Breaking up big companies and market power concentration

On March 8th, US senator Elizabeth Warren published an essay outlining her proposal to break up the country’s big tech companies, specifically naming Amazon, Google and Facebook. These companies, Warren maintains, have amassed excessive power: “Nearly half of all e-commerce goes through Amazon”, while “more than 70% of all Internet traffic goes through sites owned or operated by Google or Facebook”.

What’s more, they have used this power to stifle competitors and limit innovation.  Warren blames “weak antitrust enforcement” for a “dramatic reduction in competition and innovation” in the sector, noting that “the number of tech startups has slumped, there are fewer high-growth young firms typical of the tech industry, and first financing rounds for tech startups have declined 22% since 2012”.

This reduction in competition and innovation, the argument goes, is taking place through two channels: acquiring potential competitors; and competing in a marketplace they own. Hence, the proposal to break-up big tech companies in fact consists of two distinct proposals for separation. First, to address the conflict of interest, companies that “offer to the public an online marketplace, an exchange, or a platform for connecting third parties” and generate annual global revenue above $25 billion would have to choose between owning the “Platform Utility” or participating in it.[1] Second, to re-establish a competitive landscape, Warren would appoint regulators tasked with reversing anti-competitive mergers using existing antitrust laws. She explicitly mentions the following mergers: Amazon with Whole Foods and Zappos; Facebook with Whatsapp and Instagram; Google with Waze, Nest and DoubleClick.

“America has a long tradition of breaking up companies when they have become too big and dominant,” says Warren. She recalls that when, in the 1990s, Microsoft “was trying to parlay its dominance in computer operating systems into dominance in the new area of web browsing”, the government brought an antitrust case against it that “helped clear a path for Internet companies like Google and Facebook to emerge” in the first place.

Warren is not alone in evoking examples from US antirust history to argue her position. Tyler Cowen, however, finds more similarities to an antitrust case brought against IBM in 1969 and takes the opposite view. The US government tried to break up the company when it “controlled nearly 70 per cent” of the computing-for-businesses market. “The suit ran on for 13 years, costing IBM and the government alike millions of dollars, not to mention the attention of IBM innovators,” writes Cowen. He adds that “the antitrust case arguably made IBM less able to identify the market shift toward personal computers” and contributed to a collapse of IBM’s market share and record losses.

In Cowen’s opinion, major tech companies have been very effective innovators, and the proposed break-up could similarly distract and weaken them, thus ultimately hurting innovation. In fact, Cowen credits big tech companies for what others often accuse them: rather than using acquisitions to eliminate potentially threatening competitors, they have enabled the rise of their acquisitions through their vast means. He cites Alphabet’s (Google’s parent company) acquisition of YouTube and Android, and subsequent investment in and upgrade of their content and services, as prime examples of that.

Cowen also challenges the idea that big tech companies form true monopolies. In the social network space, Facebook still competes with a large numbers of alternatives, digital or not, and adds that “it’s easy to imagine Facebook becoming less of a major player with time”. In advertising, Google and Facebook may be leaders in the space but are still competing against each other, as well as other conventional players (e.g. television). Cowen also argues that when it comes to advertising, Google “is fundamentally a price-lowering institution for small and niche businesses that can now afford more reach for less than ever before”, which actually enables competition in other sectors. Therefore, although he is not “suggesting that all is well in the online world, and some critics make entirely valid points”, “vigorous antitrust response hasty and harmful”.

By contrast, Kenneth Rogoff sides with Warren and remarks that “the debate about how to regulate the sector is eerily reminiscent of the debate over financial regulation in the early 2000s”.  Much like in finance then, supporters of a light regulatory touch cite the tech sector’s complexity to justify it, powerful companies can afford very high salaries that attract talent away from regulatory bodies, and the role of US regulators is “outsize”. Thus, although “ideas for regulating Big Tech are just sketches, and of course more serious analysis is warranted”, Rogoff “could not agree more that something needs to done, especially when it comes to Big Tech’s ability to buy out potential competitors and use their platform dominance to move into other lines of business”.

Published only five days after Warren’s essay, the ‘Report of the Digital Competition Expert Panel’ is yet another proposal for what should be done. The Economist summarises this report, prepared for the British government by a team led by Jason Furman, former chief economist to US President Obama.  The report’s authors argue that concentration is intrinsic to the digital economy because of network effects, and that sustained dominance can lead to higher prices, less choice and innovation.

However, they reject the breaking-up of digital companies or regulating them with capped profits and tight supervision, like water or power utilities. Instead, the report suggests that government action should focus on stimulating competition and choice. One such action is introducing a code of conduct on competitive behaviour on big platforms, echoing another element of the Warren proposal. This code of conduct would, “for example, prevent an online marketplace such as Amazon from favouring its own products over those of a rival in a search result shown to a consumer”. Another measure is “data mobility”, which would reduce switching costs. With data mobility “individual customers could move their search and purchasing histories from one platform to another. Social-media users could post their messages to friends, regardless of the networking site those friends use. And anonymised bulk data gathered by one firm would be made available to new entrants with safeguards for privacy”.

The article describes the report as “balanced” and “first rate”, but also offers some scepticism. It wonders how “data mobility” will work in practice. It also questions the impact the report could have. To quote, “even if Britain were to adopt its recommendations, the tech titans are global in scope and American by nationality. Ultimately America and the EU (which Britain is due to leave soon) are the powers that will decide their destiny”.

This line hints at the bigger picture: regulation of big tech goes beyond competition and innovation, and touches on politics, both domestic and international. In tabling her proposal Warren aims, among other things, to “restore the balance of power in our democracy” by ensuring the privacy of users’ data and reducing the leverage companies have in demanding “massive taxpayer handouts in exchange for doing business” vis-à-vis local administrations.

For the tech sector, Ken Rogoff thinks “it is a problem that cannot be overcome without addressing fundamental questions about the role of the state, privacy, and how US firms can compete globally against China, where the government is using domestic tech companies to collect data on its citizens at an exponential pace”. More generally, Joseph Stiglitz remarks that “as corporate behemoths’ market power has increased, so, too, has their ability to influence America’s money-driven politics”. Since “the challenge, as always, is political”, Stiglitz expresses doubt that “the American political system is up to the task of reform” and believes that “it is clear that Europe will have to take the lead”.

In one of its briefings The Economist goes even further, arguing that “if you want to understand where the world’s most powerful industry is heading, look not to Washington and California, but to Brussels and Berlin”. Not only is the EU more likely to be more objective owing to its lack of big tech firms; it is also where Alphabet, Amazon, Apple, Facebook and Microsoft have a quarter of sales and the world’s biggest economic block, meaning that its standards are often copied in the emerging world. What’s more, the EU has what the article calls “a distinct tech doctrine” which, similarly to the approach of the “Report of the Digital Competition Expert Panel”, rejects break-ups and utilities-style regulation and is rather based on equal treatment for rivals who use a platform and individuals’ sovereignty over their data. On the latter, the EU’s General Data Protection Regulation (GDPR) is to be followed by allowing interoperability, and thus customer switching, between digital service providers.

However, rising concentration is a problem that extends beyond the digital economy according to Federico Díez and Romain Duval of the IMF. Looking at product’s price to cost ratio, or markup, for nearly 1m companies from 27 advanced and emerging-market economies, they conclude that from 2000 to 2015 it increased on average by 6%. The increase is more pronounced in advanced economies and outside the manufacturing sector, including of course the digital economy. Importantly, the bulk of the increase in markups has come from those companies that had the highest markups to begin with. This top 10% of companies are more profitable, more productive and use more intangible assets than the rest. Díez and Duval argue that the rising power of the most productive and innovative companies has been helped by their superior ability to exploit proprietory intangible assets, network effects, and economies of scale, creating a “winner-takes-most” dynamic.

The authors write that this rise in market power has had non-negligible effects in reducing investment and the labour share in income. “If markups had remained at their 2000 levels, the stock of capital goods today would be on average about 3 percent higher and GDP about 1 percent higher”. Furthermore, “increased market power since 2000 has also accounted for at least 10 percent of the overall decline (0.2 out of 2 percentage points) in the share of national income paid to workers in advanced economies”.

But above all, Díez and Duval identify weaker incentives for innovation and the potential attempt of market-dominant companies to erect barriers to entry as reasons for policymakers to be vigilant for the future. To be sure, they propose a diversified policy toolkit: lower domestic barriers to entry, lower barriers to trade and foreign direct investment in services, stronger competition law and policies, corporate tax reform and intellectual property rights that “encourage groundbreaking innovations more than incremental ones”.