Faculty of law blogs / UNIVERSITY OF OXFORD

Centros@20 Series: Two Comments on Elizabeth Pollman’s 'Tech, Regulatory Arbitrage, and Limits'

In a paper to be published in European Business Organization Law Review’s special issue on Centros@20 (summarised here), Elizabeth Pollman provides an answer to the questions of why we don’t see more regulatory arbitrage in the tech industry and what constrains it. Her paper draws upon an article by Victor Fleischer identifying a number of constraints on regulatory arbitrage, and adds three new ones: what she calls the ‘social license’, the bundling of laws and resources, and the fact that in some cases a different—and perhaps more radical—course of action is taken: ‘regulatory entrepreneurship’, or the strategy of ignoring, first, and then obtaining, changes to the law, rather than merely avoiding it.

This comment focuses on two elements in Professor Pollman’s analysis: the nature (and desirability) of regulatory arbitrage and the strength of social license constraints.

Professor Pollman’s treatment of regulatory arbitrage is unambiguously negative: it is a problem in search for a solution, or at least something which has to be constrained. In her own definition, 'the term consistently includes the notion of manipulation or strategic design of an activity to take advantage of specific legal or regulatory treatment' (emphasis added). This negative connotation is definitely consistent with the post-financial crisis world we live in, where exploitation of regulatory arbitrage opportunities has been identified as one of the many kinds of abusive behavior that almost led to a financial meltdown in 2008. The tax minimization schemes which giant tech companies have indulged in have only rightly increased skepticism about the idea that businesses may select the applicable laws and regulations.

It may be just a reflection of my having grown up intellectually in an environment with different views on the same phenomenon (the supposedly free-marketeer-dominated pre-crisis era), but I would still argue that there are virtues to granting businesses freedom to pick the laws they prefer in some areas. If that were not the case, the only explanation for the choice made by individual jurisdictions (and even by supra-national institutions like the EU) to explicitly allow for regulatory arbitrage would be that they are captive to special interests. While that may be true in some cases, one would hesitate to think it’s the case for all. The EU internal market construction itself is, in part, premised on the idea that businesses should be given some freedom to choose the EU member state that gives them the most favourable legal environment to engage in the production and sale of their products. The regulatory arbitrage opportunities unleashed with the Centros case are just a consequence of this general approach. But even within jurisdictions, a number of phenomena can be viewed as allowing firms to choose the relevant regulatory framework: think of different legal forms, or tax and accounting menus (the latter also embedded in EU accounting directives and regulations). This is just a reflection of the idea that, in many instances, one size does not fit all and, relatedly, that the lawmakers may not always be in the best position to couple an individual firm to a given regulatory framework based on one or the other characteristic it displays.

One should also not forget that the alternative to a world in which firms can pick and choose the rules they like is one where the state as a lawmaker has a monopoly over the supply of such rules. It is somewhat surprising that we tend to see private monopolists in a negative light, but we do not always take the same starting point when we think of a lawmaker in a similarly monopolistic position. Only an unfettered faith in the effectiveness of our democratic institutions can justify one’s belief that the monopolistic lawmaking power by public institutions will be abused less often and less harmfully for society than monopolistic market power by private firms.

Consider the most hideous form of regulatory arbitrage, international tax arbitrage: leaving aside its current, possibly extreme manifestations, it may be viewed as less abominable than we tend to think, if one focuses one’s mind on what the world would look like without it. In an age where political moderation can no longer be taken for granted, the total suppression of opportunities for tax arbitrage could easily lead to abuse of a state’s newly gained monopoly power in exacting taxes: abuse could for instance take the form of high, hard-to-sustain tax rates to finance short-term political projects favouring some key constituencies; or it could manifest itself as unequal treatment for industries or individual firms that are inimical to the ruling party.

Again, these vintage views of regulatory arbitrage are not meant to argue that regulatory arbitrage is always a positive phenomenon. Rather, that it can be a good or a bad one depending on circumstances that can obviously change in space and time.

One important constraint on regulatory arbitrage that Professor Pollman draws our attention to is the idea that all businesses operate thanks to a ‘social license’, an implicit permission to sell their products and services that society—or public opinion—grants them and may withdraw in the presence of serious misbehaviour. How the social license works in the tech sector, where companies often impose themselves on the market based on a 'break everything' attitude, is of particular interest. Pollman’s paper brings the example of Uber as a company that broke both laws and social conventions, which led at one point to the loss of its regulatory license in London. Whether that also had a significant impact on Uber users’ attitude towards the company is of course another story, and one that is difficult to find evidence about.

One important distinction that Professor Pollman makes is between abusive behaviour that harms users (such as higher prices for rides to the airport during protests against the executive order 13769, ie the ‘Muslim ban’) and abusive behaviour (including regulatory arbitrage) that harms others, such as drivers and competitors (such as Lyft or Waymo). The former should have a direct impact on profitability, the latter not necessarily so. Uber engaged in both: whether it also lost business as a consequence of both is another story, one that arguably, and more generally, has to do with the presence of competitors to which a disgruntled customer can switch to. If no competition is available, exploitation via abusive behaviour is just another form of monopoly rent extraction. Which leads to the final observation that with great power comes great responsibility: once a dominant position is reached, you exploit it at your own peril: the customers may be captive, but they may be unhappy too and convert their dissatisfaction into political sensitivity to the monopolist’s behaviour. Ultimately, it will be fear of a political reaction which will make the social license constraint bite. That is why the paper’s example of Uber’s travails with the London transportation authority fits as an illustration of the social license constraint.

Luca Enriques is Allen & Overy Professor of Corporate Law at University of Oxford.


With the support of