This presents economies with the ultimate dilemmaNov 15th 2018Print edition | Special report
ONE OF THE few things that most people in the West agree on is that there is some kind of problem with the big tech firms. Here is a list of common complaints. They have high market shares and concentrated ownership so that tycoons mainly benefit from their growth, cause addiction, censor free speech, do not censor free speech, are infiltrated by Russian spies, suck up to Chinese autocrats, do not pay customers for their data, give private data to third parties, refuse to give data to third parties, don’t invest much, bully their critics, underpay workers, poach too many experts from universities, pay too little tax and corrupt politics. Some of these are directly about competition. In other cases the link is tangential. For example, if customers had good alternatives perhaps they would switch from today’s firms when the companies behave badly. Often there is no link to competition at all.
Tech firms get so much flak that it is worth considering the case for the defence. It is surprisingly easy to make. Consumers love their products. Between them the big Silicon Valley platform firms have 8bn customers. They have increased choice for consumers. If you want to watch the greatest hits of Scottish curling or Arnold Schwarzenegger you no longer have to dig around car-boot sales. Amazon has 353m products on sale, 3500 times more than the typical supermarket. In one poll Americans said they would have to be paid an average of $17,500 a year to forfeit the use of their search engine, which if true means that total search revenues could be 83 times higher than the sales of Google’s parent, Alphabet, last year.
Despite the ubiquitous use of the term “giant”, today’s tech firms are not unprecedentedly large. Ranked by domestic sales Apple is 14th in America, Amazon is 15th, Alphabet 37th and Facebook 107th. Uber and Airbnb are minnows that don’t even make the top 300. The tech firms are accused of extracting giant rents from society. But the largest five have lower earnings relative to the economy than the mightiest monopolists of the past did, with a median profit of 0.16% of GDP. That compares with a median of 0.24% of GDP for four historical goliaths in the year that antitrust regulators hit them: Standard Oil and US Steel (1911), IBM (1969) and AT&T (1974). For Amazon and Netflix the rents flow in the other direction because their prices are low today: in total they subsidise their combined 240m paying subscribers to the tune of about $50 per person per year, based on the amount of additional free cashflow they would have needed to cover their cost of capital in 2017.
Their effect on the economy has been positive in many ways. Online inflation is running at one percentage point below official inflation, reflecting the bargains available on the web. Economists have criticised the firms for employing only a few tech bros and creating no assets apart from executive Koi carp aquariums. But this view is out of date. The big five tech firms have almost 1m staff, not far off the 1.5m Walmart has in America. They are investing at a massive pace: some $137bn in 2017. As a result their combined hoarding rate (their free cashflow) actually fell from a peak of 0.66% of GDP 2015 to 0.61% last year.
The tech firms can be a powerful source of competition. Think of Amazon threatening to take on America’s rotten drug-distribution industry, or Netflix’s detonation of the cable-TV racket in America. The danger of digital disrupters is forcing comfortable incumbents to raise their game, from Germany’s car firms to Walmart. Meanwhile, the perception that big tech is entrenched is itself new. Facebook almost missed the mobile revolution: in 2012 it had fewer than 20 staff working on its core mobile team. Today Apple, Facebook and Google still depend on one main source of revenue. If they ever face a serious threat they could crumble.
Finally, tech firms are belatedly improving their conduct in response to the public outcry. Next year their collective tax rate is likely to be in line with the national average in America. On October 2nd Jeff Bezos at Amazon said that it would raise the minimum wage it pays its workers to $15 (twice the national minimum). Facebook has hired an army of people to clean up its content. As they are getting older they are getting nicer. What’s not to like?
To see why tech platforms might be a threat, you have to think about a different kind of corporate concentration: the market share of the mind. Typical Britons spend 24 hours a week online. Some 77% use social media, 89% say they compare prices online, and 57% use their phone as a ticket or boarding pass. A majority shop and bank on the web. All of these figures are higher for young adults, suggesting that usage will rise over time. Only 30% of Britons are fully aware of all the ways in which their personal data are being harvested.
Seen in this light, tech firms are becoming the conduit through which people interact with the world. Their business model is to impose a levy, either by charging users subscriptions and commissions, or by manipulating their buying decisions, or by charging other firms that want to access the platforms’ captive customers through advertising. The tech sector becomes a layer that sits across the entire consumer economy. Network effects mean that big firms are hard to dislodge. As they accumulate data on their users they become more enmeshed in their lives, making it expensive for customers to switch. And as artificial intelligence and data mining get cleverer, the platforms will manipulate their users even more, until eventually it is not clear who is in charge.
This future may sound abstract, but it is implicitly what Wall Street expects to happen. Politicians and regulators rarely consider stockmarket valuations, but they imply that tech firms will acquire an alarming amount of clout. To justify its valuation, Facebook’s rate of “monetisation” will have to surge, suggesting that it extracts a bigger fee from other firms who want to reach consumers. To justify its $820bn market value, Amazon will have to increase its share of American retail to 12% (Walmart’s share today is 7%). Likewise Netflix will have to roughly double its nominal fee per user over the next ten years. Though tech firms’ profits as a share of GDP today are not extraordinarily large, Wall Street is predicting they will be in a decade’s time, with the median ratio for the five firms rising to 0.28%. That is above the 0.24% median level of Standard Oil, US Steel, AT&T and IBM when they were each clobbered by antitrust regulators. The tech firms are expected to have higher returns on capital than the oligopolies of old, suggesting that they are better at extracting income per dollar of assets.
As they expand, the West’s tech firms have the capacity to disrupt new industries. But one test of their limited appetite for competition is to compare them with China’s two big tech platforms. Alibaba and Tencent are pouring money into a wild, economy-wide price war that has escalated to include digital payments, video, retail, games, travel, home delivery, cloud computing and music. Most of these areas are also contested by speculative upstarts that are losing billions of dollars a year subsidising customers. Compared with this violent scrum, America’s tech sector looks like a genteel game of badminton. During 2018 the two Chinese firms’ combined operating margin is forecast to fall by eight percentage points as price battles and new investments weigh. For the five American firms margins will stay roughly flat.
Meanwhile the development of new technologies could cause the tech firms’ manipulative capabilities to rise. Personal assistants could spy on and deceive their users. So far the practice of price discrimination, or charging different customers different prices for identical products, does not seem to be widespread. Alberto Cavallo, a scholar at Harvard University, has analysed Amazon’s prices and found that they are identical across different geographies 91 times out of 100. Still, it is easy to see how this could change. Imagine that Alexa, Amazon’s voice assistant, were to study your shopping habits and charge you a “surge” price when you run out of milk. Or what if Siri, Apple’s equivalent, were to track your email conversations with friends, conclude that you are keen to visit them and direct you to expensive flights, taking a cut from the airline? As data analysis and platform spying become more sophisticated, every price could be bespoke and opaque. Perhaps in the future, machines will even collude with each other to rig prices without humans knowing.
It would help if consumers had the capacity to switch away from technology networks that gradually start to exploit them. In the first half of the 20th century food brands competed on safety: the comforting knowledge that you would not find a dead mouse in your Kellogg’s cornflakes is why such firms did so well. Today’s tech markets seem stable but could flip suddenly if the powerful network effects that created them were to go into reverse. There could be a sudden surge of customers towards a new search engine or social-media firm that promised to guarantee users’ privacy or pay them for their data.
All this has become less likely because the big tech firms are ruthless about buying up nascent competitors. Facebook’s purchase in 2012 of Instagram, a rising social platform, for $1bn is perhaps the most famous example. The following year it tried to buy Snap, another potential rival, and when it could not, replicated many of its features. In 2014 it bought WhatsApp. Alphabet has invested in over 300 start-ups since 2013. The fear among start-ups of being bought or being crushed has led to talk of a “kill zone” surrounding the big firms which no new firm can survive.
The tech platforms therefore present the ultimate dilemma. They are the superstars of the Western world, and yet their business models have the potential to become malign, their market values imply they will become more powerful and their conduct suggests that they will avoid major confrontations with each other and wipe out potential competitors. The list of possible solutions is shorter than the list of grumbles. There are four main categories.
The first is to break the firms up. In some cases this could be done without huge damage. For example AWS, Amazon’s cloud division, could survive happily as an independent company, eliminating the danger of data about third parties gathered from it being used to influence Amazon’s e-commerce arm. But for many firms break-ups would also eliminate the benefit that customers say they get from participating in a network. It is possible that one of the dismembered parts of Facebook or Google would grow quickly to become as dominant as the erstwhile parent. For governments in all but two countries in the world, America and China, break-up would be hard since the big firms are not headquartered there.
The second approach would be to turn platform firms such as Facebook, Alphabet, Uber and Twitter into regulated utilities. Their prices and return on capital would be capped (causing profits to fall by about 65% and 81% for Alphabet and Facebook, respectively). Economic regulation would sit alongside a broader attempt to police the firms’ behaviour, just as utilities have to promise to provide clean water and reach every home. But state supervision would douse these firms’ innovative spirit. America’s two 20th-century experiments with price regulation were disasters. The airlines between 1938 and 1978, and AT&T until 1982, were inefficient empires that did not care about customers. Because of the politicisation of America’s institutions and the rise of lobbying, the risk of cronyism today would be high.
The third approach would be to counter the platforms’ power with people power. Users might team up to form large “customer unions” that bargained collectively with the tech firms, demanding better privacy terms or even payments for users in return for their data. Or new “digital concierge firms” might be set up to act as trusted intermediaries with the tech firms, ensuring that privacy terms were met and search results were fair, and arranging any flow of payments in return for the use of data. This could work well, though it is not clear that consumers can be bothered.
That leaves the last category: creating competition by force. The idea is not as strange as it sounds. Across the world governments have opened up consumer markets for energy and telecoms. In the case of tech the big firms could be prohibited from buying smaller ones. They could be forced to share their data and intellectual property with new entrants on reasonable terms. In 1956 regulators forced IBM to give other firms access to its patents in return for a fair fee, a decision that helped new tech firms emerge. And even if the agglomerated systems and databases of the tech firms remained intact, they could be forced to give customers ownership of their own data and pay them for their use (in the case of Facebook and Google this could amount to roughly $8 per customer per year). Customers would capture a bigger slice of the pie and have an incentive to switch to firms offering better terms.
Entrepreneurs are starting to experiment. For example, Sir Tim Berners-Lee, the inventor of the web, has launched a firm called Inrupt that offers a virtual “pod” where users can keep their data.
It took years of trial and error before markets for gas and power took off around the world. As well as requiring entrepreneurs, such a change needs deft regulators operating in strong institutional frameworks. At the moment those do not exist.