something that paid more money, and wanted something that would give me more flexibility to spend
more time with my kids.
139
8
9
C.
10
Californians Who Were Not Working as Drivers Before the Existence of Uber
and Other Lead Generation Platforms
11
102.
The evidence I have reviewed and discussed above demonstrates that existing drivers
12
have found that Uber and other lead generation platforms have opened up additional options to them.
13
These options impact each persons work differently and may be valued differently depending on
14
each individuals circumstances. Importantly, though, Uber and other lead generation platforms have
15
also opened up work options for people who were not existing drivers. People who had previously
16
either not been working at all or were actively working in other non-driving jobs have started offering
17
driving services due to the flexibility and ease afforded by using the Uber App to find passengers. In
18
the United States in December 2014, over one-third (36 percent) of drivers were not actively looking
19
for a new job before using the Uber App.
20
December 2014, roughly 38 percent had no other job, 31 percent had a full-time job in addition to
21
driving referrals received through the Uber App, and 30 percent had a part-time job in addition to
22
driving referrals received through the Uber App.
140
Of drivers using the Uber App in the United States in
141
23
24
139
Declaration of Sean Aldeguer, 5/19/15, 2.
25
140
1/22/15, p. 10.
141
1/22/15, p. 10.
26
27
28
Gibson, Dunn &
Crutcher LLP
47
CV 13-03826-EMC
1.
Example 5: People employed in existing jobs who add work hours by providing
driving services using a TNC
103.
Consider an employee who likes her full-time job and has no intention of leaving. For
example, Amarachi Onuoha currently work[s] as a full-time Mobility Specialist at a hospital. In
order to make extra money [she] offer[s] transportation services as a driver through Uber (on the
uberX platform).
would have been leisure time or wasted time into additional earning power, while keeping her
existing employment. If she simply meets the above described regulatory requirements, and supplies
her own four-door car that is newer than 2000 or 2005, depending on the city,
142
The existence of uberX has created new options for this employee to turn what
143
she can take jobs
10
driving prearranged passengers. For example, a driver who uses Uber to find passengers may do so
11
in order to pick up a small number of jobs driving people while the driver is on her own way to and
12
from her full-time job.
13
situation. This is not an uncommon occurrence; in fact many drivers reported in declarations that
14
they had full-time employment outside of the car service industry.
15
advantages associated with receiving income both as an employee and as an independent contractor,
16
in particular with regards to saving for retirement. Employees are limited in the amount of retirement
17
savings they can make on a tax-deferred basis; additionally working as an independent contractor
144
Whether a driver chooses to do this or not will depend on her individual
145
Such individuals enjoy tax
18
19
20
21
142
Declaration of Amarachi Onuoha, 5/28/15, 2.
22
143
Uber requires that the car is a 4-door car and is either from 2000 or newer, or from 2005 or newer,
depending on the city. (Vehicle Requirements, http://ubersupport.weebly.com/uberx-anduberxl.html, accessed on 6/27/15).
144
Some drivers do drive passengers on their way to and from work. Declaration of Dario Grant,
5/27/15, 10; and Declaration of Paul Kirz, 5/26/15, 2.
145
For example, see Declaration of Carlos Erick Oliva, 5/26/15, 3; Declaration of Cliff Lyons,
5/26/15, 2; Declaration of Gonzalo A. Macias, 5/26/15, 3; Declaration of Alain Ntere Mani,
5/27/15, 3; Declaration of Rusty Patrick Shaw, 5/27/15, 3; Declaration of Lee Samantha
Faelnar Te, 5/26/15, 4; Declaration of Adrian Castillejos, 5/28/15, 2; Declaration of Amarachi
Onuoha, 5/28/15, 2; Declaration of Carlos Ramirez, 5/28/15, 5; and Declaration of Osama
Tamaro, 5/28/15, 2.
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
48
CV 13-03826-EMC
allows individuals to make tax-deferred retirement contributions using both their income as
employees and their incomes as independent contractors.
146
104.
Consider Dario Grant, who is a full-time employee of Los Angeles County, is
restricted in terms of jobs and hours he can work outside of this job, and drives using Uber in order to
fund his IRA and pay for his car: I work forty to fifty hours a week at [my] job. I'm an employee of
Los Angeles County. Los Angeles County restricts how much time I can work doing other things. I
can't work doing something where there is a conflict of interest. I also can't spend more than 24-
hours during a week working outside of my County employment I decided to drive using UberX
because I have two goals: 1) fund my IRA, which is $5500 a year; and 2) pay down my car which is
10
147
$4000 a year.
11
105.
Similarly, consider an employee who has found part-time work doing exactly what he
12
wants to do to build his skills (e.g., apprenticing as a carpenter, working in a theater, or working as a
13
property manager) but needs additional income to round out his earnings and may additionally value
14
flexibility. The existence of uberX has created new options for him. The flexibility of whether and
15
when to use his car to drive prearranged passengers means he can earn money taking driving jobs
16
during the exact hours that do not conflict with his existing employment. For example, Amir Afshar
17
Tavana explains: Uber does not restrict me from engaging in another occupation or business. In
18
addition to my transportation business, I have been working part-time for a marketing company for
19
the past five years.
20
Observatory in the operations department 3 days a week for 6 hours a day. Both my job at Griffith
148
Similarly, Shakeh Barseghian states: I have another part-time job at Griffith
21
22
23
146
Retirement Topics - 401(k) and Profit-Sharing Plan Contribution Limits, IRS, 10/23/14,
http://www.irs.gov/Retirement-Plans/Plan-Participant,-Employee/Retirement-Topics-401k-andProfit-Sharing-Plan-Contribution-Limits, accessed on 7/6/15; see also Ashlea Ebeling, How
Entrepreneurs Can Get Big Tax Breaks For Retirement Savings, Forbes, 3/4/13,
http://www.forbes.com/sites/ashleaebeling/2013/02/13/how-entrepreneurs-can-get-big-taxbreaks-for-retirement-savings/, accessed on 7/6/15.
27
147
Declaration of Dario Grant, 5/27/15, 79.
28
148
Declaration of Amir Afshar Tavana, 5/27/15, 4.
24
25
26
Gibson, Dunn &

Crutcher LLP
49
CV 13-03826-EMC
Observatory and Uber are okay with me having other jobs on the side. I receive a W2 from Griffith
Observatory and a 1099 from Uber.
149
106.
Consider Julia Jimenez, who is an employee of a fashion design agency, which
contracts her out to companies for specific jobs. She drives using Uber full-time when she is waiting
to be placed on a job and drives part-time once she is placed: In my work as a computer pattern
maker, I work as an employee for a fashion design agency, which contracts me out to other
companies for specific jobs that may last a couple months to a year My last job through the
fashion design agency just ended last week, and I am waiting to be placed in another full-time
position. While I wait to be placed, I am driving, using the Uber App to find customers full-time.
10
Once I receive another placement from the fashion design agency, I will go back to driving part-
11
time.
150
12
13
2.
14
107.
Example 6: People who have multiple independent contractor jobs
Consider an entrepreneurial person who has no employer, and uses the many different
15
internet platforms that connect people to jobs, taking the best options among these various jobs in real
16
time while foregoing anything less attractive. The existence of uberX has created one more set of
17
options for her, and she will take a driving job only when it is more valuable relative to the other
18
options she is considering in that moment. For example, Desirae Smidt is a stay at-home parent and
19
makes use of multiple websites and platforms to receive leads, including Uber, Care.com,
20
TaskRabbit, Thumbtack, and SitterCity. While she takes projects from each of these sources, most of
21
her leads come from Uber and Care.com.
151
22
108.
Consider Andrew Harris, who drives using Uber but also participates in multiple other
23
business ventures, including using delivery service apps and online platforms: I also use Sidecar
24
and Postmates, which are delivery service apps. However, I often log into more than one app at a
25
26
149
Declaration of Shakeh Barseghian, 5/26/15, 2.
27
150
Declaration of Julia Jimenez, 5/27/15, 2.
28
151
Declaration of Desirae Smidt, 5/28/15, 3.
Gibson, Dunn &

Crutcher LLP
50
CV 13-03826-EMC
time. I spend about half of [my] time [as a] driver for the Uber app [sic], and half of my time driving
for a combination of the Sidecar and Postmates apps. Uber does not prohibit me from using other
apps to find fares... In addition to transportation service, I also have other business ventures,
including buying and selling items on Ebay and Craigslist.
152
5
6
3.
109.
Example 7: People who were employed, but switched to taking driving jobs
Consider someone who was employed, similar to the employee driving a delivery
truck above, but did not find his current job satisfying. His employer may place a variety of
requirements on his time and work that he is not amenable to. The existence of uberX gives him a
10
new option. Does he value the flexibility of total freedom in his working hours? Does he have a
11
vehicle that is sitting idle? A person who switched from his existing employment to using uberX to
12
find passengers valued the combination of features of driving jobs he found from TNCs relative to his
13
prior employment position.
14
better means of earning money than the alternative employment available to him (including his prior
15
job).
153
16
110.
He would continue to choose to take driving jobs as long as they are a
Consider Christopher Jose Martinez, who was unhappy at his previous job at Coca-
17
Cola, and now enjoys driving full-time using the Uber App. When speaking of his Coca-Cola job he
18
says: I was often depressed when I was doing that job; I felt that my life lacked a purpose. He
19
started driving with Uber while still working for Coca-Cola. He then took the biggest risk of his
20
life; he quit his job and began driving full-time using Uber.
154
21
22
23
152
Declaration of Andrew Harris, 5/29/15, 3, 5.
24
153
For example, Masis Avdalyan states: I used to work at Access Services San Gabriel Transit. I
left the company because they put too much pressure on me to work their schedule. They don't
care what kind problems [sic] you might have at home or how hard you work I realized that
working for Uber, I could control my schedule and pick up my son when I needed to. Working
for Uber helped me avoid all of the stress of the scheduling at my old job. (Declaration of Masis
Avdalyan, 5/26/15, 8); see also Declaration of Aaron Parrish, 5/26/15, 4; and Declaration of
Majic Hindi, 5/28/15, 6.
154
Declaration of Christopher Jose Martinez, 5/27/15, 3.
25
26
27
28
Gibson, Dunn &
Crutcher LLP
51
CV 13-03826-EMC
4.
Example 8: People who were not working, but now are entering the labor
force and taking driving jobs
111.
Consider someone who was not working (e.g., a student, retiree, or someone between
jobs). The jobs available to earn money were not appealing or available for reasons that are unique to
each person. The existence of uberX gives each of these people a new option, which is to work the
exact amount she would like, and to have total freedom regarding when to start, stop and restart
working. Were someone to enter the labor force and use uberX to find passengers to drive using her
own car, it would be a sign that this option offered something new and of value. Such a person would
continue to choose to take driving jobs from uberX as long as they suit her need to earn money and
10
her taste for the activity. For example, the flexibility of driving using the Uber App for leads gives
11
retired individuals the opportunity to earn extra income, while still maintaining the flexibility to do
12
whatever they want.
13
them the time and flexibility to attend classes and do school work on their own schedule.
155
Similarly, Uber allows students to make some extra money while also giving
156
14
112.
Some drivers who use the Uber App are between jobs or are looking for jobs, as
15
shown in Exhibit 1. For example, William Gfroehrer is going through some life changes and is
16
using Uber while looking for other jobs in the meantime.
17
explains: I am between jobs right now and am actively looking for a purchasing position, possibly
18
with a manufacturing company. Right now I am just driving for [u]berX. If I get a purchasing job, Ill
19
probably just drive Uber on the weekends.
20
search process so that they do not have to take the first job they can find. The flexibility of the Uber
21
work flow is presumably particularly important in this regard, as an individual who obtains a job
157
158
Another driver, Adel Shallouf,
For such individuals, Uber may facilitate the job
22
155
For some examples of retired individuals who use Uber for the flexibility, see Declaration of
James Coltrane, 5/28/15, 2; and Declaration of Ray Lebron, 5/29/15, 12.
156
For example, Arshavir Steven Saryan states: I am a full-time student at Azusa Pacific University
and am earning my Bachelors degree in English and Communications. When I drive using the
Uber App, I usually drive in the mornings and weekend nights because that provides me the most
flexibility for school (Declaration of Arshavir Steven Saryan, 5/27/15, 67)
27
157
Declaration of William Gfroehrer, 5/18/15, 5
28
158
Declaration of Adel Shallouf, 5/26/15, 4.
23
24
25
26
Gibson, Dunn &

Crutcher LLP
52
CV 13-03826-EMC
interview would be able to stop driving and go to the interview on even short notice. In this way, the
availability of uberX driving jobs may complement unemployment insurance, allowing individuals to
eventually find the jobs that are best suited to their needs and abilities.
113.
Some drivers who use the Uber App may not have another job but they have other full
and part-time commitments, and drive around this commitment. Some drivers have to attend to
family needs. Janice Fry explains: Im a full-time mom. I have a two-year-old son and a five-year-
old son. I drive for Uber when I have a babysitter. My friend from church normally babysits for me.
When hes babysitting the kids, I can drive.
value the flexible work schedule due to restrictive personal commitments.
159
10
114.
Exhibit 2 shows additional examples of drivers who
Some individuals have emerged from retirement to work as drivers. For instance,
11
Mike Evans explains: I retired five years ago from my job working with Farmers Insurance. I was
12
my own boss and worked my own hours, which is very similar to how I use the Uber app. My
13
daughter recommended that I look into driving with Uber to help me get over the boredom that
14
follows retirement. I enjoy speaking with the passengers and getting out of the house. Honestly,
15
driving with Uber has saved my sanity.
16
Becoming a driver for Uber appealed to me because it suits my lifestyle: I am retired and only want
17
to work when I feel like it. I also thought that becoming an Uber driver would provide me with
18
pocket money for holidays.
160
Another driver, Bhupinder Singh Bhangal, states:
161
19
20
21
22
23
24
25
26
159
Declaration of Janice Fry, 5/26/15, 56.
27
160
Declaration of Mike Evans, 5/27/15, 2.
28
161
Declaration of Bhupinder Singh Bhangal, 3.
Gibson, Dunn &

Crutcher LLP
53
CV 13-03826-EMC
1
2
EXHIBIT1
EXAMPLESOFDRIVERSWHOUSETHEUBERAPPBETWEENJOBS
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Source:DeclarationofAdelShallouf,5/26/15,4;DeclarationofSangwaEmmanuel,5/20/15,3;DeclarationofSheriKellogg,5/18/15,2;
DeclarationofWilliamGfroehrer,5/18/15,5;DeclarationofVictorinoCaballero,5/26/15,3;DeclarationofAlexander"AJ"Galloway,5/27/15,4;
DeclarationofJuliaJimenez,5/27/15,2;DeclarationofCharlesSkipClose,5/28/15,3;DeclarationofRichardOrtega,5/26/15,2;Declarationof
DanaAlexander,5/26/15,2.
26
27
28
Gibson, Dunn &
Crutcher LLP
54
CV 13-03826-EMC
1
2
3
EXHIBIT2
EXAMPLESOFDRIVERSUSINGTHEUBERAPPWHOVALUEAFLEXIBLEWORKSCHEDULEDUETO
RESTRICTIVEPERSONALCOMMITMENTS
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Source:DeclarationofHiisIlmi,7/9/14,UBEOCO0000907984at82,13;DeclarationofDonR.Esperon,5/20/15,2;DeclarationofMohammad
Sattar,5/18/15,10;DeclarationofSamirMoussa,5/19/15,8;DeclarationofSelengePurevThompson,5/20/15,5;DeclarationofTesfayeHaile,
5/19/15,5;DeclarationofSauKuenChu,5/19/15,6;DeclarationofJohnA.Bischop,5/26/15,8;DeclarationofDavidL.Henning,5/18/15,2;
DeclarationofHenryOdiase,5/20/15,9.
26
27
28
Gibson, Dunn &
Crutcher LLP
55
CV 13-03826-EMC
D.
Small Businesses That Did Not Exist Before Uber
115.
Californians who have used Uber and other lead generation platforms to find one or
more driving jobs are not the only ones who have found opportunity and choice in the creation of
lead generation platforms. Entrepreneurial activities have developed due to these platforms.
example, and as described above, if a driver would like to start his own driving business, he no longer
needs to invest in a medallion or an expensive dispatch system. Additionally, if one finds driving
jobs through Uber or other lead generation platforms, the flexibility in work schedule can support
other entrepreneurial efforts. For example, Ali Bouhelal worked as an independent contractor for a
multi-driver business company using Uber before starting one himself. He stated: I used my time
162
For
10
employed with this company to familiarize myself with the Uber business model, as well as to have a
11
source of income while I began building my own business. While driving for California Limousine
12
Service, I began applying for and processing all the permit, requirements, and licenses I needed to
13
start a driving company.
14
used to be drivers of multi-driver businesses.
163
Ali Bouhelal is one of many owners of multi-driver businesses who

164
15
16
E.
17
App and Outside Options
18
19
Drivers Using the Uber App Have Different Backgrounds, Reasons for Using the
116.
There has been a large increase in the number of Californians finding driving jobs
using the Uber App. This is seen clearly in Exhibit 3, which shows the number of active drivers
20
162
Other examples of lead generation platforms are HopSkipDrive and Shuddle, two ride services
designed specifically for transporting children to and from school and different activities. Both
services are styled after Lyft and Uber, utilizing independent contractors as drivers, many of
whom are mothers and bring their own children along on rides. Claire Martin, A Ride StartUp for the After-School Set, New York Times, Sunday 6/28/15 p. B3.
24
163
Declaration of Ali Bouhelal, 8/5/14, UBE-OCO0000902025, 34.
25
164
Daniel Bisneto used to work for Jet Cars limousines as a driver and subsequently started his own
driving services business Eternity Limousine Service (Declaration of Daniel Bisneto, 9/4/14,
UBE-OCO0000906873, 35). Christopher R. Edwards used to work as an independent
contractor for a small transportation company but then decided to open his own business; he
explains: In February 2012, I decided to end my contract with the transportation company and
to open my own business with a partner, which at that time was Green Driver (Declaration of
Christopher R. Edwards, 10/7/14, UBE-OCO0000906367, 4).
21
22
23
26
27
28
Gibson, Dunn &
Crutcher LLP
56
CV 13-03826-EMC
using the Uber App between June 2010 and June 2015. But as shown above, the circumstances in
which any given driver chooses to take driving jobs are very different.
3
4
5
EXHIBIT3
THENUMBEROFDRIVERSWHOCOMPLETEDATLEASTONERIDEUSINGUBERX,UBERBLACK,
ORUBERSUVBETWEENJUNE2010ANDJUNE2015
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Source:UberMonthlyDriverData,CaliforniaDrivers
20
21
117.
In short, there is no typical Uber driver. Drivers have different education levels,
166
some are employees in other jobs,
167
some are independent contractors in other work,
165
and some
22
23
24
165
1/22/15, Table 1, p. 8.
166
For example, see Declaration of Lakisha Evans, 5/26/15, 2; Declaration of Thu-hong Catellier,
5/29/15, 2; and Declaration of Harpreet S. Gandhi, 5/19/15, 4.
167
For example, see Declaration of Dennis Saenz, 5/18/15, 5; Declaration of Alex Duque, 5/15, 5;
and Declaration of Corey Green, 5/27/15, 7.
25
26
27
28
Gibson, Dunn &
Crutcher LLP
57
CV 13-03826-EMC
168
only take jobs from lead generation platforms, such as Uber.
the course of a day,
in circumstance, reasons for and use of the platform all reflect the differences in the outside options
each driver has when she chooses to take or not take a driving job she found on the Uber App at any
given point in time.
118.
169
170
week,
or month,
171
Some drivers take very few jobs over

172
while some take jobs all day long.
These differences
173
In summary, in this section I have described the many different types of drivers who
use the Uber platforms. These drivers have many different work arrangements. Many drivers have
full or part-time jobs in addition to driving. The way in which drivers use the Uber platforms are
very different; some drivers contract directly with Uber and drive referrals from uberX using their
10
personal cars, while others work, as employees or independent contractors, for a transportation
11
company that then contracts directly with Uber. Moreover, drivers can, and do, drive referrals from
12
more than one Uber platform, creating a wide variety of possible permutations. For example, drivers
13
who usually use the UberBLACK platform may also use the uberX platform; while drivers who
14
usually use the UberSUV platform may also use the UberBLACK and uberX platforms; drivers may
15
also use many of Ubers other platforms, such as UberPOOL, UberSelect, UberLUX, and uberXL. In
16
February 2015, of drivers who completed at least one ride using uberX, UberBLACK, or UberSUV,
17
52 percent also drove at least one ride through these other platforms.
18
in this section show that there is no typical driver using the Uber App.
174
All the differences described
19
20
21
22
23
24
25
168
For example, see Declaration of Paul Cotter, 5/18/15, 4; Declaration of Alberto Dico Jr.,
5/26/15, 4; and Declaration of Kathy J. Robinson, 5/20/15, 2.
169
For example, see Declaration of Bo Lee, 5/19/15, 8.
170
For example, see Declaration of Margalit Lipski, 5/26/15, 3.
171
For example, see Declaration of Chris Keller, 5/19/15, 7.
172
For example, see Declaration of Naveed Zafar, 5/19/15, 8.
173
For some different examples of prior work experience, see Declaration of Casey Cheung, 5/19/15,
3; Declaration of Mike Evans, 5/27/15, 2; Declaration of Christopher Jose Martinez, 5/27/15,
3; Declaration of John Robert Lehman, Jr., 5/19/15, 2; Declaration of Elias Negash, 5/19/15,
34; and Declaration of Chris Keller, 5/19/15, 3.
174
Uber Monthly Trip Data, California Drivers.
26
27
28
Gibson, Dunn &
Crutcher LLP
58
CV 13-03826-EMC
VI.
UNIQUE AND INDEPENDENT
THE RELATIONSHIP BETWEEN EACH DRIVER AND THE UBER APP IS
3
4
119.
In the previous section, I described the many different ways in which workers and
their work options are affected by the Uber App. These many differences make clear that the Uber
App is used by many different types of individuals for many different reasons, and that as a result,
there is no typical driver using the Uber App. These differences are not unexpected given that for
many dimensions of the work, Uber has no policy, or to the extent it does have policies, they differ
across the many licensing agreements used by Uber.
10
11
A.
Drivers Who Use the Uber App Are Not Required to Drive Any Particular Time,
12
Amount or Place Creating Highly Individualized Work Circumstances for Each
13
Driver
14
15
1.
16
120.
Each ride is a new job
Drivers go through a number of steps in order to ultimately use the Uber App to
17
receive, and drive, a passenger referral. Each driver signs up to use the Uber App. Then she must
18
decide whether and when to turn the App on; which platform to use; and whether and when to take a
19
driving job referral.
20
121.
It is my understanding that drivers do not need to pay to sign up for the Uber App, and
175
21
they are not paid by Uber to sign up for the App.
22
order to be eligible to receive passenger referrals from the Uber App. For example, the PUC requires
23
that drivers using TNC platforms comply with a zero-tolerance policy on drugs and alcohol and
Drivers do need to meet specific requirements in
24
25
26
27
28
175
I understand that in 2014 in San Francisco, Uber required drivers to pay $40 for a 3 hour training
class for TCP drivers to start using UberBlack or UberX (Email from Uber San Francisco, Uber
SF: New Uber Black Training UPDATE, 7/9/14, COLOPY0009069 at COLOPY000907).
Gibson, Dunn &

Crutcher LLP
59
CV 13-03826-EMC
176
successfully pass a criminal background check, and their vehicle must pass a 19-point inspection.
Uber has its own requirements, but because Uber has used many different licensing agreements,
many of these requirements vary by city and have changed over time. For example, Uber may
require that a vehicle is no more than a certain number of years old and is in good condition, but the
specifics vary by city.
177
122.
One could imagine a policy whereby after signing up for the Uber service, the driver
would be committed to turn on the platform at certain times or for a certain duration in order to
receive and accept job referrals. However, it is my understanding that no such uniform policy
exists.
178
A driver is not expected to be on call at any particular time, and Uber would not contact a
10
driver to demand that she turn the platform on. Although Uber occasionally notifies drivers through
11
texts or emails when there are major events occurring or locations that are busy, it is entirely within
12
the drivers discretion to decide whether to turn on the App or drive in those areas. For example,
13
driver Barry Perelman states, Uber texts me about how to make more money, by driving in
14
15
176
Public Safety while Allowing New Entrants to the Transportation Industry, 9/19/13, pp. 2629;
California Public Utilities Commission, CPUC Establishes Rules for Transportation Network
Companies, Press Release, 9/19/13.
177
Vehicle Requirements, http://ubersupport.weebly.com/uberx-and-uberxl.html, accessed on

6/27/15. See also Rasier Transportation Provider Service Agreement, 5/15/13, p. 4, (You agree
that you shall maintain a vehicle that is a model approved by the Company. Any such vehicle
shall be no more than ten (10) model years old, and shall be in good operating condition.). Other
agreements do not include this requirement. See Uber Software License and Online Services
Agreement, 7/24/13.
178
I understand that some licensing agreements do not provide requirements that govern whether a
driver needs to turn on the App or accept a ride to maintain an active App account. For example,
see Uber Terms and Conditions, 8/11. I also understand that in some instances Uber may
deactivate a drivers account after a long period of inactivity and that the period of inactivity prior
to which a driver may be deactivated varies across licensing agreements. For example, see Rasier
Software License and Online Services Agreement, 12/10/13, p. 9 (The Agreement shall be
automatically terminated for inactivity of more than 180 days, with the date of termination being
the 180th day following the date of the last Request accepted and performed by you.) and Rasier
Software License and Online Services Agreement, 11/10/14, p. 3 (You acknowledge and agree
that you are required to fulfill a request for Transportation Services using the Driver App at least
once a month to maintain an active Driver profile, and Company reserves the right to deactivate
your Driver ID if you have not fulfilled a request for Transportation Services using the Driver
App at least once a month.). I also understand that if the driver is in compliance with the
requirements for accessing the Uber platform (e.g., the PUC requirement that the driver have
insurance coverage), then Uber will reactivate the account upon request.
16
17
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
60
CV 13-03826-EMC
Downtown or West Hollywood, places that are late night areas where they are trying to encourage me
to pick up fares. I ignore those text messages, because I will not pick up fares past 9:30 at night, and
I will only do that in my area so I ensure I will not have to pick up young folks partying. I go offline
at these times by turning off the app.
text me that certain areas are busy, but I don't have to drive in those areas if I don't want to.
a driver can be signed up for the platform and essentially never choose to drive a single passenger
referral (i.e., never do a single job).
179
8
9
123.
Another driver, Victor Mejia says, Uber will sometimes

180
Thus,
When a driver does choose to turn the App on, she does not pay, nor is she paid, for
App on time.
181
A driver could be logged on to the Uber App (and multiple other TNC platforms)
10
and receive a referral while she is sitting in a coffee shop visiting with friends or at the track running
11
laps. There are no requirements for a drivers behavior when the platform is on, but the driver has
12
not yet accepted a passenger referral.
182
13
124.
When a driver is logged on to the platform she has the ability to accept or decline
14
passenger referrals (i.e. accept or decline the job). My understanding is that Uber even spells this out
15
directly in different ways in different contracts. For example:
16
17
You shall be entitled to accept, reject, and select among the Requests received via the
183
Service. You shall have no obligation to the Company to accept any Request.
18
19
20
179
Declaration of Barry Allan Perelman, 5/26/15, 7.
21
180
Declaration of Victor Mejia, 5/26/15, 3.
22
181
I understand that at certain points in time in certain cities, Uber has offered temporary promotions
in which Uber offered to guarantee drivers a certain minimum amount of money if they surpassed
certain thresholds for time using the Uber platform. For example, see Email from Uber to
Matthew Manahan, Thank You For Partnering With Uber | Earn Even More This Weekend,
9/26/14, MAN0007723 at MAN000772.
182
I understand that if a driver is logged on to the platform and has not accepted a ride for an
extended period of time, the platform may automatically log off.
183
Rasier Transportation Provider Service Agreement, 5/15/13, pp.12. This version of the contract
even states explicitly: Each Request that you accept shall constitute a separate contractual
engagement. Ubers policy on the ability to reject requests has evolved over time and in fact it is
now stated in the contracts that the company has a cancellation policy that can be updated as often
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
61
CV 13-03826-EMC
125.
Were a driver to accept a driving job from a passenger referral, Uber has no
184
expectation or requirement that she will take another one after finishing that one.
is its own job.
Thus, each ride
4
5
2.
126.
Drivers control when and where to take jobs.
One could imagine a policy that controls drivers schedules, including how many
hours they drive during a given day and at what time. It is my understanding that no such policy
exists. Rather, drivers using the Uber App decide when and where to take jobs. At all times, each
driver can choose whether he wants to look for rides, accept rides, complete the ride, or not. This
10
flexibility is true across minutes, hours, days, weeks, and months, as I show below. Drivers can
11
decide to take an extended time off from using the Uber App to study, vacation, or pursue other
12
endeavors. While using the App, drivers can reject rides, and do so to varying degrees. Furthermore,
13
drivers implement their own strategy for when and where to turn the App on and when and where to
14
accept rides.
15
127.
In their declarations, many drivers state that driving with Uber gives them the freedom
185
16
to choose their own schedule.
17
examples of the many different ways that drivers use this flexibility to choose the hours they want to
18
drive. Some drivers have a specific full-time schedule in which they drive passengers. For example,
Declarations of drivers who use the Uber App also provide
19
20
as the company chooses. For example, see Rasier Software License and Online Services
Agreement, 11/10/14, 2.4. (You retain the option, via the Driver App, to attempt to accept or to
decline or ignore a Users request for Transportation Services via the Uber Services, or to cancel
an accepted request for Transportation Services via the Driver App, subject to Companys thencurrent cancellation policies.).
21
22
23
184
The agreements between Uber and different drivers make this point explicit in a variety of
different ways. For example, see Rasier Transportation Provider Service Agreement, 5/15/13, p.
2 (The Company will offer the Service to you during those times you choose to be available to
receive the Requests. You shall have no obligation to use the Service at any specific time or for
any specific duration. You shall have complete discretion to determine when you will be available
to receive the Requests.) and Rasier Software License and Online Services Agreement, 11/10/14,
2.4 (You retain the sole right to determine when and for how long you will utilize the Driver
App or the Uber Services.).
24
25
26
27
28
185
See Declaration of Masis Avdalyan, 5/26/15, 67; Declaration of Joseph Elias Saker, 5/29/15,
1213.
Gibson, Dunn &

Crutcher LLP
62
CV 13-03826-EMC
Jiawen Ling explains when she is driving using the Uber App: I start work around 7:30 a.m. and
8:00 a.m I will work until 6 or 7 p.m., but I will take a break around noon to rest.
concentrate their work on the weekends; for example, Norman Paul Sternfeld explains that he
typically drive[s] only on Saturday and Sunday, eight to ten hours each day, starting around 4am and
ending around noon or 2pm.
Maria Khaira states: I have no specific schedule. I always work Friday nights, Saturdays from 2
p.m. or 3 p.m. to 1 a.m. or 2 a.m. Sundays, I drive after church for 8-10 hours. But it varies and
depends. I drive when I want.
186
187
Other drivers
Yet other drivers choose a more ad hoc schedule; for example,
188
128.
The amount each driver chooses to work can also vary from week to week or month to
10
month. Again, driver declarations provide context for how drivers might use this flexibility to adjust
11
their schedules over time. For example, some drivers change the amount of time they drive based on
12
family needs; Demes Mulat explains that his hours with Uber depend on his familys schedule: I
13
don't have a set schedule for when I drive using the Uber App. It depends on whether my sons are in
14
school or not, if my wife is available to care for them or if I have to care for them, it just depends.
15
Ronie Canlas sets her hours to match up with a changing schedule of her fulltime job: My schedule
16
at the animal shelter changes every 3 months or so, which is another reason I really like to drive
17
because I can just adjust my own hours.
189
190
18
129.
I show that there is no typical approach taken by drivers regarding when to take
19
passenger referrals from Uber. There is enormous variability in the number of jobs that drivers using
20
the Uber App choose to take each month or day. Exhibit 4 shows, for each driver who completed at
21
least one ride using the Uber App in California during February 2015, the number of rides completed
22
during that month. The number by each driver ranged from 1 to 1,076. As Exhibit 4 shows, many
23
24
186
Declaration of Jiawen Ling, 5/19/15, 8.
25
187
Declaration of Norman Paul Sternfeld, 5/19/15, 7.
26
188
Declaration of Maria Khaira, 5/20/15, 4.
27
189
Declaration of Demes Mulat, 5/18/15, 9.
28
190
Declaration of Ronie Canlas, 5/20/15, 45.
Gibson, Dunn &

Crutcher LLP
63
CV 13-03826-EMC
drivers drove a small number of trips, such as 20 or fewer, and a few drivers drove a larger number of
trips.
3
EXHIBIT4
DISTRIBUTIONOFTHENUMBEROFRIDESCOMPLETEDBYCALIFORNIADRIVERS
INFEBRUARY2015
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Source:UberMonthlyTripData,CaliforniaDrivers
18
19
Note:AnalysisincludesCaliforniadriverswhocompletedatleastonerideusinguberX,UberBLACK,orUberSUVinFeburary2015.493outof50,346
drivershavecompleted500ormorerides.Theseindividualsarenotshowninthishistogrambutareusedinthecalculationofthedistribution.
20
130.
Drivers also use this flexibility to choose not to use the Uber App for extended periods
21
of time to accommodate unexpected commitments, study, vacation, or pursue other endeavors. For
22
example, Katie Guilfoyle took time off to care for sick family members: Uber doesn't require that I
23
work a set schedule or a specified number of hours. In fact, I took a couple months off of driving
24
191
when my daughter was sick, and there were no repercussions.
Thomas Walding took time off to
25
travel: Today was the first time in 3 weeks that I used the Uber app because I have been out of the
26
27
28
191
Declaration of Kate Guilfoyle, 5/18/15, 7.
Gibson, Dunn &

Crutcher LLP
64
CV 13-03826-EMC
country. I use the Uber app for additional income that lets me drop off the face of the earth for a few
weeks at a time. Uber has not suspended my account or threatened to suspend my account because of
the amount I drive.
131.
192
Exhibit 5 shows the number of rides completed each month from April 2014 through
March 2015 for ten drivers randomly selected from all drivers who completed at least one ride using
uberX, UberBLACK, or UberSUV over the course of the 12 month period (identified as Drivers A
J).
jobs. For some drivers, the number of rides completed per month is consistent; some others use the
Uber App only sporadically, while others have unpredictable periods of low use. Driver I, for
193
It is clear that drivers vary considerably in the amount they use the Uber App to find and take
10
example, worked steadily throughout the twelve month period, completing between 300 and 450
11
rides in each month. Some other drivers had more sporadic patterns in driving. For example, Driver
12
G was steadily accepting and completing hundreds of rides during the Spring, but stopped working
13
on rides entirely between August 2014 and December 2014. This driver then started again in January
14
2014 and completed over 100 rides, but then stopped in February and March 2015. Driver E
15
completed rides in eight out of twelve months, but never completed any more than 11 rides in any
16
given month. This is consistent with someone who spends most of their time engaged in activity
17
other than driving using the Uber App. The evidence presented in Exhibit 5 is consistent with the
18
evidence from the declarations provided above. Similar exhibits showing data for additional random
19
samples of drivers can be found in Appendix C.
20
21
22
23
24
25
26
27
28
192
Declaration of Thomas Walding, 5/18/15, 7.
193
Drivers A E were selected randomly from drivers whose total number of trips completed
between April 2014 and March 2015 was below the median among drivers who completed at
least one ride using uberX, UberBLACK, or UberSUV in April 2014. Drivers F J were selected
randomly from drivers whose total number of trips completed between April 2014 and March
2015 was above the median among drivers who completed at least one ride using uberX,
UberBLACK, or UberSUV in April 2014.
Gibson, Dunn &

Crutcher LLP
65
CV 13-03826-EMC
1
2
3
EXHIBIT5
NUMBEROFRIDESCOMPLETEDEACHMONTHFROMAPRIL2014TOMARCH2015FORASAMPLE
OFCALIFORNIADRIVERS
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Source:UberMonthlyTripData,CaliforniaDrivers.
24
Note:DriversAE(DriversFJ)wereselectedrandomlyfromdriverswhosetotalnumberoftripscompletedbetweenApril2014andMarch2015
wasbelow(above)themedianamongdriverswhocompletedatleastonerideusinguberX,UberBLACK,orUberSUVinApril2014.
25
26
132.
Exhibit 5 provides evidence of the variability in work patterns observed in a random
27
sample of drivers. However, in some circumstances, including the present one, a random sample is
28
not able to capture the full breadth of the heterogeneity in the data and it is helpful to display directly
Gibson, Dunn &

Crutcher LLP
66
CV 13-03826-EMC
the range of variation in the data. I do this in Exhibit 6. Exhibit 6 shows the number of rides
completed each month from April 2014 through March 2015 for five selected drivers (a different set
from those described above but also identified as Drivers AE). These drivers were chosen
specifically to highlight the substantial variability in the patterns of work flow for drivers using the
Uber App.
month to month. Although this driver was busier in some months than others, he or she completed
between 57 and 220 rides in each month. On the other hand, consider Driver E, who displays very
high variability in the number of rides completed each month. From April 2014 through September
2014, this driver completed more than 800 rides in each month with the exception of July, in which
10
he or she completed over 480 rides. Yet, in October, Driver E only completed 55 rides and did not
11
work on rides at all from November 2014 through January 2015, but then picked back up in February
12
2015, completing almost 700 rides in February and over 770 rides in March. In addition, consider
13
Driver B, who had a steady pattern of driving work throughout the twelve month period. This driver
14
completed between 2 and 30 rides per month. In some months, Driver E was completing between 50
15
and 100 times more rides than Driver B.
194
Driver C displays a relatively consistent pattern in the number of rides completed from
16
17
18
19
20
21
22
23
24
25
26
27
28
194
The analysis is limited to drivers who completed at least one ride using uberX, UberBLACK, or
UberSUV in April 2014. For each driver, I calculated the variance across months in the number
of completed rides. I then identified five drivers according to this driver specific variance.
Driver A has the minimum observed variance; Driver B has variance in the 10th percentile;
Driver C has the median variance; Driver D has variance in the 90th percentile; and Driver E has
the maximum observed variance.
Gibson, Dunn &

Crutcher LLP
67
CV 13-03826-EMC
1
2
3
EXHIBIT6
NUMBEROFRIDESCOMPLETEDEACHMONTHFROMAPRIL2014TOMARCH2015
FORSELECTEDCALIFORNIADRIVERS
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Note:AnalysislimitedtodriverswhocompletedatleastonerideusinguberX,UberBLACK,orUberSUVinApril2014.Foreachdriver,Icalculatedthe
th
varianceacrossmonthsinthenumbercompletedrides.DriverAhastheminimumobservedvariance;DriverBhasvarianceinthe10 percentile;
th
DriverChasthemedianvariance;DriverDhasvarianceinthe90 percentile;andDriverEhasthemaximumobservedvariance.
27
28
Gibson, Dunn &
Crutcher LLP
68
CV 13-03826-EMC
133.
Drivers using the Uber App are also able to choose when they want to work during
any given day. Drivers can turn on the Uber App whenever they want to contemplate accepting ride
referrals and are free to turn it off whenever they are not interested in potential jobs (i.e., to work at
another job, care for and pick up children, end their working day, etc.). How each driver uses this
flexibility will vary from driver to driver and depends on individualized factors including the other
work options available to them and personal commitments.
134.
Exhibit 7 highlights this variation. Exhibit 7 plots data from February 3, 2015, for five
drivers sampled at random and shows the number of ride referrals accepted by each driver during
each hour of the day. Driver A accepted four referrals for rides on February 3, 2015. One ride was in
10
the 10 oclock hour, another was in the 12 oclock hour, and two were in the 2 oclock hour. This
11
driver may have been accepting ride referrals between running errands, and the temporal pattern
12
might be consistent with dropping children off and picking them up from school. Driver B drove
13
more frequently, seemingly focusing on accepting referrals for rides after the morning rush hour and
14
after the evening rush hour, with a long gap in the middle of the day from the 1 oclock hour to the 6
15
oclock hour. Driver C had a night owl pattern of driving from the midnight hour until the early
16
hours of the morning. This pattern might be consistent with working multiple jobs. Driver D
17
accepted three ride referrals in the 4 oclock hour in the afternoon, and no others. This might be
18
consistent with a students schedule. Finally, Driver E began driving in the late afternoon, stopped
19
accepting referrals for two hours around the early dinner hour, and then accepted ride referrals
20
consistently from the 6 oclock hour until the 11 oclock hour. Similar exhibits showing data for
21
additional random samples of drivers can be found in Appendix D.
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
69
CV 13-03826-EMC
1
2
3
EXHIBIT7
NUMBEROFRIDESCOMPLETEDEACHHOURBYASAMPLEOFCALIFORNIADRIVERS
ONFEBRUARY3,2015
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Source:UberHourlyTripData,CaliforniaDrivers
24
Note:DriverswereselectedrandomlyfromdriverswhocompletedatleastonetripusinguberX, UberBLACK, or UberSUVon2/3/15.
25
26
27
135.
Once a driver has decided to log onto the Uber App, she still retains control over what
jobs she wants to take. While using the Uber App, drivers are not required to accept every referral
28
Gibson, Dunn &
Crutcher LLP
70
CV 13-03826-EMC
195
offered by Uber.
to work: that is, drivers exercise choice both regarding whether and when to have the Uber App
turned on, and also regarding whether to accept referrals made to them while the App is turned on.
The reasons a driver may decide not to accept a given ride are individualized and depend upon
individual circumstances and preferences. For example, some drivers reject rides because the
potential passenger has a low rating. Sharon Collins states: While Uber provides me with leads for
passengers, I am not required to accept the leads. I rejected a lead at 11:30 p.m. one evening where
the male passenger had a two star rating.... I value the rating system, because it allows me to decide
if there is a passenger that I prefer not to do business with.
This is yet another way in which drivers who use the Uber App can choose when
196
Yet other drivers reject rides because
10
they prefer not to provide the particular type of service requested by the passenger; for example,
11
Sevuloni Ratabumusu sometimes declines rides when he is too far away from the prospective
12
passenger or when an Uber Pool is requested.
197
13
136.
Exhibit 8 shows the variation in acceptance rates during February 2015 for drivers in
14
California who completed at least one ride using uberX, UberBLACK, or UberSUV during the
15
month. Rounded acceptance rates ranged from 0 percent to 100 percent.
16
17
18
19
20
21
22
23
195
For example, see Declaration of Juan Hurtado, 5/20/15, 14; Declaration of Juliana Bushore,
5/20/15, 18; Declaration of Philippe Lestienne, 5/18/15, 17; Declaration of John Hoffman,
5/27/15, 9; Declaration of Jeffrey Green, 5/26/15, 12; Declaration of Ali Bouhelal, 8/5/14,
UBE-OC000009020 25, 22; Declaration of Casey Cheung, 5/19/15, 9; Declaration of
Kimberly Henderson Strother, 5/20/15, 4; Declaration of Peter Marinakis, 5/20/15, 11;
Declaration of Tom Hakim, 5/28/15, 12; and Declaration of Leonardo Bertucci, 5/25/15, 3.
27
196
Declaration of Sharon Collins, 5/28/15, 12.
28
197
24
25
26
Gibson, Dunn &

Crutcher LLP
71
CV 13-03826-EMC
EXHIBIT8
LEADACCEPTANCERATESVARYWIDELYACROSSCALIFORNIADRIVERS
FEBRUARY2015
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Note:368outof50,313drivershaveanacceptanceratebelow40%.Theseindividualsarenotshowninthishistogrambutareusedinthecalculation
ofthedistribution.33driverswereexcludedfromthisanalysisbecausetheirroundedacceptancerateswere0ononeormoreplatforms.
17
137.
18
A big part of Ubers value as a platform matching passengers to drivers can be
19
attributed to its efficiency. The efficiency is better maintained if drivers who have the Uber App
20
running are interested in driving a potential referral. There are many reasons for this including wait
21
time and pricing. Wait times for passengers and drivers alike are kept down when the only drivers
22
logged on to the platform are those who actually want driving jobs. Uber gives every contacted
23
driver 15 seconds to decide whether he or she wants to accept a lead or not.
24
drivers who are not actively looking for jobs had inadvertently kept the App open, wait times could
25
be much higher than the estimate originally shown to the passenger given the number of nearby
26
drivers on the platform at the time the passenger opened the Uber App. As a result, prices may
27
28
198
198
If a large number of
Email from Uber San Francisco to Douglas OConnor, Uber SF: Update to Auto Logoff,
2/26/15, UBE-OCO000001702 at UBE-OCO000001701.
Gibson, Dunn &

Crutcher LLP
72
CV 13-03826-EMC
increase for rides in geographic areas in which there are fewer potential drivers for a given set of
passenger requests (i.e., surge pricing).
passenger demand in their local area (i.e., passengers are looking for rides in a low driver supply
area), and incentivizes drivers from other geographies to come to that area to drive.
199
138.
This both advantages drivers when there is higher
I understand there is some evidence that Uber has warned some drivers that their
accounts were at risk of being deactivated when the drivers had a habit of keeping the platform on,
but did not accept many rides.
drivers not to have the platform open and running when they are not looking for rides. This improves
the efficiency of the platform for passengers and drivers alike. When drivers are not interested in
200
These notifications are consistent with Uber seeking to incentivize
10
finding driving jobs they have a simple option: turn the App off. A driver can do this for any amount
11
of time: 10 minutes, an entire day, a week, or a year.
201
12
139.
While driving, drivers have different strategies for when and where to log onto the
13
Uber App. For example, Anthony Braxton drives in a certain neighborhood at a certain time of day:
14
I have a strategy for where I choose to log on to the Uber App. I live in the Mission District, so I
15
usually log on there because I don't have to go very far for a fare. In the afternoon, I head to the
16
Financial District because I know I can find a passenger who wants to go home from work. I drive
17
18
19
199
Uber SF states: As you know, surge pricing raises Ubers rates when the demand for cars is
greater than the supply on the road (Email from Uber SF to Douglas OConnor, BUSY
Weekend in SF, 3/8/13, UBE-OCO000001834 at UBE-OCO00000183). The challenge of
dealing with the supply/demand balance is not unique to lead generation platforms. Severin
Borenstein writes about the variable supply/demand challenge in the electricity market: Over
the last few years, a great deal has been written about time-varying retail pricing of electricity.
Many authors, myself included, have argued that real time retail electricity pricing (RTP) retail
prices that change very frequently, e.g., hourly, to reflect changes in the markets supply/demand
balance is a critical component of an efficient restructured electricity market (Borenstein,
Severin, The Long-Run Efficiency of Real-Time Electricity Pricing, The Energy Journal,
26(2), 2005, pp.124 at p. 1).
200
See Declaration of Helder Parreira, 11/17/14, 4.
201
I understand that in some instances Uber may deactivate a drivers account after long periods of
inactivity. I also understand that if the driver is in compliance with the requirements for
accessing the Uber platform (e.g., the PUC requirement that the driver have insurance coverage),
then Uber will reactivate the account upon request.
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
73
CV 13-03826-EMC
202
towards areas that I know tend to have a surge, such as the North Beach area or Marina District.
Casey Cheung also focuses on surge pricing: I am trying to make as much money in the shortest
amount of time possible. My strategy in trying to accomplish this depends in part on surge pricing,
which means that I get paid more for the same rides. I used to notice a lot more surge pricing back
when I first started in September 2014, but there seems to be less now. I try to target the surge
pricing fares because I make a lot more money on them, as they are almost always double digit fares
even if they are short rides.
203
8
9
3.
10
Each driver is only paid when he/she chooses to drive, and the amount of pay
depends on his or her unique circumstances
11
140.
Drivers contracting with Uber earn money piece rate when they take driving jobs.
12
Drivers who use the Uber App are paid by passengers for successfully completing a ride, based on the
13
distance driven and time spent on the ride.
14
driver.
15
themselves receive different compensation for a number of reasons. As I discuss in Section VII,
16
passengers may offer cash tips to drivers. And each driver has his or her own relationship with others
204
205
Uber provides the service of collecting this fare for the
To do this, Uber collects payment from passengers for the job of the ride. Drivers
17
18
202
Declaration of Anthony Braxton, 5/20/15, 8.
19
203
Declaration of Casey Cheung, 5/19/15, 6.
20
204
I understand that at certain times, in certain cities, there may have been temporary promotions
during which the financial arrangement between drivers and Uber was different.
205
It is my understanding that while some licensing agreements address this, others do not. For
example, see Rasier Software License and Online Services Agreement, 11/10/14, 4.1 (You are
entitled to charge a fare for each instance of completed Transportation Services provided to a
User that are obtained via the Uber Services (Fare), where such Fare is calculated based upon a
base fare amount plus mileage and/or time amounts, as detailed for the applicable Territory
(Fare Calculation).). Other versions of the licensing agreement address this point too, if in a
slightly different way. For example, see Rasier Transportation Provider Service Agreement,
5/15/13, p. 4 (In exchange for accepting and fully performing on a Request, you shall be paid an
agreed upon Service Fee for your completion of that Request.). Other versions of the licensing
agreement do not explicitly state whether Uber will arrange and collect the fare. For example, see
Uber Terms and Conditions, 8/11 p. 5 (The Company may introduce you to third parties for the
purposes of you providing transportation to them for a feeThe Company will not intervene I
[sic] any disputes between you and such third parties. We cannot and will not play any role in
managing payments between you and such third party parties.).
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
74
CV 13-03826-EMC
involved in the driving service. For independent drivers using their own car, Uber provides a
payment processing service, by which Uber collects the fare on the drivers behalf, deducts its
licensing fee, and provides the rest to the driver. While some drivers finance or own their car, others
lease their car or work at an hourly rate for another company.
driver to driver.
206
141.
Thus the driver earnings vary from
Some drivers work with a multi-driver business, which pays drivers in a multitude of
different ways. Robert Raymond, a driver for a multi-driver business, explains that he pays the multi-
driver business a weekly fee but keeps the entirety of fares I earn from my driving using the Uber
App.
207
Another driver for a multi-driver business, Daniel Bisneto, was paid a set hourly wage for
10
all hours [he] worked, regardless of whether the trips [he] completed were Jet Car's non-Uber
11
customers or passengers booked using Uber.
12
and are paid a share of their Uber fares, by the multi-driver businesses. Shevket Saatchigou, a multi-
13
driver business owner, explains: Most of the fares the drivers got using [the companys] vehicles
14
were through Uber accounts under my partner account. I would receive the fare, take out my daily
15
rate according to the car the driver was using, and they could keep whatever they made beyond
16
that.
208
Yet other drivers drive for multi-driver businesses
209
17
18
19
206
Some drivers own or finance their car (for example, see Declaration of Angel Herroz, 5/28/15,
2; Declaration of Enoch Devon Shadkam, 5/26/15, 4; and Declaration of Dean Strickler,
5/26/15, 4). Other drivers lease their car. Drivers often lease the car from either a multi-driver
business (for example, see Declaration of Keith Daniel Maddock, 5/18/15, 6; Declaration of
Sameh Shehata, 5/27/15, 3; and Declaration of John Andrews Sideserf, 5/26/15, 4) or from a
car dealership (for example, see Declaration of Robert Jones, 5/26/15, 2; Declaration of
Christopher Armstrong, 5/26/15, 9; and Declaration of Hazim Jajo, 5/28/15, 2). Other drivers
for multi-driver businesses are paid an hourly rate for use of the vehicle (for example, see
Declaration of Daniel Bisneto, UBE-OCO0000906873, 9/4/14, 8; Declaration of Rufus Fields,
10/1/14, UBE-OCO000091216, 9; and Declaration of Mark Forester, 11/20/14, UBEOCO0000955560, 10.)
207
Declaration of Robert Raymond, 5/27/15, 3. See also Declaration of Samir Moussa, 5/19/15,
4.
208
Declaration of Daniel Bisneto, 9/14/14, UBE-OC00000906873, 9/4/14, 4. See also

Declaration of Rufus Fields, 10/1/14, UBE-OC0000091216, 9.
209
Declaration of Shevket Saatchigou, 5/19/15, 5.
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
75
CV 13-03826-EMC
4.
142.
Drivers can - and some do - influence the price paid by the customer
Drivers have different ways to influence the price paid by customers. For instance,
drivers can contact Uber to propose a fare reduction or increase or reduce the fare by stopping the
fare before the ride is completed. One could imagine a policy in which drivers were expressly
prohibited from negotiating fares. It is my understanding that Uber has no such uniform policy. The
many Uber licensing agreements and contracts vary in how they address price negotiations. Often
the agreement directly states that a driver has the right to negotiate the fare, but other agreements do
not directly address price negotiation at all.
influence the price paid by customers, as I now detail.
210
10
143.
There are many reasons why drivers may choose to
Some drivers stop the fare early if they made a mistake along the way, such as
11
selecting an undesirable route; for example, William Gfroehrer explains: I have never negotiated a
12
fare but I have ended a ride early if I made a mistake so that the customer does not have to pay for my
13
mistake.
14
much: Often, I am very soft-hearted and I will give people a few miles for free. I will turn off the
15
app and let them have a few free miles if I've enjoyed the conversation and we've had a nice time.
16
So, I will turn off the Uber app early to give them a shorter ride and let them pay less.
Babikian Agop ends the ride early because he enjoyed the customers company so
212
17
18
211
144.
Drivers also can approach Uber on a ride-by-ride basis to influence the price paid by
the customer. A driver may choose to increase the fare in order to factor in additional passenger
19
20
21
210
For examples of contracts that specifically state that such negotiation is allowed, see Rasier
Transportation Provider Service Agreement, 5/15/13, p. 5 (You and the Company shall always
have the right to negotiate a Service Fee different from the pre-arranged fee. The purpose of the
pre-arranged Service Fee is only to act as the default fee in the event neither party negotiates a
different amount.) and Rasier Software License and Online Services Agreement, 11/10/14, p. 7
(You shall always have the right to: (i) charge a fare that is less than the pre-arranged Fare; or
(ii) negotiate, at your request, a Fare that is lower than the pre-arranged Fare.). For examples of
agreements that do not address price negotiation, see Uber Terms and Conditions, Pre-7/11 and
Uber Terms and Conditions 8/11.
211
Declaration of William Gfroehrer, 5/18/15, 17. See also Declaration of Kate Guilfoyle, 5/18/15,
17.
212
Declaration of Babikian Agop, 5/26/15, 12.
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
76
CV 13-03826-EMC
213
requests not reflected in the original price.
didnt properly capture the start and end point of the ride.
want to mitigate it by decreasing the fare; Reynaldo Adelantar states: I have asked for a fare
adjustment once or twice, when I overshot the destination.
Sometimes, a passenger will say that they do not want to pay the full fare because I went the wrong
way or a trip took too long. When this happens, I request a fare adjustment via the Uber app and the
price is automatically adjusted.
requests. For example, some drivers contact Uber to ask for an increased fare due to additional
stops,
10
Some drivers ask for a fare adjustment because the app

214
Sometimes drivers make a mistake and
215
216
217
Kaptan Ghimire explains:
Drivers also request a fare adjustment to factor in customer
or because the passengers ordered uberX, but more passengers get in the car than is allowed
218
for uberX and so they should have ordered uberXL.
11
12
B.
13
Supervision of Drivers on Jobs is Minimal and Drivers Have Other Full- and
Part-time Commitments
14
15
1.
Customers provide the only ratings of driver performance
16
17
18
213
For example, see Declaration of Kirill Sobolev, 5/19/15, 12.
19
214
Chris Uribe explains: I have had to request a fare adjustment a few times. For example, there
have been times when I have dropped off a passenger in an area with poor cellular service, and
because the iPhone did not get service, the Uber app was not able to record the trip mileage until I
got back to an area of better cell service. In those situations, the fare ends up being more
expensive than it should be. I would then request that Uber correct the fare based on the actual
mileage of the trip, and Uber did make the corrections (Declaration of Chris Uribe, 5/20/15,
16.)
23
215
Declaration of Reynaldo Adelantar, 5/18/15, 13.
24
216
Declaration of Kaptan Ghimire, 5/20/15, 10.
25
217
Oleg Lantsman explains: Once in a while, I request fare adjustments if the passengers require
something extra, such as an extra stop (Declaration of Oleg Lantsman, 5/19/15, 8).
218
Hovsep Sarafyan explains: In my experience so far, I've only had to contact Uber to discuss the
fares especially when there are more than four passengers but the rider request was just for an
UberX and not an Uber XL. Whenever this situation has arisen, Uber has resolved it quickly and
without a problem (Declaration of Hovsep Sarafyan, 5/26/15, 16).
20
21
22
26
27
28
Gibson, Dunn &
Crutcher LLP
77
CV 13-03826-EMC
145.
As explained above, driver and passenger ratings allow the Uber App to solve a
market failure (imperfect information and asymmetric information). Driver ratings are customer-
driven;
evaluation of drivers, nor does it give drivers significant rewards for high performance.
understanding that prospective customers use the ratings of prior customers to decide whether or not
to accept a ride from the driver. Likewise, drivers use other drivers ratings of passengers to select
riders.
219
220
8
9
Uber has no input into the ratings provided by customers. Uber neither performs any
146.
It is my
An individual drivers star rating is determined by averaging the customer-given star

221
ratings of the last 500 completed rides.
The ratings are completely customer-driven. Santos
10
Venegas explains: To my knowledge, my rating is determined completely by my passengers, and
11
Uber has no control over my rating at all. Additionally, I have the opportunity to rate my passengers
12
13
14
219
This point is made in a variety of ways throughout the different versions of the agreements that
Uber driver enter into. For example, in earlier versions, it is said that: For quality assurance
purposes, the Company has access to Ubers star rating system designed to determine the level of
service provided by the Transportation Providers contracting with the Company through Client
feedback. In a sense, the star rating is similar to a Yelp or Zagat rating, as it is based on a
continuously growing collection of star reviews submitted by Clients. Rasier Transportation
Provider Service Agreement, 5/15/13, p. 5. Other versions state that: You acknowledge and
agree that: (a) after receiving Transportation Services, a User will be prompted by Ubers mobile
application to provide a rating of you and such Transportation Services and, optionally, to provide
comments or feedback about you and such Transportation Services. Rasier Software License
and Online Services Agreement, 11/10/14.
220
Uber does have a weekly celebration of, and prize for, two of its platform users. There is a Sixth
Star award that consists of a $1,000 American Express gift card and an Uber gift package that is
awarded to two drivers a week. These awards are customer driven, with one driver being chosen
by rider nomination for doing something extraordinary and the other driver being chosen for
having the highest rating and number of trips in a given region. Finally, A Way to Give Your
Driver a Sixth Star, http://newsroom.uber.com/2014/10/finally-a-way-to-give-your-driver-asixth-star/, accessed on 6/26/15. Due to the large number of drivers, the probability of winning
the celebratory award is very low per individual driver. The Sixth Star Award Goes Global,
Uber, 4/9/15, http://newsroom.uber.com/2015/04/the-sixth-star-award-goes-global/, accessed on
7/6/15.
221
Ubers guide to the Rating System states: Only the last 500 trips are counted towards your
ratings. So, if you have completed more than 500 trips, only your last 500 trips count toward
your rating (Email from Uber San Francisco to Douglas OConnor, The Rating System, Part 1,
9/12/13, UBE-OCO000003017 at UBE-OCO00000303; Email from Uber LA to Ike Zimack,
Uber Notice: Account Status, 10/2/14, COLOPY0009867 at COLOPY000986).
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
78
CV 13-03826-EMC
222
on the App as well.
incentivizes; Harpreet S. Gandhi states: I think that the star rating system that Uber uses is great,
because it makes sure that the best drivers and best riders are connected, and it ensures the best
behavior from both sides. I believe my star rating is about 4.9 right now, and I know that is a high
rating. In order to maintain that high rating, I will offer my riders free water. No one told me I
should do this or suggested I do so, I just thought it would be a good idea.
Some drivers appreciate the informational exchange the platform
223
7
8
2.
147.
Drivers control the route taken during a ride
Drivers who use the Uber App are not required to use a particular navigation
10
application during the job and I am not aware of any claim or evidence that Uber monitors or
11
provides feedback on routes taken to complete rides.
12
preferred navigation systems while driving passengers, whether that is another GPS application like
13
Waze or Google Maps or their own personal GPS device. Kathy J. Robinson explains: When I
14
drive, I dont rely on the Uber app for navigation because I much prefer Waze. Some passengers
15
even tell me I should be the Waze spokeswoman because I tell everyone how great it is. I find its the
16
best tool I can use to help me drive.
17
instead of relying on the Uber app to provide the most efficient ride. Richard Charles Elliott explains
18
that if he is familiar with an area he will just [drive] the way [he] know[s] best.
224
225
Drivers are able to utilize their own
Drivers also note that they often use their own expertise
226
19
20
3.
Drivers choose the ambience of their car
21
22
23
222
Declaration of Santos Venegas, 5/27/15, 10.
24
223
Declaration of Harpreet S. Gandhi, 5/19/15, 11.
25
224
Deposition of Michael Coleman (Driver Operation Specialist, Uber), 7/9/14, pp. 100101. For
example, see Declaration of Kathy J. Robinson, 5/20/15, 10; Declaration of Shafqat Malik,
5/26/15, 6; and Declaration of Guy Cooper, 9/3/14, UBE-OCO0000907478, 15.
27
225
28
226
Declaration of Richard Charles Elliott, 5/18/2015, 12.
26
Gibson, Dunn &

Crutcher LLP
79
CV 13-03826-EMC
148.
It is my understanding that there is no uniform policy that controls the feel,
atmosphere and look of a drivers car. Many drivers understand that there is no dress code and hence
different drivers can wear different things.
will help their business; Yonatan Girma tries to dress professionally because [he] think[s] it is good
to be respectful to my passengers.
Charles Skip Close states: Uber did not require me to adhere to any dress code and in fact on most
days I would wear shorts and a t-shirt so I could be comfortable while I drive.
for their own convenience and will often wear what they wore to a previous job.
227
228
Some drivers dress professionally because they think it
Other drivers dress more casually for their own comfort;
229
Yet others dress
230
149.
Some drivers provide customer service on their own initiative and in their own way.
10
Uber may make suggestions on how drivers can make passengers happy (e.g., such as offering water
11
and gum to the passengers), but many drivers understand that these are suggestions and are not
12
requirements.
13
Uber gives partners tips on how to interact with passengers, but I dont recall what they are. What I
14
do for my passengers I do because Im accustomed to treating people courteously, such as greeting
15
them, letting them know my name, saying good afternoon or good evening, and the like. My star
16
rating is 4.86. I dont do anything in particular to try to butter the passengers up; I just try to be
231
Jose Diaz provides customer service because it feels natural to him: I recall that
17
18
227
For example, see Declaration of Richard Charles Elliott, 5/18/15, 11; Declaration of Rych
Dorsie, 5/19/15, 12; Declaration of Sharon Hurt, 5/19/15 4; Declaration of Naveed Zafar,
5/19/15, 13; Declaration of David Wong, 5/18//15, 9; Declaration of Muhammad Lodhi,
5/20/15, 15; Declaration of Tony Thurston, 5/20/15, 10; Declaration of David Malamud,
5/18/15, 20; Declaration of Brian Haig, 5/20/15, 16; and Declaration of Chris Uribe, 5/20/15,
19.
228
Declaration of Yonatan Girma, 5/18/15, 11.
229
Declaration of Charles Skip Close, 5/28/15, 8.
230
Christopher J. Armentrout explains: I havent made any changes in how I act as a result of the
guidance Ive gotten from Uber. For example, if Im leaving straight from work, I wear whatever
I wore to work that day (Declaration of Christopher J. Armentrout, 5/20/15, 8).
231
For example, see Declaration of Fabio Lorenzo Betancourth, 5/20/15, 8; Declaration of Hanibal
Poolisrezaeih, 5/19/15, 6; Declaration of Howard Hsu, 5/18/15, 13; Declaration of Ismail
Ezzikhe, 5/19/15, 8; Declaration of Issa Alzghoul, 5/19/15, 3; Declaration of Jiawen Ling,
5/19/15 , 13; Declaration of Jonathan Oliver, 5/18/15, 10; Declaration of Juan Hurtado,
5/20/15, 17; Declaration of Juliana Bushore, 5/20/15, 21; and Declaration of Kimberly
Henderson Strother, 5/20/15, 6.
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
80
CV 13-03826-EMC
232
myself.
transportation industry: Uber makes suggestions on how to increase my star rating from
passengers, such as having water in the car. I have experience in the transportation industry, so very
little of this was new to me. However, I do not need to follow these suggestions. I make my own
decisions on how to provide customer service to my passengers. I have a background in customer
service and I know what good customer service is.
customer service he provides needs to be anything beyond just being a good driver and being
courteous: Uber doesnt require that I provide any sort of amenities to passengers. I dont provide
any gum or candy or anything like that. My mission is to drive, not to feed passengers. I do a good
10
job driving, and thats what I focus on. In my experience, passengers expect me to be a motorist, and
11
they are happy when you do a good job at that.
12
that goes beyond Ubers suggestions to their passengers; Hanibal Poolisrezaeih explains: I was the
13
first person to provide wifi to my passengers. I have a bike rack for passengers who have bikes. I
14
provide an Ipad for my passengers. I provide all sorts of amenities for my passengers.
Jacob Dayeh provides customer service as a result of his past experience in the
234
233
Euripedes Silva does not believe that the
Other drivers provide innovative customer service
235
15
150.
While Uber offers suggestions to drivers during the onboarding process, because these
16
onboarding practices and presentations vary across many dimensions, different drivers receive
17
different information. Ubers driver operation specialist Michael Coleman explains that the
18
onboarding process in which drivers learn how to use the software varies by city, and depends on
19
the needs of the city.
20
Uber Black in Los Angeles in early 2012, the onboarding process entailed the owner of the
21
transportation company first going through some paperwork; then the driver would come into the
236
The onboarding processes have also changed over time. For example, for
22
23
24
25
26
27
28
232
Declaration of Jose Diaz, 5/26/15, 9.
233
Declaration of Jacob Dayeh, 5/28/15, 11.
234
Declaration of Euripedes Silva, 5/18/15, 8.
235
Declaration of Hanibal Poolisrezaeih, 5/19/15, 6.
236
Deposition of Michael Coleman (Driver Operation Specialist, Uber), 7/9/14, pp. 5051. See also
Deposition of Matthew Atkin (Senior Project Manager, Uber), 7/9/14, pp. 1315.
Gibson, Dunn &

Crutcher LLP
81
CV 13-03826-EMC
Uber office for an education session about the App, to provide further information, and to take a city
knowledge test.
virtualized although some drivers still come into an office.
by driver class. Ilya Abyzov, product manager at Uber, explains that the document used during
onboarding depends on which class of partner youre talking about.
237
Later, the city knowledge test was dropped and the Uber app training was
238
The onboarding process also varies
239
6
7
4.
Drivers choose and generally, but not always, supply the instrumentalities of
work
151.
To use the Uber App and accept a referral for a driving job requires having a car to
10
drive and a phone that is compatible with the Uber App. Many drivers, though not all, provide these
11
instrumentalities themselves. Regarding the phone, drivers can choose between using their own
12
phone and leasing one from Uber. Regarding the car, drivers can use their own personal car, leased
13
or bought, or they can work with a multi-driver business, which provides them with a car. Uber does
14
not provide drivers with a vehicle.
240
15
152.
Many drivers use their own personal phone, whereas other drivers use a smartphone
241
16
offered by Uber for $10 a week.
17
was provided to me via e-mail, and I reviewed it and accepted the terms online. I downloaded Ubers
18
software on my cell phone, which allowed you to start using the Uber app.
19
phone issued by Uber in order to keep their work phone and their personal phone separate; Philippe
20
Lestienne explains: I lease an i[P]hone through Uber. I pay $10 per week to use the Uber i[P]hone.
21
I only use it for the app. I have my own separate cell phone that I do not use for the Uber App. I had
James Coltrane, for example states: The licensing agreement
242
Other drivers use the
22
23
237
Deposition of William Barnes (General Manager for Los Angeles, Uber), 9/19/14, pp. 4647.
24
238
Deposition of William Barnes (General Manager for Los Angeles, Uber), 9/19/14, p. 51.
25
239
Deposition of Ilya Abyzov (Product Manager, Uber), 9/19/14, p. 95.
26
240
Deposition of Michael Coleman (Driver Operation Specialist, Uber), 7/9/14, pp. 156.
27
241
Deposition of William Barnes (General Manager for Los Angeles, Uber), 9/19/14, p. 30.
28
242
Declaration of James Coltrane, 5/28/15, 5.
Gibson, Dunn &

Crutcher LLP
82
CV 13-03826-EMC
the choice whether to use my own phone or lease a phone from Uber. I chose to lease a phone
because I prefer to keep things separate.
phones for their own convenience.
243
Some drivers even decided to have the App on both
244
245
153.
Some drivers own their own car while others lease a car to drive.
Some individuals
even have more than one car that they utilize for driving on the Uber App. For example, Lavon
Ohanian states: I have two cars that I drive when I use the UberX app a 2015 Toyota Prius and a
2010 Volkswagen CC. Usually I drive the Prius, but I have put almost 35,000 miles on the Prius
since November. As a result, I started alternating weeks between the Prius and the CC so that I can
manage the level of wear and tear on my Prius.
246
Some drivers choose what type of car they use in
10
order to reduce their business costs; E. Bond Francisco states: I use a 2014 Toyota Prius V to
11
provide transportation services. I purchased this vehicle myself in approximately October 2014. I
12
also chose the Toyota Prius V because of the gas mileage, which allows me to reduce my business
13
costs.
14
although there are variations in how drivers pay for the use of the vehicle and whether they can use it
15
for their personal use. Sameh Shehata, for example, leases a car from the business he drives for, on a
16
weekly basis,
17
Owners of multi-driver businesses also vary in whether they own or lease the cars that drivers utilize
18
for driving on the Uber App.
247
Drivers for multi-driver businesses normally lease cars from the multi-driver businesses,
248
249
whereas Brian Haig works with a business that keeps a portion of each fare.
250
251
There are even companies that own some cars and lease others.
19
20
243
Declaration of Philippe Lestienne, 5/18/15, 9.
21
244
For example, see Declaration of Jonathan Oliver, 5/18/15, 3.
22
245
For example, Hazim Jajo owns his car (Declaration of Hazim Jajo, 5/28/15, 2), whereas Fred
Blitzer leases his car (Declaration of Fred Blitzer, 5/26/15, 5).
246
Declaration of Lavon Ohanian, 5/27/15, 4.
247
Declaration of E. Bond Francisco, 5/19/15, 5.
248
Declaration of Sameh Shehata, 5/27/15, 3.
249
Declaration of Brian Haig, 5/20/15, 5.
250
Ismail Ezzikhe owns his vehicles (Declaration of Ismail Ezzikhe, 5/19/15, 3), whereas Arthur
Armstone leases his vehicles (Declaration of Arthur Armstone, 5/27/15, 3).
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
83
CV 13-03826-EMC
VII.
PROPOSED CLASS ON AVERAGE AS WELL AS CLASS CONFLICT
PLAINTIFFS TIPPING AND EXPENSES THEORY IMPLIES NO HARM FOR THE
3
4
154.
I understand that Plaintiffs claim that as a result of Ubers policies regarding tips,
drivers in California who use the Uber App are paid less than they would have been but-for the
alleged conduct.
commonly harmed because they are misclassified as independent contractors and thereby required to
pay business expenses (such as for their vehicles, gas, and maintenance).
252
I also understand that Plaintiffs argue that drivers using the Uber App have been
253
155.
In this section I show that Plaintiffs theory of harm cannot be assessed commonly for
10
all drivers who use the Uber App. Plaintiffs claims of common facts regarding tipping and expenses
11
are false. There are individualized factors that are important to consider when assessing whether any
12
given driver using the Uber App may have been harmed by Ubers tipping and expense practices as
13
claimed by Plaintiffs. As a result, damages cannot be calculated on a class-wide basis. Moreover,
14
there is meaningful potential for class conflict when one considers carefully the likely nature of the
15
but-for world. Plaintiffs have not described what the nature of the but-for world would be, but I
16
conclude that it could take a variety of forms. In some of those forms, no member of the putative
17
class experienced harm. In other of those forms, members of the putative class would be in conflict
18
with one another, in the sense that some members would have benefited while others would have
19
been harmed.
20
156.
In particular, and as I more fully articulate in a subsequent section, I take particular
21
note of the fact that Plaintiffs theory of harm is ill-posed, in the following sense: if it were decided
22
that all drivers using the Uber App were determined to be employees rather than independent
23
contractors, Uber would be expected, as an economic matter, to reorganize its business accordingly.
24
Some of these reorganizations might be substantial and might lead to harm for many members of the
25
26
251
See Declaration of Patrick Canlas, 8/11/14, UBE-OCO0000911420, 10.
27
252
SAC, 12.
28
253
SAC 3.
Gibson, Dunn &

Crutcher LLP
84
CV 13-03826-EMC
class. In this section, I limit my attention to the specific issue of tipping and expenses and
contemplate no reorganizations. Under the narrow version of what the but-for world looks like, it is
far from clear that any member of the putative class would be harmed. Under the broad version of
what the but-for world looks like, it is far from clear that all members of the putative class would
experience harm and, if some members of the putative class would experience such harm, then
there is meaningful potential for class conflict.
7
8
A.
It Is Not Possible to Calculate Potential Damages from Ubers Tipping Practices

on a Class-wide Basis
10
157.
The facts that Plaintiffs claim are common to the proposed class and upon which they
11
base their theory of harm regarding driver tips are false. There is no common practice in the car
12
service industry regarding tipping. As a result, each drivers experience is, and would be expected to
13
be, individualized. Moreover, even if one were to assume, as Plaintiffs claim, that somehow tips
14
were included in the fares collected by Uber for the driver, it would not be possible to reliably
15
determine the amount of tips that would have been received by any particular driver, let alone using
16
common evidence on a class-wide basis.
17
18
1.
19
158.
There is no common industry practice regarding tipping
I understand Plaintiffs have asserted that it is customary in the car service industry
254
20
for customers to leave approximately a 20 percent gratuity for drivers.
21
evidence shows this is not true. Certain passengers tend to tip more than others, and some passengers
22
leave no tip at all. Moreover, the amount a particular passenger may leave for a tip can vary
23
considerably from ride to ride. The reasons for this variation are numerous and individualized; tip
24
size is not simply determined by the length of the ride.
25
26
159.
However, the available
Publicly available data from New York City shows considerable variation in tipping
behavior among taxi passengers. For taxi rides that were paid for using a credit card, tips ranged
27
28
254
SAC 19.
Gibson, Dunn &

Crutcher LLP
85
CV 13-03826-EMC
from zero to as high as 40 percent. Leaving no tip at all is not uncommon. In fact, the seventh most
common tip amount was 0%, with roughly 3 million out of 51 million rides having 0% tips.
255
3
4
2.
The experience with cash tips of each driver using the Uber App is varied and
individualized
160.
Driver declarations submitted in this matter provide evidence that drivers using the
Uber App have varied and individualized experiences with cash tips. Some drivers have never
received a cash tip from a passenger for a ride arranged using the Uber App;
rarely received cash tips.
257
256
others have only
However, there are also drivers who receive cash tips more frequently.
10
For example, Corey Green says he receives cash tips almost every day, Hanibal Poolisrezaeih states
11
that he receives tips for 80 percent of the rides he completes through the Uber App, and Kindred
12
Sheppard estimates that he receives cash tips about twice a day.
258
13
161.
In addition, the size of the cash tips received by drivers using the Uber App varies
14
substantially from ride to ride and from driver to driver. For example, Patrick Canlas states that the
15
tips vary from client to client and the variety of tips [range] from $2 to $400 (which only happened
16
once). In addition, according to Mr. Canlas, [i]ts hard to say why I receive a tip or not, but in my
17
experience, certain people always tip by nature, not because its expected from the drivers
18
19
255
Eric Chemi and Ariana Giorgi, Heres How Much You Should Be Tipping Your Cab Driver,
Bloomberg Businessweek, 7/31/14, http://www.bloomberg.com/bw/articles/2014-07-31/hereshow-much-you-should-be-tipping-your-cab-driver, accessed on 6/29/15.
256
For example, see Declaration of Philippe Lestienne, 5/18/15, 28 (I have never received a tip
from a customer.) and Declaration of Eduardo Belloso, 5/26/15, 12 (Uber does not collect
tips. I never get tips from passengers. Uber tells passengers that there is no tip in the fare. There
are no tips.).
257
For example, see Declaration of Jonathan Meyers, 5/26/15, 8 (I rarely get cash tips, although it
does occasionally happen--maybe once in a hundred rides. My understanding is that I get to keep
100% of those cash tips for myself. I usually get about $1-5 on the rare occasions when I do get a
tip.).
258
Declaration of Corey Green, 5/27/15, 10 (I have received cash tips almost every day that I've
driven.); Declaration of Hanibal Poolisrezaeih, 5/19/15, 9 (I am allowed to get tips from
passengers. I tell my passengers that a tip is not required but if they offer me a tip, I take it. I get a
tip 80% of my rides.); and Declaration of Kindred Sheppard, 5/18/15, 20 (I have received
cash tips about two a day.).
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
86
CV 13-03826-EMC
259
perspective.
while using the Uber App. Corey Green says that he has received cash tips anywhere from $2.00 to
$60.00,
260
Mr. Canlas is not the only driver who reports receiving a wide range of cash tips
and Christopher Jose Martinez has received tips from $1 all the way up to $100.
261
4
5
3.
Identifying the tips a driver using the Uber App would have received is not
possible using available evidence
162.
In the previous two sub-sections, I explained that there is no single tipping practice
that can characterize tipping behavior in the car services industry and that the variability in passenger
tipping practices leads to tipping experiences for each driver that are varied and individualized,
10
including for drivers using the Uber App. As a result, even if we assume that a court determined that
11
tips were included in the fare charged for rides arranged using the Uber App, it would not be possible
12
to use available, common evidence to determine the size of the tips a particular driver would have
13
received. To do so would require substantial individualized information and inquiry. For example,
14
one might want to consider the tipping tendencies of each passenger driven by a particular driver, the
15
quality of the service provided in each ride, and how the service quality impacts the size of the tip
16
offered by each passenger. This information is not commonly available.
17
163.
In addition, there is no way to retroactively identify which drivers using the Uber App
18
received cash tips, for which rides, or the size of those tips. As I documented above, some passengers
19
pay cash tips after rides using the Uber App, but the Uber App does not track any cash tips that are
20
exchanged between passengers and drivers. There would be no way to determine whether a
21
particular passenger paid a cash tip for a given ride, did not pay a cash tip but assumed the tip was
22
covered by the fare, did not pay a cash tip because they generally do not pay tips for car service, or
23
did not pay a cash tip because they simply chose not to tip for that particular ride for whatever reason.
24
25
26
259
27
260
Declaration of Corey Green, 5/27/15, 10.
28
261
Declaration of Christopher Jose Martinez, 5/27/15, 6.
Gibson, Dunn &

Crutcher LLP
87
CV 13-03826-EMC
1
2
164. For all the reasons described above, it would not be possible to calculate potential damages
from Ubers tipping practices on a class-wide basis using common and available evidence.
3
B.
Meaningful Potential for Class Conflict
5
165.
6
7
8
9
10
11
14
15
16
charges and how that might change if Uber were required to implicitly or explicitly pay drivers for
tips. I show that under some of the but-for worlds that might hypothetically form a basis for a theory
of harm, there is no harm, whereas in other hypothetical scenarios, there is a meaningful potential for
class conflict.
166.
What if Uber changed the way it paid drivers so that each fare included a fixed tip, as
Plaintiffs claim should be the case? In this scenario, I am assuming that nothing else has changed
about Ubers business model or the relationship between Uber, drivers, and passengers. The only
change is that Uber now explicitly includes a fixed tip as part of the fare charged to passengers.
Would all putative class members be better off?
167.
17
18
In the previous section, I showed why it is not possible to calculate damages from
tipping behavior on a class-wide basis. In this section I consider the licensing fee that Uber currently
12
13
If Uber Were to Implicitly or Explicitly Pay Drivers for Tips There Is
In this scenario, Uber now faces a new costnamely that of covering driver tips.
Firms are organized to make profits. A fundamental precept of economics is that firms confronting
262
19
20
21
22
new costs take steps to minimize the impact of those costs on their profitability.
situations, the firm is able to fully avoid any such new costs. Applying these principles to the matter
at hand, I would expect that if Uber were including a fixed tip as part of the fare charged to
passengers, it would contemplate changing its referral rate so that the imposition of this new cost was
263
23
In some
revenue neutral.
As noted above, Uber might contemplate other changes as well, but I defer
24
25
26
27
28
262
For example, see Pindyck, Robert and Daniel Rubinfeld, Microeconomics, Third Edition
(Englewood Cliffs: Prentice-Hall, 1994), pp. 238273.
263
Revenue neutrality is often discussed in the tax reform literature to capture the concept that the
government can restructure the taxing procedure in a way that allows them to still receive the
same amount of revenue (for example, see, Michael Pflger, On the Employment Effects of
Revenue Neutral Tax Reforms, Public Finance Analysis, New Series, 54(4), 1997, pp. 430446
Gibson, Dunn &

Crutcher LLP
88
CV 13-03826-EMC
consideration of these until the next section. Consequentially, here I ask simply: What would happen
if Uber changed its referral rate in order to help cover the new cost of driver tips that are included in
the fare?
168.
To answer this question, suppose that for each ride, Uber leaves unchanged the total
fare charged to the passenger, but sets aside a fixed share of the total fare as the drivers tip. This
tip is paid in its entirety to the driver. Under these changes the passenger pays the same fare, the
driver receives a larger share of the fare since Uber does not take a licensing fee on the fixed tip,
and Uber, in turn, receives a smaller share of the total fare. However, Uber is organized for profit
and can charge the licensing fee it deems would put itself in the best position vis--vis its competitors
10
(e.g., other lead generation platforms such as Lyft and Sidecar). Suppose that Uber increased the
11
referral rate in an amount to exactly offset the revenue it loses by setting aside the fixed tip for the
12
driver. In turn, the drivers share of the non-tip portion of the fare would be reduced, exactly
13
offsetting the increased pay from the tip. The final costs and revenues of all three parties the
14
passenger, the driver, and Uber are unchanged.
15
169.
To see how this might work with a simple numerical example, suppose that in the
16
current world, Ubers licensing fee were a flat 20% of the fare charged to passengers. (I recognize
17
that licensing fee schedules are more complex than this; I seek merely to clarify the content of this
18
argument with a few hypothetical numbers.) To make the argument conservative, suppose the fixed
19
tip Uber provided to drivers were a hefty 20%seemingly providing drivers with a substantial
20
improvement. This improvement is a mirage, however. Because Uber is at liberty to charge the
21
referral rate it deems appropriate, under this hypothetical Uber could reduce its pre-tip fares by
22
approximately 16.7% so as to leave customer fares unaffected, and yet increase its licensing fee to
23
24%. Such a change would leave passenger costs unchanged and Uber revenues unchanged.
264
24
25
at p. 430. Similarly, Uber can restructure its licensing procedure in a way that allows them to still
receive the same amount of revenue.
26
27
28
264
In general, if the current world licensing fee rate is c and the constant tip rate Uber provided for
drivers were t, then for an actual world fare of $F, new pre-tip fares are set as $F/(1+t), and the
new licensing fee rate is c(1+t). Here, c(1+t)=0.2(1+0.2)=0.24.
Gibson, Dunn &

Crutcher LLP
89
CV 13-03826-EMC
170.
Note that in this hypothetical scenario, I have simply relabeled compensation. I have
set aside a portion of the drivers portion of the passengers fare to be called tip, I have given all of
it to the driver, and I have calculated what new fares Uber would charge to leave customer fares
unchanged and what new licensing fee Uber would have to charge to have its revenues unaffected.
171.
The only thing that the putative class members would have in common is that none of
them are harmed: their earnings are the same under the but-for world Plaintiffs seek and under the
actual world of which Plaintiffs complain.
172.
Alternatively, imagine that instead of including a fixed tip in the total fare, tips were
instead determined by riders through the App in some way. For the reasons described above, tips
10
would likely vary substantially across drivers and rides. Uber would not necessarily find it in its
11
interests to leave fares unchanged since under this scenario passengers are paying more for rides.
12
Uber might lower its fares so that passengers are paying, on average, the same total amount for a ride
13
as they are in the current world. Uber could then increase the licensing fee it charges such that it
14
receives, on average, the same revenue that it does under the current arrangement.
15
173.
To return to a basic numerical example as above, suppose that in the current world,
16
Ubers licensing fees were a flat 20% of the fare charged to passengers. (Again, I recognize that
17
licensing fee schedules may be more complex than this; I seek merely to clarify the content of this
18
argument with a few hypothetical numbers.) If passengers on average paid tips of 15%, then under
19
this hypothetical Uber would reduce the fares it charges customers by approximately 13%, allow for
20
passengers to choose their tip amount, and increase its licensing fee to 23%. Such a change would
21
leave passengers aggregate costs unchanged (in the hypothetical where they pay a 15% tip on
22
average) and Ubers aggregate revenues unchanged.
265
23
24
174.
However, in this scenario, some drivers would be better off and some drivers would be
worse off. The drivers who receive higher than average tips from passengers would be better off.
25
26
27
28
265
In general, if the current world licensing fee rate is c and the average tip rate t, then for an actual
world fare of $F, new fares are set as $F/(1+t) and the new licensing fee rate is c(1+t). Here,
c(1+t)=0.2(1+0.15)=0.23.
Gibson, Dunn &

Crutcher LLP
90
CV 13-03826-EMC
The drivers who receive lower than average tips from passengers would be worse off. Thus, in this
hypothetical scenario, putative class members are in conflict with one another.
175.
Moreover, as I described above, highly individualized factors influence tipping.
Consequently, an individual investigation would be necessary to identify if any particular driver was
harmed or made better off.
6
7
C.
It Is Not Possible to Calculate Potential Damages Incurred from Expenses on a
Class-wide Basis
176.
Plaintiffs claim that some of the main sources of expenses borne by drivers using the
266
10
Uber App are vehicle and gas expenses.
11
cost of these expenses is not common to the class. Drivers using the Uber App drive a wide variety
12
of vehicles. Driver declarations provide a sense of the variation in type and age of these vehicles.
13
These vehicles incur expenses at varying rates. Gas mileage varies considerably among cars used by
14
drivers.
However, the extent to which drivers incur, and bear, the
267
268
In addition, the value of a vehicle does not depreciate linearly; that is, new vehicles
15
16
17
18
266
SAC 23.
267
For example, see Declaration of Janet Waller, 5/19/2015, 8 (I drive a 2013 Scion XB when I
drive using Uber and Lyft Applications. I own the car as well.); Declaration of Quinn Renfro,
5/28/15, 23 (We shared one car, a 2009 Yukon Denali, which was owned by Q-Star
Charters. Within the next six months, between November 2013 and February 2014, my company
acquired two more vehiclesa 2008 Cadillac STS and a 2009 Cadillac Escalade.); Declaration of
Edmond Chouteau, 5/26/15, 6 (I drive both UberX and UberPlus. The vehicle I use for
UberPlus is a Mercedes and I have a Ford F150 that I use for UberX); Declaration of Angel
Herroz, 5/28/15, 2 (I own a 2012 Jeep Patriot that I use to drive with the Uber App.); and
Declaration of Terry Gambra Baruti, 5/20/15, 3 (I use a Town & Country LX minivan for
transportation of passengers.).
268
To best demonstrate the variation in gas mileage, consider the following three declarants. Each
declarant reported driving for UberX, but the first has a 2015 Toyota Prius, the second drives a
2015 Honda Accord, and the third uses a 2015 GMC Acadia. The Prius receives an mpg rating of
51 city/48 hwy, the Accord receives a rating of 27 city/36 hwy, and the Acadia receives a rating
of only 17 city/24 hwy. Declaration of John Nicholas Marantos, 5/28/15, 2; Declaration of
Thomas Ogu, 5/19/15, 4; Declaration of Abdulai Amidu, 5/26/15, 2; 2015 Prius, Toyota,
http://www.toyota.com/prius/, accessed on 7/3/15; 2015 Accord Sedan, Honda,
http://automobiles.honda.com/accord-sedan/, accessed on 7/3/15; and 2015 Acadia, GMC,
http://www.gmc.com/acadia-crossover-vehicle.html, accessed on 7/3/15.
19
20
21
22
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
91
CV 13-03826-EMC
269
depreciate faster, per mile and/or per day, than older vehicles.
incurred by a particular driver may depend on the age and odometer reading of her vehicle at the time
she was providing each ride referred through the Uber App.
177.
This means that the expenses
In addition, some drivers using the uberX platform use their personal vehicle (while
270
others using uberX drive liveried cars with TCP licenses).
personal vehicles likely also use their vehicles for personal driving when they are not driving a
referral from the Uber App. The degree of accumulated wear and tear on a vehicle that is incurred
during personal use would likely vary considerably across drivers, depending in large part on the type
and amount of personal driving.
10
178.
Many of the drivers who use their
The responsibility each driver currently bears for his or her own expenses is not
11
common. Drivers driving a black car for a licensed TCP, for instance, may not cover their vehicle
12
and gas expenses. For example, drivers for the TCP company Black Limousines have their gasoline,
13
car washes, and tolls covered and they only have to pay for their own parking or moving violations,
14
lunches, and transportation to get to the company to pick up the cars that they use.
15
another licensed TCP, in contrast, may be responsible for paying the majority of their own expenses,
16
especially if they use their own car to drive. For example, the driver who works for J.J. VIP Limo
17
Inc. is responsible for paying his own gas, maintenance, etc. and is not reimbursed for any
18
expenses.
271
272
19
20
Drivers for
179.
For all the reasons described above, it would not be possible to calculate potential
damages incurred from expenses on a class-wide basis using common and available evidence.
21
22
269
Q: What is Car Depreciation? Kelley Blue Book, 12/17/13, http://www.kbb.com/what-is/cardepreciation/?r=643161140864619600, accessed on 7/3/15; and Don't Get Stuck, Choose
Wisely, Kelley Blue Book, http://www.kbb.com/new-cars/about-total-cost-of-ownership/,
accessed on 7/3/15.
270
For example, see Declaration of Alexander AJ Galloway, 5/27/15, 3; and Declaration of

Ahmed Yousif Abdi, 5/28/15, 2.
27
271
28
272
Declaration of Mostapha Makki, 5/26/15, 3.
23
24
25
26
Gibson, Dunn &

Crutcher LLP
92
CV 13-03826-EMC
180.
Were Uber to implement a rate scheme to cover expenses as a proportion of the fare,
then the same argument regarding tips would apply here: the only thing members of the putative
class would have in common is that there would be no harm. Uber would predictably increase its
licensing fee to offset any flat rate expenses it paid to drivers. Uber would predictably do so in such a
way as to leave itself in a revenue neutral position, and this would leave drivers with no actual change
in their economic position.
7
8
D.
If Uber Paid for Driver Expenses There Is Meaningful Potential for Class
Conflict
10
181.
In the previous section, I showed why it is not possible to calculate damages from
11
expenses on a class-wide basis. In this section I consider the licensing fee that Uber currently
12
charges and how that might change if Uber were required to implicitly or explicitly pay drivers for
13
expenses. I show that under hypothetical scenarios, there is a meaningful potential for class conflict.
14
182.
What if Uber bore responsibility for paying, either implicitly or explicitly, for
15
expenses incurred by drivers? Assuming that nothing else has changed about Ubers business model
16
or the relationship between Uber, drivers, and passengers, would all putative class members be better
17
off?
18
183.
In this scenario, Uber now faces a new costnamely, covering driver expenses. What
19
would happen if Uber changed its licensing fee rate in order to help cover this new cost? As I
20
showed in the previous section, the expenses incurred by each driver are individualized, but Uber
21
could increase the licensing fee rate it charges so that it could cover, on average, driver expenses
22
through the increased licensing fees. This would necessarily reduce the revenues received by drivers,
23
but would be at least in part offset by the fact that their expenses would be now covered.
24
184.
In this scenario, drivers would be, on average, no better or worse off than before Uber
25
covered expenses. However, some drivers would be better off and some drivers would be worse off.
26
Drivers who incur higher than average expenses would be better off. For these drivers, the reduction
27
in revenue as a result of the increased Uber licensing fees would be more than offset by their expense
28
reimbursement. However, drivers who incur lower than average expenses would be worse off for
Gibson, Dunn &

Crutcher LLP
93
CV 13-03826-EMC
similar reasons. Moreover, as I showed above, the actual expenses incurred by each driver depend on
individual factors. Thus, individual investigation and information would be required to identify if a
particular driver was harmed or made better off. In this scenario, there is again substantial scope for
conflicts among putative class members.
5
6
VIII. MANY DRIVERS WOULD BE HARMED IF ACCEPTING A PASSENGER
REFERRAL FROM UBER MADE THE DRIVER AN UBER EMPLOYEE
8
9
185.
In the previous section, I showed (a) that damages could not be calculated on a class-
10
wide basis, (b) that under some notions of the but-for world there are harms to no members of the
11
putative class, and (c) that if there are harms for some members of the putative class then there is
12
meaningful potential for class conflict. A theme of these considerations is that the licensing fee Uber
13
currently charges would be expected to change in the wake of any change to how tips and expenses
14
are handled.
15
186.
273
In this section, I discuss the potential harm to different putative class members under
16
two different hypothetical scenarios: (1) were all drivers who use Ubers platform for referrals to be
17
found uniformly to be Uber employees and Uber agreed to keep at least some of them on as
18
employees, and (2) were Uber to change its business model to only allow companies, and not
19
individuals, to use the platform.
20
21
22
23
24
25
26
27
28
273
To be clear, in my opinion, there is nothing inappropriate about changing the referral fee in the
wake of any alteration to Ubers work relationship with drivers who accept ride referrals. Uber is
a firm organized for profit and has the right to charge the referral fee it deems appropriate in light
of the competitive environment it confronts and its costs.
Gibson, Dunn &

Crutcher LLP
94
CV 13-03826-EMC
A.
Many Current Drivers Would Be Harmed If Use of the Uber App Rendered
Every Driver an Uber Employee
187.
What if every driver using the Uber platform were found to be an employee of Uber
274
(i.e., every member of the putative class were found to be an employee of Uber)?
impact different drivers in the putative class going forward? Economic principles allow one to
assume that drivers who currently use the platform prefer it to their current outside options. Thus, the
threshold question is whether each driver in the putative class would be able to continue to use the
Uber App. An equally important question is whether each driver would be better off (i.e., prefer) the
tradeoffs that come from all platform users being found to be Uber employees.
How would this
10
11
1.
Increased costs (some fixed) relative to each driver-employee could lead
12
Uber to employ fewer drivers and many putative class members would be
13
harmed
14
188.
If use of the Uber App rendered a driver an Uber employee, then the costs, both fixed
15
and variable, to Uber of each driver accepting ride referrals using the platform could increase
16
substantially, because employers must pay in whole or in part for certain legally mandated benefits
17
and insurance coverage including for example workers compensation,
18
insurance,
275
276
277
employment training tax,
278
and paid sick leave.
state unemployment
There may also be further legally
19
20
21
22
274
For the purposes of this discussion, define Uber driver as a driver who has signed up for the
Uber platform, turned it on, and driven at least one passenger through a referral.
275
Obtaining Workers Compensation Insurance, Governors Office of Business and Economic

Development,
http://www.business.ca.gov/StartaBusiness/AdministeringEmployees/WorkersCompensation.asp
x, accessed on 7/7/15
276
Employment Taxes, Governors Office of Business and Economic Development,

http://www.business.ca.gov/StartaBusiness/AdministeringEmployees/EmploymentTaxes.aspx,
accessed on 7/3/15.
277
Employment Taxes, Governors Office of Business and Economic Development,

http://www.business.ca.gov/StartaBusiness/AdministeringEmployees/EmploymentTaxes.aspx,
accessed on 7/3/15.
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
95
CV 13-03826-EMC
mandated benefits and insurance, or compliance with other policies, which Uber would have to cover
if drivers were deemed employees. For example, if drivers were deemed employees, it is not clear
how Uber might go about verifying whether it were in compliance with minimum wage policies.
This would immediately make flexible work and multi-apping problematic from Ubers
perspective. Such considerations could substantially undercut Ubers economic incentive to allow
drivers to use the platform unless their use of it was sufficiently frequent to justify the costs and
difficulties associated with their hire. For example, consider a driver like Ray Lebron. His driving
brings very little revenue to Uber. Approximating current Uber licensing fees at 20% of fares, Mr.
Lebron generates less than $60 a month, or $720 a year, in revenues for Uber.
279
Uber would have
10
very little incentive to have a relationship with drivers of this type if they were deemed to be
11
employees, particularly in light of the increased costs described above, and would be expected to
12
choose not to hire them. As a result, it is manifestly evident that any drivers similarly situated
13
might be harmed.
14
15
2.
Drivers who do not value the benefits that come with employee status, would
16
likely be harmed if Uber raised its referral rates to cover the cost of employee
17
benefits
18
189.
Imagine that Uber, as an employer, had to provide workers compensation and
19
unemployment insurance to all its driver-employees. It would be unsurprising if Uber, in order to
20
pay for these benefits, increased the average licensing fee it charges drivers for accepted referrals.
21
22
278
See Employee Benefits, Governors Office of Business and Economic Development,

http://www.business.ca.gov/StartaBusiness/AdministeringEmployees/EmployeeBenefits.aspx,
accessed on 7/3/15: Beginning July 1, 2015, employers will be required to offer paid sick
leaveAn employee who, on or after July 1, 2015, works in California for 30 or more days
within a year from the commencement of employment is entitled to paid sick days for prescribed
purposes.
279
Ray Lebron works on average 4 hours a week (Declaration of Ray Lebron, 5/29/15, 10). The
calculation assumes he makes $16.37 per hour, which is the average rate in LA for drivers who
work 115 hours/week (Hall, Jonathan, and Alan Krueger, An Analysis of the Labor Market for
Ubers Driver-Partners in the United States, Princeton University Industrial Relations Section,
Working Paper 587, 1/22/15, p. 18).
23
24
25
26
27
28
Gibson, Dunn &
Crutcher LLP
96
CV 13-03826-EMC
Drivers who value the new benefits might find themselves equally situated or even better off than
before. But drivers who did not value the benefits would be worse off. They would earn less when
driving using the Uber App without any offsetting increase in value because the benefits do not hold
value for that particular driver.
5
6
3.
Were Uber to take an employers stance of asking for employee loyalty as a
condition of employment or some expectation of permanence (e.g., setting
hours of work), many putative class members would be harmed
190.
Were Uber to be an employer, and all users of the platform employees, Uber would be
10
free to require of those drivers many of the typical conditions of employment. Were Uber to require
11
employee loyalty (i.e., that a term of existing employment was that drivers work only with Uber and
12
not with other driving services), many putative class members would be harmed (absent
13
demonstrable and meaningful offsetting benefits). For example, and as shown above, putative class
14
members currently choose to work with and for many different transportation companies at the same
15
time. These putative class members would generally be expected to be harmed, although a
16
determination of the amount of harm would require individual investigation. One such example
17
might be a full-time driver who currently chooses to drive Uber referrals only part-time, and drives a
18
taxi part-time. Such an individual would no longer enjoy the flexibility to work in this way. If she
19
prefers her taxi job to driving referrals with her own car, she would then be limited to part-time
20
employment with the taxi company. Absent demonstrated benefits, she would be expected to be
21
harmed. As a second example, consider a driver who always uses his own car, but takes passenger
22
referrals from Uber, Lyft and other lead generation platforms, sometimes simultaneously. Such an
23
individual might no longer be allowed to enjoy this flexibility and might have to rely solely upon
24
Uber referrals, consistent with any new understanding that drivers are Uber employees. Absent
25
demonstrated benefits, this driver would also be expected to be harmed. Moreover, I understand that
26
it is additionally conceivable that under such a determination, such a driver might have breached a
27
28
Gibson, Dunn &
Crutcher LLP
97
CV 13-03826-EMC
280
duty of loyalty to Uber by previously having driven for its competitors.
conceivable could potentially expose such a driver to litigation. While in both examples, putative
class members would be expected to be harmed, the fact of harm and the magnitude of any such harm
would require individual investigation.
191.
The fact that this is
Were Uber to exercise the right to choose when and where employees workedwhich
could be helpful for Uber on the margin from a resource matching/planning perspectivethe many
drivers who value the freedom to begin, end and re-start work whenever they please would be
harmed, absent meaningful demonstrated offsetting benefits. Determining the magnitude of harm
would by necessity involve pricing how much workers value flexibility, which would involve
10
individual investigation.
11
12
4.
13
Were all drivers for black car/limousine companies to become employees of

Uber, some putative class members would be harmed
14
192.
A driver currently employed for a black car company who receives and prefers an
15
hourly wage could be harmed if the choice to accept an Uber referral turned him into an Uber
16
employee. Uber does not pay hourly, but rather by the job, which means that drivers who use the
17
Uber App for referrals bear risk (e.g., of not getting enough referrals or the right referrals and not
18
earning enough day to day). Some drivers would prefer to remain in an employment situation in
19
which they did not bear daily risk on their take home earnings.
20
193.
An owner-driver of a black car company could be harmed if he, a currently self-
21
employed business owner, and all his drivers were to become employees by accepting a referral from
22
Uber. Were he to choose not to accept referrals in order to maintain his status as a self-employed
23
business owner, he would unambiguously be harmed as he now has access to fewer referrals. On the
24
other hand, were he to choose employment status for himself and all his employees, the determination
25
of whether or not that owner-driver would be harmed would require inquiry into all the details of his
26
business and the ultimate status of his relationship with Uber.
27
28
280
The American Law Institute, Restatement of the Law Third, Agency, Volume 1, p. 33.
Gibson, Dunn &

Crutcher LLP
98
CV 13-03826-EMC
B.
2
3
Many Current Drivers Would Be Harmed If Uber Only Allowed Companies, and
Not Individuals, to Use the Platform
194.
It is instructive to consider the hypothetical of what would happen to putative class
members were Uber to choose to only allow companies, and not individuals, to use its platform. For
example, could this happen if all individual driver use of the platform were deemed to constitute
employment by Uber, but drivers who use the platform through their employment with a small
business were not? In this hypothetical Uber would no longer allow individuals to use the platform
directly, but rather they could only do so if they were employed by a business that Uber had
approved. Many putative class members would likely be harmed by this, unless it could be somehow
10
demonstrated that all putative class members would be able to obtain work for employers on that
11
basis. Barring that, any potential benefit of being determined to be an employee rather than an
12
independent contractor would have to be juxtaposed with the possible harm suffered by members of
13
the putative class unable to obtain work.
14
195.
Similarly, under such a hypothetical, individuals who currently exercise freedom over
15
their hours and are free agents in regard to for whom and when they work would be forced to obtain
16
employment with a firm in order to be able to offer driving services with their own car. The hours,
17
location and other requirements of the job would likely be different than the drivers current
18
independent relationship using the Uber App for referrals. Their new status as an employee would be
19
expected to be associated with an employers traditional insistence on hourly schedules, modes of
20
dress, and so on. Those putative class members who value the current flexibility of driving using a
21
passenger referral system with no requirements on days worked, geography, number of rides or time
22
driven in a day, when to stop using the App, and so on, would be harmed absent some demonstrated
23
offsetting benefit to being an employee. Demonstrating such offsetting benefits would again seem to
24
require individual investigation; even if the benefits were somehow able to be determined on a class-
25
wide basis, the harm from loss of flexibility cannot be determined using information common to the
26
putative class and thus would require individual investigation.
27
28
Gibson, Dunn &
Crutcher LLP
99
CV 13-03826-EMC
IX.
CONCLUSION
2
3
196.
Traditionally, the market for driving services has confronted information and
coordination problems that have kept simple market solutions from working well. The essence of
these problems is that those who would like to be driven may find it difficult to locate in time and
space those who might be willing to drive them to their destination. For many people, this may be a
central motivation for car ownership, despite its cost. Uber and other lead generation platforms
provide a new, technological solution to these problems, leveraging the prevalence of cell phones on
the part of drivers and those who might want to be driven. This technological approach complements
10
11
existing regulatory approaches.

197.
In my report, I have emphasized that Uber provides a wide array of platforms, each of
12
which is best understood as a product, that facilitate the matching of passengers and drivers. Most of
13
the economic activity taking place on Uber platforms is regulated by the PUC, but other economic
14
activity taking place on Uber platforms is regulated by cities and counties. Those activities regulated
15
by the PUC may be regarded as underneath either the TCP or TNC regulatory regimes, with the
16
distinction being drawn based on whether the driver is engaged in work on behalf of a third-party
17
entity that is regulated as a TCP.
18
198.
Drivers using Uber platforms are involved in a wide variety of work relationships with
19
companies engaged in driving services. Those accepting ride referrals through UberBLACK may be
20
sole proprietors or employees or independent contractors of third-party black car companies. Those
21
accepting ride referrals through the UberTAXI platform could own their own taxi company or be
22
independent contractors or employees for third-party taxi cab companies. Those accepting ride
23
referrals through the uberX platform could be sole proprietors or employees or independent
24
contractors of third-party driving service companies.
25
199.
Moreover, drivers may, and do, accept ride referrals on more than one Uber platform,
26
mixing and matching as they see fit the platforms for which they qualify. Every driver has the option
27
of accepting ride referrals on any platform for which she meets the qualifications.
28
Gibson, Dunn &
Crutcher LLP
100
CV 13-03826-EMC
200.
Additionally, drivers using Uber platforms are involved in a wide variety of work
relationships with companies having nothing to do with driving services. Some drivers moonlight
as drivers in the sense that they work full-time for an employer and accept Uber platform referrals
when convenient for them. Others work primarily as freelancers, cobbling together various flexible
work or part-time work and accept referrals from Uber platforms when they have the time and
inclination. Yet other drivers have no employer and only rarely accept referrals from the Uber App,
perhaps because of competing obligations at home or leisure choices. Finally, some individuals may
simultaneously be contemplating accepting referrals from one or more Uber platforms at the same
time as they are contemplating accepting referrals from competitors for Uber.
10
201.
In my opinion, there is no typical driver who uses the Uber App. I am not aware of
11
any evidence that Uber exercises control, retains a right of control, or has any uniform policy in
12
connection with control over drivers physical conduct of their work. Consistent with this
13
understanding, the evidence reviewed above demonstrates that each driver controls whether to drive,
14
when to drive, where to drive, and how to drive; that each driver elects whether, when, and where to
15
turn on the Uber App; and that each driver makes an election regarding which platform to turn on
16
from among those for which she is eligible.
17
202.
Just as work options and practices vary greatly, so do driver expenses. Drivers are in
281
18
the best position to minimize their own expenses: they choose which type of car to drive,
19
choose which route to take, how aggressively to drive, whether to purchase generic or name-brand
20
gasoline, regular unleaded, midgrade, or premium gasoline, and so on. Because of this heterogeneity
21
in expenses, determining how much members of the putative class may have been harmed would
22
require, at a minimum, individual inquiry and investigation. Moreover, I am not aware of any
23
commonly available information on expenses.
they
24
25
26
27
28
281
I understand that these choices are subject to the requirements of the Uber platform for which the
driver contemplates accepting ride referrals. For example, more cars are eligible for driving using
the uberX platform than for driving using the UberBLACK platform, which is a higher end
driving service.
Gibson, Dunn &

Crutcher LLP
101
CV 13-03826-EMC
203.
Because tipping practices vary greatly in the real world, there is similarly no method
for estimating how much members of the putative class may have been harmed using commonly
available information.
204.
Moreover, as I emphasized above, drivers would not be uniformly better off if Uber
were to change its service to add tips or pay driver expenses. To the best of my knowledge, Plaintiffs
have not articulated any notion of a but-for world that might form a basis for claims of harm. In my
report, I postulated what some of those but-for worlds might look like. Under some of them,
members of the putative class are united in having experienced no harms. Under others, there is
meaningful scope for class conflict, with some members benefiting from the status quo and others
10
being harmed by it. Further, determining which member is of which type and what magnitude of
11
harm or benefit is involved would involve individual inquiry and investigation.
12
205.
Underscoring these concerns, I conclude that drivers use the platform in such a large
13
variety of ways, under such a large variety of working relationships, that many current drivers who
14
have used the Uber App for referrals would be harmed if it were commonly found that the use of the
15
Uber App turned every driver into an employee of Uber. Uber would, under any such determination,
16
be expected to contemplate many different reorganizations of its business model, and many of these
17
would be associated with potentially substantial harms to drivers.
18
19
I declare under penalty of perjury under the laws of the United States of America that the foregoing is
20
true and correct.
21
22
Executed this 7 day of July at Kensington, California.
23
24
25
26
______________________________________
27
28
Gibson, Dunn &
Crutcher LLP
102
CV 13-03826-EMC

Declaration of Justin McCrary - Expert Report

Uploaded by

Copyright:

Available Formats

Declaration of Justin McCrary - Expert Report

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Declaration of Justin McCrary - Expert Report

Uploaded by

Copyright:

Available Formats

1

GIBSON, DUNN & CRUTCHER LLP

Attorneys for Defendant

DOUGLAS OCONNOR, THOMAS

Case No. CV 13-03826-EMC

DECLARATION OF DR. JUSTIN MCCRARY

SCOPE OF OPINIONS ........................................................................................ 7

TECHNOLOGICAL SOLUTIONS HAVE REDUCED MARKET

A. Internet Platforms Help Solve Traditional Coordination Problems ................ 11

Internet Platforms Help Solve Traditional Information Problems .................. 14

New Methods of Connecting People to Work Opportunities .......................... 16

LEAD GENERATION PLATFORMS LIKE UBER HAVE CHANGED

A. Coordination Issues in Transportation Prior to New Technologies ................ 19

The Transportation Industry Prior to Lead Generation Platforms ................... 19

New Options and Sources of Competition Emerge with the

THERE IS NO TYPICAL DRIVER WHO USES THE UBER APP................ 33

Californians Working as Drivers Before the Existence of Uber and Lead

Californians Who Were Not Working as Drivers Before the Existence of

D. Small Businesses That Did Not Exist Before Uber ......................................... 56

Gibson, Dunn &

DECLARATION OF DR. JUSTIN MCCRARY

APP IS UNIQUE AND INDEPENDENT ......................................................... 59

VII. PLAINTIFFS TIPPING AND EXPENSES THEORY IMPLIES NO

A. It Is Not Possible to Calculate Potential Damages from Ubers Tipping

If Uber Were to Implicitly or Explicitly Pay Drivers for Tips There Is

It Is Not Possible to Calculate Potential Damages Incurred from

D. If Uber Paid for Driver Expenses There Is Meaningful Potential for

VIII. MANY DRIVERS WOULD BE HARMED IF ACCEPTING A

Many Current Drivers Would Be Harmed If Uber Only Allowed

I am an economist with expertise in economic modeling, statistical methods, law and

as Assistant Professor at the University of Michigan, I became an Assistant Professor of Law at

In addition to my post as Professor, I am the Founding Director of D-Lab, the Social

faculty regarding high-performance computing, statistical software, and appropriate statistical

and with Bob Cooter and Eric Talley (20122014).

invited to become a Faculty Research Associate.

As noted, I previously worked at the University of Michigan. From 2003 through

Gibson, Dunn &

advanced econometric theory to Ph.D. students.

My research spans a diverse range of topics, including antitrust, crime, employment

discrimination, econometric and statistical methodology, education, fertility, financial markets,

Economics, and the Journal of Empirical Legal Studies.

pharmaceuticals; the potential anti-competitive implications of a proposed telecommunications

sale of retail gasoline.

Finally, I am frequently asked to give talks regarding the utilization of statistical

A copy of my curriculum vitae, including a list of previous testimony and depositions,

any other matter.

Uber Technologies, Inc. (Uber) is a defendant in a federal putative class action

have worked as [UberBLACK], [UberSUV], or [uberX] drivers in California.

information becomes available.

Gibson, Dunn &

below), Rasier-CA LLC, a wholly owned subsidiary of Uber is regulated as a TNC.

Gibson, Dunn &

licensing agreements allow this; some do not.

many of these would be associated with potentially substantial harms to drivers.

TECHNOLOGICAL SOLUTIONS HAVE REDUCED MARKET FAILURES

these technologies, with benefits to buyers and sellers alike.

information can significantly impede efficiency.

Increased coordination between buyers and sellers, perhaps through intermediaries or

Imperfect information is often described in terms of incomplete information (when a market

Gibson, Dunn &