In the first of a series on subjects where economists are rethinking the basics, we look at arguments against letting businesses grow as large as they would like
Many antitrust experts are unconcerned: industrial concentration, they argue, does not tell you how competitive the market for a particular good is. But some economists have blamed falling levels of competition for far-reaching economic ills, such as stagnant labour markets and growing inequality. In a paper published in 2019 the late Emmanuel Farhi of Harvard and François Gourio of the Federal Reserve Bank of Chicago argued that the rising market power of big companies was linked to low interest rates and weak investment, factors shaping the whole economy.
许多反垄断专家对此并不担心。他们认为，产业集中度并不能说明特定商品的市场竞争程度如何。但有些经济学家认为，是竞争减少造成了劳动力市场停滞和不平等加剧等影响深远的经济弊病。在2019年发表的一篇论文中，哈佛大学已故经济学家伊曼纽尔·法希（Emmanuel Farhi）和芝加哥联储的弗朗索瓦·古里奥（François Gourio）认为，大公司不断增长的市场支配力与低利率和投资疲软有关，而这些因素影响了整体经济。
As in the days of the Chicago school, other economists see these critiques as ignoring the role of efficiency. A recently published paper by David Autor and John Van Reenen of MIT, David Dorn of the University of Zurich, Lawrence Katz of Harvard and Christina Patterson of the University of Chicago argues that globalisation and technological advances have concentrated economic activity in a small number of “superstar firms”. Because these firms are more productive, the industries which have seen the most of this concentration have also seen the fastest productivity growth.
和在芝加哥学派的时代一样，其他经济学家认为这些看法忽略了效率的作用。麻省理工学院的大卫·奥特尔（David Autor）和约翰·范雷南（John Van Reenen）、苏黎世大学的大卫·多恩（David Dorn）、哈佛大学的劳伦斯·卡茨（Lawrence Katz）和芝加哥大学的克里斯蒂娜·帕特森（Christina Patterson）最近发表的论文认为，全球化和技术进步将经济活动集中在少数“超级巨星公司”。由于这些公司的生产率更高，因此集中度最高的行业生产率增长也最快。
It is when they are applied to technology giants that these arguments get most heated. In America the Department of Justice, the Federal Trade Commission, Congress and many states are investigating whether Amazon’s dominant position in online shopping, Apple’s immense profitability or the duopoly that Facebook and Google enjoy in online advertising can be seen as involving the abuse of the giants’ market power. Google has been the subject of three separate competition investigations by the EU and fined €8.2bn ($9.7bn).
Competition in, or competition for?
Businesses built on “platforms”, as Amazon, Facebook and Google are, raise particular issues when it comes to competition because they have two separate sets of customers. Amazon deals with both retailers and consumers, Facebook and Google with both users and advertisers. In the 2000s Jean Tirole and Jean-Charles Rochet, two French economists, laid out an economic framework for looking at such platform businesses which showed that their optimal strategy will often be to provide cheap access to one side of the platform and charge steeply on the other. Consumers enjoy free Google searching and Facebook socialising; advertisers pay through the nose to reach them as they do so.
谈到竞争，亚马逊、Facebook和谷歌等基于“平台”的企业会引出一些特殊问题，因为它们有两类不同的客户。亚马逊的交易对象同时包括零售商和消费者，Facebook和谷歌则同时包括用户和广告主。在本世纪的头十年，两位法国经济学家让·梯若尔（Jean Tirole）和让·夏尔·罗歇（Jean-Charles Rochet）提出了研究此类平台公司的经济学框架，展示了这些公司的最佳策略通常是在平台一端让用户以低廉的成本访问，而在另一端高价收费。消费者免费享受谷歌的搜索和Facebook的社交功能，而广告主则要花大价钱才能在平台上打广告吸引消费者。
Platforms existed before big technology firms: television, newspapers and credit cards are all platforms of sorts. But the internet has provided vast scale and reach. Adding users is cheap, and it is often the case that the more users a platform has the more attractive it becomes to those yet to sign up. A firm that attains critical mass becomes overwhelmingly dominant: winner takes all.
Does it matter if the winning platforms dominate the digital economy? In terms of consumer welfare it seems, on the face of it, a sweet deal: users get stuff which is of real value to them at a price—zero—to which no one can object. But on the other side of the platforms things look more worrying. A recent investigation by Britain’s Competition and Markets Authority found that the cost of digital advertising for firms was worth £500 ($650) per household per year. Were the market less concentrated, those costs might fall—and some of the savings would be passed on to consumers in the form of lower prices.
如果通吃的平台主导了数字经济，有什么要紧吗？就消费者的福利而言，从表面上看这很合算：用户以没人会反对的价格（一分不花）获得对他们有真正价值的东西。但在平台的另一端，情况看起来就更令人担忧了。英国的竞争与市场管理局（Competition and Markets Authority）最近的一项调查发现，企业的数字广告成本平均为每年每户500英镑（650美元）。如果市场集中度降低，这些成本可能会下降，而节省下来的部分成本将通过降低价格惠及消费者。
Another potential worry is that there are conflicts of interest in many big-tech business models, such as when Apple sells through its app-store software which competes with its own, or when Amazon collects data about the sales of third-party products with which it competes.
Perhaps concentration would be tolerable if the big firms lived in fear of usurpation by a hot new entrant. But startup platforms face growing barriers to entry. One is amassing the reams of data which enable firms to tailor their services to individual users. Another is that in the digital economy it is relatively easy for an incumbent to see what it is that users like about what a startup offers, provide something similar and push it out to millions (if not billions) of existing users. That reduces the incentive to innovate in the first place. A final worry is that wealthy incumbents can close off the possibility of competition by buying new entrants before they pose a real threat, as when Facebook bought Instagram in 2012 and WhatsApp in 2014. In the decade to 2019 the five largest technology firms made over 400 acquisitions with scant intervention by competition authorities.
There are, broadly speaking, two sets of ideas for reforming competition economics and antitrust enforcement in response to these worries. Adherents of the more radical call themselves “neo-Brandeisians” after Louis Brandeis, an early-20th-century American Supreme Court justice who thought the overarching purpose of government antitrust action should be to prevent any one firm from exerting too much power over the economy. Neo-Brandeisians such as Lina Khan of Columbia Law School and Matt Stoller of the American Economic Liberties Project, a think-tank, want to broaden the purpose of antitrust investigations beyond promoting consumer welfare. Governments, they argue, should not fear breaking up the tech giants; they should fear leaving them be. In this view the companies’ size and power are a threat not just to consumers and workers but to democracy itself.
面对这些担忧，在改革竞争经济学和反垄断执法上大致有两种思路。其中较激进的思路的拥护者自称“新布兰代斯主义者”。路易斯·布兰代斯（Louis Brandeis）是20世纪初美国最高法院的大法官，他认为政府反垄断措施的总体目标应是防止任何一家公司对经济有过大的影响力 。哥伦比亚法学院（Columbia Law School）的丽娜·可汗（Lina Khan）和智库美国经济自由项目（American Economic Liberties Project）的马特·斯托勒（Matt Stoller）等新布兰代斯主义者希望反垄断调查的目的不应只是增进消费者福利。他们认为，政府不应害怕拆分科技巨头。它们应该担心的是任由这些巨头发展。这种观点认为公司的规模和实力不仅对消费者和劳动者构成威胁，对民主本身也构成威胁。
Each time I roam…
To its Chicago-school critics, Neo-Brandeisianism is “hipster antitrust”, replacing a transparent and rigorous methodology with an ill-defined set of social goals. It might disempower technology firms, but it would empower regulators. If concentrations of market power should be viewed with suspicion, so should concentrations of regulatory power: they bring the risk of arbitrary and unaccountable decision-making. In America, its home territory, this debate is predictably partisan: neo-Brandeisians are listened to only by Democrats.
The second set of ideas for reform is more incremental. It seeks not to abolish the consumer-welfare standard but to reinterpret it. Carl Shapiro of the University of California, Berkeley, has suggested calling it the “protecting competition standard” to make clear that it takes into account all the harm that anti-competitive practices might do to consumer welfare, including that which is indirect or diffuse.
Applying this interpretation of the consumer-welfare standard to digital platforms means accepting that in some situations firms will naturally grow large, meaning that at any point in time there will be little “competition in the market”. But there can still be “competition for the market” if a new, better product has a chance to disrupt the status quo. That might mean blocking more of the sort of early acquisitions which snuff out potential competitors, or reversing the burden of proof in such cases, so that the merging companies have to show that their plans will benefit consumers. It also might mean forcing incumbents to share some of their data, or at least making it easier for users to switch easily between platforms.
This agenda might not do much to satisfy neo-Brandeisian complaints about the political power of tech titans today. But it could succeed at making life at the top slightly more precarious. ■