What would Sherlock Holmes make of Uber? The scene is wonderfully imagined in an episode of the US television series Elementary. Two fresh-faced co-founders of a fictional ride-sharing app, Zooss, are helping Holmes investigate a crime. But as he surveys a sea of screens, a sprawling spaghetti map of the comings and goings of Zooss passengers and drivers all over the city, Holmes becomes visibly agitated.

The entrepreneurs try to reassure him. ¡°This is literally all the information we have on our users: [who they are], where they are, where and when we pick them up, where and when we drop them off ¨C that¡¯s it,¡± they remark.

Incredulous, Holmes replies: ¡°That¡¯s it? You¡¯re describing a level of omniscience traditionally ascribed only to God and Father Christmas.¡± The Zooss founders intervene, attempting to quell rising concerns about algorithmic power, spying and tracking: ¡°For argument¡¯s sake, let¡¯s say someone here could do what you¡¯re saying. No one would. We have stock options.¡± They trail off, leaving Holmes to summarise their business model with devastating prescience: ¡°We are Big Brother incarnate ¨C but trust us, because our motives are purely financial.¡±

The question of what to do about technology platforms such as Uber, and super-platforms such as Google, Apple, Facebook and Amazon, is one of the greatest and thorniest issues now facing policymakers around the world. Behind the shiny interfaces and the hyper-efficient services lies a potent cocktail of challenges to privacy, competition, consumer protection and taxation. And, according to an increasingly urgent turn in scholarship by writers including Mireille Hildebrandt, Frank Pasquale, Julie E. Cohen, Brett Frischmann and Evan Selinger, these challenges are so potent that, if left unaddressed, they could prove fatal to law, democratic institutions and human agency as we know them.

ÁñÁ«ÊÓƵ

The supposedly steadying hand keeping this heady brew safe, consumers happy and the world rich with choice is competition in the market, backed up by the brass knuckles of competition law enforcement. But the adequacy of competition in the digital sphere has come under increasing scrutiny, particularly in the past two years. This is due in no small part to the current European Commissioner for Competition, Margrethe Vestager, who has so refreshingly and inspirationally taken on the task of challenging the likes of Google for abuse of its dominant position, and Apple for its abuse of state aid.

Unravelling the competition (or, to our American friends, antitrust) dimensions of the data-driven economy demands someone of the fearless but measured tenacity of Holmes or, indeed, Vestager. It requires penetrating a wall of rhetoric and myth, and a deep familiarity with competition policy¡¯s objectives and limitations.

ÁñÁ«ÊÓƵ

This is the task that two of the world¡¯s leading competition law scholars, Ariel Ezrachi and Maurice Stucke, have set themselves in Virtual Competition: The Promise and Perils of the Algorithm-Driven Economy. This highly readable and authoritative account sets out the ways that platforms have replaced the invisible hand with a digitised one ¨C a hand that is human-engineered, subject to corporate control and manipulation, and prone to charges of unlawfulness, on three fronts in particular. First, collusion. Second, behavioural discrimination. And third, asymmetric ¡°frenemy¡± dynamics, such as that between Uber and the super-platforms Google and Apple, which distort competition through extraction and capture.

¡°Humans have colluded on everything from turtles to packaged ice to rare banknotes,¡± write Ezrachi and Stucke. But the new digitally enabled collusion is at a remove from the smoke-filled hotel rooms of our collective consciousness ¨C it is in the silent energy-gobbling data centres that constitute ¡°the cloud¡±. The authors describe four ways in which computers may collude. Two are relatively straightforward: the ¡°messenger¡± scenario, which is the digital equivalent of a hotel room agreement, with computers shuttling exchanges; and ¡°hub-and-spoke¡±, such as the manner in which Uber¡¯s opaque surge pricing notionally optimises supply and pricing, but in a way that is inscrutable beyond the firm, and increasingly immune to competition as resources drain from underfunded alternative transport options.

The two other collusion scenarios are more elusive, subtle and difficult to ground in existing antitrust doctrine: the ¡°predictable agent¡±, where industry-wide pricing algorithms facilitate tacit collusion through interdependence and ripple effects, and the ¡°digital eye¡±, which is in effect what happens when machine learning meets the data-siphoning scene that so unnerved Sherlock Holmes in Elementary, giving ¡°a panoramic, God-like view of our state of being¡±. The question this raises, of course, is if tacit collusion is a plausible outcome of self-learning algorithms tuned to profit maximisation, is ¡°trust us¡± good enough any more? And will it protect the majority?

It is becoming increasingly apparent that widespread deployment of algorithmic tools can intensify, rather than reduce, the chasm between the wealthy and the vulnerable. This is the issue Ezrachi and Stucke address as behavioural discrimination. With ever-increasing hoards of data, firms can engage in near-perfect dynamic price discrimination, flipping our attributes, likes and fancies into individually enclosed and tailored worlds. Overall, they argue, this is corrosive to social welfare, because the more vulnerable among us end up paying more. The authors¡¯ assessment of where this is heading is of the most sober kind: absent legal intervention, perfect discrimination will likely become the new norm.

There is a lively gameness to Ezrachi and Stucke¡¯s study, marked by their willingness to call out what the world¡¯s internet users stare into every day ¨C monopoly power on a scale unlike any we have ever known ¨C and their systematic attempt to provide the language and tools needed to start tackling it. The book relies for some of its pivotal claims on an adjacent work that should be seen as a companion text: the unimaginatively titled but brilliantly executed Big Data and Competition Policy (2016). Co-authored by Stucke, this time with the distinguished US anti-trust practitioner Allen Grunes, it contains a detailed analysis of merger and antitrust cases and lucidly explores the interplay between privacy and competition in a way that neatly sets up the analysis, and fills some of the gaps, in Virtual Competition.

The constant aim of both of these works, and their clear achievement, is in exposing the facile mirage of competition in digital markets. The challenge they set down is what we might do about it. Uber, for example, running completely counter to what these scholars would demand, has continued to tweak its terms of service, continually and with impunity, and it retains data well above and beyond what is reasonable for its service. Its frenemy super-platform supporters, Google and Apple, have facilitated this intense data extraction, creating a toxic mixture of vertical and horizontal effects. But when we consider the paucity of actors with the will and might to do anything about it, Holmes¡¯ dark assessment rings true: between God and Father Christmas lies Uber.

¡°There will always be alternatives,¡± write Ezrachi and Stucke. ¡°So you can always walk at 1 a.m. on New Year¡¯s Day if you are dissatisfied with Uber¡¯s surge pricing. But that defense is available to any monopolist. And as we trudge through the snow, it becomes apparent that we are no longer in a market governed by an invisible hand.¡± In the digital world ¨C and indeed the physical one into which the ruling class of platforms and super-platforms are rapidly encroaching ¨C it seems it is the law of the fist that reigns supreme. And for us, as consumers and as citizens, it is ¡°a descent from king to slave on the data treadmill¡±.

Julia Powles is a postdoctoral researcher in the Faculty of Law and the Computer Laboratory, University of Cambridge.

ÁñÁ«ÊÓƵ

Virtual Competition: The Promise and Perils of the Algorithm-Driven Economy
By Ariel Ezrachi and Maurice E. Stucke
Harvard University Press,?378pp, ?22.95
ISBN 9780674545472
Published 24 November 2016

The authors

Ariel Ezrachi (above left), the Slaughter and May professor of competition law, University of Oxford, and director of Oxford¡¯s Centre for Competition Law and Policy, lives with his family in a small village just outside of Oxford. ¡°We very much enjoy its multicultural and diverse communities, the international atmosphere and its deep history, together with the tranquillity of the beautiful countryside.¡±?

His interest in his discipline originated in his late teens, ¡°from the realisation that it governs all aspects of our lives ¨C in particular our rights and our social and commercial relations. Law seemed to provide a master key to understanding much of what happens around us and one¡¯s possibilities in this changing world.¡±

As an undergraduate, he remembers reading Frank Easterbrook and Daniel Fischel¡¯s text The Economic Structure of Corporate Law?and being ¡°fascinated by the analysis of law and economics and their insights into bargaining and welfare maximisation. I think this book and more generally the study of corporate law stimulated my interest in economics and law and eventually in competition law and market dynamics.¡±?

Of his work to date with his co-author, Ezrachi recalls: ¡°A few years back Maurice and I began collaborating ¨C combining our work on quality in distribution and online search. We had a great time bouncing off ideas and exploring new areas of law and policy. What followed was a study leave that Maurice spent with his family at Oxford. This gave us a wonderful opportunity to spend hours discussing our work and developing it further. It also helps that we have a similar sense of humour!¡±

ÁñÁ«ÊÓƵ

Ezrachi has developed training for European judges. Does he see the judiciary ¨C and the European judiciary in particular ¨C and bodies such as the European Commission as key allies in the battle against digital giants¡¯ encroachment on civil liberties and fairer marketplaces?

¡°From a competition law perspective, the aim is to sustain a vibrant digital market dynamic with innovation and investment, while safeguarding consumers¡¯ interests,¡± he says. ¡°We presented our work to many competition enforcers. We found agencies to be extremely engaged and well versed on the challenges involved in identifying the adequate level of intervention.¡±?

Does a focus on competition law and the risks to competitive marketplaces posed by digital giants convince some people where civil liberties arguments do not??

¡°As a society, I think we are beginning to appreciate the cost of ¡®free¡¯. I hope our book and the discussion on market dynamics serve to highlight these issues. While some may find civil liberty arguments to be abstract, the misuse of data and its possible effect on the money in our pockets maybe more easily comprehended ¨C appreciating the adverse effect it may have on our welfare.¡±?

What gives him hope for 2017? ¡°The kindness of ordinary people.¡±

Ezrachi¡¯s co-author Maurice Stucke (above right), professor of law at the College of Law, University of Tennessee, was born in New York City and raised in Queens. ¡°Education, continually learning, and challenging oneself were highly valued in our family, with my father Harry a professor and dean at Long Island University, my mother Heike studying and lecturing on art therapy, and my sister Daniela Shumate teaching over the past 20 years.¡±?

He was, he says, not a particularly bookish child in early life. However, he adds, ¡°Sister?Eileen Gannon?at Archbishop Molloy High School opened the doors, suggesting supplemental books, such as Brideshead Revisited, Brave New World and A Clockwork Orange. We would discuss the books after class. She was always interested in my thoughts, encouraging, never reproachful. Reading became a rewarding pursuit.¡±

He began his undergraduate studies at Long Island University, where he launched a satirical magazine with his friends Alex O¡¯Meara and Ed Dumas. He then transferred to Georgetown University, ¡°where my parents, after witnessing years of athletic mediocrity, finally saw me win. With an inspiring coach, Pete Bautz, our undefeated novice boat won the national championships in rowing.¡±?

Of his collaboration with his co-author, Stucke says that during a sabbatical, ¡°Ariel graciously offered to host our family at Oxford. After rowing with my wife in a double, lunch with the fellows at Pembroke College was always a highlight. One afternoon, en route to lunch, we were discussing the rise of pricing algorithms, and wondered, ¡®What would happen if computers could collude?¡¯ That launched several articles and our book.¡± ??

Stucke chaired a committee on the media industry that drafted a transition report for the incoming administration of US president Barack Obama. How does he think Obama did on these issues?

¡°While campaigning, Obama promised to rejuvenate antitrust enforcement. But civil, non-merger enforcement was a disappointment, with only one monopolisation case. The rise of super-platforms, as our book discusses, raises significant concerns not only on our economy but on the marketplace of ideas. Thus the Obama administration failed on multiple levels: not checking the abuses of these giant tech firms, allowing the case law to atrophy, and ceding the intellectual leadership on antitrust to Europe.¡±?

Asked if he believes that those unconvinced by civil liberties arguments will see the danger to marketplaces posed by the tech giants, he responds, ¡°Perhaps. Historian Richard Hofstadter asked in the mid-1960s, ¡®What happened to the antitrust movement in the US?¡¯ Once the US had an antitrust movement without antitrust prosecutions, observed Hofstadter.?By the 1960s, however, there were antitrust prosecutions without an antitrust movement. Today we have far fewer antitrust prosecutions without an antitrust movement.¡± ?

He adds: ¡°Competition in the US appears to be decreasing in many economic sectors.?Witness?the decades-long decline in the number of new businesses being started and in the rate at which workers change jobs.?At the same time, many industries have become more concentrated, with profits increasingly falling into the hands of fewer firms. The anticompetitive effects we identify will only?exasperate?the wealth inequality. Barry Nalebuff, in reviewing our book, put it nicely: ¡®When the masses get mad enough, perhaps they¡¯ll elect a new trust-busting Teddy Roosevelt for the digital era.¡¯ Whether Trump will?revive antitrust remains to be seen. But we hope our book puts pressure on the next administration to start enforcing the laws.¡±?

What gives him hope for the coming year? ¡°Our four children.¡±

ÁñÁ«ÊÓƵ

Karen Shook