Abstract: This paper assesses the European Commission’s draft Merger Guidelines against the treatment of efficiencies, the failing firm defence, and the counterfactual in the 2004 Horizontal Merger Guidelines and the 2008 Non-Horizontal Merger Guidelines. My central claim is that the draft guidelines represent a welcome but incomplete correction. They rightly recognise that merger control can no longer be organised solely around static price effects, structural presumptions, and a narrow conception of consumer harm. They expand the role of efficiencies, introduce the language of theories of benefit, and give greater prominence to innovation, investment, resilience, sustainability, etc. But they remain too close to the older orthodoxy. A merger policy that sees only the loss of rivalry and not the possibility of productive reorganisation will make systematic errors of its own. A modern merger policy must be able to distinguish between mergers that entrench market power and mergers that unlock efficiencies, accelerate innovation, facilitate orderly exit, strengthen challengers, or improve resilience. The Commission’s draft guidelines move in that direction, but do not go far enough.
***
1. Introduction
For two decades, EU merger control has operated under a stable but increasingly strained intellectual settlement. The 2004 Horizontal Merger Guidelines and the 2008 Non-Horizontal Merger Guidelines recognized that mergers may generate efficiencies.[1] They accepted that merger analysis is forward-looking. They acknowledged that entry, buyer power, failing firm conditions, vertical integration, and efficiencies may affect the competitive assessment.
However, horizontal mergers were assessed primarily through the prism of unilateral and coordinated effects, and non-horizontal mergers were assessed primarily through foreclosure. Harms were often inferred from structure, closeness of competition, diversion, foreclosure incentives, or reduced rivalry. Benefits, by contrast, had to be proven with a high degree of specificity. That asymmetry is problematic. The result was predictable: efficiencies were admitted but never decisive.
The failing firm defense was recognized but confined to exceptional circumstances. It ignored that many firms are not legally failing, but are economically non-viable. They are “zombies” who survive through creditor forbearance, repeated refinancing, implicit subsidy, family ownership, state support, or the hope that someone else will exit first.[2] They need to exit because they trap capital, depress investment, sustain excess capacity, and discourage entry by more efficient firms. Of course, a zombie firm may discipline prices today, but it is likely to damage competition tomorrow.
Merger control is necessarily counterfactual: it compares the world with the merger to the world without it. The 2004 and 2008 guidelines recognized this implicitly, and sometimes explicitly. But they did not make counterfactual construction the organizing principle of the analysis. In static cases, the no-merger counterfactual may plausibly be continued independent rivalry. And yet, the no-merger world may involve a failure to scale, failure to invest, under-deployment of technology, chronic non-viability, asset fragmentation, or disorderly exit. Agencies should not accept speculative counterfactuals. But the status quo can itself be speculative. In dynamic merger control, the counterfactual is the case.
The framework of the old guidelines was administrable. It disciplined self-serving claims. It avoided transforming merger review into open-ended industrial policy. Those virtues remain important today, but they have become insufficient. Modern merger cases increasingly involve industries in which the static number of firms is not the only, and sometimes not the main, determinant of consumer, let alone, total welfare. Innovation ecosystems, high-fixed-cost network industries, declining sectors with excess capacity, start-up acquisition markets, leader–laggard structures, data-driven platforms, and fragile supply chains all force merger control to confront a harder proposition: competition is not merely the rivalry that exists today; it is the process by which firms invest, scale, innovate, enter, exit, and reallocate resources over time.
The Commission’s Draft Merger Guidelines are an explicit attempt to update the merger assessment framework after roughly twenty years of experience with the old ones.[3] Many of the changes I expected to find in the Draft Merger Guidelines can be found there. In particular, I appreciate the renewed focus on the effects of the merger on the merging firms’ and their (actual or potential) rivals’ ability and incentives to invest and innovate and on future product market competition. Also positive are (a) the recognition that assessing mergers in industries where capacity is relevant requires the quantifying the effect of capacity constraints on market power via the degree of “pivotality” of the merging firms;[4] (b) the explicit treatment of minority shareholdings and common ownership; (c) the expansion of the effects of the merger on the loss of head-to-head competition on a market for the purchase of labor.
But there are other changes which I miss, or I find lacking. The Draft also illustrates the limits of institutional reform from within an inherited framework. It modernizes orthodoxy rather than replacing it. It expands the list of cognizable efficiencies but does not fully redesign the method for assessing them. It recognizes dynamic benefits but subjects them to an evidentiary template still modelled on static pass-through. It restates the failing firm defense rather than transforming it into a broader inquiry into exit, reallocation, zombie persistence, and disorderly failure.
I focus on those missing or imperfect changes in what follows. Perhaps, not surprisingly, the reforms listed below all point in the same direction: they will make it easier to defend procompetitive mergers at the risk of approving some anti-competitive ones. Doing so would move the Draft Guidelines closer to a genuinely balanced framework—one in which efficiencies are neither presumed to flourish nor presumed to disappear after integration, but are analyzed, like possible anticompetitive effects, through disciplined economic reasoning.
I fully understand that a first-best merger policy is unrealistic. Dynamic efficiencies are inherently harder to quantify ex ante and easier to construct post hoc. Moreover, the political costs of under-enforcement and over-enforcement are not symmetric. I am fully aware that many regard under-enforcement as the key problem the new guidelines should address. However, thirty years of experience tell me that both under-enforcement and over-enforcement are real problems that should be treated just the same. This paper thus modestly seeks to re-balance the economic debate at the risk of looking too bold to some and, perhaps, too timid to others.
2. Reforms I Dream About
Somewhere over the rainbow lies not a world in which mergers are presumed beneficial, but one in which the economics of harm and the economics of benefit are judged by the same standards of evidence.[5] The reforms discussed below[6] seek to bring EU merger control closer to the blue skies, where it could help to address the key problem identified in the Draghi and Letta reports:[7] European firms’ low productivity. Europe’s challenge is to develop an ecosystem in which more productive firms can grow at the expense of less productive firms. This does not happen today for various interrelated reasons. Mergers are not simply aggregations of market shares, as those supporting a strict structural presumption would make us believe. Rather, they are exit mechanisms governing the re-deployment of human capital. A merger policy that allows productive firms to grow and inefficient firms to exit facilitates the re-allocation of human capital and thus becomes an engine of productivity growth and welfare.
2.1. A Broader View of Efficiencies
The first reform concerns efficiencies. A marginal-cost efficiency should count, of course, but a fixed-cost efficiency that enables entry or expansion should count too. Innovation complementarities should also count, and the same applies to innovation sharing and efficient capital reallocation through exit. All these efficiencies can be tested in practice, and all of them, if verifiable, increase consumer and total welfare.
A reduction in fixed costs may be tested through cost accounting, procurement data, production plans, and pass-through analysis. An innovation-sharing claim could be tested through evidence of technological complementarity, integration capability, prior internal deployment, R&D plans, and incentives to scale the innovation. A laggard-merger claim could be tested through evidence of fixed-cost barriers, failed independent expansion, differentiation from the leader, and the ability of the merged entity to become a credible constraint. A resilience claim could be tested through stress scenarios, evidence of exit barriers, persistent losses, demand volatility, and capacity fragility. Because of this, none of these claims should be merely dismissed as inherently implausible or unverifiable.
Merging parties should be required to explain how the merger changes their incentives or the constraints they face, and establish the facts that support their theories of harm. That approach requires that the agencies open their minds to efficiencies other than short-run variable cost savings. It requires that they develop the skills needed to assess dynamic efficiencies properly, and also that they contemplate the possibility of real business efficiencies that do not merely offset possible short-run price increases.
2.2. A Realistic Counterfactual
The second reform concerns the counterfactual. Authorities should recognize that a transaction between close substitutes in a stable market is not the same as a start-up acquisition, a merger among laggards, a distressed-firm acquisition, a network-investment merger, or a consolidation in a declining sector. This classification does not decide the case. It merely tells the agency which questions to ask and prevents the authority from forcing every transaction into a static unilateral-effects template. Many merger disputes are formally about efficiencies or harms, but substantively about the counterfactual. Would the target have remained a competitor? Would the start-up have scaled absent the merger? Would the laggard have entered? Would the distressed firm have survived? Would the industry have restructured smoothly? Would the innovation have been developed independently? These are questions that should determine the case. The burden of answering them should lie squarely in the agencies’ hands.
C) A Zombie Firm Defense
The “classic” failing firm defense should remain, but it should be supplemented by a broader category for persistent economic non-viability. This reallocation filter would ask three questions. First, is the firm economically non-viable as an independent competitor, even if formal bankruptcy is not imminent? Relevant evidence would include persistent losses, repeated restructurings, creditor forbearance, inability to finance necessary investment, declining demand, excess capacity, and dependence on external support. Second, is there a feasible and likely less anticompetitive alternative? This should be a realistic inquiry into whether another buyer is likely within a relevant time horizon and whether that buyer would preserve comparable rivalry. Third, would asset redeployment recreate comparable competition in time? The question is not whether machines, patents, customer lists or facilities could be sold so that they do not “leave” the relevant market. Rather, the agency should ask whether the competitive force embodied in the firm—its organization, know-how, teams, approvals, relationships and capabilities—would survive and constrain the market. Competition is not served by mistaking zombie persistence for rivalry. A merger that accelerates efficient reallocation may be consumer welfare-enhancing even if it reduces the number of firms.
2.3. From Killer Acquisitions to Innovation Ecosystems
The fourth reform concerns innovation ecosystems. Of course, I accept the “killer-acquisition” concern: incumbents may acquire nascent threats to discontinue projects, redirect innovation, or protect existing rents.[8] But many start-up acquisitions are the most efficient exit route for failed projects. The possibility of exit can increase entry and innovation incentives ex ante. Investors finance risky innovation partly because exit is possible. If merger control reduces expected exit value too much, it may reduce entry, experimentation, and venture finance. The same acquisition that eliminates a potential future rival ex post may have helped induce entry ex ante. The relevant questions are therefore: Is the target a credible future competitive constraint? Does the acquisition accelerate diffusion through complementary assets? Is the likely post-acquisition strategy continuation, scaling, integration, discontinuation, or shelving? Do alternative acquirers exist?
2.4. Recognize Resilience as a Defense
Resilience is the ability of a market to absorb demand or supply shocks while maintaining reliable access to products at reasonable prices. It matters in high-fixed-cost, declining, fragile infrastructure-like markets. A market may appear competitive because prices are low and capacity is abundant. But if those low prices reflect unsustainable overcapacity, financial distress, and delayed exit, the market may be brittle. A merger that enables orderly capacity rationalization may improve consumer welfare by reducing the probability of disorderly collapse, even if it increases prices in normal conditions. Resilience claims should not be accepted at face value. They should be tested through stress scenarios. Parties should identify plausible shocks, explain why the market is fragile absent the merger, and show how the merger changes the probability or severity of consumer harm.
2.5. Tailor Standards of Proof to Efficiency Type
The final reform I dream of finding over the rainbow is evidentiary. A proof template designed for marginal-cost savings should not be applied mechanically to every efficiency in the economy. The appropriate evidentiary question depends on the economic mechanism that allegedly gives rise to efficiencies. The standard should be high, but not blind. Static cost efficiencies should be quantified where possible. Dynamic efficiencies should be supported by internal documents, implementation plans, track record evidence, technological complementarity, scenario analysis, and credible causal mechanisms. Resilience should be tested through stress scenarios rather than point forecasts. Exit and reallocation should be assessed through evidence of non-viability, asset specificity, and realistic alternatives.
3. Assessing the Draft Merger Guidelines
3.1. Efficiencies
The Draft Merger Guidelines are not a minor update. They introduce the “theory of benefit:” a merger may benefit consumers by creating “direct” efficiencies (cost reductions, synergies, improved quality, etc.) and “dynamic” efficiencies (innovation, investment, resilience, sustainability and longer-term competitive performance). This is a genuine step forward. But the Draft remains cautious. Efficiencies must still benefit consumers in the short run, be merger-specific, and be verifiable. The burden remains on the parties. The greater the harm, the stronger and more certain the efficiencies must be. These requirements are sensible. But, unless applied with mechanism-specific discipline, they may continue to under-credit dynamic efficiencies. The risk is that dynamic competition is invoked only when it supports a theory of harm based on qualitative assessments, but discounted when it supports a theory of benefit because the alleged efficiencies may not be quantifiable.
3.2. Counterfactual
The Draft also improves the treatment of the counterfactual, as it recognizes that, while the pre-merger situation is normally the benchmark, there are exceptions. It allows for foreseeable changes in the absence of the merger, including entry, expansion, exit, innovation, regulatory change, alternative transactions, and technological development. But the Draft remains anchored in the status quo. While it is reasonable to require alternative counterfactuals to be realistic and supported by evidence, the risk is that the authority requires strong evidence to depart from continuity, but does not require comparable evidence to assume continuity.
The better rule would be symmetrical. In markets characterized by financial distress, persistent underinvestment, technological transition, high fixed costs, declining demand, scaling constraints, exit barriers, or credible evidence that one or more parties may not continue to compete effectively absent the concentration, the Commission should assess whether continued independent rivalry is a realistic counterfactual. In such cases, the pre-merger situation should not be presumed to persist merely because formal exit is not imminent.
3.3. Failing Firm Defense
On the failing firm defense, the Draft is much less innovative. It merely restates the traditional doctrine. Thus, it misses an opportunity by failing to incorporate a broader account of economic non-viability, zombie persistence, asset reallocation, and disorderly exit. The Draft should be amended to make three points. First, the “classic” failing firm defense applies where the firm would exit in the near future, and the traditional cumulative conditions are met. Second, where those conditions are not met, evidence of persistent economic non-viability may be relevant to the counterfactual and to the assessment of competitive effects. Third, the Commission will be required to examine whether alternative outcomes would preserve comparable rivalry. Asset redeployment should not be treated as sufficient merely because physical assets can be sold “within” the relevant market. The decisive question is whether the competitive constraint embodied in the firm could be preserved one way or another in time.
3.4. Innovation Defense
The Draft Merger Guidelines represent an important step forward in recognizing that competition increasingly takes place through innovation, investment, and technological progress, rather than solely through short-run price competition. Yet, the analytical architecture remains fundamentally asymmetric. Innovation is treated principally as a source of competitive harm, while innovation benefits continue to occupy the narrower category of efficiencies, where they are subject to exceptional skepticism and demanding evidentiary requirements.
This asymmetry no longer reflects either modern economic theory or the accumulated empirical evidence. Mergers may stimulate innovation by combining complementary capabilities, improving appropriability, facilitating knowledge sharing, reducing inefficient duplication of research, expanding the scale over which innovations can be deployed, relaxing financial constraints, and enabling investments that neither party could profitably undertake independently.
Competition authorities should remain appropriately skeptical of unsupported innovation claims, just as they should remain skeptical of unsupported theories of innovation harm. But skepticism should operate symmetrically. Neither side should enjoy the benefit of an implicit presumption unsupported by the economics of the particular case. Recognizing an innovation defense is not a departure from effects-based merger control, but its logical completion. If merger policy aspires to evaluate transactions according to their likely effects on consumer welfare, then dynamic efficiencies deserve exactly the same analytical seriousness as dynamic theories of harm. The objective is not to lower evidentiary standards but to apply them consistently.
3.5. Resilience
The Draft’s recognition of resilience, sustainability and security of supply is one of its most important departures from the old framework. The inclusion of these concepts reflects the realities of the present economy: supply shocks, energy transition, digital dependence, strategic inputs, geopolitical risk, and the fragility of global value chains. The danger is that resilience can become a rhetorical shield for anticompetitive consolidation. The Draft should therefore make clear that resilience matters only where it is connected to consumer welfare: reliable access, stable quality, continuity of supply, and avoidance of severe price or output volatility under plausible shocks. The rigorous approach cannot be an excuse to deny resilience arguments that are substantiated by facts merely because they rely on long-term considerations of a probabilistic nature.
3.6. Standard of Proof for Efficiencies
The Draft maintains a high evidentiary standard regarding efficiencies. That is correct. But it does not discriminate across efficiency mechanisms. For static cost savings, the Commission rightly requires quantification, implementation evidence, merger-specificity, and likely pass-through. But that framework cannot apply to other mechanisms. For example, for innovation efficiencies, the Commission should instead require evidence of technical complementarity, R&D plans, internal documents, integration feasibility, incentives to continue innovation, and the absence of a plausible less anticompetitive route. For reallocation efficiencies, the Commission should examine non-viability, exit barriers, asset specificity, alternative buyers, and the speed and quality of asset redeployment. This approach is not less rigorous than the Draft’s. It does not let dynamic claims float in abstraction, but it also refuses to strangle them with the wrong evidentiary template.
The Table below provides a practitioner’s toolkit for the assessment of different efficiency mechanisms.
4. Conclusion
The Draft Merger Guidelines are a serious and welcome attempt to modernize EU merger control. They helpfully recognize that the world has changed since 2004 and 2008. They understand that competition is not only price rivalry and, therefore, give greater prominence to innovation, investment, resilience, sustainability, security of supply, and efficiencies. They introduce the so-called “theories of benefit.” They make counterfactual analysis more explicit. These are important achievements. But, in my opinion, the Draft does not yet go far enough. It remains a modernization of orthodoxy that expands the categories of admissible benefits without fully redesigning the method of assessment. It makes the counterfactual more explicit, but continues to privilege continuity. It does not incorporate a theory of economic non-viability, exit, and reallocation. It recognizes resilience but does not yet discipline it through stress-test counterfactuals.
The risk is therefore twofold. If the Commission applies the Draft conservatively, dynamic efficiencies will remain visible in principle, but under-credited in practice. If it applies the Draft loosely, dynamic claims may become a vehicle for industrial-policy rhetoric. Neither outcome is desirable. The solution is disciplined realism. Merger policy should remain skeptical. But that skepticism should be directed at bad claims, not at categories of economic reality that are difficult to measure. Nothing in the approach advocated in this paper implies abandoning the strict character of EU merger law. What changes is the order and discipline of analysis. The authority should not first presume harm and then ask whether efficiencies can somehow rescue the merger. It should ask what dynamic process the merger changes: rivalry, innovation, investment, entry, exit, scaling, reallocation, or resilience.
A merger policy that cannot see market power will under-enforce. But a merger policy that cannot see efficiencies will over-enforce. The challenge is to make efficiencies part of enforcement when they are likely to be real. Not easy, but not necessarily more difficult than forecasting the merger’s possible future anticompetitive effects or predicting the likely effects of a merger on entry. Ultimately, consumers may suffer when the agency over- and under-enforces, and it is hard to say that one error is more likely or more costly than the other. We need to learn a lesson from central banking: its stated goal is to fight high inflation, but one needs to tread carefully because deflation typically involves much greater social costs. Likewise, consumers do prefer lower prices, but it is important to remember that they are even more averse to empty shelves or empty pockets.
Table 1. Practitioner Toolkit: Assessing Different Categories of Merger Efficiencies
| Efficiency | Rationale | Key questions | Evidence + | Evidence – |
| Variable-cost efficiencies | Lower marginal costs increase incentives to compete on price. | Are cost savings merger-specific? Are they likely to be passed through? | Engineering studies; procurement savings; production optimisation; previous integrations. | Fixed-cost savings presented as variable-cost savings; vague synergy estimates. |
| Fixed-cost efficiencies | Lower fixed costs permit entry, investment and innovation that would otherwise be unprofitable. | Do fixed costs constrain expansion? Would each firm invest independently? | Capital budgeting documents; investment thresholds; board papers. | Firms were already investing successfully without merger. |
| Innovation complementarities | Combining complementary technologies increases innovative capacity. | Are technologies complementary rather than overlapping? | Patent maps; R&D collaboration history; engineering evidence. | Extensive overlap suggesting elimination rather than combination of R&D. |
| Innovation sharing |
Innovations can be deployed across a larger installed base after integration. | Does scale materially increase returns to innovation? | Integration plans; distribution networks; installed base analysis. | Innovation could equally be licensed or commercialised independently. |
| Elimination of duplicative R&D | Coordination avoids wasteful duplication while preserving innovation output. | Is duplication socially wasteful or competitively valuable? | Pipeline comparisons; resource allocation plans. | Independent projects pursue genuinely different technological approaches. |
| Portfolio optimisation | The merged firm reallocates R&D toward higher-value projects. | Are projects complementary or substitutes? | Internal prioritisation documents; expected project returns. | Cancellation of disruptive projects benefiting consumers. |
| Capability integration | Integration combines complementary capabilities unavailable through markets. | Which complementary assets are missing absent the merger? | Manufacturing capability; regulatory expertise; AI infrastructure; data assets. | Capability could be accessed through licensing or contracting. |
| Innovation-for-buyout | Expected acquisition increases start-up investment incentives ex ante. | Would acquisition materially increase expected returns to innovation? | Venture capital evidence; financing constraints; exit valuations. | Target already has credible IPO or independent growth path. |
| Orderly exit / capital reallocation | Merger accelerates efficient redeployment of capital from declining firms. | Is independent survival economically realistic? | Persistent losses; refinancing dependence; excess capacity. | Temporary downturn only; viable restructuring alternatives exist. |
| Laggard merger efficiencies | Combining weak firms creates an effective competitive constraint. | Does the merger create a viable challenger? | Market shares; investment plans; leader-laggard asymmetry. | Parties are already effective competitors. |
| Coverage / network investment | Larger scale makes network rollout economically viable. | Does integration materially increase network coverage? | Engineering models; rollout plans; investment commitments. | Rollout would occur anyway. |
| Resilience efficiencies | Orderly restructuring reduces risk of disorderly exit and supply disruptions. | Does the merger materially reduce collapse risk? | Capacity utilisation; stress tests; financing conditions; industry forecasts. | Resilience argument merely disguises higher prices without structural evidence. |
Jorge Padilla*
* Jorge Padilla (B.A, M.Phil. (Oxon), D.Phil. (Oxon)) is Senior Managing Director at the economic consultancy Compass Lexecon and Senior Fellow of CEMFI in Madrid. He is also Founding Fellow of the Royal Economic Society in London. I wish to thank the comments of Thibault Schrepel which have contributed to make this a much better paper. This paper has not been commissioned or funded by any party, and no party had the right to review the paper prior to its circulation. The author is solely compensated by Compass Lexecon, an economic consultancy. As a consultant, he has represented many companies over the years. He holds no paid or unpaid position as officer, director, or board member of non-profit organizations or profit-making entities whose policy positions, goals, or financial interests relate to the article. The list of his clients can be found at Compass Lexecon Website. This paper does not necessarily represent the views of Compass Lexecon or its clients.
***
References:
- [1] See European Commission, (2004), Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 31, 5.2.2004, pp. 5-18.
- [2] See European Commission, (2008), Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 265, 18.10.2008, pp. 6–25.
- [3] European Commission, (2026), Draft Guidelines on the assessment of mergers under Council Regulation (EC) No 139/2004 on the control of concentrations between undertakings, and references therein.
- [4] Jorge Padilla and Roman Fischer, 2026, European Reflections on the European Commission’s draft Merger Guidelines: A view on pivotality, available at https://compass-lexecon.files.svdcdn.com/production/editorial/2026/05/Reflections-on-the-European-Commission%E2%80%99s-draft-Merger-Guidelines-A-view-on-pivotality.pdf?dm=1780581128.
- [5] “Somewhere over the rainbow, skies are blue, And the dreams that you dare to dream really do come true,” written by Harold Arlen & E.Y. Harburg and sang by Judy Garland for the 1939 film The Wizard of Oz.
- [6] For a more detailed explanation, see Jorge Padilla, 2026, “When and how to rescue efficiencies from oblivion in horizontal merger control?,” 14 Journal of Antitrust Enforcement 1.
- [7] Jorge Padilla and Roman Fischer, 2026, European Reflections on the European Commission’s draft Merger Guidelines: A view on pivotality, available at https://compass-lexecon.files.svdcdn.com/production/editorial/2026/05/Reflections-on-the-European-Commission%E2%80%99s-draft-Merger-Guidelines-A-view-on-pivotality.pdf?dm=1780581128.
- [8] Colleen Cunningham, Florian Ederer & Song Ma, 2021, “Killer Acquisitions,” 129 Journal of Political Economy 649. For argument and evidence pointing in the opposite direction, see Jonathan M. Barnett, 2024, “Killer Acquisitions” Reexamined: Economic Hyperbole in the Age of Populist Antitrust,” 3 University of Chicago Business Law Review; and Marc Ivaldi, Nicolas Petit & Selcukhan Unekbas, 2025, “Killer Acquisitions: Evidence from European Merger Cases,” 86 Antitrust Law Journal 647.
