Data privacy: why internet giants elicit antitrust critiques
The Cambridge Analytica furore vindicates fear that groups such as Facebook are not benign monopolies
For a generation, competition law has been based on the premise that intervention is warranted if companies cause or may cause significant harm to competition in the market and thereby diminish consumer welfare. This orthodoxy holds that the size of a company does not occasion concern — for several reasons.
A company may be a large conglomerate enjoying no power in any market. It may be large because it requires scale to be efficient, or because it is simply better at what it does, and the rewards of competitive rivalry should not be diminished because competitors came second or third.
This orthodoxy is starting to be challenged. There are companies of such extraordinary power, size and reach that the traditional rationale of competition law — maximising consumer welfare — is simply too limited to rectify the harms they may cause. Companies that have attracted these concerns are the giants of the new internet economy: Google, Amazon and Facebook. From an orthodox perspective, this is puzzling.
These companies have grown quickly because they have provided innovative products, used ubiquitously, widely considered essential and provided free or at competitive prices. A conventional analysis of these companies would say that their size and dominance are a function of their success.
They have greatly added to consumer welfare and fundamentally changed the way we shop, procure information and communicate. Far from warranting antitrust censure, they exemplify what innovative market economies can produce.
If Google, Amazon and Facebook have created a consumer welfare bonanza, why the concern that these companies are not benign monopolies?
One critique reprises the very origins of antitrust legislation in the US. The Sherman Act and the Clayton Act were promoted by those who considered economic concentration a threat to liberty and democracy. In the 1930s, judges such as Louis Brandeis spoke from the bench of the right of the government to use antitrust laws to regulate concentrations of economic power to safeguard democracy.
These historical judgments seem rather contemporary. Facebook has been engulfed in a furore that has placed its integrity at the centre of the political stage. It permitted the data of about 50-million people to be obtained by a company, Cambridge Analytica, which used the data to target voters at the behest of Donald Trump’s presidential campaign.
The abuse of privacy and the use of data for political ends raise important issues. But these concerns may seem to travel some distance from matters relevant to antitrust.
There is a link, though. Facebook and Google are social networks of singular power and reach, despite the fact that people can choose not to use them.
But they do not have ready substitutes: to abstain is in a very real sense to be cut off from important resources and virtual communities. And because their use is free, consumers are also the product. The network generates valuable data that is sold in ways that are not transparent and can permit abuse.
Google too marshals data on a scale hitherto unimaginable. It is the medium through which the very nature of news is defined and redefined. Like Facebook, Google permits news of entirely undiscriminating pedigree to reach users. The information conveyed has no regard at all for what is true or even an approximation of truth, and prejudice is permitted to parade as fact. The liberation of information the internet creates is also its darkest attribute.
It is not straightforward to disentangle the antitrust problem from the issues of freedom and democracy. For some the worry is that social networks can be harnessed to bring about skewed political outcomes.
It could be argued that as long as the data is available to all who would use it, political contestation would be sharpened, not diminished, and democracy enhanced. The risk is that a single company has resources that can be used to influence political outcomes. And these resources would not be available to all who would use them or would only be available at a price that would make electoral results responsive to the highest bidder.
The challenge is to consider whether antitrust intervention is the correct response or whether other forms of regulation — for example, to secure privacy rights or ensure fair access to data for campaigning — are the better guarantee of democracy.
One response is that securing democracy is too important a value to be put to this choice. If Google and Facebook are monopoly networks whose conduct is unlikely to be disciplined by competition, intervention is warranted to prevent the abuse of economic power by companies that threaten democracy. The logic of this position is that the most extreme antitrust remedy, the break-up of Google and Facebook, would avoid the dangers to democracy that the size of these companies now pose.
If Facebook and Google were cut down to size and faced competitors, would it cure the problem? It would afford consumer choice. Abuse of privacy would allow consumers to switch providers rather than the current inferior alternative of switching off altogether.
Rivalry in the market does afford some discipline, but at some cost. The loss of scale and network efficiency may very well compromise the quality of the product. If Facebook no longer had 2-billion users would it have the same utility?
It is also not clear that the abuse is simply a function of size. The incentive to exploit data or transmit fake news would inhere in the baby Facebooks as much as in the pared-down parent. The most drastic remedy of break-up may therefore not be effective. But rivalry provides choice, demanding discipline from a dominant company that is acting with impunity.
Direct regulation of privacy as well as other measures to secure the democratic process may furthermore be more effective than the market’s discipline. The political consequences of data abuse and the indiscriminate transmission of information are of great concern. Facebook or Google are rightly not trusted to regulate themselves. But should an antitrust do so, and does it have the remedial tools for the job?
The rise of nationalism has placed this commonplace principle under some challenge. The orthodox basis of merger law is not the national identity of the acquiring or target firm, but the firm’s power in the market
A second challenge to orthodox antitrust aims to secure national interest. Until recently, the refusal of a merger on grounds of national interest occurred rarely in advanced economies, and then in cases, generally, where significant defence interests were at stake. The premise of merger control was permissive: mergers, in the absence of high concentrations of market power, should be allowed because they discipline supine companies and encourage trial and error in the market.
Cross-border mergers have become a familiar feature of globalised markets. Although merger control has remained a national competence, there has been a fair measure of uniformity across jurisdictions that the primary question in deciding whether to approve a merger is whether it gives rise to a significant lessening of competition in the relevant market.
The rise of nationalism has placed this commonplace principle under some challenge. The orthodox basis of merger law is not the national identity of the acquiring or target firm, but the firm’s power in the market.
The chauvinist perspective is different: firms are national assets; their ownership matters to the well-being of the nation. Falling into foreign hands may imperil the nation, and hence the need to intervene.
This position begins with a seemingly defensible stance. There may be some assets, held privately, and so necessary to the defence of the realm that they cannot be allowed to fall into foreign ownership. This stance assumes that domestic owners will manage these assets in the national interest and that the ownership of firms has a national character.
That is less and less so. If the asset should become necessary for defence and the composition of its ownership proved problematic, it could be brought under public control when the danger arose, not prospectively.
But the rise of nationalism as the source of public policy gives rise to expansions of public interest beyond vital defence interests. The takeover of GKN in the UK sparked conditions not simply based on defence assets but on the retention of capacity and investment commitments.
The risk to antitrust is great. Once sectors of the economy are conceived of as essential national assets, they are not subject to free exchange in the market. Public interest becomes a discretionary executive override and antitrust regulation sacrifices parts of its domain.
The issue then is no longer whether intervention is required to make the market work, but rather whether intervention takes place in spite of a transaction doing no harm in the market. These issues arise most especially regarding merger control. But one can imagine public interest issues coming to play a role in the assessment of abuse of market power or even cartels. Cartels that should be condemned are permitted in the public interest and the abuse of monopoly power is allowed to continue so as to secure a national champion.
This world of national assertion is some way off, but it is no longer impossible to imagine.
• Unterhalter practised competition law in SA and the UK and is a high court judge.