Big tech over-caution is bad for all of us

One big question in policymaking at present is why the long-term growth rate of economies has slowed markedly from where it was in the past and whether there is anything that can be done about that. There are lots of explanations offered. Let us consider the potential contribution of one often overlooked factor. It is probably not the largest one, but it is still significant.

To understand what this factor is, we need to think about the interaction of three things. First, a number of innovative sectors in our global economy have become dominated by single firms – the so-called “tech giants”. Some people see that as a problem in itself. I don’t. I support the orthodox view that even when these tech giants are the single or overwhelmingly majority firm operating in some of their sectors, they are not “monopolists” in the economist’s sense of having significant market power, because they face the threat of being replaced by a new entrant or of having another large tech rival, operating in a related sector, leverage its strong position there to enter their space. Economists call a market of this sort – in which there may be only one firm but it doesn’t have monopoly power despite that – a “contestable market”.

Second, in some sectors the best way to regulate isn’t to actually impose regulation but instead to discipline the behaviour of firms by threatening to intervene if they step too far out of line. The best-known version of this was in traditional UK financial regulation where the “Governor’s eyebrows” (as in Bank of England Governor) being raised was the metaphor for the regulatory authorities explicitly or implicitly threatening to intervene if behaviour were poor. A similar approach is sometimes used for tech giants too. This can be a good form of regulation in many sectors because explicit regulatory intervention tends to be blunt, making things too black-and-white resulting in too much being permitted or too much being banned. Firms are better placed to understand the shades of grey and to fine-tune their behaviour, under regulatory threat, in ways that explicit formal regulation would be unable to achieve.

So far so good. We have described two excellent forms of liberal regulation, allowing very large majority players in contestable markets, sometimes guiding their behaviour more by regulatory threat than by explicit formal regulation. This light touch regulatory approach should be flexible, encourage economic efficiency and good service to customers.

But now consider our third factor. In many innovative sectors the very large firms operate as conglomerates, owning not only their core businesses but businesses in other sectors too. So, for example, Google has a division – as has Apple – that is researching and experimenting with driverless cars. It’s obvious why. There are “economies of scope” – some of the tech thinking relevant to the development of search algorithms and artificial intelligence, along with geographic mapping used in searches, is relevant to the development of driverless cars. But here’s where the trouble starts.

In innovative sectors, risks must and ought to be taken. If one does not take enough risks, progress will be too slow. Across an economy as a whole, if not enough risks are taken in innovation then change and growth will be too slow. But a large conglomerate subject to regulatory threat may be reluctant to take reputational risks in case that triggers regulatory intervention. For example, Google may not want to risk its driverless cars being involved in crashes that injure people in case that leads to Google’s core business being regulated. Consumer activism across sectors may exacerbate the problem – if consumers seek to boycott a firm’s products in one sector because they perceive its activities in another sector poorly, that will again disincentivise optimal risk-taking in the emerging innovative business.

That means that what is efficient as a policy for a single-sector-focused business – the “Governor’s eyebrows” approach – may, when applied to large conglomerate firms operating across multiple sectors, inadvertently lead to the suppressing of innovation and slower GDP growth for the economy as a whole. Tech giants operating across multiple sectors may be unwilling to take as much reputational risk as would be economically optimal, in case the political fall-out of something going wrong damages their core businesses. And that could be bad for us all.