In this paper, we investigate the potential crowding out of private investment by public subsidies in the deployment of broadband fiber networks. We estimate a model of fiber entry using a rich dataset on fiber deployment for more than 34,000 municipalities in mainland France from 2014 to 2019. We then assess whether private investment would have occurred in subsidized municipalities in the absence of state aid. We find that in 36% of cases, public subsidies were allocated to municipalities where private entry would have occurred within three years. We estimate that about 902 million euros of the total 2,203 million euros in total subsidies disbursed by the end of 2019 may have crowded out private investment. However, we also show that the French broadband plan accelerated fiber coverage in subsidized municipalities in the early stages of deployment.

We study the impact of horizontal mergers on the incentives of merging firms to invest in incremental innovation. We provide a decomposition of this impact that clarifies the various forces at work and the differences between demand-enhancing and cost-reducing innovation. Moreover, we derive sufficient conditions for a merger to either reduce or raise the merging firms' incentives to innovate, and show that the comparison of the price diversion ratio and the innovation diversion ratio can help screen mergers. We also uncover a useful connection between the level of production synergies induced by a merger and its impact on innovation.

In this paper, we consider two platforms that compete for the development of a new product to integrate into their ecosystems. The new product can be developed either in-house by the platforms or by an independent startup active only in the technology market. The technology of the startup can be transferred to the platforms either exclusively (startup acquisition) or non-exclusively (licensing). A platform acquires the startup’s technology either because it failed to develop the technology itself or to prevent a rival from acquiring it (a killer acquisition). The presence of the startup affects the platforms’ R&D efforts through an insurance effect, which reduces the cost of failed innovation, and a competition effect, which diminishes the returns to innovation. The magnitude of these effects depends on the merger policy decided by the competition authorities. We show that allowing acquisitions stimulates platform innovation, but at the cost of a more concentrated market structure.

Platform interoperability is considered a powerful tool to promote competition in digital markets when network effects are at play. We study the effect of interoperability on competition between two ad-financed platforms, allowing for endogenous multi-homing of consumers. When the platforms are symmetric and decide non-cooperatively on their level of interoperability, interoperability emerges in equilibrium if the value of multi-homers relative to single-homers is sufficiently low for advertisers. From a welfare perspective, the equilibrium level of interoperability can be either too low or too high. When one (“large”) platform has an installed base of customers, its incentive to make its services interoperable is lower than for the other, smaller platform. However, mandating interoperability between the asymmetric platforms is not always socially optimal.

In this paper, we study the competition between two horizontally differentiated digital platforms. Each platform can adopt either a user-funded or an ad-funded business model. The platforms also invest in service quality to attract users. We find that which business model is more conducive to quality investment depends on the relative value of the platforms' services to the marginal multi-homing user compared to the value of the marginal user to advertisers. From a consumer welfare perspective, in equilibrium, platforms tend to adopt the user-funded business model too often and insufficiently differentiate their business models. Our results also show that regulatory interventions in digital markets aiming at mitigating market power of ad-tech firms or improving privacy protection can induce shifts in business models, with potentially ambiguous effects on consumer welfare.

The report aims to contribute to an effective implementation by the European Commission of the EU Digital Markets Act, which aims to increase contestability and fairness on the European digital markets, and ultimately augment users’ choice and digital innovation.The report starts by developing five good regulatory principles at the substantive level (effectiveness, proportionality, non-discrimination, legal predictability, and consistency with other EU laws) and four procedural principles: participation, ex ante and ex post evaluation of compliance measures, due process and ex post assessment of the law). Then the report applies those principles to a series of specific DMA obligations: choice architecture, horizontal and vertical interoperability and data related obligations. Finally, the report recommends a series of output indicators to contribute to the compliance assessment as well as the evaluation of the effectiveness of the law.

In this paper, we evaluate the efficiency of the French State aid plan for broadband deployment, the Plan France Très Haut Débit. According to State aid rules, public subsidies should not be substitute for private investment and should target areas with market failures. We estimate a structural model of fiber entry using a rich dataset on fiber deployment for more than 34,000 municipalities in mainland France over 2014-2019. We then assess whether private investment would have occurred in subsidized municipalities in the absence of public subsidies. We find that between 64% and 93% of the time, public subsidies were granted to municipalities where private entry would not have occurred. Overall, we estimate the cost of "inefficient" public subsidies to be between 243 and 902 million euros, with total subsidies amounting to 2,203 million euros by the end of 2019. Finally, we find that the plan helped to increase fiber coverage in subsidized municipalities in the early stages of fiber deployment.

We study the interaction between the holder of a standard-essential patent (SEP) and two downstream firms using the patented technology to design standard-compliant products. The SEP holder approaches the downstream firms simultaneously in the shadow of patent litigation and is subject to fair, reasonable, and non-discriminatory licensing requirements. We show that the patent holder faces a litigation credibility constraint and a license acceptability constraint when setting its licensing terms. For patents of intermediate strength, there is no royalty that allows the patent holder to reconcile these constraints. Consequently, it cannot license its technology and must go to court against infringers. We show that the availability of an injunction improves the patent holder's ability to license its technology, but it tends to inflate the royalty rate for implementers.

We develop a model of strategic geoblocking, where two competing multi-channel retailers, located in different countries, can decide to block access to their online store from foreign consumers. We characterize the equilibrium when firms decide unilaterally whether to introduce geoblocking restrictions. We show that geoblocking results in a “puppy dog” strategy (Fudenberg and Tirole, 1984) for firms, which allows them to soften competition, but that it comes at the cost of lower demand. In the short term, a ban on geoblocking leads to lower prices, both offline and online. However, in the longer term, when firms can invest in increasing the demand from online shoppers, the ban may have adverse effects on investment and social welfare. We extend our analysis to account for price discrimination and investigate the role of shipping costs.

Using weekly music charts data in 10 countries over the period 1990–2015, we analyze whether digitization leads to a trend of homogenization of music content or conversely to a greater acoustic disparity. We split the digitization era in four periods that correspond to four new emblematic distribution models (Napster, iTunes, YouTube, Spotify). Our main result is that while acoustic diversity decreased during the iTunes and the YouTube periods, the period that begins with the introduction of audio streaming services, such as Spotify, represents a turning point and is marked by a significant increase in acoustic diversity.

The literature is rather inconclusive when it comes to asserting whether digital technologies tend to favor collaboration (Do-It-Together, DIT) or creating alone (Do-It-Yourself, DIY) in creative production. In this paper, we argue that providing an answer to that question implies adopting a micro-perspective, which ties individual creators’ usage of different types of digital technologies, and their choices of DIT or DIY. Using data from a sample of French musicians, we find that while the use of some digital technologies is clearly associated with artists creating alone, other digital technologies have a more ambiguous association with DIY or DIT. We then uncover the boundary conditions of the association of these ambiguous technologies with DIY and DIT behaviors by showing how individual characteristics of the creators moderate this association.

We consider a platform that carries content from two upstream content providers and presents personalized recommendations to participating customers. We focus on streaming platforms in media markets, where users pay a subscription fee to join the platform but no usage fee, and consume a mix of content originating from each provider. We characterize the bias in the user-specific recommendations offered by the platform when one content provider charges lower royalties than the other. We establish that if consumers are sufficiently insensitive to bias, the recommendation system allows the platform to credibly threaten upstream providers to steer consumers away from their content, which reduces their market power. We also investigate the effects of vertical integration by the platform and show the robustness of our results to nonlinear (personalized) streaming services.

In an effort to ensure contestability and competition in digital markets, policymakers worldwide discuss whether interoperability obligations are an appropriate regulatory tool to achieve these goals. In particular, horizontal interoperability obligations have been proposed for messenger services and social networks, which would require the incumbent platform to be interoperable with a rival platform. The key feature of such horizontal interoperability is that it allows sharing of direct network effects, and similar regulation has been in place for telecommunications networks for decades. However, we show that horizontal interoperability obligations can be a harmful remedy in the context of digital services. The main reason for this assessment is that in the context of digital markets only partial interoperability can be achieved, because innovation is faced-paced and efforts to standardize a common set of interoperable features will always lag behind. Interoperability, by contrast, requires standardization and a relatively steady environment. Absent full interoperability some proprietary network effects remain and consumers still gravitate to the larger network in order to take advantage of the full richness of features. At the same time, horizontal interoperability lowers the incentives of consumers to multihome services, which is a powerful driver for contestability. We develop a dynamic multi-period model, which formalizes the trade-off between the (imperfect) sharing of network effects through interoperability and reduced incentives to multihome, and show that mandated interoperability can impede the ability of a more efficient entrant platform to contest the less efficient dominant platform. Our results have immediate implications for the ongoing policy debate and suggest that horizontal interoperability obligations may not be an appropriate remedy for regulating dominant online platforms.

Two firms compete in prices and information disclosure levels. Firms derive revenues from two possible channels, i.e., by selling their service to consumers and by exploiting user data, sold to a monopoly data broker. A consumer signing up to one firm's service decides on the amount of personal information to provide. In a single-homing framework, firms engage in either a strict privacy regime with no information disclosure and high prices or a flexible privacy regime with positive disclosure levels and low prices, depending on consumer valuations. With the possibility of multi-homing, firms face issues in the monetization of multi-homing user data, which affects privacy regimes. On top of consumer valuations, the incentives to multi-home and product differentiation also impact firms' strategies. Firms may even end up engaging in a zero-privacy regime with maximal disclosure levels if monetization issues on multi-homing user data are not too significant.

This paper aims at investigating whether the individual attitude towards risk is a predicator of the movie piracy behavior. This issue is of interest as public policy against digital piracy, such as the Graduated Response of the Hadopi law in France, aims at making piracy risky. In this paper, the attitude towards risk of French students, elicited using an experimental measure, are linked to survey data on digital piracy of movies. Our results indicate that attitude towards risk of the sample of students does not predict their online movie piracy behavior. This result is in line with previous studies showing the low efficiency of public policy against digital piracy.

In this paper, we study the impact of co-investment by incumbents and entrants on the roll-out of network infrastructures under demand uncertainty. We show that if entrants can wait to co-invest until demand is realized, the incumbents’ investment incentives are reduced and total coverage can be lower than in a benchmark with earlier co-investment. We consider two remedies to correct these distortions: (i) co-investment options purchased ex-ante by entrants from incumbents, and (ii) risk premia paid ex-post by entrants. We show that co-investment options cannot fully reestablish total coverage, while premia can do so in most cases, though at the cost of less entry. Finally, we show that an appropriate combination of ex-ante and ex-post remedies can improve welfare.

We study the competition between a private firm and public firms on prices andinvestment in new infrastructures. While the private firm maximizes its profits,public firms maximize the sum of their profits and consumer surplus, subject to abudget constraint. We consider two scenarios of public intervention, with a nationalpublic firm and with local public firms. In a monopoly benchmark, we find that thenational public firm has the highest coverage and charges a uniform price allowingcross-subsidies between high-cost and low-cost areas. Moreover, the private firmcovers as much as local public firms. In a mixed duopoly, a stronger competitivepressure drives firms' prices up while it drives down (up) the national public (private)firm's coverage.

It is crucial for content providers (CPs) to appear prominently on dominant online platforms in order to attract consumer demand. Apart from organic search results, content providers can obtain such prominence also in return for a monetary payment to the platform, e.g., in the form of sponsored search results. In this article, we investigate some of the economic consequences, if such payment can also be made with consumers’ data instead of money. Since data is non-rivalrous, the economic effects of data sharing for prominence are more complex and differ from paying for prominence. In a game-theoretic model we show that more consumer data will be collected as soon as CPs can obtain prominence on the platform. Whether the platform is more biased under a prominence-for-money scheme or under a prominence-for-data scheme depends on the marginal value of shared (non-exclusive) data. If this value is high, prominence-for-data will yield a higher platform bias, lead to more data collection by the CPs, and ultimately lower consumer surplus. Our results therefore bear important insights for the regulation of data-rich online platforms.

Across the world, regulators and policy makers are grappling with how to establish a competitive, safe and fair online environment that also safeguards users’ fundamental rights as citizens. Ahead of the European Commission’s Digital Markets Act (DMA), this book “Digital markets and online platforms: new perspectives on regulation and competition law“, presents CERRE’s latest contribution to the debate with concrete policy recommendations. Together, the policy recommendations in this book present a roadmap that should be pursued for EU policy makers to safeguard competition and innovation in digital platform markets. They can be organised into three key areas for action: (i) More effective enforcement, (ii) increased transparency and switching easiness, and (iii) providing access to key innovation capabilities.

d'analyse économique.La reproduction ou représentation de cet article, notamment par photocopie, n'est autorisée que dans les limites des conditions générales d'utilisation du site ou, le cas échéant, des conditions générales de la licence souscrite par votre établissement.Toute autre reproduction ou représentation, en tout ou partie, sous quelque forme et de quelque manière que ce soit, est interdite sauf accord préalable et écrit de l'éditeur, en dehors des cas prévus par la législation en vigueur en France.Il est précisé que son stockage dans une base de données est également interdit.