Digital Conglomerates and EU Competition Policy

Abstract

The paper analyses firms’ motivations and the competitive effects of digital conglomerates with the relevant industrial organisation and strategic management literature. On that basis, it makes recommendations to improve the methodologies and modes of operation of EU competition policy in the digital sector. The paper first shows that some of the characteristics of the digital economy may explain digital conglomerates. On the supply-side, those include the important economies of scope in product development as product innovation and development are often modular and based on shared inputs (such as data, hardware and software). On the demand-side, those include the consumer synergies generated by product ecosystems. The paper then shows that the pro- and anti-competitive effects of conglomerates are amplified in the digital economy. Regarding the anti-competitive effects, bundling may allow big platforms to envelop their smaller competitors in adjacent markets, raise entry barriers for innovating entrants or soften competition by increasing differentiation. The control of key sharable inputs may increase the incentives to refuse access or decrease the costs of an anti-competitive product proliferation strategy. Those effects are even stronger when the digital conglomerate has achieved the position of gatekeeper for access to customers or to specific products. Those anti-competitive effects should always be balanced with the positive welfare effects of digital conglomerates that are equally amplified. Moreover, conglomerate acquisitions of innovative start-ups may in some circumstances lead to a decrease in innovation, which is detrimental to welfare. The paper finally recommends some improvements in the enforcement of EU competition policy in digital markets. (i) Dynamic efficiency should be prioritised over static efficiencies; (ii) Market power should be assessed dynamically by focusing more on potential competition and by defining markets for sharable inputs and innovation capabilities; (iii) The theories of harms should be adapted to the firms’ incentives in the digital economy, in particular the anti-competitive bundling theories need to be extended, the threshold to impose access under the essential facilities doctrine needs to be adapted to the characteristics of data and the effects of a merger on innovation need to be directly taken into account; (iv) Antitrust intervention should be quicker and more agile and the standard of proof should not only take into account the risk of type I and type II errors but also the cost of those errors.