Skip to main content

Algorithms, Digital Platforms and Competition

Periodic Reporting for period 1 - DigitalComp (Algorithms, Digital Platforms and Competition)

Reporting period: 2019-02-11 to 2021-02-10

Dealing with challenges associated to digital platforms is currently one of the top three priorities according to the European Commission’s Digital Single Market strategy. Platforms can create immense value for the economy, online commerce and drive up productivity. But, at the same time, they have been disruptive and source of regulatory controversy. By design, in the core of platforms’ business models there are algorithms which are based on machine learning principles and use personal data as input to match efficiently and at real time supply with demand. This project analyses the dynamic impact of digital platforms on markets and consumers and addresses challenges that are associated with their disruptive operation, using a novel and multi-level economic approach. At the same time, it also explores the impact of algorithmic design and automated systems in decision making, market competition and society. The research agenda is separated in 3 chapters. The first chapter deals with algorithmic competition and evaluates whether big data raise entry barriers and what the incentives of algorithmic systems to discriminate are. It also assesses policy measures to increase algorithmic transparency and accountability. The second chapter deals with the dynamic nature of digital platforms. A firm with significant market power today might not be in the position to conserve its market power tomorrow because of the entry of, or drastic innovation by competitors. The chapter develops a methodology that defines a robust measure of future potential competition. It also provides insights over the creation and expansion of digital platforms in EU and US and illustrates firms' equilibrium market strategies in fast growing markets. Chapter 3 deals with the impact of automation on employment. By estimating the impact of the introduction of robots in EU industries on labor, it identifies the associated labor displacement and productivity effects and the optimal policy response.
The first chapter deals with the implications of data for the effectiveness of algorithmic systems as well as market competition and the incentives of firms with efficient algorithms that predict individual preferences to discriminate between their clients. The questions that are addressed include: i) when algorithmic bias and discrimination is emerging in firms’ equilibrium strategies; ii) how the market power of the algorithmic system is related to a firm’s incentives to discriminate; iii) what the social welfare implications of market discriminatory strategies are.
The work carried out so far for this topic had as an output two applied theory papers. The first one is inspired by the Infrastructure as a Service (IaaS) cloud computing market. The second one studies the search engine environment.
The second chapter deals with competition in digital markets. The first paper of this work package deals with developing an empirical test of potential competition and provide guidelines how it can be used when merger cases are evaluated. The second paper deals with developing a new framework for the study of market collusion and the incentives of firms to participate in it. The third paper deals with policy challenges and concerns related to market competition in digital markets and defines a new framework for policy interventions that can be both effective and maximize the value generated in these markets.
The third chapter deals with the implications of digital technologies on European labor markets. A comprehensive EU-wide empirical exercise is developed that provides new evidence on how digital technologies affect skills, wages and employment. The empirical study also considers policy instruments like quality of education, product market regulation and labor market flexibility and studies how they affect the implications of new technologies.
The overall message of Chapter 1 from the two theory papers is that algorithmic bias can exploit consumers and foreclosure the market by distorting competition. The policy paper is currently developed with the purpose to propose robust solutions that will contribute to algorithmic transparency and accountability.
The main message from Chapter 2 is that digital platform characteristics require particular attention when we evaluate competition concerns. New tools and tests need to be introduced in competition policy analysis.
Chapter 3 addresses empirically the implications of digital technologies in European labor markets and it identifies their impact of different skills and education groups. It also illustrates the important role of education, training and social protection in reducing concerns generated from technological advancements in EU employment.
Title picture of a selected publication