Everyone seems to agree that we live in a “superstar economy” — one in which the leading companies (the likes of Apple, Amazon and Alphabet) are making dramatic strides both in productivity and in profitability. In response to this conventional wisdom the only question is: how should regulators respond?
There is a case that we should let the superstars keep doing what they do. Enjoy the fruits of their creativity. The smartphone! Internet search that works! Next-day delivery of next week’s landfill! For now, the superstars are innovative and efficient. In due course new monopolists will arise, motivated to seize the crown.
This case is not absurd, although my own instincts point the other way. Profit-minded monopolists can be innovative, but all too often they fail to adapt. More often markets deliver because they support a messy, pluralistic process of trial and error.
If that view is right, government intervention may be necessary to prevent the monopolies of today choking off the innovations of the future. One possibility is to break up the largest companies. Regulators might rule that a company cannot simultaneously own a platform while also offering products on it. This is the position taken in an artful argument published last week by US presidential hopeful Elizabeth Warren.
An alternative approach, set out this week by the UK’s Digital Competition Expert Panel, is to regulate the behaviour of the big digital groups rather than to change their structure. Regulators could insist on standards for data sharing and interconnection with established players, allowing smaller competitors to grow on or around the existing platforms.
I’m sympathetic to both approaches, although worry that it is easy for regulators to do more harm than good. Ms Warren’s rousing call to arms might easily lead to fruitless antitrust trench warfare. But set aside for a moment the argument about how to respond to the superstars and ask a different question. What if they are actually economic white dwarfs, dense but dim and fading?
That seems a strange conjecture. It appears to contradict the problem of laggard companies: research from the OECD suggests that many companies are far less productive than those at the frontier of their industrial sector, and that this gap is getting bigger. This suggests that superstars are blazing hot and the problem lies with the rest of the economy.
But much depends on what we mean by superstar companies. The OECD research focuses on highly productive concerns. Many of them are surprisingly small, with revenues in the tens of millions rather than the tens of billions.
Another alternative is to look at the most valuable 20 companies in the US economy, and the most valuable four in each of 60 or so different sectors, whether or not they are digital and whether or not they are highly productive. This is the approach taken by Germán Gutiérrez and Thomas Philippon, economists at New York University, in a new working paper. This perspective casts doubt on the very idea that the superstars of the US economy are particularly large or productive by postwar standards.
Compared with their forebears, these companies are unremarkable in both their sales and their employee numbers. They have a larger economic footprint than the leaders of 20 years ago, but smaller than 40 years ago. Alphabet, Amazon and Apple are impressive companies, but so were Intel, Microsoft and Walmart 20 years ago, or General Electric, General Motors and IBM before them.
Any large group contributes to the productivity of the economy as a whole in two ways: directly, by producing output, and indirectly, by drawing in resources from less productive players. Messrs Gutiérrez and Philippon reckon that the direct contribution of the leading companies is smaller than it used to be, while the indirect contribution has not increased by enough to compensate. Overall, they calculate that the contribution of the star companies to labour productivity growth in the US has fallen by 40 per cent since the year 2000.
This is a surprising result. Part of the explanation may lie in measurement: it is never easy to measure productivity, particularly in services and even more so in services provided free of charge in exchange for data and attention. But the “fading star” result may well be real, and surprising only because we often let the big digital groups stand as symbols of scale and market power. There is more to the US economy than Silicon Valley.
The US economy does seem to have a monopoly problem: concentration is increasing in most industries. Companies make money less by becoming more efficient, and more by exploiting their pricing power.
There is plenty for US antitrust authorities to get their teeth into. Perhaps they should take a look at Europe, where competition policy is more robust and politically independent, while EU markets have become more competitive and less profitable than those in the US. That may not be a coincidence.
Written for and first published in the Financial Times on 15 March 2019.