Big Data and Competition

Policy: A US FTC Perspective

Alden F. Abbott
General Counsel, U.S. Federal Trade Commission
Penn Wharton China Center, Beijing, China, July 6, 2019
(Views set forth herein are solely attributable to me)
 Introduction
• Delighted to be here today, my thanks to Professor Christopher Yoo
for having invited me, as well as to the three cosponsoring
institutions, the University of Pennsylvania Law School’s Center for
Technology, Innovation, and Competition; the Universität Mannheim;
and the University of International Business and Economics.
• The views expressed today are my own, they are not attributable to
the Federal Trade Commission or any Federal Trade Commissioner.
• Today I will discuss the FTC’s enforcement approach with regard to
“big data,” and comment on the way in which U.S. competition law
agencies approach matters having a “big data” component.
 Big Data: The Big Picture
• Before turning to specifics, let me take a very high level look at
potential competitive issues raised by possession and use of big data.
• A 2016 FTC report explains that the term “big data” refers to a
confluence of factors, including the nearly ubiquitous collection of
consumer data from a variety of sources, the plummeting cost of data
storage, and powerful new capabilities to analyze data to draw
connections and make inferences and predictions.
• It is a mistake to discuss “big data” in the abstract. Access to big data,
alone, should not be presumed to create market power or
competitive advantages such as entry barriers.
 The Big Picture, continued
• 3 big data characteristics caution against assuming that the
possession of bid data automatically conveys market power: (1) many
types of data are readily available and replicable; (2) multiple entities
can often collect and use the same set of data without foreclosure
concerns; and (3) data can quickly become obsolete.
• Thus, it is necessary to determine whether big data create a
competitive concern on a case-by-case basis. Whether control of a
particular type of data harms the competitive process will depend on
the specific markets at issue.
 The Big Picture, continued
• Competition authorities should carefully analyze what (if any) competitive
advantages stem from mere possession of a data set.
• Shelf life may be short, much data may become obsolete relatively quickly. The focus should
be on potential for actual competitive effects, such as the creation and strengthening of
entry barriers and market foreclosure.
• U.S. antitrust law disfavors requiring access to an “essential facility,” e.g., big data (see
Verizon v. Trinko (2004), severely limiting essential facility claims).
• See 2014 FTC Report on Data Brokers (data available and abundant from many sources)
• Treating the mere possession of data as a barrier to competition could have a
significant chilling effect on innovation. Concerns about foreclosure risks
stemming from network effects must be weighed against efficiency benefits
generated by such effects.
• Does big data holder engage in exclusionary behavior? (monopolization).
• Will combining big data holders threaten credibly to reduce competition in
particular markets? (merger analysis).
Big Picture, continued: Monopoly Leveraging
• U.S. antitrust law does not recognize a standalone monopoly leveraging
offense, i.e., use of monopoly in one market to gain competitive advantage
in a second market.
• In a leveraging scenario involving a monopolist, US law finds no violation
unless monopoly is seriously threatened (there must be a “dangerous
probability of success” in monopolizing second market, Verizon v. Trinko).
• Many forms of “leveraging,” such as technological ties, can represent an
efficient form of product intergration or product enhancement that
benefits consumer and is therefore procompetitive. Thus, as a practical
matter, monopoly leveraging highly is viewed most skeptically in the U.S.
• The EU and other jurisdictions find abuse of a dominant position in the first
market when competition is merely distorted in the second market.
 Big Picture, continued – Remedies
• AS TO PROPOSED REMEDIES, note that legally mandated data access, data sharing, or
data pooling involves significant administrative costs.
• There may be less incentive to develop a collection of data if it is likely that the collection
will be subject to forced sharing. Mandatory sharing may also cause enhanced risks of
cartelization.
• To the extent remedies are required to offset anticompetitive effects connected to the
control of a set of data, those remedies should be narrowly tailored to specifically
address the perceived harm.
• When antitrust does intervene, competition agency should ensure that (1) feasible
remedies to address the competitive concern exist; and (2) those remedies do not pose
their own prohibitive costs or other risks to the competitive process.
• Take care that any remedy does not lead to worse competitive outcomes, whether due
to a chilling effect on incentives to innovate or due to the increased risk of collusion that
information sharing presents.
 FTC Tech-Related Antitrust Enforcement
• The FTC’s antitrust enforcement wing, the Bureau of Competition, formed
a Technology Task Force in February 2019 dedicated to monitoring
competition in high tech markets, taking appropriate enforcement actions.
• The Task Force is monitoring tech products and services, including
industries within the online advertising, social media, software and
application, and mobile spaces. Along with exploring industry practices
and law enforcement investigations, the Task Force will review
consummated and proposed technology mergers.
• Big data issues will, of course, be confronted by the Task Force.
• Big data issues were also examined at recently concluded FTC Hearings on
Competition and Consumer Protection in 21st Century, stay tuned for
possible FTC policy developments based on the findings of the Hearings.
 Special Aspects of Big Data and Platforms
• Specialized data related to personal information — think real estate records or
credit data — have previously been subject to antitrust enforcement actions.
• In today’s online world, however, the antitrust debate centers around how to
treat data about or created by consumers that is collected through online
platforms and used by these entities to target ads, improve current offerings, and
create new products. This type of consumer data is often an input for other
products and services.
• For example, Waze (owned by Google) collects and aggregates the location and
speed of travel of individual users’ phones and uses it to produce dynamic trip
directions based on changing traffic conditions.
• Consumer data is also a commodity asset for advertisers, allowing them to target
their ads more precisely, which makes those ads more valuable and thus allows
the platforms that hold such data to charge a higher price for that advertising
space than other advertising channels.
 Data Issues in U.S. Merger Enforcement
• U.S. merger enforcers have examined data-related issues, examples below.
• For instance, in Bazaarvoice (2014), the Justice Department (DOJ) successfully
challenged a 2012 consummated merger involving companies that provide
software platforms for online ratings and reviews (“R&R”) of products created by
consumers that manufacturers and retailers host, share, distribute, and display.
The court found a relevant market for R&R platforms, noted that the merging
parties had called themselves duopolists in this market, and found that the
merged firm likely would be able to charge monopolistic prices. In a settlement of
the case, DOJ required Bazaarvoice to divest all of the assets it had acquired in
2012.
• And in a series of mergers involving entities with databases of public real estate
records used for title insurance underwriting (called title plants), the FTC required
the merging parties to sell a copy of their title plant.
• For other examples, see Sher & Yost, US: Digital Platforms (Ant. Rev. Am. 2019).
 U.S. Merger Enforcement, continued
• In Corelogic (2014), CoreLogic, Inc. agreed to settle FTC charges that its
proposed $661 million acquisition of DataQuick Information Systems, Inc.
from TPG VI Ontario 1 AIV L.P. would likely substantially lessen competition
in the market for national assessor and recorder bulk data. The FTC’s
settlement order required CoreLogic to license to Renwood RealtyTrac
national assessor and recorder bulk data as well as several ancillary data
sets that DataQuick provides to its customers. The order allowed RealtyTrac
to offer customers the data and services that DataQuick now offers and to
become an effective competitor in the market.
• A 2018 FTC order modification (in light of poor compliance) required
CoreLogic to provide bulk data to RealtyTrac until at least 2022, an
additional three years beyond the term in the 2014 order.
 U.S. Merger Enforcement, continued
• In 2015, DOJ sued to block Cox Automotive’s acquisition of Dealertrack. Cox owns the
AutoTrader and Kelley Blue Book brands.
• As part of its acquisition, Cox sought to purchase Dealertrack’s inventory management
solution business (“IMS”) — a business unit devoted to providing analytics and
algorithms to assist car dealers with the management of their vehicle inventory.
DealerTrack also held ownership of valuable vehicle information data.
• DOJ was concerned that Cox would not only become an effective monopolist in the IMS
market but also would acquire valuable vehicle information data that served as inputs to
IMS businesses. With control over that data, Cox could “deny or restrict access” to the
data “and thereby unilaterally undermine the competitive viability of Cox’s remaining
IMS competitors.”
• To allow the deal to go through, DOJ not only required Cox to divest the IMS portion of
Dealertrack’s business, it also required Cox to enable the continuing exchange of data
and content between the websites it owns and the divested IMS business.
 U.S. Merger Enforcement, continued
• In CDK/Auto-Mate (2018), FTC sued to block a merger of two digital
tech platforms where firms were current competitors, but one was a
market giant—close to a duopolist—while the other was far smaller.
• Complaint alleged harm to current competition, but focused even
more sharply on harm to future, or nascent competition. That harm
arose from the smaller competitor’s substantial efforts to remake
itself into a greater competitive threat going forward. The transaction
was abandoned by the parties after the FTC filed suit.
• This case illustrates FTC’s ability to deal with threats to nascent
competition through mergers employing existing legal tools.
 Monopolization and Big Data
• U.S. monopolization law can be applied when appropriate to counter
anticompetitive actions (“exclusionary conduct”) involving big data
• Professor Hovenkamp, the leading U.S. treatise writer, concludes that
exclusionary conduct involves acts that “are reasonably capable of creating,
enlarging or prolonging monopoly power by impairing the opportunities of
rivals [need not show actual harm]; and (a) do not benefit consumers at all,
or (b) are unnecessary for the particular consumer benefits the acts
produce, or (c) produce harms disproportionate to the resulting benefits”
• Key U.S. monopolization decision is Microsoft (D.C. Circuit 2001 en banc), in
which the entire U.S. Court of Appeals for the D.C. Circuit (sitting en banc)
asked “whether as a general matter the exclusion of nascent threats is the
type of conduct that is reasonably capable of contributing significantly to a
defendant’s continued monopoly power” – widely cited by U.S. courts and
scholars and readily applicable to cases involving high tech and big data
 More on U.S. v. Microsoft (2001)
• D.C. Circuit found Microsoft violated Sherman Act § 2 by commingling
computer code for its Windows operating system and its Internet Explorer
(IE) web browser, requiring all Microsoft Windows purchasers to accept
pre-installed version of Explorer as well.
• Since computer manufacturers did not want to support 2 versions of same
program, effect of commingling was virtually to eliminate chief IE rival,
Netscape, from original distribution part of browser market.
• This in turn made it much harder for Netscape to develop tools to make
computers compatible with different operating systems, thus letting
Microsoft inefficiently maintain its operating system monopoly.
• No plausible procompetitive justification for Microsoft’s commingling.
 Digital Platforms and Antitrust: AmEx Case
• In Ohio v. American Express (2018), the U.S. Supreme Court required
examination of effects on both sides of AmEx’s digital platform – services
to merchants and services to cardholders.
• It thus held that higher nominal prices to merchants due to Amex contract
clauses barring merchants from “steering” customers toward using other
credit cards did not show a prima facie anticompetitive effect – quality and
output on the merchant side of the market also had to be examined to
determine if there was competitive harm.
• The Court stressed this case involved a “transactions platform” in which a
sale required parties on both sides of platform to simultaneously agree to
use AmEx’s services. Not all platforms required a “two-sided” market
definition to carry out antitrust analysis (future impact of case uncertain).
 FTC Platform-Related Monopolization Case
• In April 2019, FTC sued health information company Surescripts (SS) in federal district court,
alleging that the company employed illegal vertical and horizontal restraints in order to maintain
its monopolies over two electronic prescribing, or “e-prescribing,” markets: routing and eligibility.
• E-prescribing provides a safer, more accurate, and lower-cost means to communicate and process
patient prescriptions than traditional paper prescribing.
• FTC alleged SS monopolized two separate markets for e-prescription services: The market for
routing e-prescriptions, which uses technology that enables health care providers to send
electronic prescriptions directly to pharmacies; and the market for determining eligibility, a
separate service that enables health care providers to electronically determine patients’ eligibility
for prescription coverage through access to insurance coverage and benefits information.
• FTC asserted that SS intentionally set out to keep e-prescription routing and eligibility customers
on both sides of each market from using additional platforms (a practice known as multihoming),
using anticompetitive exclusivity agreements, threats, and other exclusionary tactics.
• Among other things, FTC alleged that SS took steps to increase the costs of routing and eligibility
multihoming through loyalty and exclusivity contracts. According to the FTC’s complaint, SS
successfully used these tactics to stop multiple attempts by other companies to enhance
competition in the routing and eligibility markets.
 Competition-Consumer Protection Issues
• Finally, what about interrelationship, if any, between competition and
consumer protection concerns regarding big data?
• Sharing as a competition remedy has traditionally been invoked where data
is difficult or expensive to create, raising an entry barrier that keeps out
competitors who need access to such data. In the U.S., this has been
imposed typically in a merger analysis, where two holders of such a data
set want to combine (see previous discussion).
• By contrast, the concern driving privacy law, like the European Union’s
General Data Privacy Regulation, or GDPR, is that consumer data has
become too widely available, with a perceived loss of consumer control.
The GDPR’s remedy adopted for privacy concerns limits collection and
restricts sharing of data, except at the consumer’s direction.
• Tension between competition and consumer protection approaches?
 Competition-Consumer Protection: Update
• On July 23 DOJ announced it has opened a broad antitrust review of bit tech
companies (WSJ – investigation was prompted by “new Washington threats” from
Facebook, Google, Amazon, and Apple).
• On July 24 Facebook revealed it is under FTC antitrust investigation.
• Antitrust and consumer protection raise different issues/concerns.
• For example, in July 24 press conference discussing settlement order between FTC and
Facebook over Facebook’s defective privacy practices, an FTC staffer stressed that deceptive
conduct at heart of this settlement implicated consumer protection concerns, not antitrust
concerns.
• In February 2019 the German competition agency (BKT) held that Facebook
abused its market dominance in Germany by conditioning use of its social
network on the collection of user data from multiple sources, and ordered
Facebook to change the way it collects data from German users.
• In August 2019 the Higher Regional Court of Dusseldorf expressed “serious doubts about the
legality” of this decision, and suspended it pending a final verdict – BKT is appealing.
 Conclusion
• The new high tech platforms and their use of large collections of data
raise major questions under both competition and consumer
protection law, in multiple jurisdictions.
• My comments, which represent only my views, have only scratched
the surface. Expect additional enforcement actions and policy
development here.
• Big digital platforms raise various other important policy issues (of
course), such as platforms’ roles in mediating speech by users and the
supervisory (or regulatory) role of government, that are beyond the
scope of my presentation.