It is becoming increasingly common for companies to set prices for their products automatically using algorithms. One recent study found evidence of more than 500 sellers using algorithmic pricing on Amazon marketplace (see here). The Commission’s Preliminary Report on the E-commerce Sector Inquiry also reported use of automatic price adjustments (see paragraph 552).
There are two broad areas in which pricing algorithms could have the potential to cause anti-competitive effects. The first is ‘personal’ (potentially discriminatory) pricing, where for example a company quotes different prices to different people based on an algorithmic analysis of their personal data. The second is anti-competitive collusion, where companies use algorithms to automatically fix prices.
Both such examples have recently been flagged as areas of concern for competition authorities. Speaking at the 18th Bundeskartellamt IKK Conference in Berlin, Mr Justice Roth (chair of the UK’s Competition Appeal Tribunal) warned that the methods used by competition authorities will have to evolve to deal with the issues posed by personal pricing. At the same conference, Commissioner Vestager warned that “companies can’t escape responsibility for collusion by hiding behind a computer program” (here).
Is there really a threat to competition?
Competition authorities have already issued infringement decisions regarding the use of pricing algorithms, suggesting that the current competition law enforcement tools are sufficient to combat any concerns. Last year, the UK’s Competition and Markets Authority fined two companies that agreed to use algorithms to fix prices for the sale of posters and frames on Amazon (see here). In Eturas, the CJEU held that travel agents participating in a platform that implemented a discount cap could be liable if they knew about the anti-competitive agreement and failed to distance themselves from it (see here).
However, the actions of non-dominant companies in using pricing algorithms whilst acting independently do not fall within the current competition law framework, even if such use ultimately results in higher prices for consumers.
Yet it is easy to envisage that higher prices will result. If an online retailer can use an algorithm to track the prices used by another online retailer for common products, and immediately adjust its own prices to match any discounts, it can prevent the second online retailer from gaining a reputation for lower prices. The incentive for either retailer to lower its prices is removed. And although a 2013 OFT study suggested that businesses currently use personal data to offer consumers personalised discounts (see here), it isn’t hard (for the more cynical amongst us at least) to imagine businesses using such data to identify consumers, particularly those who are busy and pressed for time, who are likely to pay higher prices without questioning it.
What can competition authorities do about this?
First, it’s worth noting that the competition authorities are alert to the challenges posed by the digital economy more broadly. A joint paper by the French and German authorities, Competition Law and Data, examined the competition issues arising out of ‘big data’. The US FTC has also reported on the increasing collection, use and sale of consumers’ personal data (here). Commissioner Vestager’s recent speech made clear that the European Commission is also more specifically aware of the potential threat of pricing algorithms, warning that they “need to be built in a way that doesn’t allow them to collude”.
However, to the extent that actions by companies using pricing algorithms fall outside the current competition law framework (i.e. if a company implementing discriminatory pricing isn’t dominant, or companies using pricing algorithms have not entered into any agreements or concerted practices to do so), arguably the competition authorities should not try and stretch the existing law to cover these kinds of situations.
Instead, this could be more of a challenge for legislatures. It’s a question of policy: does the competition law framework need re-working to cover these sorts of issues?
To some extent this is already happening. Germany has already introduced significant changes to its antitrust laws to make it easier for the Bundeskartellamt to define markets and assess market power in the digital sector (see here and here), particularly where services are offered for ‘free’ and where multi-sided markets are involved. Commissioner Vestager has recently proposed a new directive designed to make national competition authorities more effective enforcers, ensuring for example that all national competition authorities have the power to search mobile phones, laptops, and tablets for evidence.
Differential (if not personal) pricing is also under the spotlight in relation to geo-blocking – where companies and online retailers apply barriers or impose restrictions on consumers on the basis of their nationality or place of residence. At present, such conduct can be examined under the competition rules only if it results from an agreement between separate undertakings or if the company holds a dominant position. The Commission has proposed a regulation designed end the enforcement gap in this area.
However, there is little concrete evidence of any action against the main issue identified in this article – namely, the potential anti-competitive effects arising from the independent use of pricing algorithms. Any change to the competition law framework designed to cover this would involve a significant (quite possibly unpalatable to many) change to the way competition law currently works around the world.
Does the competition law framework need to be changed?
Perhaps competition law isn’t, or shouldn’t be, the solution. In their book Virtual Competition, Professors Ezrachi and Stucke offer some other suggestions. For example, the government could promote market entry by companies with different economic incentives, for example consumer-owned co-operatives that redistribute profits via rebates. It could provide subsidies to companies using algorithms that actually promote customers’ interests, or sponsor ‘maverick’ firms that offer disruptive technologies or that are more likely to take the lead in cutting prices.
Perhaps no changes will be needed at all. As Professor Salil Mehra points out (here, p.52-53) the effective use of algorithms and big data have the potential to make businesses vastly more efficient and reduce their costs. This could ultimately result in lower prices for consumers, even if tacit collusion is occurring.
Either way, the pace of technological development is always likely to outstrip the pace of legislative change. It will be very interesting to see how these issues play out in the future.