Regulatory Comments

ICLE Comments on 2025 EU Merger Review Guidelines

Topic A: Competitiveness and Resilience

A.1.   Do the current Guidelines provide clear, correct and comprehensive guidance on how merger control reflects the objective of having a productive and competitive economy?

Properly understood, “competitiveness” and “productivity” are the products of well-functioning, efficient markets. In this sense, they are fully consistent with the purpose of the EUMR: to safeguard competition and consumers from excessive market power, which distorts resource allocation, creates deadweight loss, and diminishes overall economic performance. In dynamic markets, productivity and competitiveness are mutually reinforcing: greater efficiency in the use of resources drives firms to compete more effectively on the European and global markets, while robust competitive pressures from inside and out provides incentives for firms to innovate and improve efficiency. As discussed in Section F, the Guidelines could offer clearer direction on how productivity and efficiency gains are weighed under the Significant Impediment of Effective Competition Test (“SIEC Test”). This guidance would be particularly timely in light of the Draghi Report’s recognition that size and scale can generate substantial economic benefits, including sustained productivity growth.

Arguably, the greater concern lies in potential misinterpretations of an increasingly prominent policy buzzword: competitiveness. Unlike “productivity”, which can be grounded in measurable efficiency gains (Syverson, 2011), “competitiveness” is not embedded in the current EUMR framework. It therefore could be misconstrued or selectively invoked to justify policies at odds with the EUMR’s objectives. A common fallacy is to equate European competitiveness with protecting European-owned firms from foreign rivals, including by preventing the acquisition of European firms by foreign undertakings. Sustainable competitiveness instead arises from maintaining an open, contestable, and innovation-friendly market that attracts capital, rewards efficiency, and drives long-term productivity growth.

Absent demonstrable public-policy concerns (e.g., security) of the kind contemplated in Article 21(4) EUMR, prohibiting or materially impeding mergers, partnerships, or joint ventures between European and foreign companies on the premise that such combinations would harm European competitiveness has it exactly backward. Access to foreign investment, capital, and know-how is a critical driver of firm growth and innovation (e.g., Valickova et al., 2015); restricting these inputs would stifle the development and scaling of European companies and ultimately erode—rather than enhance—Europe’s competitiveness.

To avoid confusion, the new Guidelines should make this point unequivocally: that “competitiveness” is a byproduct of a well-functioning, predictable EUMR anchored in unitary standards and sound economic analysis. Such an approach is essential to maintain confidence in the EU market, attract investment, scale promising companies, and ultimately ensure that the continent remains competitive. Predictable and consistent merger review—not discretionary or selective assessments of concentration—has long been Europe’s forte and one of its chief competitive advantages. The Guidelines should build on this strength, not squander it.

A.4.   What are the characteristics of markets where scale is necessary to compete effectively? Please be as specific as possible on the level of scale needed and why.

As the Draghi Report notes, companies and markets increasingly require scale to compete effectively—particularly in global, innovation-driven technology sectors. Markets where scale is decisive typically exhibit a combination of high fixed costs and low marginal costs, substantial R&D requirements, and strong network effects that reward larger user bases. In such markets, size often confers bargaining power to secure better inputs, prices, and even exclusive agreements with suppliers or distributors. Intangible assets—such as brand recognition and trust—further reinforce the advantages of scale, making it harder for smaller rivals to catch up. In such environments, achieving sufficient scale is often crucial.

In the tech sector, for instance, startups that cannot secure adequate funding and infrastructure often fail simply because they are unable to scale. In these contexts, mergers can play a critical role by providing access to the resources needed to compete. By pooling complementary assets and capabilities (Teece et al., 1997; Cirjevskis, 2019), a merger can enable a firm to move beyond the proof-of-concept stage and develop a commercially viable product. Without such combinations, many promising startups in these fields remain trapped in the lab, unable to reach the market.

A.5    What are the benefits that merged companies’ increased scale might bring to competitiveness:

A.5.1   In a scenario where the increased scale does not create or strengthen market power (e.g. a merger between complementary players in terms of products or geography)? Please select the benefits that you believe are relevant for increased competitiveness of the merged entity.

a. Network effects (i.e., whereby a product or service gains additional value as more people use it. Combining complementary assets (products, technologies, or geographic reach) can foster network effects that improve competitiveness and consumer welfare. Empirical studies consistently show that where network effects result from complementary integration, rather than exclusionary practices, they can lower transaction costs and spur innovation. For instance, research on payment systems and two-sided markets (e.g., Rochet & Tirole 2003; Evans & Schmalensee 2007) highlights that larger integrated platforms often deliver efficiencies that fragmented competitors cannot.

A concrete example would be the combination of complementary user bases: a merger between a payment-services provider strong in Europe and another strong in Asia could connect two previously fragmented networks, thereby increasing the overall value of the network to both consumers and merchants. Similarly, in online marketplaces, merging two regional platforms can generate stronger indirect (cross-side) network effects, as more buyers attract more sellers and vice versa, increasing liquidity and choice.

A practical illustration of such network effects is PayPal’s acquisition of Xoom in 2015. PayPal’s large user base in the United States and Europe was complemented by Xoom’s strength in remittance corridors to emerging markets. By merging, the two networks became more valuable to each participant: U.S. senders could now reach recipients in markets where PayPal previously had little presence, while Xoom users gained access to PayPal’s global merchant and consumer ecosystem. This expansion of complementary user bases increased cross-border transaction volumes, deepened liquidity, and enhanced the utility of the combined platform for both consumers and merchants.

Metrics to assess the benefits from network effects may include: the number of active users/merchants cross-border; buyer-seller ratios and growth; reduction in average time to match supply and demand; and retention/churn rates.

b. Intangible capital (assets lacking physical substance, e.g. patents, copyrights, goodwill, know-how). A startup can benefit significantly in a merger that offers access to the intangible capital of an established acquirer, providing it with valuable assets such as brand reputation, industry expertise, customer trust, and established relationships. This support is especially critical in a firm’s early stages, when its brand recognition and trust are limited, thus making market penetration challenging. Additionally, the acquirer’s proprietary knowledge and operational experience can enhance the startup’s product development and go-to-market strategies, increasing the likelihood of commercial success.

A practical example is Facebook’s acquisition of Instagram. At the time of the acquisition, Instagram was a rapidly growing startup, but relatively unknown beyond its core user base. After merging with Facebook, it gained access to Facebook’s strong brand, extensive marketing capabilities, and trusted reputation. This intangible capital helped Instagram to build credibility with both advertisers and users, accelerating its growth and monetization. Additionally, Facebook’s expertise in scaling social platforms enabled Instagram to refine its product roadmap and expand its infrastructure more efficiently. By leveraging Facebook’s brand, marketing expertise, and operational know-how, Instagram rapidly expanded its advertiser base and launched innovative features like Stories and Reels.

Metrics to measure the benefits of intangible capital post-merger include user growth and engagement (active users, retention); advertiser and customer expansion (number of advertisers, revenue per advertiser); brand recognition (brand awareness, NPS); product-development efficiency (time-to-market, feature launches); and revenue and monetization (total revenue, monetization rate).

For instance, following Facebook’s acquisition, Instagram’s user base grew from 30 million in 2012 to more than 1 billion active users by 2023. Before the merger, the platform also had very few advertisers and generated minimal ad revenue (close to zero). After the merger, however, Instagram became a major monetization engine, with ad revenue estimated at $30–35 billion annually by 2023. While precise causality is hard to identify in cases like this, there is at least a possibility this growth would not have been attained absent the merger with Facebook.

c. Access to equity investment. Mergers often allow smaller firms or startups to tap benefit from the financial and managerial resources of larger, well-capitalized acquirers (Manne, 1965). By integrating into a larger entity, a startup can leverage the acquirer’s balance sheet, investor credibility, and established financial channels to secure funding and deploy those funds more effectively than would be possible independently.

This access to capital is particularly important for scaling operations, investing in infrastructure, and pursuing global market expansion—activities that typically require resources far beyond what early-stage venture funding can provide. In the European context, access to equity investment through mergers is especially valuable, given the limited alternatives for raising large-scale capital. As highlighted in the Draghi Report, poor access to capital remains one of the EU’s primary competitive weaknesses, making mergers an important avenue to bolster the financial capacity of innovative firms.

A concrete example is Facebook’s acquisition of WhatsApp in 2014. At the time, WhatsApp had raised roughly $60 million in venture capital, which was sufficient for early operations but limited for global expansion. Following the acquisition, WhatsApp gained access to Facebook’s extensive financial resources and strategic support, enabling it to scale rapidly, enhance its infrastructure, and expand its user base worldwide. While the exact amount of capital made available post-acquisition was not publicly disclosed, it is widely understood to be orders of magnitude greater than pre-merger venture funding—thereby fuelling exponential growth. Further, WhatsApp’s user base grew from 450 million in 2014 to more than 2 billion users in the 2020s, reflecting both accelerated adoption and the operational advantages unlocked by access to the acquirer’s financial and strategic resources.

One or several of the following metrics could be used to evaluate improved access to equity investment: expected total capital availability; the number and credibility of potential investors; and anticipated reductions in the cost of capital. The Commission could also consider the planned deployment of funds for scaling operations, infrastructure, or R&D, along with forecasted impacts on revenue growth, user acquisition, and product launches. Finally, projected increases in firm valuation and the likelihood of attracting follow-on investment or strategic partnerships could provide additional indicators of the merger’s potential to strengthen targets’ financial capacity, especially where those targets are startups.

d. Ability and incentives to invest (e.g., in network infrastructure). Mergers between firms that sell complementary goods or services can significantly enhance both the ability and incentives to invest (Lafontaine & Slade, 2007). Combining resources allows the merged entity to pool financial, technological, and human capital, reducing the barriers to large-scale investments that would be difficult or risky for either firm independently. By consolidating financial resources, the merged entity may also be able to fund costly infrastructure projects, R&D, or platform upgrades more easily. Access to the acquirer’s balance sheet, operational expertise, and risk-management capabilities increases the firm’s capacity to undertake long-term investments.

A merger can also enhance the incentives to invest. A larger integrated firm can internalize the benefits of investment across its complementary assets, increasing expected returns and reducing the risk of underinvestment. Network effects amplify these incentives: investing in infrastructure that improves connectivity benefits a broader user base, enhancing overall platform value and attracting more users or partners. For instance, merging regional cloud-service providers allows may enable investment in data centres that serve a larger customer base, generating higher returns than isolated investments. The impact of such investments can be measured through metrics like capital expenditure on network or infrastructure projects; deployment speed and coverage; returns on investment; and adoption rates among users enabled by the improved infrastructure.

A practical example is Google’s acquisition of Waze in 2013. Google Maps and Waze offered complementary navigation services; Google Maps provided global mapping and directions, while Waze contributed real-time, crowd-sourced traffic data. By integrating Waze’s technology and user insights, Google could invest more efficiently in infrastructure such as real-time traffic analytics, routing algorithms, and server capacity. This enhanced both the ability and incentive to invest, as the combined platform could scale innovations across a much larger user base, improving returns on infrastructure spending without materially increasing market power

Metrics to gauge increased incentives to invest following a merger could reasonably include changes in capital allocation to infrastructure or R&D; acceleration in time-to-market for new products; ROI on investment projects; growth in user adoption or network coverage; and internal incentive alignment through KPIs or performance targets that encourage risk-taking and innovation. For example, in the case of Google Maps and Waze, these metrics can be observed in the increased investment in real-time traffic analytics, faster rollout of routing improvements, and expanded coverage benefiting a larger user base, reflecting the combined firm’s greater ability and stronger incentives to invest.

e. Ability and incentives to innovate (i.e., R&D, including high-risk innovation). Mergers between complementary firms can significantly enhance both their ability and incentives to innovate, particularly for R&D and high-risk projects. By combining resources, expertise, and intellectual property, the merged entity can undertake larger, more ambitious innovation projects than either firm could independently. Access to the acquirer’s technical talent, research infrastructure, and financial resources can also mitigate the risks associated with high-cost or high-uncertainty innovation, increasing the likelihood that breakthrough products or technologies will be developed.

A practical example is Google’s acquisition of DeepMind in 2014. DeepMind was a small AI startup with world-class research talent, but limited resources. Integration with Google provided access to vast computing infrastructure, data, and financial backing, allowing DeepMind to pursue high-risk AI-research projects—such as AlphaGo and AlphaFold—that would have been difficult to undertake independently. The merger enhanced both the firm’s ability to innovate—through access to hardware, data, and collaborative expertise—and its incentive to innovate, as successes could be scaled across Google’s products, generating significant returns on research investments.

Metrics to assess these benefits include a forecast of R&D expenditure before and after the merger; the number of patents filed, or research breakthroughs achieved; the speed and scale of product or technology development; adoption of innovations by end-users or other business units; and ROI or downstream commercial impact of high-risk projects. In the DeepMind example, these metrics are evident in the development and deployment of AI breakthroughs like AlphaGo, which combined high-risk research with substantial practical impact across Google’s ecosystem.

f. Ability and incentives to derive value from aggregation of data. Mergers between complementary firms can enhance both their ability and incentives to derive value from data aggregation. By combining datasets, the merged entity can generate richer insights, improve algorithmic performance, and develop new products or services that neither firm could achieve independently. It is, however, important to note that the ability to extract meaningful insights is ultimately more important than the sheer volume of data collected (Manne & Auer, 2020). All else being equal and assuming the acquirer is capable of leveraging data effectively, access to larger and more diverse datasets can enable high-value applications like personalized recommendations, targeted advertising, and improved operational efficiency. Simultaneously, the merged entity has stronger incentives to invest in data analytics and infrastructure, as the benefits of enhanced insights can be captured across a broader platform or user base.

A practical example is Facebook’s acquisition of Instagram, which allowed the company to integrate Instagram’s user data with its existing social graph. This aggregation enabled more effective ad targeting, improved content recommendation algorithms, and better user-engagement analytics. Even before full integration, such forward-looking indicators as the projected increase in actionable insights, potential ad-revenue growth, and anticipated improvements in algorithmic accuracy all signalled that the merger would likely be procompetitive by enhancing innovation and platform efficiency.

Metrics that can be used to predict these benefits ex ante include the projected increase in dataset size and diversity; expected improvements in algorithmic accuracy or predictive performance; anticipated uplift in user engagement or retention; anticipated expansion in monetizable insights; and planned investment in data-infrastructure or analytics capabilities. These forward-looking metrics would help assess whether the merger will increase the firm’s capacity and incentives to generate value from combined data, thereby producing efficiencies that benefit consumers.

g. Improves access to market (i.e., ability to reach new customers or geographies in the internal market or outside the internal market). Mergers between complementary firms can significantly enhance their ability and incentives to reach new customers or expand into new geographies. By combining distribution channels, sales networks, and customer bases, the merged entity can enter markets more efficiently and with lower costs than either firm could achieve independently. This expanded reach also increases incentives to invest in marketing, localization, and other market-entry activities, as the benefits of customer acquisition can be realized across a larger platform.

A practical example is Microsoft’s acquisition of LinkedIn in 2016, which allowed LinkedIn to leverage Microsoft’s global-enterprise customer base, cloud infrastructure, and sales networks to expand adoption among professionals and organizations worldwide. Pre-acquisition, LinkedIn had a strong social-networking presence, but limited integration with enterprise tools. Post-acquisition, the platform gained access to Microsoft’s suite of products and global reach, accelerating growth and increasing user engagement across multiple regions.

Metrics that can be used to evaluate these benefits ex ante include projected increases in customer reach; number of new geographic markets entered; growth in user acquisition or active users in target regions; anticipated increases in sales or monetization potential; and planned investments in marketing, localization, or distribution infrastructure.

h. Ability to procure products more competitively from large suppliers. Mergers between complementary firms can enhance the ability and incentives to procure inputs or products more competitively from large suppliers. By combining purchasing volumes, the merged entity can achieve economies of scale and stronger bargaining power, allowing it to negotiate better prices, more favourable contract terms, or improved delivery conditions. This efficiency can lower costs for the merged firm and potentially benefit consumers through lower prices or improved service quality.

A practical example is the merger of Kraft Foods and Heinz in 2015, which allowed the combined company to consolidate procurement across a larger portfolio of products. By pooling demand for raw materials, packaging, and logistics services, Kraft Heinz could negotiate better terms with suppliers, streamline supply chains, and achieve cost savings that would have been difficult for either company independently. Another example is Nvidia Corp.’s acquisition of Mellanox in 2019. By integrating Mellanox’s high-performance networking hardware with Nvidia’s GPU infrastructure, the combined company was able to negotiate more competitive procurement of semiconductors, networking components, and specialized hardware. This allowed Nvidia to achieve cost savings, streamline supply chains, and improve its ability to deploy large-scale datacentre and AI solutions efficiently.

Metrics that can be used to evaluate these benefits include projected reductions in unit-procurement costs; anticipated improvements in contract terms or supplier service levels; expected increases in procurement volume or efficiency; and foreseen impact on product pricing and profitability.

i. Ability to compete in global markets outside the EU. Mergers between complementary firms can enhance their ability and incentives to compete in global markets beyond the EU. By combining resources, distribution networks, and complementary capabilities, the merged entity can overcome barriers to entry in new geographies, reduce costs, and scale more efficiently than either firm could independently. Integration may also strengthen the incentives to invest in marketing, localization, regulatory compliance, and supply-chain optimization, as the benefits of international expansion can be realized across a larger, more capable platform.

Scale that comes from combining complementary products or geographies primarily improves global competitiveness by lowering per?unit fixed costs of exporting and by pushing the merged firm further up the productivity distribution. Indeed, this exactly the margin on which firms can enter and thrive in foreign markets in heterogeneous?firm trade models. In Melitz?type frameworks, larger, more productive firms are the ones that can cover the fixed costs of exporting and penetrate more (and tougher) markets; openness reallocates output toward these firms and raises average industry productivity without requiring market power (Melitz 2003; Melitz & Ottaviano 2008). Empirically, bigger markets and tighter integration reduce markups and increase average productivity—features that enhance, rather than dampen, rivalry as firms expand abroad (Melitz & Ottaviano 2008)

A practical example were the mergers of Spotify with certain regional music-streaming startups in 2014–2015, which allowed Spotify to leverage local expertise, licensing agreements, and distribution channels to accelerate entry into North America, Latin America, and Asia. By combining these complementary assets, Spotify could scale internationally more quickly, improve content offerings, and compete more effectively with global rivals like Apple Music and Pandora.

Metrics that can be used to evaluate these benefits ex ante include projected increases in active users or subscribers in target regions; anticipated expansion into new geographic markets; planned investments in localization, marketing, and compliance; forecasted growth in revenues or market share outside the EU; and expected improvements in operational efficiency for international operations.

j. Ability to use countervailing market power vis-à-vis infrastructure providers. Mergers between complementary firms can enhance their ability and incentives to exercise countervailing market power vis-à-vis infrastructure providers. By combining demand and resources, the merged entity can negotiate better terms for access to such critical inputs as cloud services, payment processing, telecommunications, or logistics infrastructure. This increased bargaining power can reduce costs, improve service quality, and ensure more reliable access to essential infrastructure—benefiting both the firm and ultimately consumers.

A practical example is Dropbox’s acquisition of HelloSign in 2019. By integrating HelloSign’s electronic-signature platform, Dropbox increased its leverage with cloud-infrastructure providers and API partners, enabling more favourable pricing, improved service-level agreements, and better support for scaling operations. This strengthened Dropbox’s ability to deploy secure, large-scale document management and e-signature services efficiently, while retaining negotiating flexibility with key infrastructure providers.

Metrics that can be used to assess these benefits ex ante include expected reductions in input costs; projected improvements in contract terms or delivery reliability; anticipated capacity secured from infrastructure providers; foreseen impact on production timelines or scalability; and forecast gains in operational efficiency due to improved bargaining power.

A.7.   Under which conditions can scale that brings benefits but creates or strengthens market power be achieved only through a merger, as opposed to other means, i.e. organic growth or cooperation?

Under certain market conditions, a merger may be the only realistic means to achieve the scale necessary to generate efficiency benefits, even when this scale also creates or strengthens market power (Demsetz, 1983). This is particularly true in sectors characterized by high fixed costs, capital intensity, and rapid innovation cycles. By consolidating critical assets like data, infrastructure, or customer bases, mergers can create barriers that are difficult for competitors to overcome. While this market power may raise competitive concerns, it is frequently a necessary byproduct of achieving the efficiencies and capabilities essential to compete globally, especially in innovation-driven sectors.

For example, in AI markets, access to additional computing power could give startups a competitive advantage, leading to better services and lower prices for consumers. A startup may possess innovative technology but lack immediate access to the kinds of infrastructure that only a larger company can provide. In this context, a merger can facilitate the rapid consolidation of critical assets and capabilities, accelerating scale economies that otherwise would take years to acquire, if they are achievable at all.

Similarly, in fast-evolving fields like biotechnology, merging complementary capabilities can be the only efficient path to bring innovations to market promptly. For instance, when one company holds promising research but lacks clinical-development expertise, while another possesses those capabilities, a merger integrates these assets in a way that loose collaborations or partnerships typically cannot match within necessary timeframes.

Mergers may also be indispensable to achieve the scale required for commercialization. Startups developing highly specialized technologies—such as breakthrough battery innovations—may depend on established manufacturers’ production facilities and supply-chain networks to scale effectively.

From an economic and regulatory standpoint, it is crucial to recognize that blocking such mergers risks stifling innovation by preventing the realization of critical scale benefits. While alternatives like organic growth or cooperation can be effective in some markets, sectors with significant asset specificity, indivisibilities, strong network effects, and complex complementary resources may require consolidation to unlock the efficiencies necessary for global competitiveness (see, e.g., Klein, 1996).

A.8.   To what extent can scale that brings benefits be achieved through expansion into new geographic or product markets, rather than consolidation within the same product and geographic market?

Scale benefits can often be partly achieved through expansion into new geographic or product markets, which allows firms to access new customers and diversify risk without necessarily raising significant competition concerns. For example, a French company might grow by entering Spanish markets or adjacent product categories, instead of acquiring competitors in its existing market.

This strategy is not, however, always sufficient. In sectors with strong local network effects or multi-sided platforms, high fixed costs, or complex regulatory environments, achieving scale within the core geographic and product markets remains critical. For instance, a ride-hailing app with a strong user base in one jurisdiction may not be able to capitalize on users in other jurisdictions. Local expertise, integrated infrastructure, and established relationships are often difficult to replicate through expansion alone. Startups face particularly significant challenges in scaling organically across regions, due to the needs for tailored infrastructure and compliance with diverse regulations. This may make consolidation within incumbent markets a more viable path to necessary scale. This challenge is amplified in the EU, where the Single Market remains incomplete because of persistent regulatory, tax, cultural, and linguistic differences.

Ultimately, the choice between expansion and consolidation depends on market-specific factors, including network effects, asset indivisibility, and regulatory complexity. While geographic and product expansion can contribute to scale, consolidation within core markets often remains essential to unlock full efficiency and competitive benefits.

A.9.   How should the Commission take into account the negative effects of a merger on competitors’, suppliers’ or business customers’ resilience when assessing its impact on competition?

It is not clear that merger assessment is the appropriate legal tool to address broader concerns such as “resilience” or geopolitical dependence on a limited number of suppliers. Resilience, as such, is not a standalone factor under the EUMR. The EU merger-control framework is relatively narrow, designed to focus on market power and competition dynamics—areas where the Commission possesses clear expertise and well-established analytical tools.

At the same time, it is undeniable that mergers that reduce the number of viable suppliers or create excessive concentration, particularly in critical inputs, can make supply chains more fragile and susceptible to shocks, thus increasing the EU’s overall dependence on a few concentrated sources. Given that this could have critical strategic implications for the EU and its member states, the question arises whether the Commission should take such geopolitical risks into account when reviewing mergers under the EUMR.

Geopolitical risks and supply-chain resilience involve complex and multifaceted issues that extend beyond competition law and typically require targeted policy instruments. For instance, it is unclear how DG COMP could adequately analyse whether concentration of a specific input in a particular region—inside or outside the EU—exposes the EU to supply-chain shocks. While such analyses may not be impossible (indeed, they might be both possible and necessary), DG COMP is not well-positioned to conduct them. While granting the College of Commissioners a more prominent role in merger review could offer a partial solution, it risks politicizing and delaying the merger-control process to the detriment of businesses and European consumers.

To address these challenges effectively, the EU should instead consider relying on other tools, such as strategic stockpiling of critical materials; foreign-investment screening; industrial-policy initiatives; and trade-diversification strategies. These specialized instruments are better suited to manage systemic risks and ensure supply-chain security, without conflating them with merger control’s primary objective of preserving competitive market structures—an objective that already aligns with certain resilience concerns.

Ultimately, maintaining a clear boundary between competition assessment and geopolitical policy, to the highest extent possible, would safeguard both legal certainty and regulatory effectiveness. As we have argued in our answers to other questions in this survey, it is here where the EU’s competitive advantage lies.

A.15. In which type of markets/sectors are smaller or larger firms typically more innovative?

In the dynamic landscape of innovation, smaller and larger firms both demonstrate distinct innovation strengths across different markets and sectors (Acs & Audretsch, 1988; Cohen & Klepper, 1996; Akcigit & Kerr, 2018). While this is far from an exact science, we outline some general characteristics below.

Smaller firms, particularly tech startups, typically excel in new, disruptive areas and developing novel technological advancements (Acs & Audretsch, 1988; Christensen, 1997; Akcigit & Kerr, 2018). Characterized by agility, a high degree of risk tolerance, and innovative business models, they offer scalable products or services (Kerr & Nanda, 2015). Many successful startups are founded by entrepreneurs seeking a more flexible environment than large firms can offer (Gans, Hsu, & Stern, 2002). These entities often originate specific technologies that large companies don’t easily develop in-house (Christensen, 1997). Their diverse innovation is evident in acquisitions by major digital platforms, often spanning sectors outside the acquirer’s “core functionalities” (Gautier & Lamesch, 2022; Motta & Peitz, 2020). For example, Alphabet has acquired firms in cybersecurity, home automation, cloud computing, wearables, GPS navigation, and social gaming. Microsoft’s acquisitions include companies specializing in video games, social networking, and e-commerce. This highlights their contributions across specialized and emerging digital markets (Gautier & Lamesch, 2022).

Conversely, larger firms and established incumbents are often highly innovative in building on existing strengths, scaling technologies, and deploying innovations widely due to their significant resources (Cohen & Klepper, 1996; Bloom, Sadun, & Van Reenen, 2012). They are among the most productive and innovative segments of the modern economy, with high rates of innovation and substantial patent holdings (Autor, Dorn, Katz, Patterson, & Van Reenen, 2020; Hall, Jaffe, & Trajtenberg, 2005). R&D investments and productivity generally increase with firm size; in 2023, for instance, Amazon spent $85.6 billion on R&D, and Alphabet (Google) spent $45.4 billion. Large firms are uniquely capable of adopting advanced technologies, such as AI; absorbing high fixed costs; managing organizational changes; and leveraging extensive datasets (Brynjolfsson & Hitt, 2000; Brynjolfsson, Rock, & Syverson, 2019). Their strategic acquisitions of startups are vital, integrating smaller firms’ innovations into their ecosystems, as well as providing superior managerial capabilities and greater scale for broader commercialization (Bena & Li, 2014; Phillips & Zhdanov, 2013; Igami & Uetake, 2020).

Ultimately, both small and large firms play complementary roles: smaller firms pioneer disruptive ideas, while larger firms provide the scale and investment needed for widespread adoption and sustained R&D (Acs & Audretsch, 1988; Akcigit & Kerr, 2018). Mergers often facilitate this natural symbiosis, with large incumbents supplying the resources to scale innovative but inherently risky ventures (Gans et al., 2002; Bena & Li, 2014).

A.16. How do different market structures, such as tight oligopolies or markets with a leading company followed by smaller firms, influence the ability and incentives to innovate and invest?

Market structure alone is a poor predictor of innovation outcomes, as both concentrated and more fragmented markets can foster strong incentives to invest, depending on the circumstances (Aghion, Bloom, Blundell, Griffith, & Howitt, 2005; Gilbert, 2006; Sutton, 1991). In tight oligopolies, large incumbents often have the scale, resources, and diversified portfolios needed to fund costly long-term R&D, and dynamic competition can push them to innovate as a primary means of rivalry (Cohen & Klepper, 1996; Nickell, 1996; Vives, 2008). In leader–follower markets, dominant firms can commercialize and scale promising inventions, while smaller rivals and startups drive disruptive experimentation, often relying on acquisitions or partnerships with incumbents as viable exit paths (Teece, 1986; Arora, Fosfuri, & Gambardella, 2001; Gans, Hsu, & Stern, 2002; Kerr & Nanda, 2015). Many firms later deemed “innovative” were recognized as such only after being successfully scaled by an acquirer (Bena & Li, 2014; Phillips & Zhdanov, 2013). Accordingly, policymakers should avoid structural presumptions and instead assess the specific economic mechanisms at-play in each market before drawing conclusions about innovation incentives (Arrow, 1962; Aghion et al., 2005; Gilbert, 2006). These mechanisms include (among others) firms’ ability to appropriate returns on innovation; exploit economies of scale in R&D; respond to dynamic competitive pressures; access capital and bear risk; and foster entry and experimentation, in addition to the existence of viable exit strategies (Levin, Klevorick, Nelson, & Winter, 1987; Teece, 1986; Cohen & Klepper, 1996; Nickell, 1996; Hall & Lerner, 2010; Nanda & Rhodes-Kropf, 2013; Kerr & Nanda, 2015; Akcigit & Kerr, 2018; Cohen, Nelson, & Walsh, 2000).

A.18. What are the benefits companies may enjoy due to their global presence that can give them a competitive advantage in markets (with)in Europe?

First, less regulation in markets outside of Europe significantly benefits globally operating firms. As the Draghi Report confirms, Europe is characterized by overregulation, and inconsistent and restrictive regulations that create a “competitiveness and innovation gap” compared to regions like the United States and China. For instance, the compliance costs for a typical SME can be as high as €10,000. The EU’s “risk-averse regulatory approach” leads to legal uncertainty and disadvantages European firms, inadvertently favoring non-European companies not bound by these rules.

The General Data Protection Regulation (GDPR) is a prime example: within a year of it coming into force, European tech startups saw a 26.1% drop in monthly venture-funding deals relative to their U.S. counterparts, a decline that persisted through 2020 (Jia, Jin, & Wagman, 2021). This regulation disproportionately favored incumbents, leading to increased market concentration as websites relied more on large vendors like Google and Facebook and cut ties with smaller ad-tech providers. It also resulted in a “lost generation” of apps, with approximately one-third of existing apps exiting the EU market and new entries falling by roughly half (Kummer & Schulte, 2019; Peukert et al., 2022). The Draghi Report explicitly points to the GDPR as a regulation that has hindered EU companies due to its fragmented implementation.

Second, better access to financing or equity investments outside Europe provides a crucial advantage. European companies often struggle to attract sufficient risk capital, facing fragmented capital markets and unfavourable conditions for venture-capital investment within the EU, as highlighted by the Draghi Report. This “funding gap” forces many European entrepreneurs to seek financing from U.S. venture capitalists and to scale up in the U.S. market. Between 2008 and 2021, nearly 30% of “unicorns” founded in Europe relocated their headquarters abroad, predominantly to the United States (Atomico, 2021). In the AI sector, 61% of global funding goes to U.S. companies, 17% to Chinese, and only 6% to those in the EU (Draghi Report).

Acquisitions by large incumbent firms are one of the most important components for providing vitality to the overall venture ecosystem, accounting for approximately 80% of “liquidity events” for venture capital today (NVCA, 2025). Policies that restrict these acquisitions, such as those related to “self-preferencing” or stringent merger control (e.g., DMA’s Art. 14), can reduce larger firms’ motivation to make purchases, thereby deterring initial financing and the formation of new ventures. Indeed, venture investors become much less likely to invest in new startups without clear exit pathways.

Finally, the EU’s stringent environmental and social-protection standards impose substantial direct and operational costs on firms, which can weaken overall competitiveness and reduce EU firms’ incentives and ability to grow to a critical scale (Draghi, 2024). Compliance with environmental rules costs the chemical industry more than €20 billion annually, with some firms devoting up to 10% of their capital expenditures to regulatory compliance, while sustainability reporting under the CSRD can add up to €1 million per-year for listed companies (Draghi, 2024; European Commission, 2021a).

These burdens divert resources that could otherwise support innovation, expansion, or price competition. Moreover, they disproportionately affect smaller firms with limited compliance capacity, such as SMEs and startups (Draghi, 2024). Mechanisms like the Carbon Border Adjustment Mechanism further raise potential trade costs—estimated at €4–7 billion in certain sectors—which can serve as an incentive to shift production to jurisdictions with looser standards (“carbon leakage” or “social dumping”) (European Commission, 2021b; OECD, 2020).

Overall, the EU’s regulatory load increases production costs and reduces firms’ ability to compete on price and speed in global markets. While foreign companies operating in the EU are subject to the same rules, they are often not bound by them during their critical early-growth phase, allowing them to scale more easily before entering the EU market (Draghi, 2024). This puts EU firms, which often have been subject to a regulatory straitjacket since their inception, at a distinct competitive disadvantage (Draghi, 2024).

A.20. What would be pro-competitive consolidations in global strategic sectors, such as digital and deeptech markets (e.g., IoT, advanced connectivity, cybersecurity, cloud, quantum, and/or AI), clean and resource efficient technologies or biotechnologies that would benefit competition in the Single Market?

Pro-competitive consolidations in strategic sectors like AI, quantum computing, clean technologies, and biotechnologies are among those that combine complementary assets in ways that enhance firms’ ability and incentives to innovate, invest, and compete, without creating durable market power that forecloses rivals. In high-fixed-cost, scale-intensive markets—where small firms often pioneer breakthrough technologies but lack the resources to commercialize them—mergers can accelerate product development, improve interoperability, and achieve efficiencies that benefit consumers in the Single Market. Such transactions can also strengthen EU firms’ global competitiveness and encourage new market entry by providing credible acquisition pathways for startups.

The competitive benefits lie in merger-specific, verifiable efficiencies—particularly dynamic, innovation-driven ones—while the potential harms, such as foreclosure or the entrenchment of market power, should be carefully assessed and mitigated through evidence-based analysis, rather than structural presumptions.

Topic B: Assessing Market Power Using Structural Features and Other Market Indicators

B.1.    Do the current Guidelines provide clear, correct, and comprehensive guidance with regards to structural indicators / market features as well as the frameworks to assess coordination and foreclosure theories of harm?

The current EU Horizontal and Non?Horizontal Merger Guidelines offer workable screens (rather than presumptions) for structure, and a generally sound framework for coordinated and foreclosure theories of harm. In particular, the Non?Horizontal Guidelines’ “ability–incentive–effects” approach; their recognition that non?horizontal mergers are generally less likely to impede competition than horizontal mergers; and the <30% share/HHI<2000 “safe harbour” are all clear and correct starting points.

Where the guidelines are less complete is precisely where recent consultation materials contemplate moving: toward stricter indicators or rebuttable presumptions based on shares/HHI; expanded lists of “market features”; and more aggressive screens for coordinated effects (e.g., diversion, margins, pivotality, GUPPI/vGUPPI). The economic literature is clear the concentration measures are, at best, diagnostics and, at worst, misleading predictors of harm. The national–local concentration disconnect, the weak and often non?causal link between concentration and price, and heterogeneous innovation responses all caution against burden?shifting based on structure alone (e.g., Demsetz 1973; Berry, Gaynor & Scott Morton 2019).

The error?cost implications are straightforward: more rigid presumptions risk condemning efficient deals and misallocating enforcement resources. The EU’s own Topic B backgrounder fairly notes that existing texts mostly provide safe harbours, rather than presumptions, and that any SIEC finding must meet a “more?likely?than?not” standard. This includes below-dominance thresholds, as those are strengths to retain, not weaknesses to reverse.

Two clarifications would materially improve predictability, while preserving the existing balanced framework.

First, state expressly that market shares and HHIs are screens (useful first indications) and not presumptions, and that any SIEC theory—especially “important competitive force” and differentiated?products unilateral effects—must rest on transaction?specific evidence of diversion, margins, and repositioning, not on structural thresholds alone.

Second, in non?horizontal cases, reaffirm that any foreclosure theory must show (i) ability, (ii) incentive, and (iii) likely material effects, and that “foreclosure shares” without an incentive story are insufficient. At the same time, instruct staff to evaluate the elimination of double marginalization and other verifiable, merger?specific efficiencies within the same arithmetic the guidelines already endorse.

These steps would keep the guidelines anchored in effects?based analysis and consistent with the mainstream empirical finding that vertical integrations are often neutral?to?procompetitive on average, with harms arising only under specific conditions that the ability-incentive-effects test is designed to identify.

B.2. Do you consider that the current structural indicators / market features involving market shares and concentration levels and/or the broad frameworks to assess coordination and foreclosure theories of harm should be substantially revised?

1. Horizontal mergers—structural indicators. The current HMG fairly treat shares/HHIs as “useful first indicators” and set only safe?harbour guidance, not presumptions, for horizontal cases. They should continue to function as screens, not as burden?shifting rules. The economics literature finds little policy value in concentration and warns against stronger structural presumptions (Berry, Gaynor & Scott Morton 2019; Demsetz 1973). The effects are sensitive to market definition. For example, in the U.S. context, it shows that national concentration trends are a poor proxy for local competitive conditions (Rinz 2022). The revised guidelines should therefore emphasize case?specific evidence—diversion ratios, margins, bidding evidence, and UPP/GUPPI—over tighter concentration thresholds.
2. Horizontal coordinated effects—framework. The HMG’s Airtours test (ability to reach terms; monitoring; deterrence; and limited outsider disruption) remains the right organizing framework, but the revised guidelines should clarify the evidentiary burden and the type of proof that shows a merger makes coordination “more likely than not”, consistent with CK?Telecoms. In particular, the guidelines should specify what market facts (e.g., transparency, symmetry, capacities, elimination of a “maverick”) move the needle, and what empirical/ documentary evidence is required in differentiated?products settings or with algorithmic pricing. The broader empirical record supports this approach: surveys and newer studies consistently find that most vertical integration is neutral or procompetitive (Lafontaine & Slade 2007; Cooper et?al. 2005; Slade 2019).
3. Non?horizontal mergers vertical foreclosure. We support retaining and clarifying the NHMG’s “ability-incentive-effects” framework, including the role of EDM and other efficiencies. The guidelines should caution against treating “foreclosure shares” or vertical concentration as sufficient; incentives must be demonstrated with profitability?based analysis (including vertical arithmetic or vGUPPI) and credible effects evidence

B.3.    What should be the structural indicators / market features used by the Commission to assess the likelihood of anticompetitive effects in horizontal mergers?

The Commission should continue to use market shares and concentration levels as an initial screen in horizontal mergers, but these should be complemented by richer structural and competitive indicators. Market share and HHI provide a contextual starting point that large increases in concentration might warrant a closer look. They are not, however, dispositive measures of market power. We recommend the guidelines explicitly incorporate other indicators, such as diversion ratios, profit margins, price-elasticity data, and measures of competitive proximity, as well as qualitative evidence (e.g., internal documents on rivalry). These features paint a fuller picture of the merger’s competitive impact than market shares alone. The role of market shares/HHIs should thus remain as “useful first indications” (per current practice) and safe-harbour thresholds, while other structural features guide the substantive analysis of whether a merger is likely to harm competition

Market share and concentration levels are traditional barometers of market structure. A high combined market share or a big jump in HHI post-merger suggests the merged firm might enjoy substantial market power. These metrics are intuitive and backed by economic theory: in a simplistic model, fewer firms or a dominant firm can more easily sustain higher prices. But just as the current ones do in para.14 HMG, the guidelines should stress that these are starting points. There is ample empirical evidence that concentration alone does not reliably predict competitive outcomes. Industries can be concentrated due to efficiency (successful firms attracting more customers) without consumers being harmed by monopolistic behaviour.

Thus, a high share or HHI should prompt further inquiry, not a conclusive presumption of harm. The revised guidelines could reaffirm safe harbours (e.g., combined share under 25% or small HHI increases) as well as clarify that crossing certain thresholds (like combined share well above 40% or HHI above 2000 with a large delta) will typically lead to an in-depth analysis, albeit with no automatic outcome.

The revised guidelines should reflect, however, that structural indicators often have shortcomings in dynamic markets, particularly when combined with the Commission’s narrow approach to defining relevant product markets.  In such circumstances, the use of structural indicators alone risks underestimating the competitive constraints imposed on, and in turn overestimating the market power of, the merging parties and their competitors.

As an example of this, recent approaches to market definition in digital markets have tended to neglect that users—and, in turn, the advertisers that seek their attention—switch between a variety of digital services, and not only among services that share functional similarities.  Indeed, challenger service A will often seek to compete with incumbent service B by offering something slightly different.  Yet, too often, the differentiation which is the core of service A’s competitive challenge to service B is used as the very evidence to deny the existence of competition between these services.

In addition, in many digital markets, switching from one service to another is instant and cost-free, with few capacity constraints on the recipient firm’s growth.  These characteristics means that digital firms often face more intense ongoing competitive pressure from smaller rivals than is the case in other industries where, for instance, it may be costly for the customer to switch, or it may be slow and costly for the rival to increase capacity.

User switching across a broad variety of services is supported by a broad base of real-world evidence.  On several occasions, digital services have become unavailable to users, and in those cases, users have turned to a range of alternative services.  For example, in January 2025, TikTok went dark overnight in the United States for 12-16 hours on the eve of a congressional ban.  As a result, overall time spent on Instagram spiked by nearly 40% across the board, and panel data showed that the lost TikTok time went to Facebook (20% of the increased usage among all apps), Instagram (17%), and then YouTube (14%); even Meta’s Messenger app saw a 5% increase in traffic.

B.4.    Compared to the current Guidelines, should structural indicators be stricter or give rise to legal presumptions? Or should they be laxer/lower?

Structural indicators should not be made stricter or turned into legal presumptions; if anything, they should be used only as soft screens.

First, EU merger control is built around a case?by?case SIEC assessment of which shares/HHI are “useful first indications”, not dispositive rules. The current texts expressly set safe harbours (e.g., <30% share and HHI<2,000 in non?horizontal cases) and emphasize that thresholds “do not give rise to a legal presumption”. Nor do they shift the burden of proof to the parties; proof of a SIEC remains with the Commission on a “more likely than not” standard (including below dominance). Turning these indicative screens into rebuttable presumptions would invert this architecture, risk type?I errors, and conflict with the guidelines’ own “sliding scale” approach to evidence. The better course is to keep (or relax) soft thresholds, while clarifying how other evidence—diversion, margins/UPP?type tools, capacity/pivotality—will be weighed.

Second, the economic record does not support stricter presumptive use of concentration metrics. Decades of IO research finds no stable, policy?reliable relationship between concentration and competitive harm, and cautions against treating structural measures as standalone predictors of price effects (Demsetz 1973; Berry, Gaynor & Scott Morton 2019; Syverson 2019). In short, stricter structural presumptions would be administratively convenient but economically noisy, increasing false positives and chilling procompetitive deals.

Third, structural shortcuts are especially ill?suited outside horizontal theories. The EU’s non?horizontal framework rightly focuses on ability-incentive-effects, rather than on blunt “foreclosure shares”. The modern empirical literature indicates that vertical integration frequently yields efficiencies (most notably, the elimination of double marginalization) and, on balance, consumer gains—even when some rivals’ costs may rise (Lafontaine & Slade 2007; Cooper et?al. 2005; Crawford, Lee, Whinston & Yurukoglu 2018). Converting vertical “screens” into presumptions (e.g., based on capacity shares or generic foreclosure fractions) would flatten this nuance and contradict the NHMG’s emphasis on incentives and net competitive effects. The safer path is to keep structural indicators as triage tools, while requiring concrete evidence of both the profitability of foreclosure and likely downstream harm.

Topic C: Innovation and Other Dynamic Elements in Merger Control

C.1.    Do the current Guidelines provide adequately clear, correct and comprehensive guidance on how the Commission considers dynamic criteria in its assessment of the impact of mergers on competition?

No, to an insufficient extent (c). The current EU Horizontal Merger Guidelines offer limited clarity and insufficient conceptual depth when it comes to assessing mergers through dynamic, innovation-centred lenses. While the document occasionally acknowledges the relevance of innovation to competition, the framework remains grounded in static price-centric logic and lacks the necessary elaboration to guide assessments where innovation is the primary competitive parameter.

Paragraph 5 of the HMGs states that the guidelines apply to mergers between “actual or potential competitors in the same relevant market”. This formulation has grown ambiguous, however, considering recent merger practice. For instance, in Dow/DuPont, the Commission assessed the parties’ position in the same “innovation space”, an analytical move that does not easily square with traditional notions of market definition. In such cases, there is a risk that the concept of “potential competition” may be stretched beyond product markets into speculative future innovation pathways. The guidelines should therefore provide conceptual tools or limiting principles for such extensions.

In Paragraph 8, the Commission explains that it prevents mergers that would deprive consumers of the benefits of effective competition—listing innovation alongside price, quality, and choice—only to reduce all these effects into a shorthand of “price”. While administrative efficiency may justify such an approach in some contexts, innovation and price operate through distinct mechanisms. Mergers may affect innovation incentives in ways that are not analogous to pricing effects. For instance, rivals’ responses to innovation may be nonlinear, and mergers may alter the trajectory of technological development in ways that cannot be captured through marginal-pricing logic. It may be useful to treat the innovation-related harms (and benefits) of mergers under a separate section.

Paragraph 9 may aggravate this problem by presenting a primarily backward-looking counterfactual approach. The Commission notes that it will usually compare post-merger competitive conditions with those prevailing at the time of notification, resorting to forward-looking analysis only in “some” cases. Yet innovation competition is inherently dynamic. Product lifecycles, R&D pipelines, and the evolution of technology often follow discontinuous paths. Assessing such mergers by reference to present conditions risks missing the core competitive harms, or benefits, of a proposed transaction. Hence, it may be useful to qualify this statement by explaining that, when innovation concerns orient a merger assessment, the relevant counterfactual should be inherently dynamic.

Some recognition of the distinctive nature of innovation-based harm is found in Paragraphs 20 and 38, which allow the Commission to downplay market-share metrics in favour of an assessment of firms’ role as “important innovators”. These provisions are intuitive and welcome, but they suffer from under-definition. The guidelines do not clarify what it means to be an “important innovator”, nor how such status should be evidenced. In practice, this has led to increasingly expansive (and hardly predictable) interpretations, ranging from early-stage pipeline overlaps to speculative assessments of future technological positioning. In high-uncertainty settings such as pharmaceuticals or digital technology, this vagueness risks harming legal predictability and may encourage overly cautious dealmaking.

Finally, in connection to Paragraph 8, Paragraph 28 attempts to anchor innovation effects to familiar price-competition logic. It suggests that, when merging firms are close competitors, the removal of rivalry may allow them to raise prices—understood, again, as shorthand for all competition parameters. But this analogy breaks down in the case of innovation. The economics literature suggests that the relationship between competition and innovation is non-monotonic: eliminating a rival does not always reduce innovation incentives and may, in some cases, enhance them (Denicolo & Polo, 2019). The guidelines do not engage with this complexity, leaving a gap in their analytical apparatus.

In sum, while the HMGs do refer to innovation, they do so in a fragmented and conceptually limited manner. The core framework—counterfactual analysis, assessment of closeness of competition, and reliance on market structure—remains designed for static, price-based environments. A more robust and dedicated framework for analysing dynamic competition is needed if merger control is to protect and promote innovation in high-technology markets.

C.3.    In what circumstances can mergers negatively impact the ability and incentives of the merged company to innovate?

Mergers may negatively affect innovation by altering not only firms’ incentives to innovate but also their capacity to do so. The Commission’s dual focus on incentives and abilities in recent merger cases is a promising step. In practice, however, innovation analysis has been dominated by incentive-based reasoning, often grounded in structural assumptions that are ill-suited to assess dynamic competition and innovation. A more balanced and conceptually coherent treatment of innovation harms requires a firmer anchoring of both capability-based analysis and contextual assessments of innovation opportunity and market contestability.

The first limitation arises from an imbalance: while the guidelines and recent decisions refer to both “incentives” and “abilities”, actual enforcement tends to revolve around the former. That is, if a merger is seen to reduce the incentives of the merging parties—or of their rivals—to innovate, it becomes suspect, often without a corresponding analysis of how the merger alters their respective innovation capabilities. The Commission continues to rely on traditional structural indicators (e.g., market shares, concentration levels) as proxies for innovation incentives.

Yet economic theory has long shown that market structure is an unreliable predictor of firms’ innovation behaviour. While some stylized models link innovation incentives to low concentration, others—most famously Schumpeterian models—posit that large firms with market power are more likely to invest in R&D. Neither view holds universally (Cohen & Levin, 1989; Gilbert, 2006). This is why contemporary merger economics relies on fact-intensive and case-specific inquiries to discover whether, in the specific circumstances of the merger, the transaction threatens to hinder innovation (Bourreau, Jullien, & Lefouili, 2024).

A more fruitful approach would be to evaluate innovation incentives through the lens of contestability and opportunity. In dynamic markets, innovation can emerge from both incumbents and entrants, provided the market remains open to challenge and there is opportunity to disrupt (Shapiro, 2012). Monopolists may continue to innovate pre-emptively to maintain dominance, while rivals innovate to seize leadership (Gilbert & Newberry, 1982). Under such conditions, mergers that reduce contestability—for instance, by eliminating a credible future challenger or acquiring a key innovation input—can indeed chill innovation. But the mechanism must be case-specific and substantiated with evidence beyond mere structural inference. Chiefly, the notion of potential competition is inherently probabilistic: a merger that eliminates a company that is certain to be a future challenger is, other things equal, more likely to harm innovation than situations where this probability is more remote.

Equally important is the merger’s impact on the firms’ capabilities to innovate—a dimension that is often underexplored in enforcement practice. Whereas price-based harms mostly hinge on incentive effects, innovation-based harms require a richer understanding of what the firms can do, and how the merger might reshape their innovation potential. The Commission implicitly recognized this in STX/Aker Yards, noting that “the clearest evidence that the crucial factor in the cruise ship market is the know-how of managing complex projects and that a certain lack of innovation can be overcome is the case of Mitsubishi, who was able to deliver in 2004 two cruise ships of very high quality without previous experience in the cruise ship market”. This acknowledgment of capability-based competition underscores the need for merger analysis to scrutinize whether the transaction enhances or suppresses firms’ abilities to innovate, and specifically, how they are affected by the proposed merger.

A constructive step forward in the new guidelines could be to distinguish mergers that bring together overlapping innovation capabilities from those that combine complementary ones. A 2009 paper put the issue plainly: “The question should be framed not in terms of whether product-market competition will be impaired, as that is too much of an immediate concern, but in terms of whether capabilities will be brought under unitary control, thereby possibly thwarting future variety in new product development” (Sidak & Teece, 2009).

In short, the new Guidelines would benefit from a clearer analytical framework that distinguishes among incentives, capabilities, and contestability. In particular, i) keeping the case-specific approach to mergers while ii) moving away from inconclusive assumptions about market structure, and iii) emphasizing mergers’ impact on firms’ ability to innovate, would significantly improve the assessment of innovation harms in merger control.

C.3.a    What theory/theories of harm could the Commission consider (i.e. that would impede a company’s innovation post-merger, including due to the reduction of the incentives to innovate going forward or reduce access to IP licences)?

The Commission’s mandate under the Merger Regulation is to prevent concentrations that would significantly impede effective competition, including through a reduction in innovation. Any theory of harm that seeks to assess post-merger innovation effects must therefore focus on how a transaction may reduce firms’ incentive and/or ability to innovate, including by impairing rivals’ access to important resources or foreclosing future sources of innovation. This objective is not only consistent with consumer welfare but is a necessary condition to protect it in the long term. As a wealth of economic literature confirms, innovation is the principal driver of productivity growth and societal welfare over time (Akcigit & Van Reenen, 2023). Merger policy that aims to preserve a robust innovation ecosystem is thus deeply aligned with a forward-looking interpretation of consumer welfare (Teece, 2023).

As noted under C.3., the Commission historically has elaborated an “innovation theory of harm” focused primarily on the incentives of the merging parties to continue innovation post-transaction. This analysis, however, has often been anchored in such structural indicators as R&D market shares or overlaps in pipeline products. While these provide a starting point, they fail to capture the deeper mechanisms through which innovation may be furthered or impaired. As discussed earlier, the merger’s impact on capability combinations—the unique assets, knowledge bases, and routines that underlie innovation—is a more meaningful predictor of which theories of innovation harm can be constructed. In this respect, the Commission is well-supported by jurisprudence. As the General Court held in CK Telecoms, “the Commission need not limit its analysis to the theories of harm developed in previous decisions”, and may advance alternative ones, provided they are reasonably foreseeable and supported by cogent and consistent evidence.

We refer to our comments under C.3.b for a more detailed analysis of how mergers can harm or benefit innovation potential.

C.3.b    Under which conditions could this theory/these theories of harm materialise?

Horizontal Mergers. The core innovation harm in horizontal mergers arises when the transaction eliminates a direct innovation rival, thus diminishing dynamic rivalry, and where the potential harm from this decrease outweighs potential benefits, such as innovation-related synergies and increased appropriability. The canonical example is where two firms are simultaneously pursuing overlapping R&D or pipeline products in the same technological trajectory. Post-merger, the internalization of competition may reduce the incentive to push forward both lines of development, especially where the merged firm can delay or discontinue one stream without fearing competitive re-entry.

This theory of harm was intensively applied in Dow/DuPont, where the Commission analysed the innovation rivalry between the firms in broad “innovation spaces”. But the case largely hinged on R&D intensity and pipeline overlaps—tools that are coarse and vulnerable to overreach—and assumptions that may not apply everywhere, such as high levels of appropriability. A more refined approach would identify whether the firms possess similar innovation capabilities (e.g., in molecular platforms, data science, or regulatory know-how) and whether the merger would eliminate distinct approaches to innovation from the market. In other words, does the transaction not only reduce competition in outcomes (products), but also in approaches (ways of innovating)?

In addition to overlapping pipelines, horizontal mergers may also raise concerns when the merged entity consolidates R&D resources to the point of excluding rivals from key innovation inputs—e.g., datasets, testing environments, or experimental ecosystems. These may not show up in market-share metrics but could effectively eliminate viable pathways for others to innovate. The Commission briefly touched on this risk in General Electric/Alstom, where it noted that the merged entity could aggregate customer information on demand patterns, which was crucial for continuous innovation in gas-turbine markets.

Non-Horizontal Mergers. Innovation harm in vertical or conglomerate mergers arises primarily through foreclosure—not of product markets, but of innovation pathways.

A merged firm may acquire a critical input necessary for innovation by downstream rivals—such as access to a technological platform, a dataset, IP rights, or specialized manufacturing infrastructure. If access is subsequently restricted, rivals’ ability to innovate is impaired. The merger may, for instance, remove a neutral technology supplier and replace it with a vertically integrated entity that withheld cooperation with competitors. As one example, the abandoned deal between Nvidia and Arm may have produced such effects by withholding from rival firms IP libraries related to semiconductor design.

The important point to consider in foreclosure cases is how critical and irreplaceable the input withheld from competitors might be, and to weigh this potential harm against the benefits of unitary control. In particular, the Commission should inquire into whether innovating firms may use other inputs or innovate around the merged entity’s critical assets. For instance, the remaining firms might build alternative resources, make their own acquisitions, or adopt a modular innovation architecture by orchestrating assets via contracting and alliances.

Conglomerate mergers can impair innovation by bundling or cross-leveraging innovation assets. A firm with dominance in one technological sphere may acquire a firm in a related field to strengthen its control over a converging innovation trajectory. This may discourage independent development across domains, especially where interoperability, standards, or compatibility matter (e.g., IoT, AI, or cloud ecosystems).

Once again, however, the Commission should be careful in its approach to such mergers. Innovation often requires enlarging the scope of the firm (Teece, 1980). Hence, the “financial strength” resulting from a conglomerate merger, as in GE/Honeywell, or the mere addition of another element to a business ecosystem, such as in Booking/eTraveli, should not be taken as sufficient to establish an innovation theory of harm. Instead, in each case, the theory of harm must be based on mapping capabilities and dependencies across firms and innovation systems. The key question is whether the merger leads to strategic control over assets that shape the innovation potential of other market participants, and whether these potential harms outweigh the benefits of unitary control.

C.4.    In what circumstances can mergers negatively impact the ability and incentives of the merged company to invest?

Innovation and investment are inextricably linked. Firms invest in R&D, asset acquisition, and technology development in general to innovate and address profit opportunities, including by surpassing their rivals. Therefore, the comments we have provided from question C.3. remain relevant for this question as well.

C.5     How should the Commission account for the incentives to invest and innovate post-merger depending on the specific market features?

The HMGs rightly emphasize that mergers must be assessed against the backdrop of their relevant legal, economic, and technological context. This contextual sensitivity is particularly important when the primary competitive parameter is innovation. In such cases, the Commission must engage more systematically with market-specific characteristics that influence how, why, and when firms innovate.

A critical and currently underdeveloped dimension of the technological context is the relevant market’s innovation lifecycle. Economic and management literatures have long recognized that industries undergo distinct stages of technological evolution (Utterback & Abernathy, 1975). In the early phases of a product market, firms face deep uncertainty about consumer demand, technological feasibility, and regulatory responses. These conditions often result in a high degree of technological opportunity. Entry is frequent, market shares fluctuate, and innovation thrives through experimentation, variety, and learning by doing.

As the market matures, however, uncertainty diminishes. Dominant designs emerge, innovation efforts converge, and competition shifts from product innovation to process efficiency. Returns to R&D investment tend to fall, and the industry consolidates around a few leading firms. Over time, the market may reach a plateau of incrementalism, in which the scope for further innovation is low unless a disruptive technology resets the lifecycle. This “S-curve” of technological opportunity is widely recognized in the literature but remains largely unincorporated into merger-assessment practice (Cohen, 2010).

Understanding where a product or industry lies along its technological lifecycle is therefore crucial to assessing post-merger incentives to innovate. For instance, a merger between two strong innovators in a maturing market—where opportunities for innovation are dwindling—is, other things equal, more likely to dampen dynamic rivalry than a merger in a nascent, turbulent space where uncertainty and contestability are high. In the latter case, even a powerful merged firm may have strong incentives to innovate due to rapid change, new entrants, and shifting user needs.

It is equally important to recognize that not all technologies proceed through the lifecycle at the same pace. Some may remain in a prolonged phase of disruption, while others quickly settle into stability. For example, platform-based industries (e.g., digital services, AI applications) often sustain innovation over longer periods due to network effects, modularity, and ecosystem complementarities. Conversely, markets based on commoditized inputs or static consumer preferences may reach technological exhaustion more quickly and remain inert for a long time.

The General Court has implicitly recognized the relevance of broader technological trends to competitive assessment. In Cisco Systems, the Court noted that structural dominance in a narrowly defined market may be neutralized by broader industry dynamics—in that case, the growing convergence of videoconferencing and adjacent technologies. This insight can and should be extended to innovation analysis: when industries are undergoing convergence, integration, or modularization, merger-related market power may be short-lived, as the scope for innovation remains fluid.

In practical terms, this means that the Commission should ask the following:

• Is the product or market at a high-opportunity phase of its innovation lifecycle, or a mature/declining one?
• Are there signs of convergence, disruption, or emerging adjacent technologies that render market power unstable?
• Do the merging firms face credible threats from entrants or diversifying incumbents that keep the incentives for innovation high, even post-transaction?
• Conversely, is the merger likely to lock in a dominant design, eliminate residual pathways for differentiation, or consolidate critical assets at a fragile point in the lifecycle?

We commend the Commission’s commitment to case-specific and fact-intensive inquiries in merger law. Our proposal is that, for mergers implicating innovation, the market-specific features—such as technological maturity, rate of change, and the degree of uncertainty—should be considered more thoroughly for a sound analysis of post-merger innovation potential. Integrating these features into merger assessment will allow the Commission to move toward a better understanding of how innovation emerges or stalls in different competitive environments. The goal is to ensure that merger control remains responsive to the true drivers of innovation.

For efficiencies, please refer to our comments under Topic F.

C.6.    In what circumstances can the elimination of a (small) but particularly innovative player with a large competitive potential (e.g., in the case of nascent and emerging market or rapidly developing sectors) harm competition?

The elimination of a small but particularly innovative player may harm competition under a specific set of circumstances. Most notably, this could happen when the target firm possesses a disproportionately large potential to reshape the future technological or competitive landscape, and the acquiring firm has both the incentive and ability to neutralize this threat. This concern has crystallized in recent years under the notion of so-called “killer acquisitions”—a theory of harm that has gained prominence in both academic literature and enforcement practice.

The theory of killer acquisitions refers to the strategic purchase of an innovative target, such as a startup or emerging competitor, with the intent (or likely effect) to terminate or shelve its development pipeline, thereby pre-empting a future threat to the incumbent’s position (Cunningham, Ma, & Ederer, 2021). The concern is that such mergers remove organic sources of dynamic rivalry, particularly in markets where innovation is the primary competitive parameter. Under this view, the merger not only consolidates market power in the present but also forecloses future competition that might have arisen in the absence of the deal. The popularity of this theory has led to the emergence of adjacent concepts, such as:

• “Acquisitions for sleep”, where patents or technologies are quietly retired post-merger, rather than exploited.
• “Reverse killer acquisitions”, in which incumbents reduce or terminate in-house R&D in favour of acquiring external projects, thereby removing incentives for organic innovation.
• “Acqui-hire strategies”, where the true objective is not to acquire a product or technology but to absorb key human capital, often without triggering substantive merger review.

We note that, while intuitively appealing, the killer-acquisitions theory is not without controversy.

First, it risks overstating the link between market entry and innovation. As noted, economic research cautions against a simplistic correlation between the number of firms and the degree of innovation. Concentrated markets may still exhibit robust innovation if dominant firms face credible threats or are incentivized to invest pre-emptively. Thus, fewer firms do not always mean less innovation, and nor does acquisition always lead to suppression.

Second, empirical support for killer acquisitions remains limited and industry specific. A Commission-sponsored study estimates that 5–7% of pharmaceutical acquisitions may qualify as “killer”, a rate comparable to Commission interventions (European Commission, 2024). In contrast, evidence for killer acquisitions in digital markets is significantly weaker, with most acquisitions resulting in product continuation or enhancement, and acquired teams often retaining leadership roles within the buyer’s organization (Ivaldi, Petit, & Unekbas, 2025).

Third, post-merger effects are difficult to observe externally. Discontinuation may result from strategic integration, technical incompatibility, or shifting demand, and not necessarily from anticompetitive motives. Enforcement errors in this area risk chilling benign M&A activity and stifling productive combinations of talent, IP, and capital.

Finally, the so-called “kill zone” hypothesis—that active acquirers deter entry by creating zones of inevitable absorption—is also contested. Some evidence suggests that active acquisition markets instead spur entry and innovation, as startups design themselves for acquisition (Callander & Matouschek, 2022). An adjacent theory is that incumbents buying startups may skew the direction of innovation: entrants may innovate not to disrupt, but to align with incumbent needs (Mendelsohn & Breide, 2024). The long-term concern here is not the absence of innovation, but its path dependency and conformity. That said, it is questionable whether competition authorities like the Commission have the enforcement resources, or the legal mandate, to second-guess the direction of innovation a firm wishes to undertake.

C.6.a    How should the Commission account for the ability and incentives of nascent innovative companies to scale up when assessing the impact of a merger on competition?

The potential benefits and harms of mergers related to innovation in general, on which we elaborated under C.3 remain applicable to transactions where the target is a nascent firm. Perhaps one unique feature of nascent acquisitions that deserves additional consideration is funding. When accounting for nascent innovative firms’ ability and incentives to scale up, the Commission should give weight to the possibility that the merger will supply the small firm with the necessary financial, organizational, and technological resources to expand. This possibility stands as an additional point of caution when considering so-called killer acquisitions.

C.7     In what circumstances can mergers positively impact the ability and incentives of the merged company to innovate?

We welcome the Commission’s recognition that mergers are not inherently inimical to innovation. Under the right conditions, they can enhance the merged firm’s incentives and abilities to innovate, sometimes with positive spillovers for rivals and the broader market. These pro-innovation outcomes typically arise when the transaction facilitates more effective use of R&D resources, enables the combination of complementary capabilities, or changes market conditions in ways that stimulate dynamic rivalry.

One positive pathway is the coordination effect: the ability to reorganize and reallocate research efforts across the combined entity’s laboratories, divisions, or units. By eliminating duplication, the merged firm can free resources for new or higher-value projects, accelerating the development cycle. As two Commission officials have also noted, “a merger will yield the greatest economies of scale in R&D when the merging parties are conducting overlapping research programmes” (Ilzkovitz & Meiklejohn, 2003). In this scenario, the merged entity can spend less to achieve more, thereby increasing R&D productivity.

A second pathway involves learning effects and knowledge spillovers. When firms merge, they integrate not only assets but also (tacit) know-how, which is embedded in processes, teams, and organizational routines. Innovation sharing across the merged firm can generate breakthroughs that neither party could have achieved alone. These effects are often more pervasive and impactful in innovation-driven markets than the cost efficiencies typically associated with price competition. For this reason, some economists argue that innovation synergies should be “considered in the core assessment of mergers rather than relegated to a later stage of efficiency defences”. (Calvano & Polo, 2021).

Moreover, a merger between innovators with complementary capabilities—e.g., one firm with advanced design expertise and another with strong manufacturing or regulatory know-how—can expand the combined entity’s innovation frontier. In such cases, the merger enhances the merged firm’s ability to develop and commercialize new products, particularly in sectors where cumulative innovation and technology integration are important.

Finally, in some cases, post-merger innovation may expand the overall size of the market. For example, the merged entity could introduce a demand-expanding innovation that increases consumer adoption of a technology, raising the total level of investment in the market (Bourreau & Jullien, 2018). This can have a multiplier effect: rivals may have incentives to respond to the expanded market opportunity by increasing their own R&D, leading to higher aggregate innovation.

C.7.a    What elements, evidence and metrics can the Commission consider when balancing the potential positive benefits and spillovers of enhanced R&D capabilities against the potentially anticompetitive effects of a merger?

A merger is more likely to have a positive impact on innovation when:

• The parties’ R&D pipelines are complementary or overlapping in ways that facilitate redeployment and/or consolidation.
• The merger combines critical capabilities or resources that were previously siloed. Examples include consolidating fragmented IP portfolios, combining a strong engineering team with an established manufacturing division, and integrating two products under a single, compatible standard that accelerates development.
• Market conditions—such as growing demand—ensure continued contestability, so that a more capable merged firm still faces incentives to innovate.

C.9.    In what circumstances can the elimination of a potential competitor (that is likely to enter the market in a near future or already exert competitive constraints even if not in the market) harm competition?

Potential competitors can constrain markets in two principal ways:

• Actual Potential Competition: Where a firm is on the brink of entry and is expected to exert competitive pressure once active in the market.
• Perceived Potential Competition: Where the credible prospect of entry disciplines incumbents even before the entrant begins supplying the market.

In either case, the competitive harm arises from the loss of this “disciplining effect” (Schumpeter, 1942). If the potential entrant is eliminated through a merger, incumbents may face reduced pressure to innovate. The likelihood of harm is higher when:

• The potential entrant possesses capabilities or assets that position it uniquely for successful entry, such as superior technology, a strong reputation, or relevant regulatory expertise (e.g., in biopharmaceutical markets).
• The target is already engaged in pre-entry competitive activity—e.g., marketing products or conducting R&D targeted at the incumbent’s market (i.e., the target is already influencing the incumbent’s competitive behaviour)
• Industry conditions suggest low likelihood of alternative entry, such as high sunk costs, strong network effects, or regulatory barriers. Thus, the loss of this entrant significantly reduces contestability.

That said, we urge caution when intervening against acquisitions of potential competitors, as not all acquisitions of potential rivals are harmful. Where the target lacks the necessary capabilities to innovate in the relevant space, faces substantial financial impediments, or its entry is speculative, the firm’s elimination may have little competitive impact. In fact, if the merger materially increases the combined entity’s ability and incentive to innovate—for instance, by combining resources essential for successful product development—the transaction may enhance rather than diminish innovation, even when it involves potential competitors.

C.9.a    How should the Commission assess competition risks linked to situations where a merger eliminates a potential competitor, i.e., the target is likely to enter in a foreseeable future and become a competitor, or despite not yet being in the market already exerts competitive constraints due to its capabilities to enter?

Where a merger eliminates a potential competitor, the assessment should be firmly rooted in the specific circumstances of the case, with particular emphasis on the target’s capabilities, resources, and strategic positioning. The central question is whether the target is likely to enter the relevant market in a way that would materially alter competitive dynamics, either through direct participation or by exerting credible competitive pressure from an adjacent position.

As with all merger assessments, the first analytical step to evaluate the acquisition is market definition. Due to the nature of potential competition, defining a relevant product market may not be easy. In some cases, however, the merging parties may be potential competitors in an emerging market—one that is not yet established but whose contours can nonetheless be defined with reasonable confidence. This is sometimes referred to in the literature as the “future markets” approach (Kern, 2014). The Commission should, considering the relevant case-specific evidence, strive to define such a future market to better demarcate the boundaries of innovation and competition.

Next is to assess the fit between the target’s capabilities and the trajectory of the (future) relevant market. In dynamic industries, entry potential depends not only on the possession of resources but also on whether those resources align with the direction of technological and competitive change. For example, in markets evolving toward increasing specialization (as in some parts of the semiconductor industry where design and manufacturing are becoming very specialized), an entrant should possess specialized capabilities and resources to be considered a credible potential competitor. By contrast, in markets shifting toward integration, a credible entrant would need to show integrated capabilities.

Finally, the Commission could consider the merger in greater detail if it threatens to foreclose future innovation rivalry. Where both firms are likely and capable to enter or expand into the same market in the foreseeable future, the merger may remove the prospect of head-to-head innovation competition, especially where their capabilities overlap in ways that make them substitutes for each other in the innovation process.

C.9.b    Under which conditions could this theory/these theories of harm occur?

Based on our answer to the previous question, we suggest that a loss of future innovation competition could occur if the merger consolidates two firms that are likely to enter the future market with overlapping capabilities that can result in the development of largely substitutable products. By contrast, if the merger combines complementary capabilities, this could constitute evidence that the transaction can enable innovation that could not have occurred otherwise.

Paragraph 60 of the HMG provides that the first leg of the legal test for potential competition requires that the potential competitor already exerts a significant constraining influence on the market, or that there is a significant likelihood it would grow into an effective competitive force. In our view, this standard should not be substantially amended to accommodate the mere threat of potential competition—whether real or perceived by the incumbent—as sufficient in itself.

It is important to distinguish between recognizing the relevance of perception and equating perception with actual competitive constraint. The Court of Justice has, in similar contexts, acknowledged that the perception of competitive threat can influence incumbent behaviour. Indeed, how an incumbent assesses the capabilities and intentions of a potential entrant may shape its own innovation strategies. However, this does not mean that perception alone should be determinative. Doing so risks lowering the evidential threshold in a way that is unlikely to satisfy the Court’s demanding standard for merger prohibitions. Without objective corroboration, reliance on perception could invite speculative findings and potentially chill benign transactions.

EU courts have repeatedly stressed that merger decisions must be grounded in cogent and convincing evidence that can support a finding on the balance of probabilities. Absent tangible indicators of timely entry, elevating the mere threat of potential competition to the level of “significant constraining influence” could be rejected on judicial review as insufficiently substantiated. Accordingly, while the perception of threat can and should be a relevant factual consideration, it must be supported by objective evidence that the potential competitor is genuinely capable of entering and competing effectively within a reasonable timeframe.

C.10.  How should the Commission assess situations where the presence of a potential competitor will exert sufficient competitive constraints to countervail the merging parties’ market power?

Where the Commission establishes the presence of a credible potential competitor—as defined in our earlier answers—this should be treated as a relevant disciplining factor on the merging parties’ market power. Contestable markets theory shows that even a monopolist can be constrained if entry is both feasible and credible (Baumol, 1982). In such cases, the prospect of entry can sustain competitive pressure on innovation. If, as the theory suggests, there exists a capable and incentivized entrant, the merged entity is likely to remain attentive to innovation and responsive to market conditions, knowing that failure to do so could invite entry and erode its position.

C.11.  How should the Commission consider the pre-merger situation in the counterfactual assessment, i.e. when assessing what would have been the situation prevailing absent the merger?

Both the Horizontal and Non-horizontal Merger Guidelines acknowledge that mergers should be assessed against an appropriate counterfactual. While the default counterfactual is static (based on conditions at the time of the merger), both guidelines allow for a dynamic approach where future developments can be reasonably predicted.

For mergers where innovation effects are central, a dynamic counterfactual should be the rule rather than the exception. Innovation unfolds over time and cannot be meaningfully assessed through static models designed for price effects. The Commission should therefore examine how the relevant technological, economic, and legal context is likely to evolve absent the merger. This includes considering technological developments (such as whether the implicated technology is undergoing rapid change or approaching maturity with diminishing innovation opportunities); economic conditions (e.g., expanding or contracting demand in the relevant market), and legal factors (for instance, regulatory developments that can ease or constrain entry).

C.12.  What constitutes the right counterfactual for the Commission where crises, such as the COVID 19 pandemic, wars, or trade measures may have led to short-term shocks of potential temporary rather than permanent nature?

Crises are a relevant and often decisive element in a dynamic counterfactual analysis. They represent periods of discontinuity in which firm strategies shift, competitive conditions realign, and innovation may become both riskier and more rewarding. The European Court of Justice has already acknowledged the relevance of crises in antitrust assessments, confirming that such events can legitimately inform the analysis.

While most crises are temporary, they can trigger lasting structural changes in market conditions, firm behaviour, and customer demand. Where the Commission can predict such changes with a reasonable degree of certainty, the counterfactual should reflect them. For example, post-COVID changes in working patterns have had a durable impact on videoconferencing and related technology markets. Similarly, in the current climate, geopolitical tensions and renewed trade barriers may alter supply-chain strategies, leading firms to consolidate capabilities for resilience—as in the case of European semiconductor design, where domestic capacity-building could be a merger driver in anticipation of hostile trade measures from key supplier jurisdictions.

In short, crises should be incorporated into the counterfactual not merely as temporary shocks, but as potential catalysts for structural market shifts, where such shifts can be identified and evidenced.

C.13.  What should be the right counterfactual in cases of acquisitions of firms in financial difficulties?

In cases involving the acquisition of a firm experiencing financial difficulty, the counterfactual should remain dynamic and forward-looking, particularly where innovation effects are at-stake. Financial distress does not necessarily imply competitive irrelevance. Many innovative firms operate at a loss for extended periods, as significant R&D investments and high-risk projects may take time to yield returns. The Commission should therefore assess whether the firm is still an active innovation player despite its financial position—as those innovation-related capabilities could transfer to a new owner in case of bankruptcy. The Commission’s analysis should also determine whether the financial difficulties are linked to ongoing innovation efforts. If so, the counterfactual should include the realistic possibility that these efforts could succeed and generate competitive benefits absent the merger.

Where innovation projects are underway, the Commission should examine the firm’s capabilities to develop and commercialise them. Such capabilities include product-related capabilities—such as development capacity, technology integration skills, and commercialization channels, as well as financial capabilities—in addition to the ability to raise capital, attract investment, or otherwise secure liquidity to sustain innovation until market launch.

C.13.a  Under which conditions should a failing firm defence be accepted?

Where a financially distressed firm is the subject of a merger, the failing-firm defence should only be accepted under conditions that balance the potential competitive harm from increased market power against the costs of asset and capability loss if the firm were to exit.

If the distressed firm is still innovating (as outlined in the previous answer), the Commission should conduct a deeper, case-specific analysis to determine whether the loss of its innovation potential absent the merger would be more detrimental to competition than the market power effects of the transaction. This requires examining whether the firm possesses both the product-related capabilities (e.g., R&D, technology integration, commercialization) and financial capacity to sustain its innovation until commercialisation.

If the firm is not innovating (or incapable of sustaining innovation), the Commission should weigh the harm from any resulting market power increase against the harm from asset and capability destruction. This is particularly relevant in the EU context, where bankruptcy proceedings are unharmonized and often inefficient—thus raising the risk that valuable assets or know-how could be lost permanently, rather than reallocated to productive use.

C.14   What should be the right counterfactual in cases of acquisitions of firms in declining markets where there is clear evidence that the market size or total demand in a market is shrinking on a permanent basis?

In declining markets, where total demand is shrinking on a permanent basis due to technological change or lasting shifts in consumer behaviour, the counterfactual should still reflect the possibility of dynamic competitive responses.

Economic theory recognises demand growth as a strong driver of innovation, but the reverse is not necessarily symmetrical. While diminishing demand reduces static profit opportunities, it can also spur firms to innovate in order to counteract decline. In such contexts, firms may pursue demand-expanding innovations—e.g., by adapting products to new uses, improving performance to attract switching consumers, or developing adjacent offerings that revitalise the market.

Accordingly, the Commission’s counterfactual should not assume that a declining market inevitably leads to reduced incentives for innovation. Instead, it should examine whether firms have capabilities and strategies aimed at reversing or slowing decline. Relevant evidence could include recent or ongoing product adaptations targeting new or residual demand, and the likelihood that innovation could redefine the market or create new submarkets.

C.15.  Please explain whether you would consider justified to counterbalance the higher level of uncertainty related to the assessment of more distant future market developments also with a more significant impact of the expected effects.

The Court of Justice’s position should be understood as a call to calibrate the quality and intensity of evidence to the degree of uncertainty inherent in the theory of harm. The further into the future the alleged effects are expected to materialise—or the more contingent they are on multiple uncertain events, as in Tetra Laval—the more demanding the evidentiary standard should be.

Accordingly, in cases involving distant harm, it is justified to require more persuasive, multi-sourced evidence to substantiate the Commission’s assessment. This could include triangulation of internal documents, market studies, and third-party testimony; independent expert assessments of technological and market developments; and ex-post studies of previous cases that suggest plausibility under similar conditions.

C.16.  How far in the future should the Commission look when assessing the impact of a merger on competition? How and under what circumstances should the Commission’s assessment consider long investment cycles in a given industry?

The appropriate time horizon for assessing a merger’s impact on innovation should be determined case by case, reflecting the investment and development cycles of the industry concerned. For example, the semiconductor industry cycle resets every four to five years, while some pharmaceuticals may require a decade to progress from clinical trials to commercialisation. In Cisco Systems, the General Court considered a three-year horizon “relatively long” for innovation in video-communications markets. A uniform timeframe would therefore risk arbitrariness.

With that said, the Commission should generally be more willing to assess longer-term effects when innovation is at-stake. Unlike price effects, which may emerge and dissipate quickly, innovation effects typically materialise only after significant lags—from identifying an opportunity, through R&D and prototyping, to market launch and subsequent incremental improvements.

This means that, where credible evidence supports longer investment or development cycles, the Commission’s assessment horizon should extend accordingly, ensuring that potential competitive harm or benefit to innovation is not overlooked simply because it lies beyond the short-term view often applied in price-effect analysis.

C.17.  How should the Commission’s assessment take into account systemic trends and developments unrelated to the merger that may (indirectly) impact the relevant product market and thus the competitive assessment within that market?

Systemic trends and developments—whether technological, regulatory, or geopolitical–shape the broader socioeconomic context in which firms operate. Even where their link to the merger is indirect, they should be incorporated into the competitive assessment where that relationship can be reasonably substantiated.

Economic literature recognises that firms often innovate to mitigate or adapt to systemic changes in resource availability. When systemic trends increase the scarcity or strategic importance of a resource or technology, they can trigger sector-wide innovation incentives. For example, carbon-pricing mechanisms in the EU have spurred innovation in energy-intensive industries to reduce emissions. In a similar vein, the rapid diffusion of AI across sectors is altering production processes, product design, and competitive dynamics.

In merger analysis, the Commission should examine the role of the merging parties in the systemic trend. If, for instance, the target is actively shaping the trend—such as developing sustainable production methods in a resource-constrained sector—the transaction’s impact on that trend becomes relevant. The key question becomes whether the acquirer intends to use the target as a springboard for innovation within the systemic change, or to neutralise a competitive threat arising from it.

By embedding systemic trends into the substantive assessment, the Commission can ensure that merger control remains sensitive to long-term, structural drivers of dynamic competition, rather than focusing solely on immediate market conditions.

Topic E: Digitalisation

E.1.    Do the current Guidelines adequately reflect the evolutions linked to the digitalisation of the economy?

The existing guidelines do not sufficiently reflect the dynamic realities and rapid evolution of digital markets. Given the rapid pace of technological innovation, static-market definitions have become outdated and increasingly inappropriate. The current analytical framework inadequately recognizes the ease and speed with which digital solutions can shift their use cases or functionalities—often accelerated significantly by developments in AI. Moreover, it underestimates how quickly market structures themselves can transform, rendering previous assessments rapidly obsolete. This is particularly evident in markets characterized by strong network effects, where apparent market power or dominance held by a single firm at a given moment can swiftly be eroded by competitors (Evans & Schmalensee, 2016).

Additionally, contemporary enforcement often errs by failing to appreciate that, in two-sided or multi-sided markets, firms with substantially differentiated business models or distinct market positions can still impose significant competitive constraints on one another. Indeed, firms operating one-sided and two-sided business models can, and often do, compete (Ramos & Broos, 2017). Consequently, the current regulatory approaches risk mischaracterizing competitive markets as highly concentrated, potentially harming innovation and consumer interests.

E.2.    Should the revised Guidelines better reflect the evolutions linked to the digitalisation of the economy?

The revised merger guidelines should carefully balance their assessment of competition in the digital economy, recognizing that the highlighted aspects—such as “tipping” dynamics, network effects, data-driven competition, and ecosystem interdependencies—are not inherently harmful. Rather, they often provide significant competitive benefits that enhance consumer welfare, innovation, and market efficiency.

First, while “tipping” or “winner-takes-most” dynamics can theoretically lead to market concentration, they also encourage fierce competition for the market, fostering rapid innovation and investment. In other words, concentration tendencies may coexist with robust innovation incentives—especially in fast-moving sectors.  Indeed, the high stakes involved in “winner takes most” may strengthen competitive incentives, potentially accelerating innovation. By contrast, a permissive regime in which multiple participants can share rewards may lead to a “waiting game”, reducing the incentive to compete vigorously (Denicolò  & Franzoni, 2010).

Second, network effects, frequently perceived negatively, are in fact crucial for creating consumer value in digital markets. The value of many digital services—such as social-media platforms—increases as more users participate. While network effects can reinforce market positions, they also heighten competitive pressure, pushing companies to continuously innovate and improve service quality to maintain user engagement. Thus, “competition for the market” or “life-cycle competition” can replace ordinary compatible competition, and can even be fiercer (Farrell & Klemperer, 2007).

One reason for this tendency is that network effects are a two-sided sword: When people start leaving the platform, each departure makes the network less valuable to those who remain, potentially triggering a mass exodus that feeds on itself. MySpace demonstrates this perfectly: once users began leaving, network effects operated in reverse, hastening the platform’s collapse. The emergence of platforms like TikTok demonstrates how even markets perceived as dominated by powerful network effects remain contestable.

Third, customer inertia or switching costs, while often viewed negatively, can reflect consumer satisfaction, rather than competitive harm. High switching costs sometimes represent consumer preferences or the integration of complementary services that consumers value highly. Indeed, markets subject to lock-in (very high switching costs) may still see fierce competition for users. Companies compete upfront to attract such consumers through tactics like penetration pricing, introductory offers, and price wars.

As explained earlier, this “competition for the market” can effectively substitute for standard compatible competition and might even be more intense, as it reduces differentiation (Farrell & Klemperer, 2007). It is not a simple linear relationship, where lower switching costs are always better for consumers. Accordingly, consumer harm in lock-in cases is possible, but unlikely (Shapiro, 1995). Overlooking these nuances risks penalizing successful businesses that achieve loyalty through excellence rather than anti-competitive practices.

Fourth, data-driven competition can, rather than stifling new entrants, empower innovative companies to rapidly scale by leveraging existing data sources or partnerships (Auer. & Manne, 2024). Companies in generative-AI fields are currently demonstrating how startups can quickly disrupt incumbents despite significant data accumulation by existing market players. Access to data often lowers entry barriers, rather than raising them.

Fifth, ecosystems and interconnected products or services generate substantial consumer benefits through convenience, efficiency, and integration. The seamless interoperability within ecosystems like Apple’s iOS or Google’s Android significantly enhances user experience. Misinterpreting these benefits as competitive threats risks undermining the very innovation that drives market success and consumer satisfaction.

Finally, while interoperability is beneficial, the degradation of interoperability between services, a focus of the revised guidelines, is not inherently anticompetitive. Firms may limit interoperability to protect legitimate business interests, consumer privacy, or enhance security. Forced interoperability risks diminishing incentives for proprietary innovation.

In conclusion, the EU’s merger guidelines should avoid broadly categorizing these aspects as theories of harm. A nuanced understanding acknowledges these features can—and often do—foster significant competitive advantages and consumer benefits. Therefore, a balanced and evidence-based approach that carefully evaluates the context and actual market effects is essential. The guidelines should be structured to identify genuinely harmful mergers without stifling competition, innovation, or consumer welfare in the digital economy.

E.3.    How should the Commission take into account the following competitive dynamics in its assessment of the impact of mergers on competition?

E.3.a    “Tipping”/“Winner takes most” dynamics

The prospect of “tipping” is real in digital markets, but the economic impact of tipping is profoundly two?sided (see, e.g., Denicolò & Franzoni, 2010; Farrell & Klemperer, 2007). On the one hand, direct and indirect network effects, high fixed (but low marginal) costs, data?driven learning curves, and single-homing users can propel a single platform to a high?share equilibrium. On the other, tipping is the mechanism that often delivers the largest welfare gains, by enabling firms to unlock scale efficiencies and coordinate communities of users and complementors around superior technical standards. Seen in this light, tipping is not itself the harm; it is an outcome whose welfare effects depend on (1) why it occurred, (2) whether entry barriers remain surmountable, and (3) whether the merged entity will plausibly relax competitive pressure once tipping occurs.

Accordingly, the Commission’s merger-assessment framework should incorporate four guiding principles:

1. Focus on competition for the market, not only competition in the market. Where tipping is likely, the most relevant competitive dynamic is the battle to become the focal platform. This competition can come from current market players (be they large incumbents or smaller rivals) or firms outside the market that have the right set of dynamic capabilities and assets to compete.
2. Differentiate defensive from offensive tipping. Empirical evidence shows that tipping can be driven by superior product quality, faster iteration, or better ecosystem governance—not merely by foreclosure or predation. A merger that allows the parties to realise scale?dependent quality improvements, to reduce stand?alone cost duplication, or to pool complementary datasets can enhance offensive competition and accelerate tipping in favor of the better platform, benefitting users. In such cases, the Commission should weigh dynamic-efficiency gains at least as heavily as short?run output effects.
3. Even seemingly strong winners can be disrupted where switching costs are falling, users multi?home, or radical innovations emerge (g., MySpace being outcompeted by Facebook, and Yahoo being outcompeted by Google). Enforcers should therefore only conclude tipping is likely to occur when there is strong evidence that the merging firms would acquire an unassailable competitive advantage whose consumer-welfare costs outweigh its benefits.
4. Use effects?based screens, not structural presumptions. Tipping risks should not be treated as a per?se theory of harm. The Commission should require a plausible causal chain: the merger must materially increase tipping probability, and the post?tipping environment must be predictably anticompetitive (g., via higher prices and reduced innovation). Experience with markets that tipped but remained benign (e.g., Blu?ray, cloud gaming, ride hailing in several member states) confirms that scale alone does not foreclose dynamic rivalry. In the more severe cases, remedies targeted at interoperability or data access may resolve concerns without prohibiting welfare?enhancing transactions.

Some examples are useful. The VHS/Betamax, GSM/CDMA, and Blu?ray/HD?DVD “format wars” each culminated in tipping, yet ex-ante rivalry delivered rapid innovation, plummeting prices, and widespread adoption of the superior standard. Merger policy that chilled such standard rivalry would have harmed consumers. Likewise, in mobile OSs, iOS and Android overtook Symbian and BlackBerry despite these rivals’ early lead. These episodes counsel caution before assuming that current market leaders will inevitably perpetuate dominance post?merger.

In short, the Commission should eschew an automatic scepticism toward tipping.  Instead, it should ask whether a challenged merger (i) truncates a live race for the market via tipping, and (ii) leads to consumer harm after the tipping point. Where these conditions are not met, tipping dynamics may be a symptom of robust, welfare?enhancing rivalry.

E.3.b.   Network effects

Network effects are sometimes portrayed as intrinsically anticompetitive because they can amplify incumbency advantages. But that same mechanism may reinforce competition and innovation ex ante, as firms compete for the market. To a first approximation, network effects also benefit users who, by definition, attach greater value to large platforms when network effects are present (Katz & Shapiro, 1985; Farrell & Saloner, 1986). Proper merger analysis must therefore carefully question whether reinforced network effects would be beneficial or detrimental to consumers, while accounting for a rich body of economic evidence establishing that network effects rarely create permanent monopolies absent additional exclusionary conduct (Liebowitz & Margolis, 1999; Evans & Schmalensee, 2016). There is, indeed, scant data to support the notion that churn rates are generally lower in digital-network industries than elsewhere throughout the economy.

It is important for policymakers to recognize the dual nature of network effects. Indeed, while network effects may sometimes cement existing market positions, they can also drive dynamic competition for the market, as well as delivering significant welfare gains when market fragmentation is reduced. Likewise, it is important to acknowledge that strong installed?base effects can boost disruptive R&D by offering innovators a large addressable market. For instance, a game like Fortnite benefits from the strong installed bases of the Xbox, Play Station, iOS, or Android networks (Clements & Ohashi, 2005). Similarly, a service like WhatsApp arguably benefits from the strong networks of incumbent mobile ecosystems. In turn, digital mergers may enable new players to better capitalize on existing userbases thanks to merger-related synergies.

It is also important to note that network effects rarely confer perpetual dominance. In the social-media industry, Facebook supplanted MySpace, and TikTok eroded Instagram’s engagement. In search, Google displaced Yahoo!/AltaVista despite their head start and (potential) data advantages, and vertical search engines (such as Amazon and Booking) now capture a large share of commercial queries. In messaging, iMessage, WhatsApp, Telegram, and Signal each command strong positions in different regions and demographics, despite being textbook examples of services with strong network effects.

Given these complex tradeoffs, it is important to avoid per-se condemnation of mergers involving large network?effects. Instead, articulate safe harbours where empirical evidence shows (i) high multi?homing, (ii) low marginal network returns, or (iii) robust adjacent?market entry. And when network effects are problematic, it is important to encourage remedies like data sharing or other access remedies over structural breakups, which risk destroying consumer value by fragmenting networks. Likewise, it is important to use “dynamic” counterfactuals that recognise potential entry catalysed by technological shifts—such as the emergence of generative AI—rather than assuming static post?merger dominance.

In short, while network effects can sometimes accelerate tipping and raise switching costs, they are equally a primary driver of lower prices, higher quality, and faster innovation. The Commission should therefore treat them as an analytical lens—not a theory of harm in itself. Merger assessment must weigh the magnitude of pro?competitive efficiencies against any credible foreclosure story.

E.3.c    Customer inertia

Customer inertia is often portrayed as a source of durable market power, especially in digital ecosystems where users appear “locked?in” by personalised settings, social graphs, or data repositories. Yet inertia, like network effects, is fundamentally ambivalent: it can reflect frictions that impede efficient switching or the fact that users have found a product that meets their needs better than available substitutes. It is also important to recognize that the prospect of user lock-in encourages firms to compete more aggressively for users in the first place, and this competition may counterbalance the higher prices consumers potentially pay when locked in (Klemperer, 1987). Customer inertia is therefore not, in and of itself, an indicator of likely anticompetitive harm. Overly aggressive intervention that erodes beneficial lock?in may undercut investment incentives and reduce the very consumer surplus the guidelines aim to protect.

E.3.d    Data-driven competition

Economic theory and recent empirical evidence—including the work we published in “Antitrust Dystopia & Antitrust Nostalgia” (Manne & Auer, 2021) and “From Data Myths to Data Reality” (Manne & Auer, 2024)—undermine the common assumption that large troves of data automatically translate into insurmountable market power. Data are best understood as non?rivalrous information goods: they can be copied at almost zero marginal cost and, in many cases, are available through public repositories, licences, or synthetic?data generation. This means that control over a particular corpus rarely forecloses rivals who can obtain comparable inputs elsewhere.

Moreover, the value of data displays sharply diminishing marginal returns. Machine?learning research shows that, beyond a certain threshold, additional data add little predictive power relative to improvements in model architecture or domain expertise. The explosive growth of generative?AI services illustrates the point: despite having far less behavioural data than Google or Meta, newcomers such as OpenAI and Anthropic have matched or surpassed the incumbents’ performance by innovating on model design, reinforcement?learning techniques, and developer tools.

Because the marginal gains from ever?larger datasets taper off so quickly, so?called data network effects are usually weaker than is commonly alleged. Freshness often matters more than depth—search?click logs lose relevance rapidly, for example—while new technologies (cloud computing, open?source ML frameworks) and regulatory changes (GDPR, CCPA) continually erode any short?lived scale advantages. In parallel, formal data?portability mandates and voluntary initiatives—from open?banking APIs to repositories like Hugging?Face—have made it progressively easier for challengers to replicate or bypass incumbent data stores.

Far from dampening competition, these dynamics intensify innovation races. The fear of losing ground to data?rich rivals spurs incumbents to reinvent themselves; think of Google’s pivot from Bard to Gemini, Meta’s release of LLaMA, or Amazon’s multibillion?euro partnership with Anthropic. Confronted with nimble challengers, leading firms cannot afford to rest on historical data advantages.

Against this backdrop, merger analysis should focus less on static measures of data volume and more on four questions: (1) whether genuinely unique, non?substitutable data confer a material and durable advantage; (2) whether rivals can access close substitutes through public sources, partnerships, or synthetic generation; (3) whether the proposed combination would actually raise rivals’ costs of obtaining comparable inputs; and (4) whether any foreclosure risks outweigh concrete efficiencies, such as better model accuracy, lower false?positive rates, or stronger cybersecurity. Unless a transaction can be shown to satisfy each of these conditions, intervention would sacrifice real consumer benefits, while doing little to foster additional competition.

The recent trajectory of generative AI, digital maps, and streaming recommendations all bear out this analysis: dynamism and disruption have flourished precisely because data advantages are contestable and because innovation, not size alone, determines competitive success. The guidelines should therefore discard blanket presumptions about “data monopolies” and instead adopt an evidence?based, effects?oriented approach that recognizes both the limits of data?driven barriers and the substantial benefits that well?executed data combinations can deliver to European consumers and businesses.

E.3.e    Privacy protection-driven competition

Claims that dominant digital platforms entrench their power by exploiting user data often overlook the competitive pressure that privacy itself exerts (Acquisti, Taylor, & Wagman, 2016). Differences in data?governance models have indeed become a salient dimension of competition: some services monetize behavioural data to subsidize zero?price products; others compete precisely by promising not to harvest personal information. Far from being a one?way ratchet toward ever?greater data extraction, the market reveals a rich spectrum of privacy/utility tradeoffs that users actively arbitrate.

The past decade offers multiple illustrations. WhatsApp’s end?to?end encryption, Apple’s App Tracking Transparency (“ATT”), DuckDuckGo’s no?tracking search, and Signal’s open?source protocol each gained traction by differentiating on privacy. Likewise, established players have repeatedly pivoted—Google’s Privacy Sandbox, Meta’s encrypted messaging, Amazon’s Ring opting into local?storage video—to accommodate rising consumer demand for data protection. These strategic responses are evidence that privacy is a competitive variable: firms that fail to match users’ evolving expectations lose engagement and revenue.

Economic theory aligns with these observations. Where consumers are heterogeneous in their privacy preferences, platforms maximize profits via horizontal differentiation and tiered services—ranging from ad?supported free tiers to paid, data?light subscriptions. Mergers that combine complementary assets (for instance, a content library with a secure identity layer) can therefore increase product variety and allow more efficient matching of privacy attributes to user segments. Conversely, it cannot be excluded that transactions that eliminate a high?privacy rival without creating a comparable alternative (potentially via commitments) or bring together close competitors could harm welfare.

Accordingly, the Commission’s analysis should test three propositions rather than presuming privacy degradation.

Does the merger diminish privacy options? If the acquired firm’s data?governance model is unique and the parties cannot credibly replicate it in the combined entity, loss of an important competitive constraint is plausible. If, however, rival services (or the merged platform itself) already offer similar or superior privacy assurances, incremental harm is unlikely.

Are scale?driven efficiencies privacy?enhancing? Larger platforms can invest in advanced encryption, differential?privacy tooling, and secure multiparty computation, spreading fixed costs over more users. Such improvements often exceed what smaller standalones could afford. This is one of the reasons why privacy regulation such as the GDPR has reduced competition from small players (Jia, Jin, & Wagman, 2021; Johnson et al., 2023; Peukert et al., 2022). Integration may therefore raise average data protection.

Finally, it is important to ask whether behavioural commitments can resolve residual concerns? Wholesale prohibition ignores the demonstrated ability of technical and contractual safeguards to reconcile privacy with pro?competitive integration.

In sum, privacy concerns warrant diligent scrutiny, not blanket scepticism. The Commission should approach privacy-protection?driven competition as a two?sided inquiry: (i) could the merger realistically erode consumer welfare (in the form of lower privacy choice or quality), and (ii) might it instead galvanize investment in privacy?enhancing technologies and widen the menu of options available to users (assuming this is what they indeed desire)? With this in mind, an evidence?based framework—grounded in careful examination of the actual merger effects—would better safeguard both consumer welfare and European competitiveness than structural presumptions that treat any data consolidation as suspect.

E.3.f    Multi-sidedness of markets

Modern digital platforms are, almost without exception, multi?sided businesses that create value by orchestrating mutually reinforcing interactions among heterogeneous user groups (Rochet & Tirole, 2003; Armstrong, 2006; Evans & Schmalensee, 2016). Because any change in the conditions offered to one group is immediately transmitted to the others through cross?side elasticities, competitive analysis that homes in on a single “relevant market” is liable to miss the forest for the trees. The correct benchmark is the overall price–quality bundle generated by the system of transactions, not the standalone terms faced by any one category of users.

This has at least two important consequences for policymakers. First, what matters for welfare is the net price charged by the platform once cross?subsidies are taken into account (Rochet & Tirole, 2006; Weyl, 2010). A merger that enables the parties to pool fixed costs, to internalise externalities across sides, or to eliminate double marginalisation can therefore reduce the aggregate prices even if it raises fees on one side of the market.

Second, multi?sided rivalry is often orthogonal: platforms that look like imperfect substitutes (or even complements) on one face can exert intense competitive pressure on another. Search engines, social?network feeds, and video?sharing sites all covet revenue from, largely, the same advertisers. In this sense, Facebook, TikTok, YouTube, and Snapchat compete fiercely in advertising markets, even though the consumer experiences they offer are often highly differentiated. Any attempt to assess market power based solely on the consumer?side overlap—or to read concentration from HHI measures calculated in that narrow space—risks condemning transactions that, in fact, intensify competition for ad dollars.

This logic also implies that foreclosure strategies are rarely profitable unless cross?side gains exceed losses from excluding participants. With this in mind, the cost of closing an interface often outweighs the benefit. It follows that theories of harm premised on input foreclosure or self?preferencing must demonstrate not merely the technical possibility of exclusion but its economic rationality in a multi?sided setting.

European experience bears this out. Interchange?fee caps shifted card?network revenues onto consumers without spurring new entry; Android “choice screens” delivered inferior search defaults, while doing little to erode Google’s share of advertising; and television “must?carry” rules reduced broadcasters’ incentives to invest in exclusive content (Anderson & Coate, 2005; Crawford, 2012). Each episode illustrates how remedies that ignore price?structure neutrality can backfire, raising effective prices or depressing quality on the user side that the Commission is bound to protect.

Finally, a multi?sided lens explains why apparent incumbency advantages can dissolve quickly.  Because user groups can switch asymmetrically, an entrant that secures just one side—for instance, short?form video creators in the case of TikTok—can unravel an incumbent’s equilibrium as advertisers and complementors follow the migrating audience. Put differently, differentiation on the consumer side does not shield a platform from competition on the monetisation side. The merger guidelines should therefore resist calls to treat loosely overlapping ecosystems as separate silos; they should instead ask whether the deal will materially dampen cross?platform rivalry, thereby raising the aggregate (multi-sided) prices charged by these platforms.

In sum, multi?sidedness is not a mere technicality, but the organising principle of the digital economy. The Commission’s merger analysis must internalise this by defining markets around the platform, by evaluating concentration cognizant of the potential for multi?homing and cross?substitution, and by measuring harm and efficiencies at the platform level. Only where robust evidence shows that a transaction will raise the aggregate price charged by the platform on both sides—taking full account of countervailing scale economies and cross?side benefits—should intervention be contemplated.

E.5.    Considering modern competitive dynamics, do you consider that having different frameworks of analysis for horizontal relationships and for non-horizontal relationships is still relevant?

Competition enforcement has traditionally distinguished between horizontal (competitor) and vertical (supplier-buyer) mergers, assuming the former are more problematic for consumers. Recently, however, this distinction has been challenged. Some scholars and agencies argue that many vertical or conglomerate mergers have significant horizontal effects, often by claiming the acquirer would have entered the target’s market if not for the merger (“potential competition”).

This theory seeks to reframe nearly any merger as horizontal. This trend is perhaps best illustrated by recent enforcement actions in the United States. It was central to the FTC’s failed attempt to block Meta’s acquisition of Within, where a court found the claim that Meta would likely develop its own competing VR fitness app to be too speculative. This new perspective is also evident in the FTC’s ongoing lawsuit against Meta concerning its earlier acquisitions of Instagram and WhatsApp. Though initially cleared by global regulators who saw them as operating in separate markets, the FTC now argues these were acquisitions designed to unlawfully neutralize competitive threats. This reflects a modern shift where agencies are more willing to assume potential competition and anticompetitive intent.

Despite this trend, fundamental economic differences between merger types persist. Horizontal mergers combine competitors producing substitutes. By definition, they automatically eliminate a competitor and reduce competition, although this harm may be offset by efficiencies that must be proven.

Vertical mergers, in contrast, combine firms in a supply chain who produce complements and do not directly eliminate a competitor. They are often pursued for efficiencies, such as eliminating “double marginalization”, where successive markups in a supply chain are removed after the merger. This integration can automatically lower costs for the merged entity and lead to lower consumer prices. The likelihood of such pro-competitive efficiencies is generally higher in vertical mergers than in horizontal ones (Cooper, Froeb, O’Brien, & Vita, 2005; Lafontaine & Slade, 2007).

Critics counter that vertical mergers have “intrinsic” horizontal harms. The primary theory is foreclosure: a newly integrated firm gains an incentive to harm its downstream rivals by raising their input costs or cutting off supply. This effect is sometimes described as being exactly parallel to a horizontal merger by eliminating an indirect competitor (Salop & Scheffman, 1983; Riordan & Salop, 1995; Rey & Tirole, 2007).

This theorized harm is, however, neither “intrinsic” nor automatic. It is a contingent strategy that depends on complex factors. The merged firm must have significant upstream market power, and foreclosure must be more profitable than continuing to supply rivals. Furthermore, real-world complexities—such as rivals vertically integrating themselves, or firms diverting products to other markets in response to price signals—make the outcome of a foreclosure strategy highly uncertain. What one observer calls anticompetitive foreclosure, another may see as a rational business response to market conditions.

The core distinction lies in the nature of the competitive impact. In a horizontal merger, the elimination of a direct competitor is an automatic consequence. Any potential harm from a vertical merger, however, relies on a subsequent, non-automatic strategic choice to foreclose rivals—a choice that may not be feasible or advantageous. Moreover, a key efficiency from vertical integration—the reduction of double marginalisation—is also an automatic effect that benefits consumers, often occurring precisely in scenarios where foreclosure risk is highest.

While vertical mergers can produce horizontal harms, the mechanism is more complex, and the probability of a negative outcome is lower. The amalgamation of vertical and horizontal mergers is therefore misleading. It ignores that, in the case of the former, the path to consumer harm is indirect, uncertain, and often countered by inherent, automatic efficiencies.

E.6.    How should the current frameworks of analysis for horizontal and for non-horizontal relationships be adapted to assess the effects that digital and tech mergers can have on competition?

The existing EU frameworks for horizontal and non-horizontal merger analysis already contain all the conceptual machinery needed to evaluate transactions in digital and tech markets. What is required is not a bespoke legal test, but disciplined, evidence-based use of the present tools—market definition, competitive-effects analysis, entry and expansion, and efficiencies—tempered by the error-cost logic that underlies modern competition policy. Proposals to toughen digital-merger enforcement, by reversing burdens of proof or lowering structural thresholds, would likely raise the risk of false positives without demonstrably improving consumer welfare, particularly in markets that evolve as quickly as software and online services (Manne, G.A. 2020).

Within the existing merger-analysis framework, enforcers should fold in the economic particularities of digital businesses, rather than treat them as grounds for a new regime. Multi-sided platforms, for instance, often face competitive constraints that cut across functional market boundaries. Likewise, zero-priced services complicate the use of traditional price metrics. And data-related concerns (such as data security and privacy) can matter for product quality. But none of these features changes the fundamental questions the guidelines already ask. They simply mean that market definition may rely more on diversion ratios or attention measures, that competitive-effects analysis must look past static concentration indices to dynamic rivalry, and that entry analysis should recognize how frequently successful entrants accumulate or purchase the data they need to compete.

For example, the empirical literature gives little support to the claim that incumbent data troves invariably foreclose rivals or create insurmountable “data network effects”. Studies of search engines, ridesharing, forecasting models, and generative-AI systems find diminishing returns to additional data, and frequent cases of entry by firms that began with negligible stores of user information (Auer & Manne, 2020). That evidence cautions against presumptions of harm and underscores why enforcement should remain effects-based and grounded in existing frameworks.

Likewise, when a leading platform acquires a complementary business—say, a mobile app or AI startup—the established vertical (non-horizontal) framework is the right lens. The ability–incentive–effect test asks whether the merged firm could deny rivals critical inputs, whether it would profit by doing so, and whether any such strategy would harm consumers net of efficiencies. Efficiency evidence remains vital because vertical and conglomerate deals often lower transaction costs, eliminate double marginalisation, and accelerate innovation. In that respect, the weight of empirical work from Lafontaine & Slade (2007) and others shows that voluntary vertical integrations tend, on average, to benefit consumers. Where the facts demonstrate credible foreclosure risks that outweigh such benefits, the current framework already permits prohibition or tailored remedies. Where they do not, new presumptions would only chill welfare-enhancing investment.

In short, digital and tech mergers call for the same analytical progression that the EU applies in every sector but populated with market-specific evidence rather than conjecture. By retaining an effects-based methodology, grounding theories of harm in verifiable facts, and giving due credit to efficiencies, the Commission can police genuinely anticompetitive deals without throttling competitive dynamism.

E.7.    How should the Commission assess competition risks of non-horizontal mergers that are not based on a foreclosure conduct by the merged entity?

When a non-horizontal merger raises no classic input-foreclosure issues, the Commission’s task is to decide whether the transaction might nonetheless blunt future rivalry—most famously through a so-called “killer acquisition”. The academic record acknowledges that this is possible, but it also identifies a range of alternative, often pro-competitive explanations for the same observable facts (Barnett, 2024). Treating “killer” theories as presumptively dominant risks misdiagnosis and the attendant error costs that the existing framework is designed to avoid.

First, many targets supply complements that blossom only once they are folded into a broader ecosystem. Facebook’s purchase of Instagram and Google’s purchase of Android did not suppress those services; the acquisitions turbo-charged them, allowing both products to scale globally and to introduce features (such as Instagram Stories) that the standalone firms lacked the resources to deliver. That pattern is consistent with the traditional efficiency logic of vertical or conglomerate integration, not with predation.

Second, acquisitions can reflect the ordinary operation of the market for corporate control. Large buyers may simply believe they can redeploy the target’s assets, talent, or intellectual property more effectively than alternative owners—a judgment borne out, for example, by the rapid post-merger improvement in Instagram’s product strategy and Facebook’s ability to leverage that success across its wider portfolio.

Third, a vibrant exit market is itself a crucial ingredient of venture capital and entrepreneurial activity. For many startups, the possibility of being bought is the principal path to liquidity. Empirical work shows that most young technology firms never reach an initial public offering (IPO), and that the prospect of acquisition is often what makes early-stage innovation financeable in the first place. Deterring these deals would therefore chill exactly the kind of experimentation that policymakers want to encourage.

Fourth, some deals amount to project rescue or acceleration. Google’s cash, engineering capacity, and brand credibility turned Android from a small developer platform into a viable competitor to Apple’s iPhone—a transformation unlikely to have occurred had the startup remained standalone.

Finally, the empirical frequency of true “killer” transactions is low and hard to detect ex ante. Even in pharmaceuticals—where pipelines are unusually transparent—recent estimates put the share of genuine killer acquisitions at roughly five-to-seven percent of deals, leaving the overwhelming majority benign or beneficial (Cunningham, Ederer, & Ma, 2021). In technology markets, where overlap is nebulous and pivoting is common, the signal-to-noise ratio is lower still (Gautier & Lamesch, 2021).

Against that backdrop, the Commission should continue to apply the established ability–incentive–effect logic. It should ask whether the merged firm could profitably shelve or degrade the target’s offering, whether doing so would raise long-run profits once efficiencies and reputational costs are accounted for, and whether consumers would be worse off in the plausible counterfactual where the startup struggles to scale or even fails outright. That inquiry must be grounded in concrete documents, market evidence, and realistic alternative scenarios, not structural conjecture. Where the facts show that elimination of nascent competition is both feasible and profitable, the current framework already permits prohibition or targeted remedies. Where alternative explanations dominate, blocking the deal would sacrifice efficiencies, undermine entrepreneurial incentives, and increase the risk of costly false positives.

In short, the danger that an acquisition might “kill” a budding rival is one legitimate theory of harm—but it is only one, and it is often the least probable. A balanced assessment will test that story against the full range of alternative motives that the economic evidence highlights and weigh the merger’s prospective consumer benefits against the realistically demonstrated risks. That evidence-rich, error-cost-conscious approach is the best way to protect competition without strangling the very innovation ecosystem the Commission seeks to foster.

E.9.    How should the Commission assess competition risks of non-horizontal mergers linked to having a broad range or portfolio of products or services that are interrelated or part of an “ecosystem”?

An examination of the EU legal framework and case law reveals that, despite the prominence of “ecosystems” in contemporary antitrust discourse, the novelty and practical significance of the concept remain debatable (Colangelo, 2025). Rather than constituting a new legal category, “ecosystem” in competition law largely operates as a descriptive label for dynamics already examined in other fields. In competition policy, the term has been repurposed to describe multi-product organisations in which a single firm offers an integrated bundle of interconnected goods and services to end users. Moreover, because many digital ecosystems develop around multi-sided platforms, competition law and economics often conflate “platforms” with “ecosystems”. Platforms are better understood as multi-actor ecosystems that create value by enabling interactions among distinct user groups, rather than as mere multi-product firms.

European case law supports this view. The General Court in Google Android (Case T-604/18) described a digital ecosystem as one that “brings together several categories of supplier, customer and consumer and causes them to interact within a platform”, a formulation later cited by Advocate General Kokott. In Android Auto (Case C-233/23), Advocate General Medina drew on computing literature to define an ecosystem as “an operating model in which data and services are shared by a digital platform’s owner with external developers in service to a community of users”, while the Court itself avoided the term. And in Booking/eTraveli (Case M.10615) the Commission relegated a reference to a “multi-product ecosystem” to a footnote.

Against this background, the theories of harm the Commission should consider are the established non-horizontal ones—vertical/input or customer foreclosure (including raising rivals’ costs), tying or bundling (including technical tying and defaults), and classic conglomerate “portfolio effects”. Potential-competition concerns, when they arise, are best analysed under the horizontal/potential-entry framework. There is, in short, no basis for incorporating a freestanding “ecosystem” theory of harm into the guidelines, given that the concept remains undefined from both legal and economic perspectives and the existing framework already provides a robust basis for assessment.

E.9.b    Under which conditions or market circumstances could this/these theory/theories of harm or concerns materialise.

These concerns could materialise only under conditions that make traditional non-horizontal harms plausible. While numerous decisions address scenarios that could plausibly involve digital ecosystems, theories of harm centred on their construction or reinforcement have been largely absent from decisional practice.

In Meta/Kustomer (Case M.10262) the Commission considered the interconnectedness of markets and Meta’s broader ecosystem as context for possible foreclosure. In Microsoft/Activision (Case M.10646), authorities examined whether the deal would strengthen Microsoft’s multi-product ecosystem in ways that could harm competition, particularly in cloud gaming, with the UK ultimately blocking the merger on a classic vertical input-foreclosure theory. Other recent cases pursued more conventional concerns: Meta/Giphy focused on a “killer acquisition” theory of harm and Adobe/Figma (Case M.11033) on what the Commission termed a “reverse killer acquisition”, both applying potential-competition analysis rather than any distinct ecosystem theory.

The Commission’s prohibition in Booking/eTraveli is widely viewed as the first to invoke an ecosystem theory expressly. Some reasoning is novel—network effects might be reinforced and lead to a significant impediment to effective competition even without any increase in sales; multi-homing and active search may not eliminate inertia; and increased user engagement after product improvements can still be framed as harm if it strengthens a dominant platform. Nevertheless, beyond the vocabulary of “ecosystems”, the substance closely resembles the “portfolio effects” approach developed in Guinness/Grand Metropolitan (Case M.938) and later debated in GE/Honeywell (Case M.2220). Notably, the UK CMA cleared Booking/eTraveli, finding eTraveli was not a particularly significant channel for customer acquisition or retention in accommodation OTAs. In practice, then, materialisation requires the classic ingredients of non-horizontal harm: control over a gateway or “must-have” product, limited switching or effective multi-homing, and a credible, profitable foreclosure strategy that would lessen competition rather than merely reflect efficiencies.

E.9.c    What are the elements, including evidence and metrics, that the Commission could use to assess the potential competition risks linked to having an increased portfolio of interrelated products and services.

The elements and metrics the Commission should use are those already embedded in existing doctrine and guidance. The Commission’s Notice on the Definition of the Relevant Market suggests that a digital ecosystem may, in some cases, comprise a primary core product and several secondary products linked technologically or through interoperability; in such scenarios, it may be appropriate to apply principles similar to aftermarket analysis when defining the relevant product market, and where secondary products are offered as a bundle, the bundle itself may constitute a separate relevant market. The Notice also recognises that not all digital ecosystems can be addressed through an aftermarket or bundling lens and that additional factors such as network effects, switching costs, and multi-homing may need to be considered. Yet these references do not offer a distinct analytical framework.

Consistent with decisional practice, assessment should remain effects-based and organised around the familiar ability–incentive–effect logic: whether the merged firm would control a genuinely “must-have” element or gateway product; whether multi-homing, interoperability and contractual contestability undermine the ability to foreclose; whether “vertical arithmetic” shows that foreclosure would be profitable after lost sales are weighed against any downstream recapture; and whether predicted or observed outcomes indicate a significant impediment to effective competition—prices up, quality or innovation down, rivals’ costs raised—rather than efficiencies from integration.

Because mergers involving complementary products or services are generally regarded as pro-competitive and ecosystems can generate efficiencies through innovation, quality improvements and seamless user experiences, the analysis should remain case-by-case, resist new structural presumptions keyed to “ecosystem size” or portfolio breadth, and avoid converting the language of ecosystems into a shortcut for speculative harms.

E.10.  How should the Commission assess competition risks linked to the merged entity’s accumulation of data?

Data accumulation should not stand on its own as a merger concern because—economically—data is just information: it is largely non-rival, often easy to replicate or purchase, and its value is realized only after firms apply know-how and analytics. Those characteristics make durable data moats rare and hoarding difficult. As we have explained elsewhere, multiple actors can use the same or similar information at the same time, and exclusion is hard. Instead, what tends to matter for competitive performance is not sheer volume, but the quality of models, engineering talent, and compute—all complements that diminish the role of raw data scale. The recent evolution of generative-AI markets underscores this point: despite incumbents’ vast data troves, newer firms have repeatedly set the pace, which is inconsistent with claims that data holdings alone entrench incumbents or foreclose entry.

The empirical record that is typically invoked to support “data network effects” is thin and contested, while policy discussion too often treats conjecture as evidence. Even reports that advance strong claims about data advantages cite little empirical work and, where they do, the findings conflict; yet these conjectures have nonetheless been used to justify far-reaching interventions. Experience in merger control likewise offers little support for data-based doomsday scenarios. High-profile transactions that were predicted to entrench dominance have not borne out those fears ex post—Instagram being the most obvious case—suggesting that scale in data does not reliably translate into lasting competitive harm.

Entry dynamics also cut against treating data as a barrier. Successful platforms typically created the very datasets that later became valuable—after they entered by offering a better product. This is the familiar “sequential entry” pattern in two-sided markets: firms attract users with product improvements and only later monetize the resulting signals. That history makes it a mistake to assume rivals must match incumbents’ datasets before they can compete. It would therefore be a mistake to label such advantages as “barriers” in a way that would condemn the essence of competition: superior products and reputations that entrants can (and do) overcome.

Finally, much of the anxiety about data reflects a broader nostalgia/pessimism bias in digital-market enforcement—presuming harm from novel business models and shifting burdens on thin evidence. That stance risks large error costs while doing little to improve case accuracy. For these reasons, “data accumulation” is unlikely to constitute a credible, standalone theory of harm in merger cases—digital or otherwise. Where data genuinely matters, traditional analyses of ability, incentive, and likely effects already capture the relevant risks; the default assumption should be that data advantages are contestable and that efficiencies from combining datasets are real.

E.11.  How should the Commission assess the relevant standard and criteria determining the value of the target’s data in the context of data aggregation?

The value of a target’s data is neither fixed nor uniform across firms. It depends on the use case, on how quickly the underlying signals decay, and—crucially—on the acquirer’s complementary assets (models, engineering talent, compute, product design). Treating today’s “important” data as a lasting moat, or assuming that a dataset that is valuable to one firm is automatically valuable to rivals, is the very static fallacy that has misled much recent policy debate. Empirically, data advantages tend to be fragile: performance gains exhibit rapid diminishing returns once a model is adequately trained, and better curation, features, and architecture often substitute for sheer scale. That is why small, high-quality, task-relevant corpora and even synthetic data can rival or outperform massive but noisy datasets.

Against that backdrop, the criteria that matter most are data quality, difficulty of replication within commercially relevant horizons, accessibility/governance, and use-case value. Quality dominates because coverage, cleanliness, labelling, and signal-to-noise drive model performance more than raw volume. “Uniqueness” matters only insofar as rivals cannot obtain close substitutes—via public sources, brokers, partnerships, experimentation, or synthetic generation—at reasonable cost and speed. Accessibility (legal rights, contractual terms, interoperability, portability) determines whether any notional advantage can be exercised or diffused. And “value” is contextual: the same clickstream or telematics feed may be pivotal for one workflow or firm and trivial for another. By contrast, volume, velocity, and variety are weak predictors of durable advantage because learning curves typically flatten quickly, and many high-velocity streams are perishable and readily recreated by instrumenting one’s own user interactions.

This dynamic, complements-based view aligns with the economic literature on data-enabled learning. As Hagiu & Wright (2020) show, data can confer an edge only under narrow conditions, and even then, the advantage depends on firm-specific complements and can erode as technologies and business models evolve; it is not a general, transferrable “moat”. It also fits the broader record that “data” is just information: rivals can often assemble functionally similar signals, and the real constraint is organizational capability, not hoarding.

Authorities should also keep sight of the flipside. We want firms—including merging firms—to have strong incentives to assemble novel, high-quality datasets, because these combinations often power meaningfully better products (relevance, safety, fraud prevention, diagnostics) and lower costs. That a post-merger dataset is hard to replicate does not make it suspect; it can be precisely what delivers consumer benefits, and those merger-specific efficiencies should not be assumed away. We would therefore caution against raising evidentiary bars in a way that effectively discounts efficiencies, which are both common and often merger specific.

In short, it is important to avoid the static presumption that data which matters today will matter tomorrow or that its importance carries over across firms, while preserving the incentives to create and responsibly aggregate data that leave consumers better off.

Topic F: Efficiencies

F.1.    Do the current Guidelines provide clear, correct and comprehensive guidance on how the Commission assesses merger efficiencies?

The main provisions relating to efficiency defences under the Horizontal Merger Guidelines, namely paragraphs 76-78, as well as the three cumulative criteria on cognizable efficiency defences (consumer benefit (paras. 79-84); merger-specificity (para. 85); and verifiability (paras. 86-88)) could be revised.  

F.3.    How should the Commission assess whether merger efficiencies will benefit consumers that would otherwise be harmed by the loss of competition resulting from the merger?

The Horizontal Merger Guidelines (para. 79) provide that “the relevant benchmark in assessing efficiency claims is that consumers will not be worse off because of the merger. For that purpose, efficiencies should be substantial and timely, and should, in principle, benefit consumers in those relevant markets where it is otherwise likely that competition concerns would occur”.

Several clarifications may strengthen this benchmark. First, by reference to established jurisprudence, the Commission interprets “consumers” broadly to include both final and intermediate consumers (i.e., business customers). The guidelines would benefit from clarification on how the Commission intends to weigh situations where a merger may harm intermediate customers in the short term but promises longer-term benefits for final consumers.

Second, the requirement that efficiencies be “timely” is often difficult to satisfy in practice. In particular, dynamic efficiencies stemming from R&D synergies or complementarities may take longer to materialize, flowing from research collaboration to marketable results and eventually to consumer benefits (Coninck, 2016). The Commission may wish to consider a longer timeframe for such benefits to be cognizably assessed.

Third, demonstrating that efficiencies are “substantial” is also a challenge for notifying parties. This reflects both the scarcity of robust empirical studies on merger-specific efficiencies and the inherent uncertainty in projecting post-merger outcomes. Parties may also hesitate to rely heavily on efficiency arguments, for fear of signalling weakness in their case. To address this, the Commission could provide further guidance—such as a “best practices” document like its notice on the submission of economic evidence—setting out what forms of evidence and methodologies will be regarded as persuasive when substantiating efficiencies.

For our comments on the “location” of efficiency benefits, see our answers to F.3.d.

F.3.a.   For which types of efficiencies and under which conditions those efficiencies will likely be passed on to consumers?

The Horizontal Merger Guidelines distinguish between two broad categories of efficiencies. On the static side, mergers may yield lower prices through cost savings in production or distribution, with the Commission generally attaching greater weight to reductions in marginal or variable costs rather than fixed costs. On the dynamic side, efficiencies may stem from R&D and innovation, leading to new or improved products for consumers. This framework is sufficiently broad to capture the main sources of merger-related efficiencies.

Whether such efficiencies are likely to be passed on to consumers depends critically on post-merger market conditions. As the guidelines note, the greater the competitive pressure the merged entity continues to face, the stronger its incentive to transmit efficiency gains to consumers. Conversely, where a merger significantly increases market power, it is difficult to assume that the merged firm will voluntarily share those gains with consumers.

In assessing this pass-on mechanism, the Commission may wish to place more emphasis on the role of entry and potential competition. In fast-moving, technology-driven markets, competitive dynamics are often shaped as much by emerging players and rapid innovation as by incumbent rivalry. These features may both discipline firms to pass on efficiency gains and explain why efficiency-seeking transactions are undertaken in the first place. The Commission could therefore attach greater weight to efficiency claims in markets undergoing rapid technological change, where dynamic pressures make pass-on more credible.

F.3.b    Whether there are some types of transactions that, due to their nature, or the characteristics of the products or markets at hand, are more prone to efficiencies?

Not all mergers are equally prone to efficiencies. As the ECJ made clear in CK Telecoms, efficiencies cannot be presumed across the board; their existence must be shown in the specific circumstances of the transaction and market at hand. Indirectly, this statement means that some types of mergers are more likely than others to yield credible efficiency gains.

This is particularly the case where dynamic efficiencies are at-stake. As a former chief economist has argued, transactions with a significant innovative dimension may deserve a more favourable assessment than purely “static” mergers, given the additional sources of efficiencies that only arise through innovation (Regibeau & Rockett, 2019). For example, when merging parties pursue overlapping R&D programmes, consolidation can generate economies of scale in research: the combined entity may achieve higher R&D productivity, conducting the same volume of research at lower cost or producing greater output from the same expenditure (Ilzkovitz & Meiklejohn, 2003).

Similarly, where the parties’ assets are complementary rather than overlapping, their integration may unlock synergies that neither could achieve independently. In such cases, the transaction is more likely to deliver efficiency benefits than mergers that are merely consolidating market shares without a complementary dimension.

F.3.d.   How should the Commission trade off benefits and harm between different consumer groups when efficiencies benefit only a certain group of consumers?

It is challenging to assess a merger that benefits some groups of consumers while harming others. One dimension is the tradeoff between intermediate and final consumers. This issue is particularly salient in multi-sided markets, where a transaction may raise costs for business users but simultaneously lower prices or improve services for end-users. Given the ECJ’s consistent emphasis on protecting consumers as the aim of EU competition rules, such tradeoffs are best resolved in favour of final consumers. To do otherwise risks edging towards a “trading partner welfare” approach that sits uneasily with the logic of EU competition law.

Another dimension concerns tradeoffs between consumers in different product or geographic markets. The Horizontal Merger Guidelines currently give priority to “in-market” efficiencies, expressing reluctance to offset harm within the relevant market against efficiencies occurring outside it. That said, competition law has not been entirely blind to out-of-market effects. For instance, in cases involving sustainability initiatives, authorities have grappled with whether harm to present consumers can be justified by benefits accruing to future consumers.

While there is no established framework for such assessments, past Commission practice has linked the consideration of out-of-market efficiencies to the presence of “considerable commonality” between the consumer groups involved (Mohan, 2014). Building on this precedent, the Commission could clarify the circumstances under which different groups of consumers should be regarded as sufficiently common to justify balancing harms and benefits across them.

F.3.e    How should the Commission trade-off benefits that may materialise already short-term (e.g., product improvements) and harm to consumers that could materialise in the longer run (e.g., entrenchment of an already strong or dominant market position, raising barriers to entry)?

A similar tradeoff scenario arises when efficiencies are expected to materialise in the short term, while potential harms are projected further into the future. Product improvements or cost reductions may be immediate, but concerns about entrenchment of market power or the raising of entry barriers may only emerge over time. These situations undoubtedly merit scrutiny, but the Commission should take care not to create asymmetry in its treatment of efficiencies and harms.

The guidelines already state that efficiencies expected to arise only in the distant future should be given less weight. By the same logic, harm theories premised on long-term conjectures should similarly be discounted unless supported by robust evidence. To do otherwise risks creating what was termed an “innovation paradox”, where parties are asked to substantiate benefits within a narrow timeframe while the Commission is allowed greater room to devise long-term harm scenarios (Gurkaynak, 2023). That would also be incompatible with the ECJ’s message that the burden of proof resting on the Commission must rise with the degree of uncertainty involved in its theory of harm.

This does not mean that long-term harms should be ignored; only that they should be treated consistently with long-term efficiency claims. A more coherent approach would be for the Commission to enlarge its temporal horizon across the board—evaluating both efficiencies and harms within a broader timeframe, while calibrating the weight given to each according to the evidence available.

F.4.    What metrics, evidence and factors should be used to assess whether cost efficiencies are likely to be passed on to consumers in the form of lower prices?

When it comes to cost efficiencies, the central question is not simply whether synergies exist, but whether they will be passed on to consumers. The Commission’s methodology for assessing harm offers a useful parallel here. In building a theory of harm, the Commission considers a range of evidence and constructs a coherent, plausible account of how a merger could reduce competitive pressure and thereby harm consumers. A similar philosophy should govern the evaluation of efficiencies: the evidence should be weighed holistically, with the aim of producing a symmetric and reasonable narrative about how the merger could benefit consumers.

In both cases, the decisive issue is the degree of residual competitive pressure. Just as consumer harm materialises when pressure is weakened, consumer benefit materialises only if sufficient pressure remains to compel the merged entity to share its efficiency gains. This assessment is necessarily case-specific and can draw on a variety of evidence. For instance, if the rationale for the transaction is to facilitate entry into a new market or to compete more effectively against a larger incumbent, the merged firm will have strong incentives to convert efficiencies into lower prices or improved offerings. The task, therefore, is to identify the conditions under which such incentives are credible, and to articulate the circumstances in which cost savings are likely to translate into tangible consumer benefits.

F.5.    What metrics, evidence and factors should be used to assess whether consumers benefit from improved goods or services that may result from increased investment and innovation (‘innovation efficiencies’)?

Assessing whether consumers benefit from innovation efficiencies is inherently challenging, as these gains are uncertain and often unobservable at the time of decision-making. Yet this uncertainty is not unique to efficiencies; it also characterises the Commission’s assessment of potential harms, which likewise rests on forward-looking and hypothetical analysis. It would therefore be inconsistent to demand a higher evidentiary threshold for efficiencies than for theories of harm (Padilla, 2019).

The starting point should be the merging parties’ assets and capabilities. Evidence such as R&D budgets, staff and facilities, and strategic plans can serve as a starting point to identify whether the firms possess the resources necessary to translate a merger into innovation gains. Particular attention should be paid to whether the parties’ research programmes overlap, enabling productivity gains, or are complementary, enabling synergies. In either case, the existence of concrete capabilities at the time of assessment provides a tangible basis for evaluating potential consumer benefits.

The Commission should also consider market context. Innovation cycles vary by sector: in biotechnology or pharmaceuticals, benefits may only emerge after long development and approval periods, whereas in digital markets, faster cycles mean innovation efficiencies can materialise sooner. Structural features such as network effects or “winner-takes-all” dynamics can further amplify the consumer benefits of innovation-driven mergers. These factors provide useful metrics for assessing the credibility and timing of claimed efficiencies.

Finally, while ex-ante analysis will always involve uncertainty, this can be mitigated by incorporating ex-post evaluation into the Commission’s practice (Komninos & Petit, 2021). Periodic reviews of past mergers would generate empirical evidence on when and how innovation efficiencies have materialised. Over time, such data could strengthen ex-ante assessments, improve transparency, and provide a more robust foundation for balancing present and future consumer welfare.

F.5.a    Consumers’ willingness to pay as measured by actual purchasing behaviour.

It is unclear whether willingness-to-pay analyses can be useful to assess demand for innovation. Innovation is, by definition, a future-oriented concept. Thus, measuring the existence or probability of demand for future goods/services should only be taken as a starting point and supplemented by other evidence.

F.6.    What would be an appropriate timeframe for efficiencies to be considered timely?

The guidelines currently state only that efficiencies must arise in a “timely” manner, without prescribing a strict time limit. This open-ended formulation is valuable and should be preserved, as it allows the Commission to calibrate its assessment flexibly. In practice, however, timeliness has often been a stumbling block for efficiency defences, with claims rejected on the basis that the benefits would take too long to materialise.

A more balanced approach would be to apply symmetry between harm and efficiencies. If the Commission is willing to accept a theory of harm that involves, for example, the elimination of a pipeline project several years from market entry, then efficiencies expected over a comparable timeframe should be deemed admissible. This would align the evidentiary standards for harms and benefits (Todino, van de Walle, & Stoican, 2019).

In addition, timeliness should be assessed on a case-by-case basis, reflecting industry dynamics. The Commission itself has recognised divergent time horizons in innovation cases—ranging from one to two years in Novartis/GSK Oncology to nearly a decade in Dow/DuPont. The same logic should guide the evaluation of efficiencies: where industry lifecycles are short and innovation rapid, efficiencies may emerge quickly; in longer-cycle sectors, a longer horizon may be both realistic and necessary.

Finally, the Commission should consider the position of the industry in its technological lifecycle. Mergers driven by emerging opportunities in a new technological paradigm may yield efficiencies relatively swiftly, while in mature sectors, efficiency gains may depend on longer-term, research-intensive processes. A dynamic approach that recognises these differences would improve both the realism and the credibility of efficiency assessments.

F.7.    How can competitive benefits and harms accruing in the near future be balanced with competitive benefits and harms accruing in the more distant future?

Balancing near-term harms against more distant benefits is one of the most challenging aspects of efficiency analysis. Static efficiencies may arise quickly but often yield relatively modest welfare gains, whereas dynamic efficiencies—typically linked to innovation—take longer to materialise but can generate far greater benefits in the long run. This asymmetry raises the question of whether competition law should prioritise the certainty of “one bird in the hand” over the promise of “five in the bush”.

A pragmatic way forward is to borrow from the methodology of risk regulation, which evaluates tradeoffs along two dimensions: likelihood and impact. Short-term harms may be highly likely but of limited scope, whereas long-term efficiencies may be less certain but potentially transformative in their impact on consumer welfare. Rather than favouring one timeframe over the other, the Commission could weigh both dimensions explicitly, calibrating its assessment of efficiencies and harms according to their probability of materialising and the magnitude of their expected effects.

This approach would have the advantage of transparency: parties would know that long-term benefits are not discounted solely because of their timing but are instead assessed with reference to their plausibility and potential impact. It would also allow the Commission to capture the true value of dynamic efficiencies, without disregarding the real costs of short-term consumer harm.

F.8.    How should the Commission assess whether efficiencies are a direct consequence of the notified merger?

Whether efficiencies are truly a direct consequence of the notified merger is often described as “merger-specificity”. In principle, this test asks whether the claimed efficiencies could have been achieved by less restrictive means, such as licensing agreements or joint ventures. If so, the efficiencies are not considered merger specific.

While this criterion has an intuitive appeal, both economic and management literatures suggest it should be applied with caution. From the perspective of transaction cost economics, so-called “less restrictive alternatives” are not always less restrictive in practice. Contracts such as licensing arrangements or joint ventures are costly to negotiate, implement, and monitor (Williamson, 1985). They can also generate lock-in effects, as firms become dependent on jointly created assets or specialised know-how, creating disputes and opportunities for rent extraction. Anticipating these risks, firms may simply forego the arrangement altogether—meaning that the efficiencies would never materialise.

Management studies further indicate that partnerships and mergers are not functional substitutes but distinct organisational modes (Hagedoorn & Sadowski, 1999). Partnerships often rely on complementary specialisation and capability-sharing, whereas mergers involve fuller integration and a more comprehensive deployment of resources. Both forms have their place in an innovation-driven economy, but they are not always interchangeable. For this reason, the Commission should be careful in assuming that efficiencies achievable through merger integration could necessarily be replicated through contractual alternatives.

F.10.  How should the Commission make sure that the efficiencies claimed by the parties are verifiable and likely to materialise?

Ensuring that claimed efficiencies are verifiable and likely to materialise is essential if they are to play a meaningful role in merger assessment. In current practice, this requires that efficiencies be supported by precise, convincing, and, where possible, quantified evidence. A useful way forward would be for the Commission to define minimum standards of proof by considering the totality of the evidence in a case. This should include, where feasible, direct evidence linked to the efficiencies claimed, provided by the parties. For instance, cost data showing expected reductions in production costs or the merger’s impact on achieving minimum efficient scale (thereby reducing costs) can be helpful to quantify efficiency gains from a transaction. Taken together, these measures would make it possible for the Commission both to demand rigorous evidence and to provide a structured pathway for parties to demonstrate that efficiencies are real, verifiable, and ultimately beneficial to consumers.

F.12.  Based on which evidence and metrics can the Commission alleviate uncertainties as to the implementation of efficiencies, in particular when they will not materialise in the very short term?

Uncertainty about the timing of efficiencies is not unique; the Commission already confronts similar challenges when assessing innovation theories of harm, which by their nature often play out over extended timeframes. In those cases, the Commission seeks to substantiate its claims with rigorous and coherent evidence, even where the effects will only materialise in the long run. The same standard should apply symmetrically to efficiencies.

Uncertainties can be alleviated by grounding efficiency claims in tangible evidence: strategic planning documents, sector reports that track likely market developments, or investment commitments that lock parties into future action. These sources can demonstrate that efficiencies, while not immediate, are nonetheless credible and likely to emerge over time. By applying to efficiencies the same evidentiary discipline it applies to harm theories, the Commission would both enhance consistency and avoid systematically undervaluing long-term consumer benefits.

Topic G: Public Policy, Security, and Labour-Market Considerations

G.2.   In your experience, have there been interventions by Member States which resulted in mergers that would have otherwise happened, not taking place?

No. Though we are aware of cases in which national foreign investment laws were used to block or re-design deals (e.g., GE/Alstom in 2015). The Commission could take the opportunity to clarify when such interventions are in line when Art. 21(4), and when they are not. As a matter of principle, we think the Commission should take a narrow, conservative reading of Art. 21(4) public policy considerations.

G.4.   Do the current Guidelines provide clear, correct and comprehensive guidance on how the EU merger control assessment takes into account democracy and media plurality considerations?

“Democracy” has never been a goal of EU merger control, nor has it been value protected by the EUMR. No mergers have ever been blocked because they “harmed democracy”, nor would the EUMR—or the Treaties—allow for such intervention on the sole basis of harm to democracy. By the same token, no anticompetitive merger has ever been allowed because it would benefit “democracy”. As a more general point, “harm to democracy” is a highly contentious, politically charged issue with multitude of possible interpretations. For some, such as the Neo-Brandeisians in the US, protecting democracy means stifling large firms and replicating a Jeffersonian system of small producers. From a more classically liberal perspective, on the other hand, “democracy” means abiding by the rule of law and eschewing arbitrary interference in private matters, including voluntary mergers between consenting parties.

Furthermore, there are important theoretical constraints surrounding the connection between mergers, lobbying and democracy. First, the literature on the nexus between mergers and lobbying is far from settled, and the studies that do exist (e.g., Valletti and Broso, 2024) are limited by modelling assumptions that blur their real-world implications. Second, it is not clear that lobbying is contrary to democracy. In fact, several EU laws and institutional frameworks implicitly or explicitly permit and structure lobbying as a legitimate part of democratic governance. This includes Arts. 10-1 TEU, which establish a foundational right to engage with EU institutions, but also interinstitutional agreements on mandatory transparency, a transparency register and other frameworks which strongly suggest that structured lobbying is viewed as a legitimate form of democratic participation in the EU.

Given the potential for ambiguity, error and the lack of a legal mandate to do so, the Commission should not incorporate such vague and ill-defined goal as the direct protection of democracy into its merger analysis. Doing so goes well beyond the remit of guidelines and would in any case require a reform of the EUMR and, arguably, of the treaties. The best way to protect “democracy”, in this case, is to abide by the EUMR and the goals established therein.

As for media plurality, it should be borne in mind that, according to Art. 21(4) EUMR, “pluralism of the media” is an exception which, by its own admission, reflects exogenous considerations that can be invoked by national authorities only in exceptional circumstances. As such, it is not part of the ordinary framework of merger assessment under EUMR. That being said, “media plurality” could, in theory, be understood as a function of product variety and consumer choice and treated accordingly under the EUMR. In this respect, there are several things that could lead to confusion, and which could thus benefit from some clarification on the Commission’s part.

First, not all media mergers result in less media plurality: some media companies see diversification as a rational profit-maximizing strategy. To the extent that consolidation can result in efficiencies, media mergers therefore need not be at loggerheads with pluralism.

Second, as with the question of consumer choice in competition law more generally, the real, and deeper issue is whether more is always better. For instance, media plurality can lead to clutter, disinformation, and low quality, unverified reporting. As can “more” apps on app stores on any give operating system. At the same time, demarcating between “worthy” and “unworthy” media is a much more complex, normative-laden task than demarcating between “good” and “bad” apps. It is a task for which the Commission is ill-equipped, and one which is better left to European citizens who, given the heterogeneity of cultures across the continent, can — and often do — have contrasting views on these sensitive matters.

Third, measuring media plurality and viewpoint diversity is far more complex than evaluating product choice. It involves defining, gauging, and ranking different opinions; distinguishing those deemed “more” important from those considered trivial; and ultimately making deeply normative judgments. The Commission—tasked with assessing the effects of market power on competition and consumers—is not equipped to make such value-laden determinations. Other EU and national instruments are better suited to address these concerns.

In conclusion, the current guidelines do not mention “media plurality” or “democracy” — and rightly so, for the reasons outlined above. However, given growing pressure to incorporate such goals into merger review, the Commission could briefly explain why doing so would be inappropriate.

G.6.   In which circumstances and under which conditions can the Commission consider that a Member State is taking appropriate measures against a merger that is justified to protect its media plurality in the sense of Art. 21(4) EU Merger Regulation?

The Commission should clearly distinguish between situations where a member state avails itself of Article 21(4) to build up a “national champion” or for protectionist reasons, and those in which media plurality is genuinely threatened. Given the widespread availability of competing media platforms, channels, and sources of information — both online and offline — the Commission should, as a matter of principle, approach Article 21(4) claims with scepticism and adopt clear principles for filtering out those that pursue objectives divergent from the aims of the EUMR.

The Commission has largely done a commendable job of seeing through pretextual invocations of “public policy” arguments by member states in the media sector. For example, in 2020 it recognized Italy’s attempt to block Vivendi’s investment in Mediaset as an infringement of EU free movement rules. The same critical lens should be applied to claims made under Article 21(4).

Below we outline some tentative principles that the Commission could reflect in its new guidelines.

• Substance over slogans: Claims should be grounded in demonstrable risks to media diversity, not in abstract or speculative appeals to cultural identity, national sovereignty, or industrial resilience.
• Market-wide impact, not firm-specific interest: The assessment should focus on whether the merger would significantly reduce the diversity of viewpoints available to the public, rather than whether it disadvantages a particular domestic firm.
• Platform abundance principle: In an environment where audiences can access a wide range of news and content via multiple digital platforms, the burden of proof should be on the member state to show that the merger would meaningfully limit citizens’ access to diverse information.
• Consistency with EU values and competition principles: Invocations of Article 21(4) should not be used to circumvent the objectives of EU merger control by shielding inefficient firms or entrenching state-favored players under the guise of protecting pluralism.
• Transparency and proportionality: The national measure must be clearly articulated, evidence-based, and proportionate to the risk allegedly posed to media plurality — rather than broadly formulated or pretextual.

By articulating and applying such principles, the Commission can safeguard the integrity of the EU merger framework while respecting the legitimate interests of member states. Clear, law-based guidelines also help prevent Article 21(4) from creeping beyond its intended scope — an exceptional provision meant to be used only in truly exceptional circumstances.

G.7.   How should the Commission take into account the consequences of increased market power not only vis-à-vis customers but also vis-à-vis public authorities that may also affect customers?

The Commission does not have a legal mandate to take such effects into account. As such, there is no need for the new guidelines to explain them.

G.9.   Under which circumstances and in which conditions should the Commission consider diversity, including in the sense of diversity of opinions, in its assessment of the impact of mergers on competition?

The EUMR is fundamentally an instrument for the control of market power, not a tool for regulating speech, editorial diversity, or viewpoint pluralism. The Commission’s mandate under the EUMR is to assess whether a concentration would significantly impede effective competition in the internal market or a substantial part of it, particularly as a result of the creation or strengthening of a dominant position (Article 2(3) EUMR).

While the concept of “diversity of opinions” is normatively important in a democratic society, it falls outside the legal remit of competition law as defined by the EUMR. One might attempt to analogize it to product choice, but the comparison is flawed and should be used cautiously. The dynamics of viewpoint diversity in a pluralistic society are far more complex than consumer’s product preferences. As such, they raise a set of issues the Commission is ill-equipped to resolve.

For example, how could the effects on “diversity of opinion” be measured? Would a merger between two newspapers with opposing stances on key policy issues (key for whom?) be seen as promoting diversity, since both viewpoints might be represented within the post-merger entity? Or would it be regarded as harmful to diversity because one perspective would be expected to override the other?

Which views and opinions would the Commission consider when assessing “diversity” of perspectives, and by what criteria? How would it balance diversity in some areas against conformity in others? For instance, would differing opinions on the Serbian elections carry more or less weight than conformity regarding the Italian elections? How would these be measured alongside media stances on vaccines, taxes, digital regulation, gay marriage, or public healthcare? Would all topics be included and treated equally? Why or why not? Would some opinions be considered irrelevant—or even undesirable—while others are given priority? Would substantive agreement on an issue, albeit for different reasons, count as diversity? And what about agreement on the same issue for the same reasons, but expressed in a different style? Would that count towards diversity and media pluralism?

The bottom line is this: Assessing whether a merger reduces viewpoint diversity requires normative judgments that risk politicizing merger control and exceeding the Commission’s legal mandate under the EUMR, which is grounded in consumer welfare, not democratic pluralism. Such concerns are better addressed through media-specific regulation, not through reinterpretation of competition law.

The Audiovisual Media Services Directive and national media plurality laws are more appropriate tools for addressing concerns about concentration of editorial control or political influence in the media sector. Merging those concerns into EUMR analysis risks undermining the predictability, legal certainty, and neutrality of merger review. Alternatively, if the EU legislator wishes to broaden the mandate of merger control to include non-economic goals such as pluralism or democratic discourse, this would require a formal amendment to the EUMR, not informal reinterpretation.

As it stands, the EUMR deliberately excludes “media pluralism” from the standard scope of merger control, reserving it instead for the narrow mechanism under Art. 20(4), which member states may invoke only in exceptional circumstances. As with other industries, the Commission should continue to focus on price and foreclosure effects when assessing media mergers—both because this falls within its legal mandate and because it is best equipped to do so. The guidelines should not and, indeed, cannot, deviate from this principle.

G.12. How should the Commission assess the impact of a transaction on wages/working conditions through increased buyer power in labour markets?

G.12.b What theory/theories of harm could the Commission consider?

A theory of harm should only be considered if there is a clear and measurable link to downstream harm or if the merger causes a significant loss of competition beyond standard restructuring effects. It’s important to separate any potential negative effects related to market power from normal concerns about job losses due to restructuring or offshoring, which fall outside the scope of the EU Merger Regulation, and which are often the result of increased productivity.

G.12.e How can the Commission demonstrate that the impact of a merger on wages and working conditions translates into harm to customers? Is it necessary under the legal mandate of the EU Merger Regulation to demonstrate harm to customers in addition to a negative impact on wages and working conditions?

It is more likely the other way around: a decrease in output is likely to lead to a decrease in demand for labour.