This chapter assesses existing and potential international co‑operation instruments that could strengthen the resilience of EU pharmaceutical supply chains, organised under three objectives: expanding manufacturing capacity and diversifying sources; facilitating continuous trade in medical supplies and inputs; and managing short- to medium-term supply shocks. Existing instruments yield uneven results: trade agreements can diversify supply sources but may also facilitate offshoring, while mutual recognition agreements reduce regulatory burden yet remain limited in scope. None provides a credible mechanism for co‑ordinated crisis response when national and commercial incentives diverge. The chapter assesses three potential instruments in depth: a multilateral supply-chain monitoring mechanism (MedMIS), a joint procurement arrangement open to non-EU partners (MedPPA), and a pooled pharmaceutical reserve (RescPool). Each addresses a distinct failure mode, but operational use would require resolving governance design and legal safeguards. For all three, it sets criteria for deciding whether to scale up, narrow or discontinue.
Strengthening the EU’s Medical Supply Chains
Abstract
This chapter assesses existing and proposed international co‑operation instruments against three analytically distinct objectives for pharmaceutical supply chain resilience: 1) increasing manufacturing capacity and diversifying supply sources; 2) ensuring the continuous and seamless trade of medical supplies and inputs; and 3) managing short- to medium-term supply shocks. These objectives are not fully aligned: instruments that advance one may constrain another. For instance, non-discriminatory procurement rules that widen options during tenders also increase competitive pressure on EU manufacturers, which may create tension with strategic autonomy objectives. While the existing set of instruments reduces routine trade and regulatory frictions, it lacks credible mechanisms for co‑ordinated crisis response when the incentives of individual states and firms diverge sharply from collective interests.
Broad trade agreements can diversify supply sources and expand manufacturing capacities, although they may also facilitate the offshoring of existing capacity. Because the EU is a signatory to the 1994 WTO Agreement on Trade in Pharmaceutical Products, it already applies zero tariffs on pharmaceutical products and approximately 7 000 active pharmaceutical ingredients (APIs). Further tariff waivers could be envisaged for APIs not covered by the agreement, as well as for key starting materials. Amid current geopolitical tensions, international co‑operation can help maintain a trade regime that facilitates the movement of medical supplies while addressing security of supply.
Strategic partnerships appear more tractable than broad trade agreements. Globally, in the sector of critical raw materials (CRM), “targeted trade deals” have intensified trade between partners by 12%. Results for EU strategic partnerships on CRM, however, are not yet available. Current pharmaceutical initiatives in Africa and Latin America are designed to build regional and local manufacturing autonomy in partner countries rather than to diversify supply sources for EU markets. Their longer-term contribution to global supply diversification may be substantial. Nevertheless, these initiatives should not be evaluated against an EU supply-security criterion that they were not originally designed to meet.
Mutual Recognition Agreements (MRAs) reduce regulatory burdens and facilitate market entry, but they do not directly increase trade volumes. MRAs for medicines signed between the EU and several partner countries (Australia, Canada, Israel, New Zealand, Switzerland and the United States) primarily cover the recognition of Good Manufacturing Practices (GMP) standards and inspections. According to an industry survey, the EU-US MRA is associated with a reduction of approximately 70% in EU Member States’ inspections of US-based facilities. Further extensions could include the recognition of GMP inspections conducted by parties in third countries, the inclusion of additional product categories and the digitalisation of border processes. These digital upgrades would be particularly valuable during crises, when administrative delays often compound physical bottlenecks.
Three potential instruments are assessed in depth to address short- to medium-term supply shocks: a multilateral market monitoring mechanism, a joint procurement instrument open to non-EU partners, and a common stockpile. Each addresses a distinct failure mode. Monitoring improves situational awareness and early warning; joint procurement secures supply commitments before scarcity peaks; and pooled reserves bridge the immediate period between a disruption and production ramp-up. While each instrument would have standalone value, their effectiveness would be greater if they were connected through pre‑defined governance linkages: monitoring intelligence should inform procurement activation and stockpile release, while drawdown and procurement data should feed back into the monitoring baseline. All three instruments are feasible only under specific legal, institutional and participation conditions. They also entail distributional trade‑offs, including potential effects on routine‑period prices, equity of access and market structure.
A multilateral market monitoring mechanism for critical medicines and medical countermeasures, referred to as MedMIS, could improve situational awareness across a defined basket of high-priority products. Modelled on the governance logic of the Agricultural Market Information System, it would not be a shortage management tool in the narrow sense, but a mechanism for validating signals, classifying disruption risks and informing proportionate responses. Its feasibility would depend on two conditions. First, commercially sensitive data would need to be protected through governed aggregation, potentially using a neutral trustee to prevent firm-level identification. Second, the mechanism would require ex ante legal assurance to manage competition-law risks under Article 101 TFEU. Given these constraints, MedMIS should begin as a proof-of-concept pilot covering around 20 to 30 high-priority products, with staged review points to determine whether to expand, narrow or discontinue the mechanism.
A joint procurement instrument open to non-EU partners, referred to as MedPPA, could add resilience value where it goes beyond the existing EU Joint Procurement Agreement by securing binding supply commitments, bringing in partners that host critical upstream nodes, and establishing pre‑agreed allocation rules across participating jurisdictions. Its value would be greatest for medical countermeasures and essential medicines that are vulnerable to disruption and for which uncoordinated purchasing would risk intensifying scarcity. Because voluntary procurement mechanisms are vulnerable to defection when supply becomes scarce, MedPPA would require a dual-mode governance structure: consensus-based decision making under routine conditions, and delegated authority with binding exclusivity during declared crises. The mechanism would also need to internalise a resilience premium, potentially in the range of 5% to 20% above the lowest unit price depending on product category, to purchase capacity reservation, surge options and diversified manufacturing commitments.
A common stockpile open to third countries, referred to as RescPool, could provide the fastest operational response of the three proposed instruments to acute disruptions by ensuring physical continuity of supply during the days-to-weeks window before procurement or production responses take effect. Its main feasibility constraint is not warehousing capacity but legal deployability. For regulated medicines, stock held in one jurisdiction may be physically available but legally unusable elsewhere because of national rules on batch release, packaging, labelling, serialisation, patient information and dispensing. RescPool would therefore depend on a regulatory passport protocol allowing emergency stock to be legally dispensed across participating jurisdictions. It would also require pre‑agreed activation triggers, allocation rules and lifecycle management arrangements, including rotation and, where appropriate, keep-warm contracts. Access should be governed by a deductible principle, under which countries must demonstrate adequate national preparedness before drawing on the pooled reserve, to reduce moral hazard and avoid crowding out domestic buffers.
2.1.1. Identification of international co‑operation instruments
The OECD’s 2024 report on medical supply chains put forward a set of policy options both to enhance the security of medical supply chains in routine circumstances, and in preparation for the next crisis. Because of the globalised nature of medical supply chains, many of these policy options involve international co‑operation. In another report published in 2025, the OECD listed all existing international co‑operation initiatives (including those within the EU) aiming to enhance medical supply chains and proposed an assessment of these initiatives (OECD, 2025[1]).
This study aims to go further by considering, in addition to existing policies, international co‑operation instruments that are not yet used to support pharmaceutical supply chains but were considered in the initial proposal for a Critical Medicines Act (European Commission, 2026[2]) or otherwise suggested in public debates. The approach adopted to list all available instruments was to 1) define broad policy objectives; 2) list mechanisms with the potential to help achieve these objectives, and 3) make an inventory of international co‑operation instruments used to activate these mechanisms in the pharmaceutical sector or in other sectors (notably, but not only, the EU Critical Raw Materials Act – CRMA).
Three broad objectives have been defined to enhance security of medicines supply:
Increasing capacity and diversifying sources of supply (O1): tools that could be employed to expand or reinforce the capacity of medical supply manufacturing through resource mobilisation and long-term investment. These tools can vary in their specific objectives, such as diversifying and increasing manufacturing capacities for APIs or finished products, technology transfer and skill development, incentivising Foreign Direct Investments (FDI) to build manufacturing infrastructure, etc.
Ensuring continuous and seamless trade of medical supplies and inputs (O2): these tools aim to minimise or remove obstacles to the continuity of global supply chains in essential medicines. The tools include removing barriers to trade (tariffs, duties, quotas, etc), standardising regulatory requirements between countries, and ensuring high-quality standards of products and manufacturing processes.
Enhancing resilience to short- and medium-term shocks (O3): these tools aim to prepare for unanticipated perturbations in supply and demand in global supply chains. The tools include cross-border collaboration in procurement and stockpiling; information sharing to anticipate shortages and/or redirect supplies.
For each of them, several “mechanisms” have been identified, as well as international co‑operation instruments already in use or with the potential to be implemented. The full list of broad objectives, mechanisms and instruments is presented in Table 2.1.
The selection of instruments to evaluate was guided by the following considerations:
“Instruments” are policy measures involving partners from one, several or all EU Member States and partners from non-EU countries, aiming to achieve a specific set of objectives among the three defined above (O1, O2, O3);
The list includes instruments already used in the sector of medical goods, instruments envisaged by the initial EU Critical Medicines Act proposal, as well as instruments used by the European Union in other sectors, notably critical raw materials, to distil the first lessons we can draw from these, before their transposition to the pharmaceutical sector.
Note: The term “targeted trade deals” was borrowed from Dufour et al. (2025[3]) and is not used by the European Commission.
2.1.2. Evaluation framework
The evaluation framework is outlined in Figure 2.1. While the overarching “need” to be addressed by all policies is to enhance supply chain security for essential medicines and/or MCM, each instrument has been assessed against its capacity to reach specific objectives as defined earlier (O1, O2 and O3).
Source: Adapted from SGMAP (2015[4]), Évaluation des politiques publiques (EPP): principes, processus et méthode, https://www.eval.fr/wp-content/uploads/2018/09/epp_map_memo.pdf.
Each instrument is described and assessed according to the following template:
General objective/mechanism(s) to which the instrument refers.
Description: brief description of the instrument and its essential characteristics, such as the type of agreement/co‑operation under which it is implemented.
Current use: an overview of current utilisation in the EU context. Use in pharmaceutical-related initiatives was prioritised. However, where relevant, other areas where such tools have been implemented more extensively were also considered.
Expected Impact: a description of the anticipated impact. Criteria for impact assessment were defined by broad objective category. In cases where more precise criteria for assessment have been developed, these were described. While the capacity of each instrument to achieve its objectives determines its efficacy, the relationship between long term impact and the overarching need (i.e. enhancing security of supply chains) determines the utility of each instrument.
Risks: identification of potential unintended negative effects.
Resource use: as far as possible, financial and human resources needed for the implementation of the instruments were identified. The relation between short-term results and resource use determines the efficiency of each instrument.
Methods: Literature review, data analysis (where possible), and key stakeholder interviews were undertaken to inform the analysis of each instrument. While rigorous quantitative impact assessment is the gold standard, data were unavailable or insufficient in many cases, for several reasons. These include lack of methodologically sound formal or informal impact assessments; difficulty in attributing observed trends to specific instruments (e.g. in trade flows or foreign direct investment); the establishment of the instrument was too recent for effective impact assessment. In addition, authoritative information may be unavailable (unpublished or redacted) or sparse for some elements (e.g. resource use). Interviews were conducted to collect such information.
Limitations of the analysis: limitations are specific to each instrument and evaluation, but the applicability/suitability to the pharmaceutical sector of instruments mobilised in other sectors (e.g. food markets or critical materials) will be the subject of specific attention.
As many of the assessed tools are relatively recent, or have not been specifically applied to pharmaceuticals, definitive assessment of impact was not possible. In these cases, expected results were assessed through analysis of their implementation in other relevant areas (e.g. securing supply of essential raw materials), and augmented with information drawn from interviews with selected stakeholders. Data on resource utilisation were challenging to obtain from institutions involved in the implementation of certain instruments, either because it is considered confidential, or because it is not assessed or reported systematically.
Our ability to determine quantitative indicators for efficacy, efficiency and utility has been limited by the availability of data.
Internal and external consistency has been defined at the level of each broad objective.
Internal consistency refers to the consistency of the range of instruments envisaged to address the overall objectives (O1, O2, O3). Key questions include: do these instruments complement each other without duplication (which may indicate waste of resources) and without contradiction, i.e. do all instruments unambiguously address the same objective? Are there some gaps in instruments needed to reach a specific objective?
External consistency refers to the degree of alignment of a given instrument with policies, objectives, or commitments beyond its immediate scope or controlling institution. This could refer to broader external frameworks such as national strategies, international standards, Sustainable Development Goals (SDGs), or actions by other actors (governmental or non-governmental) in the same context.
The current framework does not address an important aspect of international co‑operation instruments, which is assessing the suitability of partner countries with which the EU should engage in different types of partnerships. On that matter, the Critical Medicines Alliance proposed a multi-criteria decision analysis framework to prioritise countries for partnerships (Critical Medicines Alliance, 2025[5]). A set of 18 criteria were identified and classified into four categories: production capacity (four criteria), ease of trade (six criteria), third-country policy (four criteria), and geographical/geopolitical factors (four criteria). Some of these criteria, for example those related to regulatory co‑operation or ease of trade, are directly linked to existing co‑operation instruments, which are considered in the proposed selected list of instruments to evaluate.
In 2024, the European Commission launched a consultative mechanism bringing together relevant stakeholders from EU Member States, key industries, the civil society, and the scientific community, with the aim of identifying priorities for action and proposing solutions to strengthen the supply of critical medicines in the EU.
In February 2025 the Alliance published its strategic report with key findings and recommendations to enhance the security and resilience of the EU’s critical medicines supply chains. The report directly fed into and informed the Critical Medicines Act proposal.
Source: European Commission (n.d.[6]), Critical Medicines Alliance – Public Health – European Commission.
2.1.3. Definition of criteria by objective category
Objective 1 – Increasing long-term capacity and diversifying sources of supply
International co‑operation instruments considered in this category and listed in Table 2.1 primarily aim to increase manufacturing capacity in different geographical areas, to prevent excessive concentration in the production of medicines or other medical products (and related inputs) sold in the EU market, and to further expand worldwide production where it is unlikely to meet increasing global demand. They encompass a range of instruments with the potential to incentivise foreign direct investment (FDI) in pharmaceutical manufacturing capacities, such as facilitation of permit granting, subsidies from the public sector or preferential bank loans. Expansion of manufacturing capacity is not only about building factories and includes knowledge transfer and skills training.
The initial set of criteria for assessment of instruments in this category are presented in Table 2.2. In many cases, measurement of selected indicators has proven impossible due to a lack of granularity of data available in information systems.
Objective 2 – Ensuring continuous and seamless trade in medical supplies and inputs
This category includes instruments aiming to facilitate trade in medical goods through regulatory harmonisation (O.2.1); reduction or removal of trade barriers and protectionism (O.2.2); and transport facilitation (O.2.3). They are listed in Table 2.1.
A first step was to assess (i) the extent and nature of tariff and non-tariff barriers to trade (NTB) and (ii) the existence and depth of co‑operation intended to reduce or eliminate such barriers (approached through mutual recognition agreements). Beyond tariffs, the broad category of non-tariff barriers (NTBs) includes regulatory requirements. The goal was to measure the level of alignment between EU and partner countries, assess the extent of existing NTBs in the pharmaceutical sector, and the potential of each tool to address them. NTB measurement has previously been explored by the OECD. Table 2.3 below displays the list of criteria for Objective 2.
Note: “MRI” stands for “Mutual recognition instruments”. The OECD launched a country survey of MRIs in 2025 to get an exhaustive list of MRIs in all sectors (i.e. not only medical goods). Results to be published in OECD (forthcoming[7]), Trade facilitation beyond the border: The evolving role of mutual recognition instruments.
Objective 3 – Managing supply / enhancing security of supply to absorb short- and medium-term shocks in demand
Three potential instruments have been identified to help supply chains absorb short- and medium-term supply shocks. The first, supply monitoring (O.3.1), would generate early-warning signals enabling co‑ordinated responses. The second and third – collective bulk purchasing open to third countries (O.3.2) and joint stockpile management open to third countries (O.3.3) – would address the consequences of those shocks directly, by securing supply volumes and bridging gaps until production can be ramped up. Extending both instruments to third-country partners is analytically significant: countries that participate in a joint purchasing or stockpiling arrangement forgo the option to compete unilaterally for scarce products, converting a potential race‑to-hoard dynamic into a co‑operative one. Table 2.4 presents assessment criteria for objective 3.
Note: AMIS: Agricultural Market Information System, EITI: Extractive Industries Transparency Initiative, FSB: Financial Stability Board, “AMC” stands for “Advance Market Commitment”; “MEAT” stands for “Most economically-advantageous tender”.
While bilateral, regional, and multilateral trade agreements have been growing in number and scope in the past decades (Mattoo, Rocha and Ruta, 2020[8]), often covering investment and complementing existing bilateral investment treaties (BITs), new forms of bilateral agreements with narrower scope, generally affecting a single sector, emerged in the 2000s, sometimes referred to as Targeted Trade Deals (TTDs) (Claussen, 2022[9]; Cernat, 2023[10]; Dufour et al., 2025[3]). Many of these targeted agreements are built on more comprehensive agreements with established partners. Others are more opportunistic and seek to remedy a specific issue (e.g. getting access to raw materials) through agreements with new partners.
Broad and targeted agreements typically invoke trade facilitation provisions to increase trade between partners and will be assessed against this objective in Section 2.3 of this report.
2.2.1. Impact of trade agreements and investments treaties on expansion of manufacturing capacity and diversification of supply
Free trade agreements (FTAs) and investment treaties (ITs) may have divergent impacts on medical supply chains. On the one hand, their main objective is to ease trade and facilitate investment, thus creating opportunities for supply chain diversification and new capacity to produce medicines and medical devices. When signed with like‑minded partners or alternative suppliers, these agreements can mitigate geopolitical risks and improve security of supply. On the other hand, such agreements may potentially have adverse impacts on the domestic supplier base when partner countries have a highly competitive pharmaceutical industry (for example in off-patent markets) and the trade agreement further facilitates market entry.
As of early 2026, the EU has agreements in place with more than 80 individual countries (European Commission, 2026[11]). A full and systematic review of the impact of FTAs and ITs signed between the EU and partner countries on the expansion of pharmaceutical manufacturing capacity lies beyond the scope of this study. Instead, this section draws on illustrative cases to examine potential impact of FTAs or ITs on medical supply chains.
Table 2.5 below presents a typology of provisions typically included in FTAs and ITs and elaborates on the potential impact of these provisions on security of supply chains in the EU. The impacts described are theoretical and direction-dependent: their materialisation depends on the specific provisions negotiated, the competitive characteristics of the partner country and broader market dynamics.
Note: TRIPS: Trade‑Related Aspects of Intellectual Property Rights.
Source: This table uses a typology of trade agreement provisions developed by Gleeson et al. (2019[12]), adapted to the objectives of this study and to the context of EU trade agreements and potential impacts on EU supply chains.
The theoretical impact described in Table 2.5 can be illustrated with two recent EU Free trade agreements. The Mercosur Agreement -signed in 2026 and not ratified yet- is not expected to affect pharmaceutical supply chains in the EU immediately, given the current imbalance in trade between the two regions. While nearly half of the finished products and more than 20% of APIs imported into Mercosur countries come from the EU, the respective shares of Mercosur APIs and finished product exports to the EU are 0.06% and 0.04% of the EU’s total volume of imports respectively (Cimini et al., 2024[13]). In the longer run, however, the ambitions of Mercosur countries to expand regional production of pharmaceuticals may offer opportunities for EU Member States to diversify their sources of supply (ibid.).
The trade agreement recently negotiated between the EU and India includes provisions for the removal of Indian tariffs on pharmaceuticals, of which the average value is currently 11% (EU tariffs on pharmaceutical products are already at or near zero under the 1994 WTO Agreement on Trade in Pharmaceutical Products). An ex-ante evaluation of the impact of the EU-India FTA on trade in pharmaceuticals estimated that imports from India could increase by EUR 2 836 to 6 158 M, while EU exports to India could increase by EUR 1 930 to 3 322 M according to a range of scenarios (Trade Impact B.V., 2023[14]). This translates to reduced output in the sector in the EU and growth in India. The report specifically addresses the impact of the FTA on supply chains security. It shows that the current EU concentration of supplier-countries1 for chemicals and pharmaceuticals is relatively low (by comparison to sectors such as minerals, fossil fuels, fish, or clothes) and will only be slightly impacted by the EU-India FTA by 2032, although import dependency will increase in the two sectors by 1‑2% (ibid.).
As indicated with these two cases, large free trade and investment agreements can potentially contribute to the diversification of supply, but this depends on the way increased market access plays in favour or not of increased capacity and trade for medical products. When considering scale economies, there is still a potential benefit for the country which is a net exporter of essential goods, as new destination markets can consolidate its competitive advantage. It should however not lead to further concentration of supply. Two-way trade with both partners expanding their manufacturing capacity (and exchanging varieties of products) is the best outcome from the point of view of diversification.
The rest of the section will focus on targeted trade deals between EU or EU Member States and third countries that explicitly aim to expand manufacturing capacity and diversify sources of supply. They would do so by incentivising outward or inward foreign direct investment in manufacturing capacity, either on top of existing FTAs or ITs, or with trade partners with whom such agreements have not been signed.
2.2.2. The potential of strategic partnerships
The Critical Medicines Act (CMA) proposal published in 2025 included provisions for the Commission to explore the possible inclusion of strategic partnerships to enhance the security of supply chains (European Commission, 2025[15]). The Critical Medicines Alliance (see Box 2.1) further developed the types of potential partnerships in its strategic report (Critical Medicines Alliance, 2025[5]), and recommended four different categories:
Leveraging existing partnerships with established trade partners, by typical trade facilitation measures (see O2) but also by promoting innovation in production processes, through skills development and voluntary technology transfer; promoting scalable manufacturing capacity expandable beyond domestic needs; and examining possibilities to co‑invest in strategic projects.
Leveraging existing partnerships with large producer and spare capacity countries, in which beyond trade facilitation and harmonisation of regulatory standards, partners would exchange information on manufacturing capabilities; and strengthen commitment to uphold and advance environmental and social standards, while preserving affordability of costs.
Intensifying and building new partnerships with neighbouring and strategically positioned countries. The goal of these will be to foster future pharmaceutical supply chains that will enhance resilience and diversification by collaborating with selected countries with moderate but potentially growing pharmaceutical production capacity and significant geopolitical importance to the EU. Partners could include accession countries or candidates to accession and be targeted by the New Growth Plan for the Western Balkans.
Intensifying and building new partnerships with capacity-development countries. These would typically seek to identify future opportunities to cultivate and expand pharmaceutical production capacity in low- and middle‑income countries (LMICs) in a mutually beneficial manner. These could build on current and previous EU initiatives, including actions under the Global Gateway and Team Europe to build local manufacturing capacity. Given the nature of this partnership, most objectives and associated actions will occur within a long-term timeframe.
The Report further suggests a list of criteria to select countries for different types of partnership according to 18 criteria, clustered in 4 broad categories: production capacity, ease of trade, third-country policy, and geographical/geopolitical factors.
The CMA text on which the EU Council and Parliament reached a provisional agreement in May 2026 confirmed the potential interest of this strategy (Council of the European Union, 2026[16]).
Ongoing initiatives to expand pharmaceutical production capacity abroad
As part of the EU Global Gateway initiative, the EU has launched several strategic, often region-wide, partnerships aimed explicitly at expanding pharmaceutical manufacturing capacity in partner countries (European Commission, 2023[17]). These initiatives, however, primarily aim to establish or expand local manufacturing, as well as foster an enabling environment to meet regional demand for pharmaceuticals, rather than to diversify EU supply chains and thus cannot be fairly assessed against this objective. They may nevertheless contribute in the longer term to the expansion and diversification of global supply chains, which are much needed in a context of increasing demand.
The Team Europe Initiative on Manufacturing and Access to Vaccines, Medicines and Health Technologies (“MAV+”) works with partner countries to strengthen their pharmaceutical system and encourage local production of vaccines in Africa. Funded by loans and other financial instruments from the European Development Financial Institutions (EDFIs), grants, budget support and blended finance, this initiative covers multiple aspects, such as contributing to the establishment of the African Medicines Agency (AMA), the WHO’s mRNA technology transfer hub, and the Partnership for African Vaccine Manufacturing (PAVM), hosted by the Africa Centres for Disease Control and Prevention (ACDC). In 2 024, a list of all projects supported by G7 countries – including EU members – to enhance pharmaceutical manufacturing capacities in Africa was presented at the meeting taking place in Italy (G7 Italia, 2024[18]).
The EU’s “Partnership on manufacturing vaccines, medicines and health technologies in Latin America and the Caribbean” is another strategic initiative under Global Gateway, explicitly aimed at boosting regional manufacturing capacity and resilience. The partnership was established in 2021 with the Community of Latin American and Caribbean States (CELAC) (European Commission, 2021[19]). In 2023 the initiative facilitated exchanges between pharmaceutical companies from both regions, with the end goal of attracting investment into the region’s sector (Cimini et al., 2024[13]).
What is the experience with strategic partnerships and strategic projects initiated to improve supply chains of critical raw materials?
To the best of our knowledge, no partnership has yet been established to diversify supply chains of critical medicines or MCM. This section therefore aims to look at EU’s experience with partnerships in other sectors. The EU Critical Raw Materials Act, adopted in 2024, for example, included provisions to enhance supply chains of raw materials through strategic partnerships and strategic projects, which makes it an interesting reference. This section does not seek to provide a thorough empirical assessment of the cost-effectiveness of these partnerships, but to extract some relevant insights from this experience.
According to the CRMA, companies with projects recognised as strategic by the EC, may benefit from advantages, such as expedited permit granting (in the EU), possible public funding from the EU or EU Member States, and preferential loans. In March 2025, the EC selected 47 EU projects for extraction, processing or recycling of CRM within EU (European Commission, 2025[20]). All but 3 of them are supported by companies with headquarters in EU Member States, with 2 originating from Canada and 1 from the United Kingdom. In June 2026, the EC selected a further 13 strategic projects outside the EU (European Commission, 2025[21]). The European Commission launched a second call and received 160 applications by early 2026 (European Commission, 2026[22]).
A recent WTO study looked at the design and impact of all Targeted Trade Deals2 for Critical Raw Materials, 80% of which were signed between 2022 and 2024. These TTDs take different forms, including formal agreements, exchanges of letters, joint statements, declarations, memoranda of understanding (MoUs), side letters and work plans (Dufour et al., 2025[3]). They aim to address specific regulatory barriers or supply chain vulnerabilities. Their negotiation and implementation are rapid. Compared to FTAs, they attract less scrutiny, oversight, and analysis, and they are not notified to the WTO. A mapping of 55 TTDs, often MOU or Memoranda of Co‑operation, reveals that “CRM TTDs seem to rely heavily on voluntary cooperation and non-committal language, often without formal monitoring structures or implementation frameworks.” Only 3 agreements included binding language and 90% did not include any dispute settlement provisions. These agreements often complement an existing Regional Trade Agreement between parties, and this is the case for about 47% of the CRM TTDs. The study demonstrated that CRM TTDs have a statistically significant impact on mineral trade flows between partner countries (ibid.), but this result is not specific to EU TTDs. Impact on FDI was not assessed.
Changes in flows of inward and outward Foreign Direct Investments (FDI) reported by the EU or EU Member States for mining and quarrying activities could provide an interesting indicator of the effectiveness of strategic partnerships and/strategic projects if data were available. However, OECD collects FDI by partner country and FDI by industry but does not collect a cross-classification of FDI by individual partner country, industry, and type of activity. The main reason is that such level of details would be subject to important confidentiality restrictions. The commercial fDi Markets database of the Financial Times includes detailed data on cross-border investment at the project level (Financial Times, 2026[23]).
In a paper published in 2024, Blot looked at all EU international relationships on CRM, from FTAs to strategic partnerships and strategic projects. For the latter, she listed a number of criticisms. She pointed out that strategic projects established outside EU, especially extractive projects, may not comply with human rights and environmental best practices. Criteria for Strategic Projects or Strategic Partnerships do not guarantee that projects and third countries’ regulatory frameworks are aligned with international agreements, including ILO conventions and recent Multilateral Environmental conventions (MEA). This is a concern in the mining sector, where companies’ performance on ESG indicators is uneven, especially on environmental responsibility and community well-being (Blot, 2024[24]).
In summary, the impact of strategic CRM partnerships could only be assessed through the evolution of trade flows between partners and access to commercial databases would be required to measure the impact on relevant FDI. Strategic partnerships and projects are much lighter to activate than broader trade and investment agreements. However, when they are not signed with established trade partners and backed by FTAs or investment treaties, manufacturing activities abroad may not comply with the highest standards of ESG.
2.3.1. Regulatory harmonisation
Regulatory harmonisation is expected to facilitate trade. In the sector of medical goods, which is highly regulated, harmonisation can be pursued through a range of international initiatives, from international discussion fora involving experts, regulators and sometimes industry, to formal agreements between authorised parties. This section summarises the state of play of regulatory harmonisation, first for pharmaceuticals and then for medical devices.
Pharmaceutical products
For pharmaceutical products, regulatory harmonisation is promoted in international fora, where regulators meet regularly to share expertise on evaluation standards and methods. Examples include the International Coalition of Medicines Regulatory Authorities (ICMRA, 2026[25]), the International Pharmaceutical Regulators Programme (IPRP, 2025[26]), or the OPEN framework European Medicines Agency (European Commission, 2026[27]). The International Council on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH, 2026[28]) has 25 members, and 41 observers among national and regional regulatory authorities, industry bodies, and international organisations. ICH works on the establishment of guidelines for good practice, which are then adopted and implemented in participating countries.
The Pharmaceutical Inspection Co‑operation Scheme (PIC/S) is a forum for close international co‑operation between about 60 pharmaceutical inspection authorities in the field of good manufacturing practice (GMP) and in which the EMA and national competent authorities are involved. The PIC/S’ mission is to lead the international development, implementation and maintenance of harmonised GMP standards and quality systems of inspectorates in the field of medicinal products. Members’ compliance with harmonised GMP inspection standards is assessed before they can join PIC/S (PIC/S, 2025[29]).
Mutual recognition agreements (MRAs) are formal instruments by which signatory authorities in one jurisdiction recognise the evaluation performed by another foreign authority and consider the result valid. This requires the highest level of standardisation of evaluation methods and criteria.
In the EU, regulatory agencies – either the EMA or national competent authorities (NCAs), depending on product characteristics – assess the safety, quality and efficacy of all pharmaceutical products before granting marketing authorisation. The EU has signed MRAs with several foreign countries, namely Australia, Canada, Israel, New Zealand, the United States and Switzerland. These agreements, initially signed at the end of the 1990s or early 2000s, and often amended since then, cover a wide range of pharmaceutical products, with varying exceptions, depending on the date of signature and difference in countries regulations. Many MRAs exclude Advanced Therapy Medicinal Products (ATMPs) from their scope and some of them exclude investigational products, medicines derived from human blood or blood plasma (e.g. Israel, Japan) or API (New Zealand, Canada). MRAs generally cover recognition of Good Manufacturing Practice and batch certification, which means that regulatory agencies recognise and accept assessments and inspections performed by the counterparty, and may also agree to share confidential information on manufacturers, products, shortages, to enhance safety in partnering countries.
These agreements do not aim to increase trade between partners immediately, although they allow inspections to be expedited that might otherwise delay completion of approval processes and generate costs. Trends in the volumes of inspections conducted by countries could help evaluate the impact of MRAs on the total number of inspections performed and the proportion undertaken by each participating agency. According to EU officials, the adoption of the EU-US MRA led to a 70% reduction in EU inspections in the United States. This was confirmed by an EFPIA survey of its research-based company members, likely to represent a large proportion of pharmaceutical manufacturing sites in the United States (EFPIA, 2025[30]). The survey also showed the benefits of inspections conducted by PIC/S members. MRAs for pharmaceuticals, however, do not directly facilitate trade between countries since each product must be assessed for marketing authorisation in each jurisdiction.
MRAs take time to implement as parties to the agreements generally want to assess each other’s capacity to conduct inspection and compliance controls before signing the agreement. The EU has already signed MRAs with its main trading partners and PIC/S membership may work as an alternative to reduce the time and costs of GMP certification and inspection.
Another avenue to facilitate trade in pharmaceuticals across countries is what is referred to as regulatory reliance. This is a process by which a national regulatory authority (NRA) takes into account and gives significant weight to evaluation reports and decisions made by another trusted authority when making its own regulatory decisions for medical products. It streamlines approvals, reduces duplication, and optimises resources while the relying authority retains full independence, accountability, and responsibility for its final decision. Reliance can be unilateral (one‑way), mutual (two‑way), or involve work-sharing, where multiple authorities share the assessment workload.
Unlike recognition, where an NRA fully accepts or adopts the decision of another NRA without further review, reliance uses the external information as a significant input, but the agency still conducts its own assessment. Recognition has not been considered by the EU. Apart from raising issues of autonomy, accountability and sovereignty, benefit/risk assessments that underpin marketing authorisation decisions may vary for a number of reasons, including differences in the burden of disease or ethnic differences in responses to treatment, even when based on the same body of evidence.
No information has been found on the amount of resources (budget and staff) involved in the preparation of MRAs or in international co‑operation initiatives such as PIC/S. The preparation of MRAs for pharmaceuticals is likely to involve trade specialists and negotiators (EC DG Trade), as well as EC DG Sante, experts from EMA and the national competent authorities. Participation in the PIC/S initiative entails membership annual fees, as well as mission and staff time costs of attending meetings and seminars, as well as joint visits (4 to 15 days per year) (PIC/S, 2025[31]).
According to the EFPIA survey, in 2024, each on-site inspection for their affiliate mobilised on average 7.7 to 10.3 inspector-days (domestic vs. foreign inspection, without travel time) and EU dedicated a little less than 200 inspector-days for all on-site foreign inspections (EFPIA, 2025[30]). The total number of GMP inspections (for all companies) and corresponding costs are not readily available. It is worth noting, however, that competent authorities lose the fees charged for GMP inspections when they do fewer inspections.
Medical devices
To be placed on the EU market, medical devices must undergo conformity assessment and obtain CE‑Marking from an authorised Notified Body. In some countries, this conformity assessment is undertaken by the medicine regulator.
The International Medical Device Regulators Forum (IMDRF) is a voluntary group of medical device regulators from around the world who have come together to work on harmonisation of standards, with the aim to accelerate international medical device regulatory harmonisation and convergence. The European Commission is member of this forum (IMDRF, 2026[32]).
The EU has signed MRAs with several countries to facilitate trade in medical devices. These MRAs often cover multiple sectors subject to conformity assessment, such as machinery or telecommunication equipment. The object of recognition is the certified good itself, which means that a medical device certified abroad can be imported in the EU market and vice‑versa. This potentially facilitates trade between jurisdictions. In practice, however, these MRAs have seen limited operational use, according to stakeholder interviews, including with DG SANTE. Assessing the effectiveness of these MRAs on trade in medical devices would require information on the number of MDs placed on the EU market while certified abroad, as well as the measurement of trade flows before and after their implementation.
Assessing the cost of negotiating MRAs on medical devices would also be challenging since these agreements most often cover a wide range of sectors. Trade negotiations might be long and resource intensive and negotiators will seek inputs from qualified experts from competent authorities to define the details of MRAs on medical devices. Despite the thousands of products potentially affected by these agreements, however, expertise is likely to rely on a very small team within the European Commission. Participation in international co‑operation initiatives such as PIC/S entails membership contributions.
2.3.2. Reduction or removal of trade barriers through trade agreements
The reduction or removal of tariffs and non-tariff measures (NTMs) on exchanges of medical goods and services aims to facilitate the movement of goods.
On the tariff side, the 1994 Agreement on Trade in Pharmaceutical Products and subsequent updates removed tariffs on finished pharmaceutical products, as well as about 7 000 APIs and other inputs for exchanges between countries participating in this agreement, i.e. Canada, the European Union, Japan, Macao (China), Norway, Switzerland, the United Kingdom and the United States (World Trade Organization, n.d.[33]). According to WTO “Most favoured nation” rule, these countries are not expected to apply tariffs on imports of products targeted by this agreement to their trade partners, even those which did not sign it. This means that EU Member States cannot do much more to reduce tariffs on these products.
Tariffs may remain on some API not included in the scope of the 1994 agreement, on key starting materials and on medical devices. These cases could be investigated to identify further actions, especially on inputs to the manufacture of critical medicines and medical devices.
On the other hand, recent geopolitical tensions have also created a high level of uncertainty regarding the imposition of tariffs and quotas.
Non-tariff measures (NTMs) may also act as barriers to trade, and they did so during the COVID‑19 pandemic, mainly through export restrictions. Divergences in intellectual property rights and regulatory requirements for the commercialisation of medical products are other examples of NTMs. Countries with the highest standards, including the EU, however, are unlikely to lower these standards, except temporarily in emergency and crisis situations, where the risk for public health is very high (Lee and Prabhakar, 2021[34]).
2.3.3. Trade facilitation
Trade facilitation policies for goods aim to make trade and customs laws, regulations, documents and procedures simpler and more efficient. According to OECD’s monitoring of trade facilitation indicators (TFIs), European countries are among the best performers in terms of trade facilitation (OECD, 2025[35]). One avenue for improvement, however, resides in improving co‑ordination, automation and interoperability of border control of increasing sustainability regulatory requirements. The 2025 OECD report on supply resilience also recommends building on the COVID‑19 inter-agency co‑operation structures to enhance crisis responsiveness and resilience, as well as international co‑operation more broadly (OECD, 2025[36]).
Three potential instruments, involving third-party countries, were evaluated under Objective 3 – supply-chain monitoring (O.3.1, “MedMIS”), joint procurement (O.3.2, “MedPPA”), and pooled reserves (O.3.3, “RescPool”). Each would address a distinct functional gap in pharmaceutical supply chain resilience. MedMIS would provide diagnostic capacity by monitoring supply conditions and classifying disruption drivers; RescPool would ensure short-term continuity through pre‑positioned reserves; and MedPPA would enable co‑ordinated procurement to restore sustainable supply. While each instrument would have standalone value, the three could be mutually reinforcing if connected through pre‑defined governance linkages.
The three instruments evaluated under Objective 3 would operate at the intersection of two distinct policy domains: crisis preparedness for medical countermeasures (MCMs), which falls primarily within the mandate of HERA and the EU Emergency Framework, and structural security of supply for critical medicines on the Union list, which is addressed by the proposed pharmaceutical legislation and Critical Medicines Act. This chapter evaluates each instrument for its potential contribution to both objectives, since pharmaceutical supply disruptions do not respect the administrative boundary between crisis and routine contexts. Product scope is defined instrument-by-instrument: MedMIS draws its pilot basket from the Union list filtered for supply-chain vulnerability; MedPPA’s product scope spans both the Union list and the HERA MCM list; and RescPool targets the Union list filtered for “bridge value” and stockpiling feasibility. Where existing or proposed EU instruments address similar functions – notably the European Shortages Monitoring Platform (ESMP) for shortage monitoring, the Joint Procurement Agreement (JPA) for joint procurement, and rescEU for stockpiling – the evaluations explicitly positionned the proposed instruments as complements.
2.5.1. Objective, decision context and scope
This section assesses the feasibility and potential operational value of an international market information and co‑ordination mechanism for critical medicines and medical countermeasures – hereinafter “MedMIS”. The model draws on the governance logic of the G20 Agricultural Market Information System (AMIS), adapted to the specific constraints of pharmaceutical and medical countermeasure markets. Although MedMIS is assessed in this report as a potential EU initiative, reflecting the scope of the study, its design would be compatible with a wider international configuration, should participating countries choose to extend it beyond the EU framework.
Short- and medium-term disruptions in the supply of critical medicines present a recurring governance challenge. Decisions on procurement, stockpile deployment, regulatory flexibility, demand management or trade restrictions often need to be taken under significant uncertainty about upstream manufacturing capacity, inventory levels and true demand. In this setting, the operational question is not simply whether scarcity exists, but what type of scarcity is emerging and what constitutes a proportionate response.
MedMIS is therefore evaluated as a crisis-operability tool designed to support this diagnostic function. Its purpose would be to improve situational awareness, validate signals and support co‑ordination when disruptions emerge. Specifically, the assessment considers whether a MedMIS architecture could:
Reduce information asymmetry by producing a trusted, decision-relevant picture of evolving supply and demand conditions (i.e. a “single version of the truth” for participants) (Zhao et al., 2023[37])
Detect emerging constraints earlier through structured monitoring of disruption events and market stress indicators;
Support faster and more consistent crisis decisions by providing an institutionalised forum to validate alerts, align assumptions and co‑ordinate policy intentions; and
Mitigate destabilising feedback loops by improving confidence in shared assessments and providing a platform for timely peer co‑ordination (OECD, 2024[38]).
MedMIS is primarily assessed as a tool for addressing agency-related failures, including incomplete information, misaligned incentives and uncoordinated responses, rather than binding physical availability constraints. Its core value would lie in enabling participants to i) classify disruptions as availability-driven, agency-driven or mixed early in their evolution (see Box 2.2); ii) match mitigation strategies to the dominant driver, avoiding disproportionate or misdirected responses; iii) reduce “phantom demand” by replacing uncertainty with credible, validated information; and iv) dampen destabilising spillovers through co‑ordinated policy signalling.
Scope boundaries
The scope of this feasibility assessment is deliberately bounded along four dimensions:
Product scope: the analysis assumes a narrowly defined “critical basket” aligned with public-health emergency relevance and high supply-risk characteristics (clinical criticality, limited substitutes, and supply-chain vulnerability).
Temporal scope: MedMIS is assessed for short- and medium-term shocks, where long-lead-time measures (capacity expansion, reshoring, major diversification) cannot prevent acute scarcity. The assessment recognises complementarities with longer-term resilience measures but treats them as outside the primary functional remit of Instrument O.3.1.
Functional scope: MedMIS is evaluated as an information-sharing and co‑ordination mechanism. It does not presume authority to mandate allocation, compel stock release or override national regulatory decisions. Its value is assessed through its ability to improve the timing and quality of decisions by strengthening the evidence base and enabling structured co‑ordination.
Geographic and governance scope: because the most consequential information gaps arise upstream and outside the EU’s jurisdiction, MedMIS is assessed explicitly as an international co‑operation instrument involving EU institutions and Member States plus partner countries and (critically) private data holders.
Distinction from existing EU shortage monitoring
MedMIS should be distinguished from the European Medicines Agency’s European Shortages Monitoring Platform (ESMP), established under Regulation (EU) 2022/123. The ESMP is a regulatory monitoring tool designed to collect and process shortage notifications from marketing-authorisation holders within the EU, supporting the EMA’s Executive Steering Group on Shortages (MSSG) in managing regulatory responses. MedMIS would differ in three respects. First, it would integrate both supply-side and demand-side signals – including procurement behaviour, inventory proxies, and upstream input indicators – rather than relying on regulatory shortage notifications alone. Second, it would operate as an international mechanism extending beyond EU borders to engage major production hubs and third-country partners, addressing the upstream visibility gap that EU-internal systems cannot close. Third, it would include a structured escalation protocol – a Rapid Response Forum convened by a designated intergovernmental body – enabling co‑ordinated policy dialogue during crises, whereas the ESMP supports regulatory crisis management within the EU through the MSSG. These instruments would be complementary: the ESMP would serve as one data input into MedMIS’s broader situational picture.
It should also be noted that an EU-level IT platform for stock monitoring was considered during preparation of the CMA proposal but was not retained, for reasons set out in the CMA Staff Working Document published in 2025 (Section 5.7). MedMIS as evaluated here differs from the discarded option in geographic scope (international rather than EU-internal), data focus (upstream production capacity and disruption drivers rather than downstream stock levels), and institutional hosting (intergovernmental body rather than Commission platform).
Actors and institutional setting
A functional MedMIS would depend on the engagement of four key actor groups, each contributing essential capabilities and incentives:
EU institutions (notably the Commission and EMA) could serve as conveners and system integrators. Their roles would include ensuring interoperability with existing EU-level monitoring and crisis-response mechanisms, supporting the development of common escalation protocols, and anchoring the governance framework;
EU Member States would play a dual role: they provide downstream signals (e.g. observed or announced shortages, procurement activity, demand surges) and act through designated focal points in the processes of data validation, verification and co‑ordinated escalation;
Third-country authorities, especially those in major production hubs (e.g. India, China) and large consumer markets, would be critical for upstream visibility. Their participation would be essential to avoid a regionally confined system that functions as a lagging indicator, reacting only after global shocks are already in motion;
Industry and private data holders, including manufacturers, contract suppliers, wholesalers and logistics operators, possess the granular operational data that enables credible early warning. This includes information on facility-level disruptions, production forecasts, inventory levels and sub-tier constraints. Without their engagement, MedMIS would not function as a reliable situational awareness mechanism.
Given these dependencies, MedMIS’ feasibility is evaluated not only in technical terms (data pipelines and indicator design), but also in institutional terms – whether credible mandates, safe harbour protections, participation incentives and information governance mechanisms can be established at a level sufficient for routine operation and crisis-time credibility.
The design and evaluation of MedMIS rest on a distinction between two modes of scarcity that demand different policy responses. While real-world disruptions often combine elements of both, the analytical separation is essential because instruments that resolve one mode can be ineffective when applied to the other:
Availability-driven crises arise when the physical capacity to produce or deliver a medicine is genuinely insufficient to meet clinical demand. Causes include destruction or shutdown of manufacturing facilities (fire, contamination), structural loss of production capacity (market exit by the dominant manufacturer), upstream input failure (API or key starting material shortages that cannot be substituted within the relevant timeframe), and biological constraints (yield failures in biologics). In these crises, the supply shortfall is real: no amount of improved co‑ordination or information-sharing will produce doses that do not exist. The appropriate response space lies in industrial policy, capacity expansion, alternative supplier qualification and, in the short term, rationing and clinical substitution. Monitoring can help diagnose and communicate the constraint but cannot resolve it.
Agency-driven crises arise when sufficient supply exists somewhere in the system but fails to reach patients due to co‑ordination failures, behavioural responses to uncertainty or institutional friction. Characteristic drivers include precautionary over-ordering by buyers uncertain about future availability (“phantom demand”), fragmented or uncoordinated national procurement that creates artificial competition for the same limited pipeline; export restrictions or defensive stockpiling by producing countries; regulatory barriers that prevent redistribution of available stock across jurisdictions (e.g. labelling, or marketing-authorisation constraints); and information asymmetry that prevents decision makers from distinguishing manageable disruption from genuine structural failure. In these crises, the aggregate supply-demand balance may be adequate, but misallocation, hoarding or friction converts a manageable disruption into a systemic shortage. The appropriate response is co‑ordination: validated information, aligned procurement behaviour, regulatory flexibility and structured reallocation.
The core value of a monitoring mechanism lies in its ability to distinguish between these two modes early in a disruption, before defensive behaviours cascade.
2.5.2. Evidence and lessons from the Agricultural Market Information System (AMIS)
Operational logic: Market monitoring as a governance cycle
The Agricultural Market Information System (AMIS), launched by the G20 in 2011 in direct response to the 2007‑2008 and 2010‑2011 food-price spikes, is best understood not as a data repository but as an institutionalised governance cycle. Those crises exposed serious gaps in reliable, timely information on crop supply, demand, stocks and export availability, and demonstrated how such gaps contributed to panic buying and destabilising policy reactions such as export bans. AMIS’s core function is to aggregate fragmented market data from participating countries, validate it against independent sources and, when market conditions deteriorate, provide a structured forum for governments to co‑ordinate policy responses (Drechsler, 2023[39]). This operational logic can be visualised as a governance cycle linking data collection, validation, public outputs and crisis activation (Annex Figure 2.B.1), the components of which are described in the paragraphs that follow.
AMIS focusses on four globally traded crops – wheat, maise, rice and soybeans – and operates through two interconnected mechanisms: i) a routine technical cycle encompassing data collection, triangulation, validation and publication, and ii) what can be characterised as an escalation protocol that enables rapid policy dialogue when acute market volatility emerges. Participants comprise the G20 countries, Spain and seven additional major producing, consuming and trading countries. Together, these participants account for over 80% of global production and trade in the four target crops (AMIS, 2017[40]).
AMIS operates through three institutional bodies: a Secretariat hosted at the FAO producing market outlooks and managing data quality; a Global Food Market Information Group gathering technical focal points from each participant; and a Rapid Response Forum (RRF) convening senior officials when abnormal market conditions warrant enhanced dialogue. A key design feature directly relevant to MedMIS is differentiated information access: national submissions are stored in a restricted database visible only to focal points and the Secretariat, while public products are openly accessible (AMIS, 2016[41]). The detailed AMIS operational architecture, including data collection processes, the publication cycle, and the escalation protocol, is described in Annex 2.A (Section B.1).
Evidence of impact: Co‑ordination effects in food markets
Quantifying AMIS’s causal impact on prices or volatility is methodologically challenging, and the available evidence is largely qualitative. Official FAO and AMIS reporting suggests several channels through which the mechanism may have helped stabilise expectations and dampen over-reactions after 2011. The FAO Director-General reported to a 2012 Ministerial Meeting on Food Price Volatility that AMIS was “functioning well” and had contributed to better international co‑ordination, information sharing and greater market transparency, helping the international community to react quickly to the July 2012 price rise and to avoid panic and unilateral measures (FAO, 2011[42]). The AMIS five‑year progress report likewise argues that enhanced transparency and policy dialogue can reduce the likelihood of sudden price spikes and panic-driven policies, even if structural shocks to supply cannot be prevented (AMIS, 2016[41]).
Analytically, AMIS impact is most pronounced in what might be characterised as co‑ordination-failure or “liquidity” crises – situations where supply exists, but information gaps or uncertainty drive volatility, panic buying and restrictive trade policies that further destabilise markets. Post-event analysis is consistent with the view that AMIS-supported co‑ordination contributed to faster policy convergence than during the 2007‑2008 food-price crisis (AMIS, 2016[41]). Separately, OECD analysis finds that during COVID‑19, 28% of export restrictions on food products were lifted within one month compared with 2007‑2008, when roughly one‑quarter of restrictions remained in place for over a year (OECD, 2024[43]). While isolating AMIS specific contribution from other factors is not possible with available evidence, the pattern is consistent with the mechanism’s transparency and co‑ordination objectives.
At the same time, independent academic work notes that persistent uncertainty around key fundamentals – notably stocks data – has limited the extent to which AMIS has fully resolved information gaps (Brockhaus and Kalkuhl, 2014[44]). Some countries remain reluctant to share sensitive data, especially on reserves; AMIS itself acknowledges that stocks data remain problematic, and has invested in expert consultations, methodological work and country-specific support to improve measurement (AMIS, 2016[41]).
The more fundamental insight from AMIS for a MedMIS design lies in its temporal contribution: its value consists not merely in generating better data, but in accelerating the system’s convergence toward a shared understanding of market conditions. For a prospective MedMIS, this suggests that a key metric of success would be whether it reduces the decision lag between the emergence of a disruption signal and co‑ordinated response – though it should be noted that no empirical study has yet quantified this lag reduction for AMIS itself. Even when structural supply constraints persist, faster validation of disruptions and earlier co‑ordination can improve system agility, reduce overreactions, and limit harmful spillovers.
Co‑operation dynamics: Participation incentives and lessons for MedMIS design
Despite the absence of binding legal obligations, AMIS has sustained participation through several interlocking incentive mechanisms. Understanding these dynamics is essential for MedMIS design, where co‑operation must be secured under conditions of greater legal asymmetry, commercial sensitivity and regulatory fragmentation:
For major agricultural exporters (e.g. Argentina, Russia, Ukraine), participation addresses the risk that credible information gaps amplify precautionary buying and policy-induced scarcity beyond what supply fundamentals would imply. The 2010 Russian drought and wildfires illustrate this dynamic: initial uncertainty about the scale and timing of production losses contributed to export bans that heightened ambiguity about global availability, in turn fuelling price spikes in import-dependent markets. By enabling faster validation of supply conditions and contributing to a shared narrative about availability, AMIS can dampen these self-reinforcing dynamics and lower the risk of policy-induced scarcity (FAO, 2012[45]).
For import-dependent countries, AMIS reduces the information asymmetries that drive panic buying and over-ordering. When credible assessments of global supply are absent, importers default to worst-case assumptions. AMIS’s Market Monitor provides an authoritative, multi-source assessment designed to anchor expectations. In crisis settings, the RRF provides a structured forum where senior officials can share information about policy intentions, providing political cover for more measured national responses and helping to avoid beggar-thy-neighbour spirals. The value lies not in legal enforcement but in co‑ordination benefits: restraint is framed as evidence‑based and reciprocated.
AMIS’s early design envisaged that data submissions would be publicly assessed as “Good”, “Fair” or “Poor” on criteria of timeliness, completeness, coherence and comparability. The Secretariat now systematically classifies participants into performance groups (green, yellow, orange, red) based on data-submission quality for internal monitoring and capacity-building purposes. While participation is not organised as a formal “give‑to-get” conditionality – public AMIS products are openly accessible regardless of reporting performance – countries that provide complete and timely data have their national balances more fully integrated into AMIS’s global outlooks and are consequently better positioned to shape the shared assessment. This creates reputational pressure: persistent late or incomplete reporting reduces a country’s influence in the collective narrative and may weaken its credibility with peers.
These co‑operation dynamics yield three core design requirements for a credible and functional MedMIS:
1. Routine validation cycles: align monitoring with regulatory and production rhythms (inspection cycles, batch-release schedules, tender calendars) rather than operating as an episodic “alert-mode” tool.
2. Independent “triangulation” capacity: enables the Secretariat to challenge submissions using independent signals (customs/trade flows, commercial datasets), limiting strategic reporting and anchoring trust.
3. Sequenced crisis protocols: replicate AMIS’s graduated protocol so that crisis meetings focus on policy co‑ordination rather than disputing baseline facts.
Critical caveats: Limits of the AMIS analogy
While the governance logic underpinning AMIS offers valuable design lessons, the structural characteristics of pharmaceutical markets differ from those of agricultural commodities in ways that impose key constraints on instrument design:
Regulatory non-fungibility: In agricultural markets, commodities are fungible once quality thresholds are met: a tonne of wheat is functionally equivalent across origins. Pharmaceutical supply is structurally fragmented by regulatory geography: what appears as “global stock” is, in practice, a set of partitioned inventories differentiated by manufacturing site, dosage form, pack size, labelling and certification. A monitoring system that tracks only aggregate volumes risks signalling abundance while specific markets face scarcity. MedMIS would therefore be designed to monitor regulatorily compatible availability, not global stocks in the abstract.
Lack of observability: AMIS benefits from partial external verifiability: crop conditions can be triangulated through satellite imagery, and some public reporting on trade flows and reserves exists. Pharmaceutical production occurs within closed facilities, and inventory is diffusely held across manufacturers, wholesalers and hospitals with limited real-time visibility. MedMIS would need to compensate through administrative triangulation – cross-referencing shortage notifications against tender outcomes, customs data, batch-release timelines and wholesaler allocation patterns.
Lastly, information sensitivity: In commodity markets, much relevant information (harvest forecasts, export data) is publicly aggregated. In pharmaceuticals, the most decision-useful information is granular (facility-level capacity, batch status), commercially sensitive and legally constrained, particularly in concentrated market segments where disclosure may raise competition-law concerns (Årdal et al., 2021[46]). The key feasibility constraint is therefore not merely political willingness, but corporate legal exposure.
AMIS itself recognises confidentiality concerns, particularly regarding stock data, and has developed mechanisms to protect commercial confidentiality while encouraging co‑operation. These constraints define the adaptation frontier: a credible MedMIS should be built around regulatorily compatible availability as the monitored object, robust triangulation mechanisms to close the verification gap, and bounded-disclosure architectures that make participation legally safe without stripping decision utility.
2.5.3. Structural feasibility of an AMIS-like system in pharmaceuticals
Shock characteristics: Agricultural losses vis-à-vis industrial bottlenecks and quality shutdowns
AMIS was designed for agricultural markets, where disruptions typically stem from geographically localised natural events, such as droughts, floods, or pest outbreaks. These shocks translate directly into reduced yields and observable inventory depletion. While forecasting remains uncertain, the causal chain is physically transparent: weather affects crops, harvests decline and global supply contracts.
Pharmaceutical disruptions follow a different logic. The binding constraint is rarely the destruction of productive capacity, but rather throughput bottlenecks at a small number of highly specialised facilities. These disruptions typically stem from equipment failures, input shortages, or capacity reallocations, propagating non-linearly across multi-tiered networks (European Commission et al., 2021[47]). Crucially, the most consequential failures are quality-driven: good manufacturing practice (GMP) deviations, contamination incidents or failed inspections can trigger immediate batch rejections, facility shutdowns and extended remediation periods.
In this context, “production” is inseparable from regulatory status: a facility may be technically capable of producing a medicine, but if it is non-compliant, its effective output is zero. Consequently, batch release dynamics such as testing, certification and qualified person (QP)3 approval are as decisive for availability as physical manufacturing itself. For MedMIS, this implies that monitoring could not rely solely on market-balance logic (production → utilisation → stocks) as a proxy for stress. Instead, it would need to treat regulatory quality events as first-order determinants of effective supply. To do so, the system should monitor not just output volumes, but:
The operational status of capacity (planned vs. disrupted);
The timing and outcome of batch release processes; and
The lag structure linking upstream events to downstream availability.
In practical terms, this means the analytical core of MedMIS would need to align closely with operational risk surveillance.
Information governance implications
The caveats outlined in the previous section make clear that the feasibility of MedMIS hinges less on analytical design than on the institutional conditions under which sensitive private sector data can be shared, validated and circulated without creating legal or commercial risk. To convert legal permissibility into sustained, high-quality participation, MedMIS would require a credible governance architecture centred around five key elements: scope, access, safeguards, auditability and incentives:
Scope: MedMIS could launch with a narrow product basket selected using risk-based criteria – e.g. products with i) high clinical criticality, ii) cross-border interdependence and iii) low clinical substitutability. This basket could be drawn from the Union list of critical medicines, selected for their supply chain vulnerability (see Article 131 of the proposed Regulation forming part of the pharmaceutical package (Council of the European Union, 2026[16])). A set of indicators to be collected from industry, including manufacturing capacity and supply projections, could be defined, ocusing on a small number of variables to minimise reporting burden and avoid diluting analytical value.
Tiered access and controlled circulation: a two‑tier access model would represent an essential mechanism for credible, bounded disclosure. The system could distinguish between restricted operational intelligence accessible only to designated public authorities under stringent confidentiality protocols (Tier A) and curated public outputs (time‑lagged capacity indices, aggregate risk levels) that inform policy decisions without enabling harmful behaviour (Tier B).
Verification and auditability: because there is no “satellite imaging” analogue for pharmaceuticals, MedMIS would need to incorporate a triangulation and challenge function within the Secretariat. At minimum, the Secretariat should be able to cross-reference member submissions with administrative and transactional signals (tender failures, order fill rates, customs/trade flows, batch release timelines, wholesaler allocation patterns, and regulator-held signals where lawful) and to run a structured “query-and-resolution” process for anomalies,
Legal safety and purpose limitation: to mitigate participation risks, the framework should incorporate: i) minimum-necessary data collection principles, ii) appropriate aggregation thresholds and time lags, iii) strict purpose limitation and access controls, iv) explicit constraints on onward sharing and data retention, and v) procedural firewalls preventing unilateral repurposing for enforcement or litigation. Where warranted, independent data trustees or secure environments can provide additional protection.
Incentives: MedMIS should offer concrete benefits, including priority access to early warning intelligence, member-grade analytical tools, and, where feasible, regulatory reciprocity arrangements. Reporting must represent a rational exchange for both public authorities and industry participants.
Absent this enabling architecture, a pharmaceutical-focussed monitoring system would risk devolving into either a low-value repository of outdated public indicators or a high-friction reporting regime characterised by selective engagement, either outcome undermining the credibility and responsiveness essential to its mandate.
2.5.4. Operational feasibility: Minimum viable dataset and architecture
Routine monitoring as a feasibility condition: Baselines, calibration and trust
An alert-only MedMIS is unlikely to be decision-useful: crisis periods are precisely when incentives to delay, withhold or strategically frame information are strongest, and when authorities have the least time to validate competing narratives. Without established baselines and validation workflows, an alerts-only mechanism risks propagating unverified claims or consuming scarce time in factual disputes.
While WHO and regional authorities have advanced harmonised definitions and shortage‑notification frameworks, existing systems remain predominantly downstream and event-driven, with limited visibility into upstream production constraints and significant cross-country heterogeneity in reporting thresholds and data granularity (OECD, 2024[38]; WHO, 2018[48]). Moreover, broad public disclosure of granular shortage signals can trigger destabilising market reactions, including defensive ordering and price escalation. Effective crisis-time co‑ordination therefore needs to sit atop an established routine monitoring architecture. Routine operation would be required for three reasons:
1. Standardisation: It aligns definitions, units, and reporting conventions, ensuring cross-country and cross-firm signals are comparable.
2. Calibration: It establishes what “normal variance” looks like (e.g. seasonal fluctuations or typical backorder rates), allowing deviations to be interpreted as meaningful signals rather than noise.
3. Trust: It builds credibility through repetition. A mechanism that routinely reconciles partial inputs into a consistent baseline, without triggering market disturbance, accumulates the trust required for high-stakes crisis co‑ordination.
Ultimately, “routine mode” determines escalation discipline. Pharmaceutical supply chains exhibit endemic operational noise; without longitudinal baselines and pre‑defined thresholds, alerts risk becoming high-frequency false positives. This erodes confidence and can itself induce destabilising responses (defensive procurement, stockpiling and policy escalation). A minimum viable MedMIS should therefore be designed to be sufficiently light to sustain participation, yet robust enough to generate the longitudinal baselines needed to make outputs interpretable, auditable and usable at speed. The data architecture should also minimise raw-data movement: a federated approach, where analytics run at source and only aggregated outputs flow to the Secretariat, would reduce both the volume of commercially sensitive data in transit and the infrastructure burden on participants.
Building a minimum viable indicator set: From measurement to key signals
A functional MedMIS must prioritise decision-useful intelligence over exhaustive data collection. The minimum viable dataset should answer three core operational questions:
what is happening to effective supply?
what is happening to demand signals?
where do the binding constraints sit along the supply chain?
Given the complexity and frequent commercial sensitivity of supply chains, direct measurement of supply is often infeasible. Consequently, MedMIS should rely on triangulated proxies rather than comprehensive stock accounting. As no single indicator is sufficient for decision making, the system should track a selected bundle of inventory proxies and service‑level indicators – such as backorder rates, order fill rates, and the activation of allocation regimes. When interpreted over time, these metrics can signal risk of allocation stress.
Demand monitoring should likewise be tailored to the specific market structure. In hospital-dominated and centrally procured markets, procurement indicators (e.g. tender volumes, participation and failure rates) could serve as leading signals. In retail-dominated markets, consumption and dispensing indicators could be incorporated; however, as these are often lagged and heterogeneous, they are more effective for calibration and confirmation than for real-time early warning. The design objective would be to combine complementary demand signals rather than privileging a single metric.
Before operational indicators can generate decision-grade signals, two structural preconditions must be secured. First, a stable critical basket reference list is required to ensure like‑for-like comparability across jurisdictions (INN/form/strength) with a defined baseline period. Without this, apparent volatility may reflect definitional drift rather than genuine disruption. Second, products must be coded with a market/regulatory segmentation flag, distinguishing theoretical global supply from legally usable supply in a given market (e.g. marketing authorisation/site approvals, pack/labelling constraints, cold-chain requirements). These elements would be the semantic infrastructure that makes monitoring meaningful.
The full minimum viable operational indicator set, covering five indicator families (capacity and continuity, inventories, flows and trade, demand signals, and shortage manifestation), is specified in Annex 2.B (Section B.2). Within this framework, three categories of signals are particularly relevant for early warning and disruption diagnosis:
1. An “input market stress index” (for input-intensive generics and chemically synthesised medicines): Deviations from baseline in quoted lead times and spot prices for key starting materials can provide upstream early warning, particularly for low-margin generics where input cost volatility often precedes manufacturing slowdown or withdrawal (Frank, McGuire and Nason, 2021[49]);
2. Tender stress signals (for hospital and centrally procured medicines): an increase in zero-bid or failed tenders offers a sensitive signal of upstream stress. This typically reflects manufacturing constraints, the withdrawal of marketing authorisation holders or a reduced willingness to serve a market – often visible before downstream stockouts occur;
3. Regulatory segmentation (for products with limited substitution or site specificity): flags that identify stock incompatible with local regulations (e.g. unregistered sites or unapproved pack formats) highlight barriers to substitution. This prevents false reassurance from aggregate volume indicators by distinguishing between physical inventory and legally available supply.
The objective is not exhaustive measurement, but a cost-efficient diagnostic stack that supports early identification and provisional classification of disruption types and likely propagation paths, sufficient to move decision makers from general alarm to targeted mitigation.
Aggregation architecture: Managing the confidentiality-utility trade‑off
Extracting meaningful signals from pharmaceutical supply chains requires managing a basic constraint: the most operationally relevant information (e.g. facility disruptions, batch-release status, and node‑level service performance) is also the most sensitive to disclose. Operational viability therefore hinges on an aggregation architecture that preserves directional information while limiting identifiability and legal exposure:
Aggregation devices: In a minimum viable design, sensitive data should be translated into member-grade signals through a disciplined toolkit: i) banded ranges (e.g. >90%, 70‑90%, <70%) rather than point estimates; ii) pooled reporting at country, region, or supply-chain-node level rather than firm attribution; iii) index normalisation relative to a routine baseline; and iv) trusted intermediary validation, where a neutral Secretariat performs anomaly checks and releases only outputs that meet pre‑defined disclosure thresholds.
Avoiding false precision: overly granular outputs, especially in the public layer, can increase risks of strategic behaviour and transparency-induced hoarding without improving decision quality. The design principle is sufficient aggregation: coarse enough to prevent re‑identification and market distortion, yet informative enough to change crisis decisions relative to the counterfactual.
This trade‑off is product-specific, which has a direct implication for scope. The minimum viable basket should therefore prioritise products where aggregation thresholds can simultaneously support participation viability and operational utility, i.e. where the system can produce a credible diagnosis without requiring disclosures that actors will not provide.
2.5.5. Testing the MedMIS proof-of-concept across tracer product categories
To test whether the proposed MedMIS could function across heterogeneous product types, the design was tested against three tracer categories: injectable corticosteroids, as small-molecule generics with concentrated API markets; PCR diagnostic tests, as products dependent on complex multi-sectoral assembly networks; and seasonal influenza vaccines, as biologicals with inelastic short-term supply. The purpose of this exercise was to assess how far a common monitoring architecture could generate actionable early-warning signals across different supply-chain structures. The detailed stress test, including the full assessment table, is presented in Annex 2.B, Section B.3.
The stress test yields three principal findings. First, injectable corticosteroids are the strongest fit for an early-warning monitoring system. Concentrated upstream manufacturing, limited surge capacity and high sensitivity to quality events create recurrent failure modes where MedMIS could help distinguish temporary disruptions from structural shortfalls. Second, PCR diagnostics expose the limits of a pharma-centric monitoring scope: these products depend on cross-sectoral inputs outside pharmaceutical regulation, meaning that MedMIS would serve less as a shortage predictor than as a policy synchronisation tool. Third, seasonal influenza vaccines illustrate the limits of early warning in markets with inelastic short-run supply: in such cases, the stabilising function shifts from industrial monitoring to allocation co‑ordination (Box 2.3).
For biologicals with inelastic short-run supply (e.g. seasonal influenza vaccines), MedMIS could not function as a production early-warning system, since lead times are fixed. Its value would instead lie in allocation co‑ordination: validating aggregated demand signals to defuse defensive over-ordering, monitoring the ancillary inputs (glass vials, stoppers, lipids) that can strand available bulk antigen, and supporting reallocation of surplus doses once national saturation is reached. Annex 2.B, Section B.4, details these three functions and the COVID‑19 Clearing House precedent on which they build.
From archetypes to pilot scope: Risk-based selection for the minimum viable basket
The stress test confirms that MedMIS should not attempt to apply a single monitoring protocol across all product types: the data architectures required for corticosteroids, PCR diagnostics and seasonal biologicals differ so substantially that a universal approach would be neither efficient nor credible. This heterogeneity has a direct operational implication: the pilot basket should be selected not only on the basis of clinical criticality and supply-chain vulnerability, but also on monitoring feasibility and supply-chain tractability.
The Union list of critical medicines, comprising over 200 active substances as of January 2026, provides the relevant universe from which to draw candidates for a MedMIS pilot.4 However, the list should be treated as a starting point rather than as a pilot basket in itself. Criticality, as defined in the Union list methodology, should not be conflated with supply-chain vulnerability. The Union list is grounded in therapeutic indication and the availability of alternatives, whereas vulnerability assessment of the supply chain is a separate analytical layer focussed on exposure, fragility and disruption pathways (EMA, 2023[50]). The initial pilot vulnerability assessment conducted by HERA, the subsequent methodology prepared by the MSSG as well as the Critical Medicines Alliance reports could provide methodological inputs for selecting a focussed MedMIS pilot basket.
HERA’s first pilot supply-chain vulnerability assessment provides both a precedent and a methodological starting point (European Commission, 2024[51]). The pilot assessed 11 medicines selected from the then Union list (216 medicines) through a stepwise process: filtering to medicines with shortage notifications reported to EMA between 2019 and 2023, reducing the pool to 90, then combining quantitative ranking with qualitative sampling to ensure diversity of manufacturing processes, geographic profiles and product specificities. Crucially, the pilot report noted that the selected medicines “are not to be regarded as “at most risk of shortages/most vulnerable” but present a diversified sample” suited to testing the assessment methodology. This logic could apply to the choice of any future MedMIS pilot basket.
The pilot assessment also identified three structural constraints that any monitoring exercise must address at scale: i) the absence of a legal basis for mandatory data collection on medicines supply chains; ii) the lack of harmonised data formats and standards, which created interoperability problems across reporting entities; and iii) firms’ hesitancy to disclose commercially sensitive information. MedMIS would translate these lessons into three selection criteria for pilot inclusion:
1. Monitoring tractability: The supply chain structure must generate observable, standardisable signals at sufficient frequency to support early warning. Under the safe harbour architecture described in Section 2.5.8, where a neutral trustee collects raw data and participants receive only aggregated outputs protected by ex-ante competition-law guidance, the tractability test asks whether the product’s key supply-chain nodes (API sourcing, finished-dose manufacturing, distribution) can yield indicators on capacity status, disruption events and demand signals that are operationally meaningful once aggregated. Products where supply chains are too opaque, too fragmented across non-pharmaceutical sectors or too commercially concentrated for even aggregated outputs to avoid re‑identification should be deferred to later phases.
2. Standardisability: The operational indicators that MedMIS requires – capacity status, inventory proxies, demand signals – must be reportable in comparable units across firms and jurisdictions. While product classification systems (ATC codes, route of administration) exist for all medicines, the European Commission’s first pilot supply-chain vulnerability assessment found that reporting practices varied substantially: some firms provided precise volumes, others relative shares, others declined to disclose. The binding constraint is therefore not product identification but measurement comparability. Pilot candidates should be drawn from product categories where existing regulatory reporting conventions, procurement data infrastructure, and standardised pack/strength configurations enable consistent aggregation without extensive ad hoc harmonisation.
3. Actionability: Clear escalation pathways and decision levers must exist once signals deteriorate. A monitoring signal has value only if it connects to a response: regulatory flexibilities, reserve activation, demand-side co‑ordination or re‑sourcing. Products where no credible mitigation pathway exists – either because substitution is impossible, reserves impractical, or cross-border reallocation prohibitive – should not absorb scarce monitoring capacity in the pilot phase.
On this basis, MedMIS Phase 1 could target a narrow basket of about 20‑30 products for the pilot, drawn from the intersection of the Union list and the expanding supply-chain vulnerability assessment programme. Selection should prioritise medicines where the vulnerability assessment methodology has already generated, or can readily generate, baseline vulnerability data, and where the three criteria above are jointly satisfied. The pilot basket should be reviewed at the 24‑month mark and adjusted in light of operational experience. The number of products in the pilot is also constrained by human and financial resources which could be allocated to the monitoring system (see Section 2.5.9).
Phase 1 would deliberately exclude:
Seasonal biologicals (where fixed production lead times and inelastic short-run supply limit the utility of conventional early-warning monitoring; the appropriate function for these products is allocation co‑ordination rather than production surveillance – see Annex 2.B, Section B.4);
Diagnostics where cross-sectoral indicator development remains difficult; and
On-patent products with vertically integrated, multi-site global manufacturing networks and established firm-level monitoring systems, where the marginal value of an additional pooled early-warning layer is limited relative to thin-margin generics.
These exclusions reflect sequencing rather than scope limitations: demonstrated performance on the pilot cohort would generate the credibility required to expand coverage in Phase 2.
2.5.6. Impacts and standards
Quality/GMP/traceability
MedMIS could potentially strengthen continuity decisions by distinguishing disruption types (e.g. GMP remediation/quality events vs. logistics failures vs. input constraints) and by structuring alerts with duration bands and intended mitigation actions. Because some tracer categories are explicitly vulnerable to quality-driven disruptions, “event taxonomy” and disciplined validation would be crucial. Operationally, the standards question is primarily one of information governance: sensitive inputs should be transformed into banded/aggregated indicators, verified by a neutral Secretariat/trustee, and released only when disclosure thresholds are met. Where relevant, MedMIS should align with existing track-and-trace and security expectations (secure data handling; auditable workflows), with certification and assurance treated as non-negotiable.
Environmental implications
MedMIS would have minimal direct environmental impact, as it is an information system rather than a physical infrastructure. Its indirect environmental effect would depend on the quality of its signal governance. If shortage signals are released prematurely or without adequate validation, they can trigger precautionary over-ordering and redundant emergency shipments, increasing transport emissions and pharmaceutical wastage. Conversely, validated and properly sequenced alerts would support more targeted responses (such as co‑ordinated reallocations and timed stock releases) reducing both unnecessary logistics and the volume of medicines that expire unused. On balance, it is reasonable to assume that a well-governed MedMIS would reduce the net environmental footprint of crisis response relative to uncoordinated national reactions.
Consumer/patient access and equity
The primary consumer-facing pathway is improved continuity: MedMIS would reduce policy-amplified scarcity by compressing diagnostic lag and improving the coherence of response (procurement posture, stock releases, regulatory flexibilities) rather than by creating physical supply. However, the assessment should also acknowledge distributional effects: MedMIS would be designed as a “coalition/club” with preferential co‑ordination benefits for members; smaller jurisdictions and non-Members may remain more exposed to allocation blindness and volume volatility. To make equity effects explicit, MedMIS could track (in the member layer) differential time‑to-restoration proxies by market size/region and require that any public communications be framed to avoid demand shocks and inequitable runs.
2.5.7. Participation incentives and geopolitical constraints
The incentive architecture: From opacity to structured disclosure
The feasibility of MedMIS would turn less on technical capacity than on whether actors have credible incentives to disclose commercially sensitive data. Pharmaceutical supply chains operate under conditions of strategic confidentiality: firms fear revealing competitive intelligence, governments fear admitting scarcity, and crises amplify these risks. Consequently, a MedMIS mechanism could not rely on aspirational “transparency norms”: it would need to operate as a transactional disclosure system where participation reduces the downside of sharing while delivering operationally relevant benefits.
For industry, participation is rational only if the mechanism offers value commensurate to the commercial risks of disclosure. The architecture should therefore offer three types of “currencies”:
1. Intelligence arbitrage (“give‑to-get”): firms would contribute proprietary data in exchange for privileged access to aggregated system-wide intelligence – specifically, validated demand signals that distinguish genuine clinical need from “phantom” panic-ordering, reducing their inventory risk. Access to this “information premium” would be strictly conditional on data quality: firms contributing granular, timely inputs would receive earlier warnings, while “free‑riders” would only receive lagged public summaries.
2. Regulatory reciprocity: the most credible incentive would not be preferential treatment in routine authorisations (which would compromise independence), but predictable crisis-time processes within existing regulatory frameworks. Participating regulators would commit to a limited set of service‑level and co‑ordination measures that reduce avoidable friction when speed matters, for example:
Expedited procedural handling of predefined shortage mitigation variations (e.g. pack/label changes, batch-release sequencing, alternative presentations) using pre‑agreed templates and prioritisation rules;5
Regulatory reliance/work-sharing where legally feasible (mutual recognition of inspection outcomes, co‑ordinated QP/batch-release arrangements), to avoid duplicative steps;
Aligned use of existing emergency flexibilities (e.g. exceptional release, labelling waivers) through predictable escalation channels.
These benefits could be structured as response‑triggered incentives rather than standing membership privileges: where a participant provides a validated early signal that enables co‑ordinated mitigation, the Secretariat could activate expedited regulatory pathways or targeted import flexibilities for that participant’s market. This links signal-sharing directly to operational benefit at each reporting cycle, rather than relying solely on membership-level access tiers.
3. Competition-law assurance for mechanism-specific data exchanges: firms would require legal certainty that submitting data to the neutral MedMIS Secretariat, and receiving only aggregated outputs, would not, in itself, constitute prohibited co‑ordination under Article 101 TFEU. This assurance would be specific to data shared through the mechanism for shortage‑mitigation purposes and would not extend to any other commercial conduct between participants. This necessitates explicit competition-law guidance (e.g. a block exemption, formal comfort letter, or published guidelines from relevant competition authorities), confidentiality protocols, and firewalled access rules that prevent data shared for shortage mitigation from being repurposed for pricing or enforcement actions.
Coalition logic, minimum participation and compliance
To be operationally credible, MedMIS would require coverage of three critical supply chain nodes: (i) major demand centres, (ii) systemically important producer hubs (for upstream API visibility), and (iii) critical private data holders (large finished-dose manufacturers). However, incentive alignment across these nodes is asymmetric: while demand centres have immediate incentives to co‑ordinate, major producer hubs often benefit from strategic opacity to manage domestic pressure without triggering international scrutiny. Consequently, the coalition will likely launch as a demand-anchored bloc, leveraging its collective market power to secure upstream engagement over time.
Even with incomplete global participation, MedMIS could generate value as a co‑ordination mechanism whose benefits improve with scale but remain preferential to participants:
For importing countries and large public buyers: consolidating and validating demand signals would reduce precautionary double‑ordering within the coalition and improve allocation discipline. The core benefit is access to validated intelligence – delivery expectations, allocation constraints and escalation timelines – that non-Members do not receive.
For producer-country governments: participation would provide a neutral, evidence‑based validation of disruption causes (e.g. distinguishing GMP remediation or input shortages from strategic withholding). This gives producer governments political cover to respond proportionately rather than defaulting to export restrictions under domestic pressure, reducing the risk of escalation dynamics with buyer countries.
For manufacturers and large suppliers: participation would replace fragmented, panic-driven demand signals with validated system-wide intelligence. Knowing whether a demand spike reflects genuine clinical need or precautionary ordering supports production planning, inventory positioning and allocation decisions, while reducing exposure to sudden cancellations and ex-post blame during shortages.
Because these “goods” are informational and non-rival, their value depends on reciprocity: without credible mechanisms to deter free‑riding, participants could consume high-quality intelligence while contributing little or low-quality data, undermining the system’s integrity. Since MedMIS would lack legal enforcement powers, compliance would rely on two practical mechanisms:
Tiered access linked to contribution quality: access to confidential dashboards (Tier 2) would not be a permanent entitlement but a privilege renewed at each reporting cycle. Participants that repeatedly submit late, incomplete or excessively aggregated data would be automatically downgraded to the public alert layer (Tier 1), losing access to the operational intelligence that makes participation valuable. This ensures that the quality of what each participant receives remains proportionate to what they contribute.
Diplomatic visibility: while MedMIS would not impose formal sanctions, a standing high-level forum would make non-co‑operation visible. A government identified by its trading partners as withholding data or engaging in unilateral hoarding would face reputational costs (strained bilateral relationships, reduced credibility in future negotiations) that can outweigh the short-term gains of non-co‑operation.
Through this architecture, MedMIS would make sustained participation the rational choice: participants that contribute reliable data would receive better intelligence and stronger diplomatic standing, while those that do not would be progressively excluded from the system’s most valuable outputs.
2.5.8. Risks and negative externalities
Signal distortion: The risk of self-fulfilling shortages
A central paradox of supply monitoring is that increased transparency can, if poorly governed, amplify the very volatility it seeks to mitigate. In pharmaceutical markets, where short-run substitution is limited and supply chains are rigid, an unverified “scarcity” signal can trigger competitive acquisition spirals. Buyers shift from ordering needed quantities to securing whatever they can obtain, depleting buffer stocks and converting manageable disruptions into system-wide shortages. This behaviour is not merely psychological: rigid procurement rules and stockout penalties can make precautionary ordering a rational response to perceived risk (Espitia, Rocha and Ruta, 2020[52]).
The risk is therefore not transparency per se, but unmanaged transparency. A MedMIS mechanism releasing raw or poorly contextualised signals would generate false positives and destabilising procurement responses. “Signal governance” (information hygiene) is therefore a feasibility condition, requiring concrete design controls such as:
Validation and sequencing: A short embargo window (24‑48 hours) for regulator/secretariat verification and co‑ordinated interpretation before broader release, ensuring signals are validated and contextualised.
Masking by design: Using banded indices (e.g. capacity >90%, 70‑90%, <70%) rather than firm-specific alerts, preventing identification of individual manufacturers.
Narrative discipline: Ensuring alerts distinguish event types (quality remediation vs. logistics disruption vs. input constraint), include plausible duration bands, and specify intended mitigation responses.
A tiered information governance model, distinguishing between restricted operational intelligence accessible only to the Secretariat and designated authorities (Tier A), and curated public outputs designed to stabilise expectations (Tier B), would address the panic externalities inherent in pharmaceutical transparency. The detailed governance model, including escalation triggers for moving information between tiers, is presented in Annex 2.B, Section B.5.
Ensuring data sharing does not breach competition law
Industry participation in MedMIS would require a high degree of legal certainty. The mechanism would depend on commercially sensitive indicators – including capacity utilisation rates, disruption events and inventory proxies – in markets that are often highly concentrated. In such settings, sharing information on production constraints or anticipated shortfalls could reveal firms’ strategic vulnerabilities. Without robust safeguards, firms would have strong incentives to restrict reporting to low-value, highly aggregated or lagged indicators, weakening the mechanism’s early-warning function.
MedMIS would need to be competition-safe by design. This would require a neutral trustee to collect and manage raw data, ensuring that no market participant can access disaggregated competitor information; clear exclusions for information related to prices, tendering strategies and commercial negotiations; strict controls over data dissemination and documentation; and formal ex-ante guidance from competition authorities confirming that the trustee model is compatible with Article 101 TFEU, which prohibits agreements and concerted practices that restrict competition. The European Commission’s COVID‑19 Comfort Letter (April 2020) to Medicines for Europe provided a precedent for this type of arrangement for a temporary crisis response, under defined safeguards including neutral intermediation, purpose limitation and restrictions on competitively sensitive exchanges (see Annex 2.B, Section B.6). As a permanent standing mechanism, MedMIS would require analogous assurance. A further EU precedent is provided by Regulation (EU) 2024/2747 establishing the Internal Market Emergency and Resilience Act (IMERA), whose Article 11 creates a monitoring framework with explicit safeguards to ensure that information sharing remains consistent with competition law obligations.
Upstream blind spots: Monitoring what matters beyond finished medicines
A further risk is that MedMIS would detect problems only at the finished-product level, missing root causes that lie upstream. Disruptions in API supply, key starting materials, packaging components or cross-sector inputs may not be visible in downstream monitoring until they cascade into shortages at the point of care. During COVID‑19, monitoring focussed on diagnostic kit availability failed to capture vulnerabilities in upstream inputs such as plastics and reagents. Comprehensive upstream mapping is not feasible across all products. MedMIS should instead focus on a narrow set of high-impact chokepoints, such as concentrated API sources for corticosteroids or specialised inputs for mRNA vaccines, and explicitly flag blind spots rather than allowing “no signal” to be mistaken for “no risk”.
Ultimately, the net value of MedMIS would depend on managing three trade‑offs: avoiding transparency-induced acquisition spirals, ensuring legal operability that enables participation, and preventing false confidence from upstream blind spots. The feasibility assessment should therefore weigh the cost and credibility of safeguards – governance protocols, legal comfort, verification capacity, and secure IT architecture – alongside the technical feasibility of data collection.
2.5.9. Implementation burden and value‑for-money
Cost profile and budgetary benchmarks
MedMIS costs would be front-loaded: they would reflect investments in institutional credibility, secure handling of sensitive data and verification protocols capable of withstanding acute crisis pressures. A multinational monitoring mechanism faces a “sustainability cliff” – below a minimum resourcing threshold, outputs are not decision-grade in crisis timeframes and therefore do not shift behaviour, rendering the investment sunk.
Benchmarking against three established international mechanisms – AMIS, the Extractive Industries Transparency Initiative (EITI) and the Financial Stability Board (FSB) – brackets the plausible cost envelope. These initiatives, as well as detailed cost benchmarks and indicative cost decomposition are presented in Annex 2.B, Section B.7. MedMIS would sit between AMIS and EITI, reflecting higher verification intensity than AMIS but lower institutional authority than FSB.
Two scenarios are assessed:
Scenario A (“targeted pilot”): EUR 6‑8 million per year, 10‑15 experts, covering a basket of approximately 20‑30 products for around 20 countries. The fixed institutional base (Secretariat, secure data infrastructure, legal safeguards, verification capacity) absorbs EUR 4.7‑5.9 million regardless of product scope; the variable per-product layer adds approximately EUR 80 – 120 000 per product. Staff costs dominate at approximately 70% of the total.
Scenario B (“global hub”): EUR 12‑16 million per year, 30‑40 FTE, covering 40‑50 products with 24/7 surge capacity. The scaling from Scenario A to B is driven by deepening verification infrastructure and operational redundancy, not primarily by product count.
MedMIS should be framed as tail-risk insurance: its benefits arise in infrequent but high-impact disruption episodes. Documented fiscal costs of major medicines shortages – including emergency procurement premiums, therapeutic substitution, crisis logistics and operational disruption – have reached tens of millions of euros per episode (Dayan et al., 2024[53]; Napier M., 2024[54]). A mechanism costing EUR 6 – 16 million annually faces a low break-even threshold: alleviating a single major co‑ordination failure every two years would plausibly offset operating costs.
To prevent institutional drift, MedMIS should follow a pre‑committed conditional timeline with two hinge points: a month‑24 feasibility review and a month‑48 operational utility review. Failure at either point would trigger transition to a lighter co‑ordination forum. For systemically important producer hubs unlikely to fund the Secretariat directly, participation incentives could be framed as regulatory reciprocity rather than extracted contribution. The detailed hinge‑point criteria, cost-sharing model, and sustainability architecture are presented in Annex 2.B, Section B.7.
2.5.10. Policy conclusions: Feasibility pre‑requisites and operational conditions
The guiding conclusion of this assessment is that governance maturity must precede analytical ambition. The analysis identifies five binding constraints, six viability conditions and a sequenced implementation pathway, consolidated below.
Summary of binding constraints
MedMIS feasibility hinges on a small number of binding constraints:
1. Safe harbours would determine participation quality: the instrument’s stabilising value would depend on timely access to decision-relevant information that is largely held by private actors and is commercially and legally sensitive. Without arrangements for confidentiality, purpose limitation and explicit ex-ante competition-law comfort, reporting would predictably converge towards low-sensitivity indicators and lagged signals. This negates the early-warning function.
2. The international dimension (beyond EU) would be crucial: the most consequential visibility gaps arise upstream and beyond EU jurisdiction. A mechanism that would not secure participation from at least some major upstream hubs and systemically important private data holders risks consolidating downstream “symptoms” with limited capacity to anticipate disruptions or validate causal narratives early.
3. Monitoring should capture quality-driven discontinuities, not only volume fluctuations: pharmaceutical shortages are frequently triggered by GMP deviations, batch-release failures and site‑specific compliance events. A monitoring logic that cannot incorporate these dynamics would systematically misclassify the type and persistence of scarcity.
4. Transparency can amplify the scarcity it seeks to mitigate: in a low-substitutability market with decentralised procurement, poorly governed transparency triggers precautionary ordering and competitive acquisition. Information governance, tiered access, disciplined external communication and validation protocols are therefore feasibility conditions, not design refinements.
5. MedMIS should operate as a voluntary, sovereignty-preserving mechanism: it should not override national procurement or regulatory decisions but only improve the information base upon which sovereign decisions are made. This would preserve political feasibility but also caps the instrument’s effectiveness: voluntary co‑ordination is vulnerable to defection precisely when collective action matters most. The mechanism would therefore ‘live or die by its ability to demonstrably improve the quality of decisions rather than mandate outcomes.
Taken together, these constraints imply that MedMIS would only be feasible as a narrowly scoped, governance‑intensive mechanism. Designs that assume broad disclosure, extensive product scope or reliance on purely public indicators are unlikely to deliver decision-relevant value commensurate with their fixed institutional costs.
Conditions for instrument viability
Where the binding constraints above identify what can go wrong, the conditions below specify what should be secured. A viable MedMIS would require all of the following elements:
1. Legal operability under competition law: a competition-law safe harbour – secured through an ex-ante instrument (such as a block exemption, a formal comfort letter or published guidelines with case‑specific assessment) from relevant competition authorities confirming that the tiered data-sharing model and trustee architecture fall outside the scope of Article 101 TFEU prohibitions or satisfy the conditions for exemption under Article 101(3).
1. Functional critical mass: participation defined by market impact, requiring a) jurisdictions representing at least 50% of OECD pharmaceutical sales volumes and b) representation of the main API manufacturing jurisdictions for each product category in the basket, or alternative sites of comparable capacity.
2. Scope discipline and product tractability: a narrowly defined critical basket, with product groupings that reflect regulatory specificity (form/strength/site constraints) and are stable enough to support routine baselines.
3. Minimum viable dataset under tiered access: the mechanism can routinely collect and validate, at least in banded or indexed form, a minimal set of indicators covering disruption events, expected duration, effective capacity status and downstream shortage manifestation. The design must not depend on granular firm-level disclosure as a precondition for usefulness.
4. Host mandate and governance: an explicit intergovernmental decision (or equivalent instrument) conferring the necessary authority and budget on the Secretariat to act as a neutral data trustee (including, where applicable, the legal capacities required for secure data handling across jurisdictions).
5. Verification and escalation protocols: standing secretariat capacity and agreed procedures for validation, escalation thresholds and crisis convening. Outputs should be produced within crisis-relevant timeframes.
If these conditions were not met, the realistic fallback would be a “demand transparency and alignment coalition”: a limited co‑ordination mechanism among willing members focussed on demand-side visibility and procurement alignment. This would differ from existing EU shortage reporting – notably the European Medicines Agency (EMA) European Shortages Monitoring Platform (ESMP), which is primarily designed for regulatory shortage monitoring within the EU – in three respects. First, it would aggregate demand-side signals across participating jurisdictions, not only supply-side shortage notifications. Second, it would operate under a structured escalation protocol for crisis periods. Third, it would be convened by a designated intergovernmental body rather than a medicines regulator. The convening institution would need to be agreed among participants, with a standing secretariat function even if analytically lighter than the full MedMIS design.
Sequencing pathway
If the minimum proof-of-concept proves operationally useful, expansion could be sequenced rather than assumed. A pragmatic pathway would be:
Phase 1 – Narrow basket (months 0‑24): Establish routine baselines and tiered crisis-escalation protocols, using banded indicators and strict confidentiality safeguards. Focus on products where allocation dynamics and policy co‑ordination can plausibly alter outcomes. The month‑24 hinge review would determine whether essential preconditions have been met; failure would trigger transition to the fallback coalition.
Phase 2 – Coverage deepening (months 24‑48): Expand participation to additional upstream hubs and private data holders. Refine indicators for upstream inputs and tier‑2/3 constraints. Improve verification through structured triangulation agreements. The month‑48 review would evaluate demonstrated decision-utility in at least one crisis or near-miss.
Phase 3 – Functional strengthening (months 48+): Rather than attempting “predictive” scoring of stochastic risks, focus on propagation modelling and protocol integration. Link MedMIS alerts to pre‑agreed national response protocols (e.g. automatic stockpile‑release triggers, regulatory-flexibility deployment), transforming situational awareness into actionable commitment.
Across phases, the guiding principle should be that governance maturity precedes analytical ambition. A high-functioning baseline with disciplined information governance is more valuable than a technically advanced model with weak participation or fragile confidentiality safeguards.
Notes: Conditions for proceeding: i) competition-law safe harbour secured; ii) participation of >50% of OECD procurement volume and top‑3 API jurisdictions per tracer; iii) host-institution mandate and funding confirmed; iv) month‑24 hinge‑review clause included.
2.6.1. Objective and rationale
Definition and purpose of pooled procurement
This section evaluates a potential international pooled procurement mechanism for medical supplies, designed to be open to participation beyond the EU (i.e. selected “third countries”). Pooled procurement is defined here as an arrangement whereby multiple jurisdictions aggregate demand and contract collectively (through shared tenders, framework agreements or advance purchase contracts) rather than conducting parallel national procurements. The mechanism’s primary objective is buyer-side co‑ordination: by replacing competitive procurement with a single, consolidated buyer interface, the mechanism aims to:
Limit market fragmentation;
Reduce “phantom demand” (overlapping orders placed by multiple buyers to hedge scarcity), and
Improve predictability for suppliers by establishing agreed rules of engagement.
Crucially, such a mechanism – referred to hereafter as “MedPPA” (Medical Pooled Procurement Arrangement) for convenience – is defined not merely as a cost containment tool, but as a demand-side co‑ordination instrument. Pooled purchasing power can also be used to negotiate tangible security-of-supply commitments – such as capacity reservation, multi-site production requirements and enforceable delivery schedules – and to provide participating countries with pre‑agreed allocation rules that guarantee access to available supply during scarcity, rather than focussing solely on achieving the lowest unit price.
Decision context and rationale
The strategic rationale for MedPPA addresses specific market failures and co‑ordination risks observed during health emergencies. In emergencies, countries face a co‑ordination problem: behaviour that is rational for each buyer – side deals, over-ordering, aggressive tendering – undermines the collective outcome. The result is price spikes and inequitable distribution. Countries could co‑operate, however, if they believed others would stay committed. Pooled procurement therefore should build credibility through three mechanisms:
consolidating demand through a single buyer interface, strengthening the coalition’s ability to negotiate supply security commitments (not just price reductions);
establishing transparent, pre‑agreed allocation rules, so each participant understands its claim rights; and
implementing material costs for defection – loss of priority access, forfeiture of reservation fees and exclusion from future participation – that make bypass more costly than pool participation.
These provisions signal to suppliers that pooled volumes are firm commitments, and to participating governments that the pool is credible. In markets for essential countermeasures, the policy objective is reliability and surge readiness rather than short-run savings. A joint mechanism allows the coalition to negotiate structural supply assurances that individual buyers cannot command, such as reserved surge capacity, diversified manufacturing footprints, and priority delivery clauses. This implies a “resilience premium”: participating states accept some reduction in autonomous procurement flexibility (and potentially higher unit costs) during routine periods in exchange for guaranteed priority access during supply shocks.
The EU already operates a Joint Procurement Agreement (JPA) under the framework governed by Regulation (EU) 2022/2371 on serious cross-border threats to health, which provides a voluntary legal framework for collective purchasing among EU Member States, associated countries and candidate countries. The JPA’s geographic scope already extends beyond the EU‑27 to include EEA/EFTA states and accession candidates. The question this section addresses is therefore not whether joint procurement with non-EU partners is feasible in principle (it demonstrably is), but whether a more structured mechanism with deeper integration, binding commitments and explicit security-of-supply objectives would generate additional resilience value sufficient to justify the institutional complexity involved.
The rationale for broader third-country inclusion is both geopolitical and operational. Pharmaceutical supply chains depend on non-EU jurisdictions for critical API manufacturing, fill-finish capacity and logistics corridors. A procurement mechanism that excludes these nodes risks fragmenting the very supply networks it seeks to secure. During acute shortages, uncoordinated purchasing by countries outside the pool can drive the same escalatory dynamics – phantom demand, competitive tendering and price spikes – that the mechanism is designed to prevent. MedPPA would go beyond the current JPA’s geographic extension by establishing tiered participation structures, product-specific binding commitments, and pre‑agreed allocation rules that the current framework does not provide.
A further design feature would be dual-mode operation. At EU level, the Emergency Framework Regulation already provides for crisis-mode activation and determines the type of procurement action. MedPPA would build on this by embedding pre‑negotiated surge clauses and delegated procurement authority into the standing framework, so that the transition from routine contracting to crisis execution is operationally automatic rather than requiring ad hoc political decisions for each procurement cycle. In routine mode, MedPPA would focus on maintaining active contracting infrastructure and supplier relationships for products vulnerable to chronic shortages. In crisis mode, pre‑agreed triggers would activate surge call-offs and centralised executive authority. This dual-mode capability would significantly enhance effectiveness, but it is a design choice to be evaluated on its merits rather than a precondition for viability.
MedPPA would not operate in isolation. Supply-chain vulnerability assessments (as piloted by HERA in 2024) would inform product-specific opt-in decisions. Where deployed alongside the monitoring mechanism (MedMIS) and the stockpiling mechanism (RescPool) evaluated elsewhere in this chapter, MedPPA would gain activation intelligence and bridging capacity respectively. Each instrument retains standalone value, but establishing governance linkages would enhance the response architecture.
2.6.2. Institutional models and global practice
To ground the feasibility assessment, this section reviews existing multilateral and regional procurement mechanisms. These examples illustrate various governance models and provide empirical lessons on effectiveness, highlighting the trade‑offs between broad inclusion and operational agility.
European mechanisms: EU JPA and HERA
The EU Joint Procurement Agreement (JPA), signed in 2014 and now governed by the framework of Regulation (EU) 2022/2371 on serious cross-border threats to health, provides the primary institutional reference point for this assessment. The JPA established a voluntary legal framework for collective procurement of medical countermeasures, with participation open to EU Member States, EEA/EFTA states, candidate countries and certain associated countries (OECD, 2024[38]).
During the COVID‑19 crisis, the JPA’s legal framework did not permit the procurement of vaccines prior to EMA authorisation. The Commission therefore concluded Advance Purchase Agreements (APAs) with vaccine manufacturers through the Emergency Support Instrument (Regulation (EU) 2016/369, activated by Council Regulation 2020/521), signed by the Commission on behalf of EU Member States, committing EUR 2.7 billion from the EU budget as advance payments to de‑risk vaccine development (European Parliamentary Research Service, 2023[55]). While the EU’s vaccine procurement strategy succeeded in securing early access for Member States, the experience highlighted that crisis-speed execution required a financing and decision making architecture distinct from the JPA’s consensus-based framework.
The creation of the Health Emergency Preparedness and Response Authority (HERA) marked a shift from reactive committees to permanent preparedness capacity. Within the EU, a central element of this shift is EU FAB, a network of ever-warm vaccine manufacturing facilities contracted to maintain operational readiness for activation during public health emergencies (European Commission, 2023[56]). This approach reflects the practical constraints of extending EU governance standards, liability provisions, budgetary regulations and audit requirements – to non-EU legal systems (Välikangas, Luistro-Jonsson and Jarvenpaa, 2022[57]). For MedPPA, the implication is that third-country participation is unlikely to take the form of direct integration into EU procurement contracts; instead, it would require a purpose‑built framework with tiered participation structures adapted to different legal and institutional contexts.
Lessons from global pooled procurement and foundational requirements
Existing multilateral and regional procurement mechanisms – the PAHO Revolving Fund, Gavi (including the International Finance Facility for Immunisation – IFFIm – and the Advance Market Commitment), the COVAX facility, the Global Fund’s Pooled Procurement Mechanism, and three regional initiatives (AVAT, the African Pooled Procurement Mechanism and MAV+) – provide operational lessons directly relevant to MedPPA’s feasibility conditions. The detailed institutional profiles are presented in Annex 2.C, Section C.1. An analysis of these mechanisms indicates that the feasibility of a MedPPA mechanism depends on three foundational requirements:
First, operational credibility requires a financial backbone: as demonstrated by PAHO, durable mechanisms are never mere co‑ordination platforms but possess the “hard infrastructure” of liquidity funds or guarantees that reduce supplier credit risk and prevent procurement failure due to national payment delays. Such a payment guarantee does not exist in the existing JPA or APAs, the inclusion of third-country participants with heterogeneous creditworthiness and payment reliability may raise the question of its utility. Requiring full sovereign guarantees would risk excluding precisely the smaller economies that would benefit most from pooled procurement; the financing architecture must therefore balance payment credibility with accessibility. The financing architecture would also need to address currency risk: if third-country participants remit in local currency while framework contracts are denominated in euros, the mechanism requires either a currency hedging facility or a requirement that participants pre‑fund in the contract currency – each carrying distinct cost and accessibility implications.
Second, market-shaping capacity is essential when supply is constrained: pooling, when not focussed on price minimisation, can be used as an instrument to secure supply, through advance purchase agreements, capacity reservations and resilience‑weighted contract criteria. This approach requires accepting a resilience premium – paying above minimum unit cost to secure guaranteed supply, diversified production or priority delivery commitments.
Finally, coalition discipline is critical. Under severe scarcity, participants face strong incentives to bypass collective arrangements and negotiate bilateral deals. For a buyer coalition like MedPPA, defection risks would be highest for products with inelastic supply, limited substitutes and high political salience. These are precisely the cases where pooled procurement would require the strongest ex ante commitments: binding participation rules, transparent allocation criteria and structured consequences for unilateral defection. For products with more elastic supply or lower scarcity risk, lighter preparedness contracts may be sufficient, providing supplier assurance without locking participants into rigid volume commitments.
Cross-sector experience confirms both the importance and the difficulty of maintaining coalition discipline. The strongest compliance record belongs to mechanisms with automatic, material penalties: the EU (greenhouse gas (GHG)) Emissions Trading System (ETS), for example, achieves near-full compliance through non-discretionary financial sanctions that make non-compliance more expensive than compliance regardless of market conditions (European Commission, 2022[58]).
In the absence of a supranational enforcement jurisdiction covering non-EU participants, MedPPA’s enforceability would therefore need to operate through material rather than judicial mechanisms. These could include forfeiture of reservation fees, loss of priority access in subsequent procurement cycles and exclusion from future participation.
Governance, funding, legal and operational arrangements
This section outlines how MedPPA could be structured, covering governance, financing, legal basis and operational processes. The feasibility constraint is institutional: maintaining coalition discipline under scarcity, ensuring regulatory operability across jurisdictions, allocating liability and establishing supplier credibility. A robust architecture must therefore specify participation criteria, attendant rights and obligations and mechanisms for transitioning from “routine” preparedness to crisis execution without collapsing into bypass and ad hoc bilateralism.
Participation and opt-in structures
A pooled procurement mechanism open to third countries is feasible only if participation is legally enforceable, operationally compatible and credible to suppliers. In practice, a single undifferentiated “third country” category is operationally unworkable: prospective participants differ materially in procurement-law compatibility, regulatory reliance pathways, liability doctrines, auditability and ability to pre‑commit funds. Treating them as one class would force the mechanism either to adopt lowest-common-denominator rules, slowing decisions and weakening enforcement, or to produce a framework so flexible that participation becomes largely symbolic when scarcity peaks. A minimum viable design therefore requires i) a tiered participation typology and ii) a two‑layer opt-in architecture.
Participation typology
A feasible architecture distinguishes three tiers with distinct legal obligations, access rights and activation assumptions. All tiers would operate through product-specific framework agreements with defined contract periods – the distinction would not lie in time‑boundedness (which would apply throughout) but in the breadth of standing eligibility and the degree of bespoke arrangement required for participation.
Tier 1: “Plug-in participants”: Jurisdictions with sufficient procurement and regulatory compatibility to participate across the full range of MedPPA product categories under near-seamless conditions. This tier would largely mirror the existing JPA participation model, which already enables EEA/EFTA states and candidate countries to operate under common tender templates, quality assurance protocols and dispute‑resolution arrangements. MedPPA would build on this foundation by adding product-specific binding commitments, pre‑agreed allocation rules under scarcity, and exclusivity or functional-exclusivity provisions that the current JPA does not include. Tier 1 would be where pooled procurement most closely resembles a single buyer interface, enabling standardised contracting and rapid activation.
Tier 2: “Associate participants”: Jurisdictions participating for a narrower set of product categories under bespoke conditions negotiated on a product-specific basis. Entry conditions would include:
A regulatory basis for deployment of procured products: in practice, this would most likely take the form of the third country accepting EU regulatory outcomes (EMA marketing authorisations) as sufficient for national deployment, rather than mutual recognition, which remains a long-term aspiration without established frameworks for most non-EU jurisdictions; the specific reliance pathway would need to be agreed per product category and documented in the participation agreement;
A dedicated liability and indemnification instrument (e.g. national indemnities aligned to common terms, pooled or no-fault compensation, or ring-fenced liability arrangements);
Acceptance that pooled contracts are tendered and audited under the procurement agent’s governing framework (EU public procurement rules and applicable audit standards), or demonstrated equivalence of the participant’s national procurement and audit regime, ensuring that all parties to a pooled contract operate under consistent transparency, integrity and accountability requirements;
Tier 3: “Co‑operation partners”: Jurisdictions or regions for whom integrated joint contracting is not currently viable, but where resilience gains could be achieved through structured co‑ordination short of pooled procurement. Concretely, Tier 3 participation would involve:
co‑ordinated tender calendars, aligning national procurement timelines to avoid simultaneous competing tenders for the same scarce products;
demand forecast and stock position sharing, enabling participants to identify surplus capacity and reduce phantom demand; and
mutual acceptance of quality assurance certificates or batch-testing results, reducing duplicative testing that delays deployment. These measures would not require pooled liability or integrated contracting but mitigate the destructive competition dynamics that uncoordinated national procurement produces during shortages.
Exclusivity options: Addressing bypass risk explicitly
Under scarcity, the primary threat is defection: participants face domestic incentives to bypass the pool and pursue bilateral deals, undermining pooled bargaining power. The design of MedPPA should therefore make an explicit choice among three exclusivity options for opted-in products:
Hard exclusivity: Participants would be contractually prohibited from parallel bilateral procurement for designated products during defined trigger periods. This raises significant legal and political challenges: prohibiting a Member State from procuring essential medicines through bilateral channels may conflict with national constitutional obligations to protect public health, and the political cost of ceding this sovereign prerogative is likely to exceed what most Member States would accept. Hard exclusivity is therefore included here as a theoretical benchmark rather than a recommended design option.
Functional exclusivity: Preferential access to reserved capacity, pricing and delivery slots is conditional on compliance, backed by enforceable penalties (e.g. forfeiture of reservation fees, loss of priority status, exclusion from future options). This should be calibrated so that the cost of defection exceeds the political benefit of bilateral deals.
No exclusivity: The mechanism functions as a co‑ordination platform during routine periods with limited crisis value. If chosen, this should be described candidly as a lower-ambition variant, because it cannot reliably prevent bidding wars when scarcity peaks.
A minimum viable crisis-relevant model therefore requires at least functional exclusivity (i.e. the recommended design option) where the cost of defection (forfeited fees, lost priority access) exceeds the political benefit of bilateral deals, without legally prohibiting national procurement action.
Practical participation conditions
To avoid symbolic membership, participation conditions should include:
Regulatory operability: Agreed reliance or recognition pathways for relevant product classes;
Financial credibility: Ability to pre‑commit funds or guarantees consistent with the chosen financing model;
Data and confidentiality compliance: Willingness to share demand and stock indicators under secure governance;
Readiness and logistics: Demonstrated ability to receive and deploy products (cold chain, customs pre‑clearance where relevant);
Governance acceptance: Acceptance of allocation rules and dispute resolution provisions tied to product-specific opt-in.
To address these three vulnerabilities, the governance architecture should operationalise binding commitment mechanisms, explicit allocation rules, and regulatory reliance pathways. The subsections that follow detail each.
Governance, decision velocity and allocation authority
Governance constitutes the primary feasibility constraint for a MedPPA mechanism because it is not merely a purchasing platform: under scarcity conditions, it becomes an allocation and enforcement system. This demands careful institutional design to manage three structural vulnerabilities that become acute when supply tightens (Box 2.4). During routine periods, co‑ordination problems are largely administrative: specifications, contracting, payment flows. In crises, the same mechanism must make rapid, high-stakes decisions about who receives limited supply, under what rules, and with what consequences for defection. Governance must therefore be designed explicitly for two operating regimes, with a clear transition mechanism between routine contracting and crisis execution.
Pooled procurement mechanisms that perform well in stable markets often fail when supply becomes constrained. Three structural vulnerabilities recur:
1. Strategic defection and the credibility gap: When scarcity peaks, participants face domestic incentives to bypass the pool for bilateral deals. Without material penalties for defection, suppliers cannot treat pooled volumes as binding commitments. The mechanism loses credibility precisely when it is needed.
2. Allocation conflicts and the fiduciary dilemma: Scarcity converts procurement into rationing. Without ex ante allocation rules and transparent claim hierarchies, distribution negotiations become slow, politically charged and legally contestable. For mechanisms open to third countries, this creates acute tension: the procurement agent may face legal or political mandates to prioritise domestic needs, undermining partner confidence in neutrality and reducing willingness to pre‑commit funds or accept exclusivity terms.
3. The “deliverability” gap: Large contracts do not guarantee delivery if upstream capacity, inputs, or regulatory compatibility fail. Even delivered goods can become “stranded” if participants lack agreed indemnity/compensation arrangements or regulatory reliance pathways, creating delays, expiry risk and politically toxic reallocation disputes.
Two-speed governance and ex ante decision rights
The failure modes described earlier share a common structural cause: decision making that remains committee‑driven at the point of execution. A minimum viable design for MedPPA would therefore separate ex ante strategic decisions from time‑critical execution:
Strategic board: Sets scope (product basket), participation tiers, financing architecture, and approves ex ante templates for allocation rules, indemnity models, regulatory reliance conditions, and exclusivity and penalty regimes. Decisions here can require consensus or qualified majority because they occur outside crisis conditions.
Procurement agent: A standing executive entity empowered to conduct tenders, sign framework agreements and execute emergency call-offs without reopening political negotiation once triggers are met. Its discretion is bounded by templates approved by the Board.
Crisis activation protocol: A predefined trigger set that transitions the mechanism from routine to crisis mode and automatically activates delegated authority, emergency procurement procedures and (where relevant) exclusivity and allocation rules.
To operationalise this architecture and prevent the governance paralysis that enables defection, the framework must specify in advance: product inclusion criteria and eligibility rules; volume commitments and call-off rules (whether volumes are indicative, firm or option-based); allocation methodology under scarcity (the allocation key and rules for exceptional cases); bypass penalties and enforcement mechanisms (how exclusivity is enforced, what triggers penalties, and what sanctions apply); and transparency and auditability requirements (what is disclosed, what remains confidential, and safeguards for supplier trust). Deferring these distributional decisions to crisis time guarantees allocation conflicts: a resilient MedPPA would resolve them ex ante, under calm conditions to preserve speed and cohesion when urgency arises.
Once third countries participate, the allocation question becomes acute: if the procurement agent is an EU body, third-country claims must be explicitly defined – as equal, subordinate, or separately tranched. The governance model would need to specify delegated execution authority, pre‑approved contractual templates, decision thresholds, dispute escalation procedures and ex-post accountability mechanisms. The detailed allocation architecture is presented in Annex 2.C, Section C.3.
Financing and legal framework
Where the policy objective is assured access rather than lowest price, MedPPA would need to purchase resilience through three complementary financing layers: framework contracting with national call-offs (for products with elastic supply), central guarantee capacity (to underpin advance commitments and reduce payment-risk pricing), and a surge option layer providing paid rights to call-off volumes within defined activation windows. The financing architecture would also need to address currency risk for non-euro participants and include an affordability channel for eligible Tier 2 partners.
The legal framework would require a treaty-level agreement defining eligibility tiers, opt-in commitments, and exclusivity provisions, combined with authorisation for the Procurement Agent to contract on behalf of participants under specified procurement rules. A pre‑established liability architecture, covering manufacturer indemnification, adverse event compensation, and jurisdictional handling of claims, is essential to prevent novel product procurement from stalling. The detailed financing architecture, legal base options, liability protocol, and currency risk analysis are presented in Annex 2.C, Section C.4.
Cost estimates, benchmarks, and value‑for-money justification
This section assesses the economic feasibility of MedPPA by quantifying the “resilience premium” – the incremental cost of purchasing security of supply rather than minimising short-run unit prices. The analysis defines the cost components of this premium, benchmarks them against existing multilateral mechanisms, and provides a counterfactual analysis against alternative security instruments. Finally, it outlines the value‑for-money justification required to satisfy fiduciary and procurement-law standards.
Definition and cost components
In standard procurement, value for money is often equated with the lowest unit price. In a security-of-supply context, this metric is flawed because it treats supply reliability as an exogenous variable. MedPPA would internalise this variable, explicitly pricing the reduction of non-delivery risk. The “resilience premium” is therefore not an inefficiency, but an insurance payment comprising three distinct cost drivers:
1. Capacity reservation fees: Annual or multi-year availability payments to manufacturers to maintain surge‑ready production lines (warm-base manufacturing) or inventory buffers. These payments decouple revenue from immediate volume delivery, ensuring capacity exists even when routine demand is low.
2. Option premia and minimum-revenue guarantees: Prepaid rights to call off volumes at pre‑agreed prices during defined trigger windows, with non-utilisation costs borne by the purchaser rather than the supplier.
3. Diversification costs: Accepting higher unit costs from geographically dispersed or backup suppliers to reduce single‑point-of-failure risk rather than consolidating purchasing with the lowest-cost provider.
Empirical evidence from health procurement suggests that resilience premiums typically range from the low single digits to approximately 20% above minimum-cost procurement. This variance depends on market concentration and the severity of the supply risk being hedged. To validate these estimates, this evaluation benchmarks MedPPA against established multilateral mechanisms that successfully integrate security premiums.
Empirical benchmarks from established global mechanisms confirm that resilience premiums in the low-to-mid single digits are consistently lower than the spot-market inflation, wastage costs, and emergency procurement premiums incurred when security-of-supply measures are absent. During the COVID‑19 pandemic, average medicine price increases following shortages exceeded 16% (HHS, 2023[59]) while UK PPE procurement recorded price increases of 166% for respiratory masks, with the Department of Health and Social Care writing off GBP 8.7 billion in PPE‑related losses (NAO, 2020[60]; Department of Health and Social Care (DHSC), 2022[61]).
The detailed empirical benchmarks – covering the PAHO Revolving Fund, Gavi AMC, AVAT, and crisis-period cost data – are presented in Annex 2.C, Section C.2.
Comparative cost-effectiveness: MedPPA vs. alternatives
The value for money of the resilience premium is best understood through counterfactual analysis. Table 2.7 below compares the structural costs and crisis performance of MedPPA against the four primary policy alternatives available to governments seeking supply security. The comparison is necessarily qualitative: no controlled study has compared crisis-period performance across jurisdictions employing different security-of-supply strategies for equivalent product baskets.
This framework suggests that MedPPA would occupy the efficient frontier in the policy space: it avoids the deadweight costs of stockpiling, the crisis-period exposure of price‑minimisation pooling, and the inequity of bilateral deals, while remaining fiscally viable across a broader product range than domestic autarchy permits.
This comparative assessment is derived from ex ante economic logic rather than controlled observation of crisis mechanisms. While comparable platforms demonstrate that pooled procurement can achieve security-of-supply objectives, direct empirical evidence quantifying the crisis-period return on investment for a resilience premium in a European context remains limited; the resilience premium should therefore be treated as a design parameter to be refined through pilot implementation. A rigorous evaluation framework for MedPPA should incorporate prospective counterfactual analysis: tracking the avoided cost of spot-market purchases during disruptions and the health-system value of uninterrupted supply against the cumulative premium paid during routine periods. Until such longitudinal data is available, the efficiency rankings in Table 2.7 reflect reasoned inference from incentive structures, i.e. that paying a capped insurance premium is superior to retaining potentially unlimited exposure to catastrophic market failure.
Regulatory and value‑for-money justification
Implementing a resilience premium requires navigating public procurement rules that have traditionally prioritised short-term price minimisation. This section outlines how MedPPA could operationalise security of supply within the EU public procurement legal framework currently in force, drawing on precedents from other critical sectors to establish a defensible value‑for-money proposition.
The most economically advantageous tender (MEAT)
The EU Public Procurement Directive (2014/24/EU) provides the necessary legal latitude for a MedPPA mechanism. Under Article 67, contracts are awarded to the “most economically advantageous tender” (MEAT), which contracting authorities may identify on the basis of price alone, of cost (for example, lifecycle costing) or of the best price-quality ratio. Contracting authorities are therefore not required to award on price alone. The best price-quality ratio explicitly permits the inclusion of qualitative criteria, including quality, technical merit and security of supply, provided they are i) linked to the subject matter of the contract, ii) do not confer unrestricted discretion on the contracting authority, and iii) are objectively verifiable and measurable (European Parliament and Council of the European Union, 2014[63]) Table 2.8 outlines an indicative weighting framework that translates “security of supply” into objectively verifiable award criteria. The weights are illustrative and are intended to show how MedPPA could operationalise security of supply within a MEAT framework rather than to reproduce an existing EU-wide procurement template. This approach is consistent with evidence from EU pharmaceutical procurement practice, where MEAT criteria allow price and non-price considerations to be combined, although price‑only awards remain common. A European Commission study found that MEAT criteria accounted for 24% of publicly listed pharmaceutical procurement procedures in the study countries over 2008‑2021, and that security of supply is already used as an award criterion in several countries, including through stockholding requirements, bank guarantees, delivery-time specifications, contract splitting and penalties for non-delivery.
Note: The weighting ranges are indicative and do not represent observed EU-wide averages. They illustrate a possible MedPPA tender design in which security of supply is treated as a core component of value for money. Actual weights would need to be calibrated by product category, market structure, supplier base and legal review, including against the Public Procurement Act proposed by the European Commission in September 2026 (see text).
Source: Based on Directive 2014/24/EU, Articles 67‑68, and Vogler, Salcher-Konrad and Habimana (2022[64]), Study on Best Practices in the Public Procurement of Medicines.
This weighting reflects the principle that security of supply can constitute a core dimension of value. By standardising these criteria, MedPPA would ensure that the resilience premium remains subject to competitive pressure: suppliers would compete not only on price, but also on their ability to provide reliable, deployable and auditable supply capacity. This approach aligns with Article 68 of the Directive, which authorises lifecycle costing methodologies that account for costs over the life cycle of a product, including costs associated with supply disruptions and the consequences of non-availability (European Parliament and Council of the European Union, 2014[63]).
The legal basis for this approach is in transition. On 9 September 2026, the European Commission proposed a Public Procurement Act, a regulation on public contracts and concessions that would repeal Directives 2014/23/EU, 2014/24/EU and 2014/25/EU (European Commission, 2026[65]). The proposal would make the best price-quality ratio the default award method, subject to a minimum weighting for quality criteria on an “apply or explain” basis, and would explicitly list security and resilience among the considerations that public buyers may reward. If adopted in that form, the proposal would reinforce the framework set out here: the indicative weights in Table 2.8 already place well over half of the award on non-price criteria. The analysis in this section is based on the Directive in force at the time of writing; the specific provisions cited would need to be mapped to the final text of the regulation once adopted, and the proposal may change substantially during the legislative process. Sectoral precedents and market-shaping as a “public good”
The value‑for-money approach proposed for MedPPA mirrors established practices in other security-critical sectors within the EU. Although the legal frameworks differ across sectors, the underlying policy principle is comparable: continuity of essential services may justify paying for standby capacity, redundancy and security attributes beyond the price of delivered units:
In the energy sector, Member States may use capacity mechanisms to remunerate electricity generators for maintaining reserve capacity that may not be used under routine market conditions (European Parliament and Council of the European Union, 2019[66]). Similarly, strategic gas storage obligations similarly allow costs above spot-market procurement to be incurred in order to ensure resilience during periods of heightened demand or supply disruption (Council of European Energy Regulators, 2024[67]);
In telecommunications and critical communications infrastructure, procurement in defence and security contexts may prioritise redundancy, continuity of service and control over sensitive infrastructure rather than price alone (European Parliament and Council of the European Union, 2009[68]).
Although the analogy is not legal equivalence, the relevant point is instead functional: in sectors where non-availability carries high social and economic costs, public authorities already recognise that value for money cannot be assessed solely by reference to the cheapest unit price. It must also account for the cost of maintaining capacity, reducing exposure to disruption and ensuring continuity of supply.
Beyond immediate crisis preparedness, the resilience premium would serve a structural economic function by correcting market failures in fragile pharmaceutical supply chains. For products characterised by thin margins and high manufacturing concentration, such as off-patent antibiotics and sterile injectables, aggressive price competition can accelerate supplier exit, underinvestment and capacity erosion. By offering multi-year commitments with predictable revenue streams, even at modest premiums, MedPPA could help stabilise markets, incentivise quality investment and support manufacturing diversification that individual tenders would be unlikely to achieve. This would represent a legitimate use of collective purchasing power, provided the mechanism is designed consistently with public procurement and competition law principles.
To satisfy oversight requirements, MedPPA’s governance framework would need to ensure:
Ex ante transparency: award criteria, the methodology for calculating the premium and allocation rules must be defined and disclosed prior to tendering, ensuring the mechanism is not used to shield uncompetitive suppliers.
Competitive process integrity: the process must remain open to multiple suppliers capable of meeting the security criteria; it cannot create closed or preferential arrangements without objective justification.
Performance verification: the intangible “product” being purchased (supply security) must be audited through objective performance metrics. This requires monitoring of reserved capacity, stress-testing of delivery schedules and penalties for failure to meet readiness standards.
Lifecycle cost documentation: reporting should calculate the total cost of ownership, demonstrating that the cost of the premium is offset by avoided crisis costs (e.g. spot-market inflation, emergency logistics) and reduced wastage from expired stockpiles.
In conclusion, the resilience premium is a functional prerequisite for converting pooled demand into secure supply. It represents a deliberate policy trade‑off: accepting modestly higher routine procurement costs in exchange for priority access, greater reliability and more stable pricing during supply shocks, when crisis-time price escalation can be severe and availability may not be assured at any price.
2.6.3. Quality assurance, logistics and the “single buyer / single approver” logic
Pooled procurement improves resilience only if it delivers products deployable in all countries, on time. For medical countermeasures, execution constraints typically arise downstream of the tender: quality assurance and batch release, import clearance and cold-chain logistics, and regulatory authorisation and liability prerequisites. This section specifies the operational conditions under which pooled contracting translates into real-world access, consistent with the tiered participation model and two‑speed governance architecture.
Regulatory operability and deployment readiness as opt-in conditions
For any product class included in pooled procurement, product-specific opt-in should be conditional on an operational “deployment readiness package” that combines: i) a defined regulatory pathway for rapid use, ii) the corresponding adverse‑event compensation/indemnification arrangements, and iii) agreed documentation and pharmacovigilance obligations. This avoids fast approval, slow deployment failure and ensures that accelerated release does not create unpriced or unmanaged liability exposure. Regulatory pathways should be specified ex ante for each product class and tier, using a limited set of standardised routes:
Lead-authority pathway: reliance on the designated lead authority decision for rapid deployment, with clearly bounded national administrative steps;
Emergency pathway: provisional acceptance arrangements for crisis-first-wave deployment, with completion of full national/WHO processes continuing post-deployment where required;
Mutual recognition / reliance agreements: defined recognition of inspections and authorisations across a compatible subset of participants.
WHO mechanisms (Prequalification and Emergency Use Listing Procedure (EUL)) could be used where appropriate, but the design should ensure that early-wave activation can rely on emergency acceptance triggers with post-hoc completion of fuller review processes where required.
Matching regulatory reality to participation tiers
The tiered approver model described below sets out the regulatory architecture that MedPPA would require to be operationally effective. Several elements, particularly Tier 1 reliance and Tier 2 accelerated processing, would require regulatory adaptations that do not exist under current frameworks, and the mechanisms for using non-nationally-authorised products remain complex and differ across Member States. The feasibility of these arrangements would need to be tested during the pilot phase before being assumed at scale.
The mechanism’s speed advantage depends on whether procurement centralisation is matched by a workable release model. A single uniform approver rule across heterogeneous participants is not feasible. A tiered approver model, consistent with participation tiers, would therefore apply:
Tier 1: reliance‑based single approver: for Tier 1 participants, the “single approver” logic is closest to full reliance: authorisation and (where relevant) batch release decisions of the lead authority are accepted, with minimal national formalities. This is the setting where pooled procurement can deliver maximal speed and harmonisation.
Tier 2: accelerated parallel processing: for Tier 2 participants, the feasible model is accelerated parallel processing anchored to a shared dossier and lead-authority assessment. Tier 2 participants would commit to i) accept the common technical dossier and quality documentation without reformatting, ii) run a time‑bounded national procedure triggered by lead-authority approval or emergency listing with pre‑defined decision deadlines, and iii) use common pharmacovigilance templates and signal-sharing. This would preserve sovereign decision making while preventing open-ended national reviews that strand pooled deliveries.
Tier 3: interoperability rather than unified release: for Tier 3 partners, the mechanism would prioritise documentation interoperability and aligned standards (shared specifications, Good Manufacturing Practice and Good Distribution Practice (GMP/GDP) recognition where feasible, common traceability and Quality Assurance (QA) documentation) rather than integrated contracting that presumes unified release. This would allow co‑ordinated procurement behaviour without assuming that regulatory sovereignty can be pooled.
This tiered approach also manages product heterogeneity. Where national requirements differ materially (e.g. mandatory language and labelling, pack configurations, traceability system compatibility), the mechanism could use regional lotting: procurement lots defined for participant subsets with compatible requirements, while preserving common supplier qualification and contracting templates across lots.
Quality assurance and supplier qualification
A credible pooled mechanism requires standing quality capacity that goes beyond paper compliance. This should include:
Supplier and site qualification: verification of GMP/GDP status and quality systems, relying on recognised inspections and mutual recognition where available, and standardising the evidence package required for all participating tiers.
Common documentation standards: harmonised requirements for certificates of analysis, batch documentation, stability data where relevant, and traceability identifiers, so the same evidence package supports release across participating jurisdictions, consistent with the tiered approver model above.
Quality incident protocols: predefined escalation procedures for deviations, quality signals and recalls, including who has decision authority and how information is shared across participants (with operational roles aligned to the Procurement Agent mandate set out under Governance, decision velocity and allocation authority, and national competent authority responsibilities).
Integrity and anti-counterfeit safeguards: chain-of-custody requirements and minimum traceability expectations for diversion-prone or high-value products.
In crisis mode, the quality model should preserve safety while enabling speed through reliance, standardised evidence, and rapid signal-sharing rather than duplicative pre‑distribution reviews in every jurisdiction. The detailed specifications for cold-chain and transport validation, distribution model design, and readiness-linked product selection including the principle that product-specific opt-in should reflect verified delivery capacity rather than nominal participation are presented in Annex 2.C, Section C.5.
2.6.4. Implementation challenges and risks
This section identifies the principal implementation risks that can prevent an “open-to-third-countries” MedPPA mechanism from functioning as an effective security-of-supply instrument in practice. The risks map directly onto the three scarcity failure modes outlined in Box 2.4: i) strategic defection and credibility gaps, ii) allocation conflict and fiduciary dilemmas, and iii) deliverability gaps where contract does not match capacity. The underlying implementation risk is therefore not that pooled procurement cannot be established, but that it becomes credible in routine periods yet brittle under scarcity, or administratively sound yet operationally non-deployable.
Market, industry and supplier dynamics
Supplier prioritisation and capacity allocation: under scarcity, suppliers allocate limited output across buyers based on margin, enforceability, payment credibility and operational simplicity. A pooled mechanism that seeks low prices without credible surge options may be deprioritised relative to buyers offering higher margins, simpler contracting or stronger enforceability. Even with large nominal volumes, manufacturers may discount pooled demand if they anticipate participant defection, delayed payments, or regulatory fragmentation.
This tension between buyer power and supplier engagement defines the mechanism’s design constraint. Where pooled procurement targets price minimisation on thin-margin products (off-patent antibiotics, generic essential medicines), it may accelerate market exit dynamics already observable in European generics markets (Panteli et al., 2024[69]). Where the mechanism instead contracts for security of supply – offering demand predictability, multi-year commitments, and option premia in exchange for capacity reservation – the incentive structure shifts: suppliers gain revenue certainty that individual national procurement cannot provide. The resilience premium is therefore not merely a fiscal question but the primary instrument for managing “monopsony risk”. This distinction should inform product-specific opt-in decisions: products amenable to security-of-supply contracting are better candidates than high-volume generics where price competition dominates:
Monopsony perception and strategic withdrawal: pooled procurement can be perceived as a monopsony that compresses margins. If the mechanism is framed primarily as price leverage rather than resilience contracting, suppliers may respond through i) reduced tender participation, ii) minimal volume offers, iii) prioritisation of other markets, or iv) gradual exit from low-margin segments. This risk is highest for generics, where extending a “lowest-price” logic across a broader coalition can accelerate consolidation and fragility dynamics already observed in key markets (e.g. antibiotics).
Overconcentration and single‑award fragility: large, pooled tenders can unintentionally concentrate purchasing into a small number of suppliers or sites, increasing systemic fragility to single‑point disruptions. Where feasible, multi‑award frameworks, diversification requirements, and performance‑linked allocation are needed to prevent the mechanism from “trading price for resilience”.
Information asymmetry on upstream constraints: contracts can fail if upstream inputs constrain production (API, vials, filters, fill – finish capacity). Suppliers may be unable to commit credibly to delivery schedules if their own sub-suppliers are constrained. Without transparency provisions and complementary market intelligence (including potential links to O.3.1 monitoring), the mechanism risks contracting on assumptions that later prove non-deliverable.
Contract non-performance and limited leverage at peak scarcity: in acute crises, legal remedies against non-performance can be slow, uncertain, or politically impractical. Unless contracts include operational performance instruments (service levels, escalation clauses, substitution rights), pooled procurement may have limited leverage once shortages materialise. This risk is partially mitigated by the financing architecture proposed in the previous section: central guarantee capacity and option structures convert legal remedies into prepaid economic leverage, strengthening supplier incentives to honour priority delivery commitments when supply is constrained.
Political economy and trust barriers
Bypass incentives and defection under pressure: the central political risk is that when scarcity peaks, participants, especially those with greater purchasing power, face domestic incentives to secure bilateral deals. This undermines pooled bargaining power and signals to suppliers that pooled volumes are not firm. Exclusivity choices matter: hard or functional exclusivity mechanisms can mitigate bypass risk, but only if the penalties are credible and materially costly (e.g. forfeiture of premia, loss of option rights, loss of priority access) and exceed the political benefits of defection for priority products.
Distributional conflict and allocation legitimacy: in crises, procurement becomes allocation. If allocation rules are ambiguous or perceived as unfair, coalition cohesion erodes quickly. “Open-to-third-countries” participation magnifies this risk: third countries may distrust EU-led allocation if they fear being crowded out, while EU actors may resist arrangements perceived to place non-EU claims on par with EU populations for scarce countermeasures. Without pre‑defined tranches/entitlements and clear crisis activation protocols, allocation negotiations can paralyse execution.
Trust and confidentiality constraints: Pooled procurement often requires sharing sensitive information on demand forecasts, stock positions, and consumption trends. Participants may withhold or distort information for strategic advantage (overstating need, understating capacity) unless governance and confidentiality rules are credible. Weak data discipline undermines demand aggregation and can generate over-ordering, misallocation, and supplier scepticism about the reliability of pooled demand.
Sovereignty costs, tier mobility limits, and the “flexibility trap”: MedPPA’s most valuable features – regulatory reliance, pooled liability models, binding opt-in and credible exclusivity – carry sovereignty and legal costs. Many potential partners may accept interoperability (Tier 3) but not deeper commitments (Tier 2), limiting the practical scope of “open-to-third-countries” participation. Overpromising broad Tier 2 participation risks creating an instrument that is formally inclusive but operationally narrow, or diluting enforcement and operability standards to accommodate incompatible sovereignty constraints. This “flexibility trap” can erode the very credibility conditions that make suppliers willing to reserve capacity.
Legal base options and institutional pathways
Four legal base options exist for establishing MedPPA: EU public health measures under Article 168(5) TFEU; economic emergency measures under Article 122 TFEU; an international agreement under Article 216/218 TFEU; or an intergovernmental agreement outside the EU Treaty framework. The most feasible pathway is a hybrid approach: an internal EU framework under Article 168(5) establishing governance and procurement rules for EU participants, combined with an international agreement under Article 216 enabling structured third-country participation. The detailed legal analysis of each option is presented in Annex 2.C, Section C.4.
Legal, competition and regulatory hurdles
Regulatory fragmentation and deployability gaps: even when procurement succeeds, products can be stranded if national authorisations, batch release, labelling, or traceability requirements diverge. The tiered “single approver” model (see Regulatory operability and deployment readiness, above) mitigates this, but Tier 2 and Tier 3 participation remains exposed to deployability delays unless accelerated reliance/parallel-processing pathways are agreed, operationalised in templates, and tested.
Liability and indemnification deadlocks: for vaccines and novel therapeutics, emergency contracting often hinges on indemnities and compensation schemes. If liability architecture is not pre‑established, suppliers may refuse to contract or require terms that some participants cannot accept. This can exclude third countries de facto even when the mechanism is nominally open, undermining both credibility and coalition discipline.
Procurement-law challenges and audit paralysis: cross-border procurement increases exposure to bid challenges, procedural disputes, and ex post audit scrutiny. In emergencies, a specific risk is ex post audit disallowance of accelerated procurement decisions (e.g. findings by supreme audit institutions) which can incentivise procurement agents to behave defensively, slowing execution. Emergency-ready timelines, pre‑approved templates, and a clear mandate for the Procurement Agent are therefore essential to reduce paralysis risk (see the governance and financing subsections above).
Competition-law uncertainty and information risks: pooled procurement can raise antitrust sensitivities if it is perceived to co‑ordinate prices or allocate markets beyond a clearly defined public-interest perimeter. Ex ante clarity, through statutory safe harbours or advance guidance from competition authorities (see Financing and legal framework, above) and strict information governance are needed to avoid legal uncertainty during activation periods.
International obligations and eligibility constraints: extending participation selectively across third countries can interact with international procurement obligations and non-discrimination principles. The mechanism therefore needs objective eligibility criteria and legally robust participation instruments to minimise avoidable exposure to challenge during crises, particularly where procurement access rights imply preferential treatment or priority delivery.
2.6.5. Impacts on consumers, industry and environment
Consumer impact
MedPPA would be designed to improve patient access to critical medical countermeasures during supply disruptions, when spot-market prices can exceed routine levels by 100% or more and availability cannot be assured at any price. By contracting for security of supply rather than solely minimising unit costs, the mechanism aims to stabilise patient-facing prices during crises and reduce the frequency of therapeutic substitutions driven by shortages rather than clinical appropriateness.
However, the resilience premium may translate into modestly higher routine‑period prices at the point of care, depending on national reimbursement and pricing policies. This trade‑off is explicit: accepting marginally elevated baseline costs in exchange for reduced price volatility and assured access during emergencies. The net consumer impact is therefore positive where supply disruptions are frequent or severe, and neutral-to-marginally-negative during extended periods of market stability. Monitoring of patient-level price impacts should be incorporated into the mechanism’s evaluation framework.
Industry impact
The mechanism’s effects on manufacturers vary by market segment:
“Innovative” manufacturers (patented products, new vaccines): The resilience premium model offers revenue predictability and reduced demand volatility. Multi-year capacity reservation agreements can support investment decisions that individual national tenders, subject to annual budget cycles and unpredictable off-take, cannot underwrite. For products where development/production costs are high and demand uncertain (e.g. pandemic vaccines, novel therapeutics), pooled procurement with credible volume commitments may be market-enabling rather than market-distorting.
Generic manufacturers and SMEs: Effects are more ambiguous. Where the mechanism prioritises security contracting over lowest-price tendering, smaller manufacturers capable of meeting quality and capacity standards may gain market access that price‑only tenders foreclose. However, where pooled buyer power is exercised primarily for price minimisation on thin-margin products, the risk of accelerating market exit, already observable in European off-patent antibiotic markets, is acute. Product-specific opt-in decisions should therefore distinguish between categories where security contracting is supplier-positive and those where price pressure dominates.
API and input manufacturers: To the extent that MedPPA would contract incorporate geographic diversification requirements or multi-site production clauses, the mechanism may stimulate investment in manufacturing capacity outside current concentration points. This may align with broader EU strategic autonomy objectives but would require that contract terms are predictable enough to support capital investment decisions.
Environmental impact
Pooled procurement at scale has countervailing environmental effects:
Logistics consolidation: Co‑ordinated procurement can reduce transport emissions through consolidated shipping, rationalised distribution networks and reduced emergency air freight during crises (when carbon-intensive expedited logistics are otherwise unavoidable).
Diversification costs: Geographic diversification of supply sources, a core resilience objective, may increase routine transport distances and associated emissions relative to single‑source procurement from lowest-cost producers.
Manufacturing standards: Procurement specifications can incorporate environmental criteria (GMP facilities meeting defined emissions standards, sustainable packaging requirements, cold-chain efficiency). However, such criteria must be balanced against supply security objectives; overly stringent environmental requirements that exclude otherwise‑capable suppliers could undermine the mechanism’s primary purpose.
Ultimately, the net environmental impact depends on contract design. The mechanism’s governance framework should include environmental criteria as permissible award factors under Article 67 of Directive 2014/24/EU (or the corresponding provisions of its successor), while ensuring that environmental objectives do not override supply security in crisis periods.
2.6.6. Product scope and alignment with the Union list of critical medicines
MedPPA’s product scope could logically draw on two existing reference frameworks. The Union list of critical medicines6 identifies medicines essential for EU health systems. A sub-set of this list targeting medicines withsupply vulnerabilities could provide a policy-validated starting point for product selection. In parallel, HERA is developing a Medical Countermeasures List under the MCM Strategy, covering medicines and medical devices necessary for preparedness and response to priority cross-border health threats – a scope that is distinct from the Union list and includes products that are not classified as critical medicines but are operationally essential during health emergencies. MedPPA’s product basket could draw on both lists, with selection governed by procurement suitability rather than automatic inclusion from either.
Three considerations would govern product-specific opt-in:
Products amenable to security-of-supply contracting: products where supply risk derives from market concentration, demand unpredictability or investment disincentives are strong candidates for MedPPA inclusion, whether they appear on the Union list, the MCM list, or both. For these products (e.g. selected essential antibiotics, anaesthetics, hospital injectables), the resilience premium model could address root causes of fragility by offering the revenue certainty that individual national procurement cannot provide.
Products where pooled procurement adds limited value: certain products face supply constraints that procurement mechanisms cannot resolve: patent-protected products with single‑source manufacturing (where the binding constraint is production capacity, not demand aggregation), or products where shortage drivers are primarily regulatory (GMP non-compliance, batch failures). For these categories, MedPPA would offer co‑ordination benefits but not transformative supply security; complementary instruments (stockpiling, manufacturing incentives) may be more appropriate.
Products requiring expedited inclusion during emergencies: medical countermeasures for emerging threats (pandemic vaccines, novel therapeutics) and products with cross-border interdependencies (diagnostic inputs, medical devices) may warrant inclusion through defined crisis-activation procedures even if not on either list at the time a threat emerges. The mechanism’s governance should permit expedited product addition through pre‑agreed criteria and fast-track decision procedures.
For operational purposes, MedPPA product selection should be informed by, but not mechanically derived from, either reference list. A mapping exercise would classify candidate products by: i) suitability for security contracting (high/medium/low); ii) value added from demand aggregation versus alternative instruments; and iii) tiered participation feasibility (which products are viable for Tier 1 integration versus Tier 2/3 interoperability).
2.6.7. Implementation timeline and critical path
Operational readiness for MedPPA would depend on sequential milestones across legal, institutional, and operational workstreams. The timeline presented in Table 2.9 below assumes a hybrid legal pathway (Article 168(5) TFEU for internal framework, Article 216 for third-country agreements) and distinguishes between achievable interim arrangements and full operational capability.
The binding constraints are i) legal framework adoption, which depends on EU legislative cycles and MFF timing; ii) procurement agent designation and resourcing, which requires institutional commitment and budget allocation; and iii) supplier engagement, which requires credible demand signals before manufacturers will commit capacity. Delays in any of these three areas cascade through subsequent phases. Pending full MedPPA operationalisation, the existing JPA framework might be used for expanded third-country participation on a product-specific basis, subject to current legal constraints.
2.6.8. Opportunities and strategic value
The preceding sections have established MedPPA’s design parameters, governance architecture, and implementation risks. This section synthesises these elements into a net assessment, identifying the conditions under which the mechanism would deliver value and the constraints that would bound its feasibility.
Opportunities and binding constraints
A threshold question for any net assessment is whether a new instrument is necessary. The EU already operates a Joint Procurement Agreement under the cross-border threats to health framework, and HERA has demonstrated operational procurement capacity for medical countermeasures. MedPPA’s rationale would therefore not rest on the principle of pooled procurement (which is already proven) but on whether extending participation to third countries generates resilience benefits that the existing EU-only framework cannot deliver.
Three such benefits are structurally unavailable under the current JPA. First, demand aggregation across a broader coalition reduces the negative externalities of uncoordinated procurement by neighbouring countries during shortages – competitive escalation, phantom demand and supply fragmentation. Second, inclusion of jurisdictions hosting critical upstream nodes – API manufacturers, fill-finish capacity, logistics corridors – could potentially enable supply-security commitments reflecting the actual geography of pharmaceutical production rather than the EU’s institutional boundaries. Third, pre‑agreed allocation rules across a wider coalition would reduce the risk that crisis-driven bilateral deals by non-pool countries destabilise supply for pool members. These benefits are conditional on sufficient participation depth and governance quality – conditions assessed below.
MedPPA would deliver greatest value when designed as an instrument for security of supply rather than simply a tool for price minimisation. However, each opportunity the mechanism would offer is paired with a binding constraint that determines whether value is likely to materialises in practice. Table 2.10 below maps these pairings across the mechanism’s core design dimensions.
The net assessment reveals a consistent pattern: MedPPA’s rationale is robust in principle but conditional in practice. Relative to the existing JPA, the mechanism would add greatest value where third-country participation would be deep enough to aggregate demand beyond the EU’s institutional boundaries, where supply chain geography would make non-EU nodes operationally critical, and where pre‑agreed cross-coalition allocation rules would prevent the destructive bilateral scramble that undermined equitable access during COVID‑19. Where these conditions were absent – fragmented participation, incomplete regulatory reliance pathways, or insufficient financial pre‑commitment – the mechanism would converge toward Tier 3 co‑ordination and information-sharing: useful, but not a step-change from what bilateral arrangements and the existing JPA already provide.
Effective demand aggregation under MedPPA would require reliable forecasting inputs. Objective 3 mechanisms, including demand forecasting models and consumption trend analysis, should be integrated into MedPPA’s operational planning. Specifically: i) routine‑period demand forecasts could inform framework agreement volume commitments and option sizing; ii) crisis-period demand projections (scenario-based) would calibrate surge capacity requirements and trigger thresholds; iii) post-crisis consumption data would validate forecast accuracy and improve future planning cycles. Without this integration, MedPPA would risk either over-committing (purchasing options that expire unused) or under-committing (insufficient surge capacity when needed).
Scope boundaries
The mechanism would achieve maximum strategic value where:
Supply becomes constrained during shocks and demand commitments can credibly purchase surge optionality;
Product specifications are harmonisable or can be segmented into compatible regional lots;
Regulatory and liability pathways can be operationalised through tiered participation; and
Participating jurisdictions demonstrate sufficient “sovereignty cost tolerance” – willingness to accept reliance‑based or accelerated parallel pathways and associated liability/compensation arrangements.
These conditions most commonly hold for vaccines and select medical countermeasures, as well as for a narrow set of essential medicines where supply vulnerability stems from concentrated manufacturing, thin margins, and quality-related disruptions. For these product classes, MedPPA would address a genuine market failure: uncoordinated national procurement cannot secure the supply commitments that pooled demand enables.
Conversely, the mechanism would offer limited value where:
Products are highly jurisdiction-specific due to labelling, pack configuration, or traceability constraints, with limited prospects for harmonisation;
Supply is highly elastic and shortages primarily reflect local distribution failures rather than upstream scarcity; or
Sovereignty constraints are absolute – constitutional requirements for national batch release or non-delegable customs authority that cannot be accommodated within tiered participation.
In these cases, value would be largely limited to Tier 3 interoperability regardless of supply conditions.
Comparative evaluation of institutional settings
MedPPA’s strategic value would depend on whether it is hosted by an entity possessing genuine procurement competence, legal authority to contract on behalf of participants, operational capacity for quality assurance and logistics (including the capacity to administer pre‑qualification requirements for GMP compliance and regulatory readiness), and political legitimacy to manage allocation disputes. Three configurations are plausible; Table 2.11 below assesses each against core feasibility criteria.
No single configuration dominates across all criteria: the EU-based procurement agent would be most achievable for a core coalition of closely aligned participants but would face legitimacy and legal constraints for broader third-country integration. A multilateral vehicle would offer greater flexibility for heterogeneous coalitions but require substantial institution-building. The interoperable network of existing platforms would sacrifice procurement integration for political feasibility, capturing co‑ordination benefits while preserving platform autonomy.
A phased approach may prove optimal: begin with the interoperable network to establish shared qualification standards and co‑ordinated tender windows, while developing the legal-operational architecture for deeper integration. This sequence builds demonstrated co‑operation before requiring institutional commitment, reducing the risk of ambitious frameworks that cannot attract sufficient participation.
Conditions for proceeding
Based on the net assessment, MedPPA should proceed to detailed feasibility work and pilot implementation only if the following conditions can be satisfied:
1. Critical mass secured: participation commitments from jurisdictions representing >50% of target market volume for the pilot product basket, with binding opt-in rather than indicative interest.
2. Legal-operational architecture validated: confirmation that (i) the proposed legal base supports the intended exclusivity and crisis-activation provisions; (ii) liability and compensation frameworks are achievable within participating jurisdictions’ constitutional constraints; and (iii) regulatory reliance pathways – including unilateral reliance arrangements for Tier 2 participants – are operational for the pilot basket.
3. Fiscal sustainability confirmed: multi-year budget commitment or ring-fenced funding mechanism sufficient to sustain resilience premium contracting through at least one procurement cycle, avoiding reversion to price minimisation that would undermine supplier confidence.
4. Governance arrangements agreed: host institution mandate confirmed; two‑speed governance model (routine/crisis) operationalised through pre‑approved templates and delegated authority; allocation rules and prioritisation hierarchy legally binding on participants.
5. Hinge‑point review scheduled: commitment to structured evaluation at Month 24, assessing whether participation, legal architecture, and operational readiness meet thresholds for scaled implementation. If conditions are not met, explicit fallback to Tier 3 interoperability or mechanism termination.
These conditions are demanding but not prohibitive. The existing EU Joint Procurement Agreement demonstrates that pooled procurement is legally and operationally achievable within the EU framework. MedPPA’s added value would lie in extending this proven model to deliver capabilities that the JPA structurally cannot: security-of-supply contracting with a resilience premium, crisis activation with delegated authority, and third-country participation that reflects the actual geography of pharmaceutical supply chains. The conditions above provide a testable framework for determining whether these extensions are achievable – and an explicit fallback to Tier 3 interoperability if they are not.
2.7.1. Purpose, scope and strategic context
This section presents the evaluation of “RescPool” – a proposed EU-led stockpiling mechanism open to third countries, designed to provide pooled reserves that could be rapidly deployed to bridge temporary supply disruptions affecting critical medicines and selected medical devices. RescPool would be distinct from existing EU medical countermeasure reserves under rescEU, which focus primarily on CBRN threats, personal protective equipment and emergency equipment. Its added value would lie in extending reserve architecture to pharmaceutical products whose cross-border deployment depends on jurisdiction-specific regulatory handling.
The policy problem RescPool would address is not “shortages” in the abstract, but the acute governance gap that emerges when clinically critical products become unavailable faster than markets, procurement systems and regulatory processes can adjust. In such episodes, health system harm is driven by the interaction of three constraints:
First, patients and providers may need replacement supply within days, while restoring normal supply can take weeks or months. For essential medicines, diagnostics and key ancillaries, even well-functioning procurement channels may require time to identify alternative suppliers, qualify substitute products, reroute distribution or scale production. RescPool would therefore operate as a catastrophe‑layer buffer, preserving service continuity during the interval in which other corrective mechanisms are necessarily slower.
Second, uncertainty can convert a local disruption into a wider scarcity episode. Under incomplete information and high accountability pressures, actors may adopt defensive strategies, including over-ordering, parallel stock-building and export restrictions. These behaviours can amplify a supply shock across jurisdictions. A pooled reserve can dampen this dynamic only if it is perceived as credible, rules-based and rapidly deployable.
Third, regulatory friction can prevent available stock from becoming usable supply. Unlike generic emergency goods, medicines are often jurisdiction-locked through labelling, patient information and dispensing rules. A stockpile can therefore fail even when inventory exists, because the product cannot legally be dispensed in the receiving jurisdiction within standard pharmacy law. For RescPool, the core policy challenge would therefore be deployable inventory – stock that is both physically available and legally dispensable at the point of need.
An existing EU-level solidarity instrument already addresses part of this problem. The EMA Voluntary Solidarity Mechanism (VSM), operational since late 2023, enables Member States facing shortages to request assistance from peers through EMA’s Executive Steering Group on Shortages and Safety of Medicinal Products (MSSG). The VSM was established as a voluntary arrangement agreed by the MSSG in June 2023, while the proposed Regulation forming part of the pharmaceutical package would provide the first formal legal basis and procedural framework for the mechanism. It was activated seven times in 2024 and had been used 13 times in total by early 2025 (European Commission, 2025[70]).
The VSM is, however, a last-resort co‑ordination channel rather than an operational reserve. It works by asking marketing authorisation holders active in responding Member States’ territories to redirect supply to the requesting country. It does not draw on pre‑positioned public inventory, does not operate through pre‑agreed allocation rules, and does not itself resolve regulatory deployability constraints such as serialisation handling, labelling adaptation and dispensing authorisation.
RescPool would therefore operate at a different level. Where the VSM co‑ordinates voluntary sharing of existing national or market-held stocks, RescPool would maintain dedicated pooled reserves with pre‑agreed activation triggers, allocation rules and regulatory passport arrangements enabling cross-border deployment. The VSM would remain available as a complementary channel for products or situations outside RescPool’s scope.
The proposed RescPool is therefore best understood as a short-term continuity instrument: it would be most valuable when a disruption is temporary and a time‑limited release can prevent immediate service disruption. However, the mechanism must also be evaluated against a realistic edge case: some apparently temporary disruptions, particularly in essential generics, may reveal deeper structural fragilities, including thin markets, API manufacturing concentration, recurrent quality-remediation problems or market exit. In such cases, the reserve would function as a strategic timeout, maintaining continuity while parallel mechanisms are activated to restore a viable supply route.
Scope boundaries: Keeping the mechanism operable
RescPool’s feasibility would depend on strict scope discipline. A pooled reserve is operationally viable only if it is treated as a managed capability with clear boundaries:
Time horizon: RescPool should be designed for short to medium disruptions (days to weeks). It is not intended to solve long-duration structural shortages by itself; rather, it buys time for other measures.
Products: a narrow basket anchored to the Union list of critical medicines, filtered for stockpiling feasibility – shelf life compatible with reserve rotation, and cross-border deployability under emergency regulatory frameworks. Non-drug medical countermeasures (CBRN agents, PPE, emergency equipment) fall outside RescPool’s primary scope, as these are already covered under rescEU’s existing reserve framework. Products where cross-border regulatory barriers cannot be resolved within crisis-relevant timeframes through emergency authorisation frameworks would be directed to “keep-warm” or surge procurement arrangements rather than physical stockpiles, regardless of clinical importance.
Function: RescPool is designed to function not just as warehousing, but as the full operational chain – readiness auditing, activation protocols, distribution handoffs and replenishment rules. This would include “regulatory passport” arrangements (emergency packaging solutions) and manufacturer indemnification protocols that make stock legally and commercially usable.
Governance scope: RescPool would act as a backstop. Access should therefore follow a deductible principle: participants retain responsibility for maintaining minimum national buffers, with RescPool triggered only when those buffers are demonstrably overwhelmed or when verified supply failure is present.
RescPool would fill a distinct gap in a toolkit of other available or potential mechanisms:
MedMIS (monitoring/information-sharing – O.3.1): would diagnose stress but would not aim to provide physical inventory.
MedPPA (pooled procurement – O.3.2): would aggregate demand but operate on procurement timelines too slow for acute gaps.
The VSM: co‑ordinates voluntary redistribution of available stocks but lacks pre‑positioned assets, activation rules or regulatory deployability infrastructure.
Product eligibility framework: From the Union list of critical medicines to stockpile basket
The proposed RescPool product basket would derive from, but would not be coterminous with, the Union list of critical medicines. The Union list establishes clinical criticality; RescPool would apply additional filters for supply-chain vulnerability, stockpiling feasibility and cross-border deployability. This dual-governance approach would ensure that eligibility decisions respect therapeutic priorities, while reflecting both supply-side risk and the logistical realities of reserve management. Not all Union list products would be suitable for pooled reserves; conversely, some non-Union list products may warrant inclusion based on their fit with reserve‑based risk mitigation. (Council of the European Union, 2026[16]).
RescPool’s product governance should also be understood in the context of the proposed Critical Medicines Act (CMA) (Council of the European Union, 2026[16]). The proposed CMA (Article 20) provides a framework for Member States’ contingency stock requirements, establishing that any obligations on supply-chain actors to hold stocks must be proportionate, transparent and must not result in negative impacts on other Member States’ supply. These national contingency stocks would form the preparedness layer upon which RescPool’s deductible principle would build: participating states would maintain national buffers in accordance with CMA provisions, with RescPool operating as the pooled catastrophe layer above national reserves, adding cross-border regulatory passport and rapid-deployment infrastructure that national stocks do not by themselves provide. The EU Stockpiling Strategy sets out the broader co‑ordination framework within which RescPool would sit.
RescPool would apply five criteria to determine which products could belong in the mechanism’s basket:
1. Clinical criticality and substitutability, RescPool would prioritise Union list categories with the highest clinical consequence of stockout (e.g. essential anaesthetics, critical care medicines and narrow-therapeutic-index drugs).
2. “Bridge value”: products where short-to-medium-term reserve release (days to weeks) would provide meaningful clinical continuity. Products where disruption recovery typically exceeds reserve duration (e.g. if the designed reserve covers 4‑6 weeks but alternative supply requires 3+ months to activate) would be lower priority for physical stockpiling. Products with very short disruption windows, where commercial channels typically resolve shortages within 24 hours, would likewise receive low priority.
3. Stockpiling feasibility: products should be amenable to reserve holding, considering shelf life, storage requirements, obsolescence risk and regulatory deployability:
a. Favourable: stable small-molecule medicines with 2+ year shelf life, standard storage conditions and broad regulatory harmonisation.
b. More costly and operationally demanding: products with shorter shelf lives (12‑24 months), cold-chain or controlled-temperature requirements, or moderate obsolescence risk (e.g. certain biologics, selected vaccine ancillaries). These remain stockpilable, but require active rotation management, dedicated storage infrastructure and higher per-unit holding costs, making them candidates for hybrid arrangements combining limited physical reserves with vendor-managed inventory or rolling commercial stock agreements.
c. Unfavourable: products with <12‑month shelf life or high platform-specificity (some diagnostics) – these may be better suited for “keep-warm” arrangements (i.e. contractual commitments to maintain surge manufacturing capacity rather than holding physical inventory).
4. Cross-border deployability: products should permit rapid distribution across countries under stress conditions, including regulatory interoperability (deployable under emergency authorisation frameworks) and logistical harmonisation (pack sizes, labelling and serialisation permitting cross-border distribution).
5. Supply chain vulnerability: products where the probability of disruption is elevated due to concentrated manufacturing (single‑source API, limited fill-finish sites), or documented recurrent shortages.
Other products for potential inclusion
Three product categories may warrant inclusion despite absence from the Union list, subject to separate governance review:
1. Medical countermeasures for emerging threats: products not currently on the Union list of critical medicines but relevant for pandemic or CBRN response (antivirals, antidotes, specific therapeutics). HERA’s Medical Countermeasures List, under development as part of the MCM Strategy, would provide the reference framework for this category. Some MCMs are already held under rescEU’s CBRN strategic stockpile; for these products, RescPool’s added value would lie in pharmaceutical-specific reserve management.
2. Enabling products (diagnostic consumables and medicine administration devices): universal diagnostic inputs (e.g. nucleic acid extraction reagents, viral transport media) and devices critical for medicine administration (specific syringes, infusion sets) may warrant inclusion because supply disruption in either category can neutralise medicine stockpiles. Both would score high on bridge value and stockpiling feasibility (long shelf lives, standard storage), and their cross-platform compatibility avoids lock-in risk.
Third-country access: Resilience through reciprocity
RescPool’s “open to third countries” feature is evaluated here as a resilience design choice; the rationale is threefold:
1. Risk pooling and geographic smoothing: supply disruptions may be correlated but not perfectly synchronised across countries. A wider pool would improve the probability that usable surplus capacity exists somewhere in the system, especially for logistics capacity and certain product categories with uneven demand profiles.
2. Operational interdependence with non-EU nodes: critical supply chains often depend on non-EU jurisdictions for inputs, packaging, fill-finish capacity, and logistics corridors. Structured third-country participation would strengthen emergency pathways by enabling pre‑agreed corridors, hosting and reciprocal logistics support, reducing the risk that cross-border frictions become the binding constraint.
3. Reciprocity as a credibility condition: the third-country dimension would be politically and operationally viable only if access rights are linked to pre‑committed reciprocal contributions – financial contributions or access bonds, hosting capacity, surge logistics commitments, and alignment with the mechanism’s operational standards. Without this, free‑rider dynamics would predictably erode the willingness of core participants to pool meaningful quantities.
Third-country eligibility and allocation rules would need to be pre‑cast, not negotiated during a crisis. RescPool would require a tiered participation model – distinguishing full participants, associate partners and pre‑authorised corridor arrangements – with ring-fencing protocols and reciprocal commitments defined ex ante. Eligibility criteria would include regulatory compatibility (capacity to implement passport arrangements for relevant product categories), demonstrated operational capability, and willingness to accept binding reciprocal obligations (financial contributions or access bonds, hosting capacity, exercise participation). Section 2.7.3 specifies the tier definitions, access modalities and legal instruments in detail. Where pre‑commitment would not be impossible, co‑operation could still occur through swap lines or corridor agreements, but such arrangements should not be counted as core deployable capacity.
2.7.2. Instrument concept
Core mechanism
The proposed RescPool would be a health-sector-led reserve mechanism, governed by health preparedness authorities but designed to leverage existing civil protection logistics infrastructure, notably rescEU’s deployment capacity and the Emergency Response Co‑ordination Centre (ERCC), for physical distribution. Conceptually, it would function as a catastrophe‑layer buffer sitting above national minimum preparedness: participating countries would retain responsibility for routine buffers and procurement, while RescPool would provide an additional layer of protection when disruption severity, verified supply failure or systemic allocation breakdowns overwhelm domestic capacity. Health-sector governance would be essential because pharmaceutical reserves, unlike commodity or emergency equipment stockpiles, require regulatory decisions (marketing authorisation, batch release, pharmacovigilance, dispensing authorisation) at every stage of the deployment chain. Civil protection infrastructure would provide the logistics backbone, with release and deployment decisions sitting with health authorities.
RescPool’s defining feature would be treating “stockpile” as an operational capability, not a warehouse. The mechanism would combine a managed reserve (physical and non-physical elements), pre‑defined activation and prioritisation rules designed to function under time pressure, and a legal-operational architecture intended to make cross-border deployment dispensable in practice. In feasibility terms, RescPool would stand or fall on whether it can avoid the “white box” failure mode – inventory that cannot legally be dispensed, and whether it could make release decisions quickly enough to preserve the bridging window.
At minimum, RescPool would require three institutional building blocks:
1. Regulatory passport: pre‑agreed emergency stock-keeping units (SKUs) and dispensing conditions that would allow cross-border use of reserve products. This would include multilingual patient information solutions, such as e‑leaflets, agreed emergency derogations for labelling and leaflets, and operable traceability and serialisation recognition across participating jurisdictions. This regulatory-operational architecture would distinguish pharmaceutical reserves from commodity stockpiles, where products are more easily interchangeable and can therefore be deployed with fewer jurisdiction-specific requirements.
2. Rapid activation governance: an activation model designed to prioritise speed, objectivity and accountability. Release decisions would be taken by a technical authority operating under pre‑defined triggers, while political oversight would focus on replenishment, scope, financial contributions and ex-post accountability rather than case‑by-case approval of each deployment.
3. Lastly, a stock rotation and capacity-maintenance model: lifecycle mechanisms designed to ensure that the medicines held in reserve remain usable when needed. Expiry is one of the greatest operational risks for pharmaceutical reserves, since medicines that expire on the shelf represent both wasted public expenditure and a false sense of security. For products suitable for physical stockpiling, the primary tool would be structured rotation through first-in-first-out (FIFO) partnerships. Under such arrangements, wholesalers, hospital pharmacies or humanitarian distributors would absorb reserve stock before expiry into routine distribution channels, while the reserve would be replenished with fresher inventory at regular intervals. This would keep the reserve operational without relying on periodic write‑offs. For products where physical warehousing is inefficient – for example, because of short shelf lives, high obsolescence risk or platform-specific technologies – the mechanism would instead use capacity-reservation contracts, paying manufacturers to maintain surge production readiness rather than holding finished products.
Lessons from comparator reserve systems
RescPool’s design draws on four strategic reserve systems that have been tested under real-world disruption conditions: the International Energy Agency’s oil security system, the EU gas storage framework, the US Strategic National Stockpile and rescEU. None of these mechanisms is a direct analogue for pharmaceutical shortage management. Oil, gas, emergency equipment and medical countermeasures differ substantially from medicines in terms of regulation, substitutability and deployment conditions. They are nevertheless relevant because they illustrate institutional design features that are central to any pooled reserve mechanism: pre‑committed contributions, objective activation rules, delegated release authority, allocation protocols and replenishment obligations. The detailed institutional profiles are presented in Annex 2.D, Section D.1.
Four converging principles emerge across these mechanisms. First, credibility requires pre‑committed contribution formulae and objective activation triggers, rather than crisis-time negotiation. Second, rule‑based allocation helps prevent politicised distribution when demand exceeds available stock. Third, regular exercises and formulary reviews are essential to prevent a gap between nominal stockpile levels and actual deployable capacity, a vulnerability exposed during COVID‑19 when some reserves proved inadequate in practice despite appearing sufficient on paper. Fourth, solidarity obligations need default rules for countries or participants that lack sufficient reserves of their own.
Pharmaceutical reserves, however, face regulatory complexity that is largely absent in these comparator domains. Each medicine is a regulated product requiring jurisdiction-specific authorisation, batch release, traceability, serialisation and dispensing approval. The operational benchmark from emergency reserve systems (i.e. rapid dispatch once release has been authorised) is logistically achievable within existing civil protection frameworks. For medicines, however, the full deployment chain also includes regulatory and clinical steps before dispatch can be authorised. RescPool would therefore require a dedicated pharmaceutical architecture that does not yet exist.
Operational cycle: Stock, monitor, activate and replenish
RescPool’s operational cycle would be designed to keep the reserve continuously deployable – both legally and logistically – rather than optimised only for peak crises. The cycle would comprise four linked functions (Figure 2.2):
Stock: RescPool would maintain inventory through a deliberately mixed architecture: a narrow set of physically held items where certainty dominates, complemented by rolling components (vendor-managed or distributor-integrated stock) and, for categories where finished-goods warehousing is fragile, keep-warm surge arrangements (capacity reservation and input buffers). Stocking is governed by standardised quality requirements, batch/lot control and audited storage conditions. For medicines and certain diagnostics, stocking also includes the creation of passport-ready Stock Keeping Units -SKUs (or the capacity to convert into such SKUs rapidly), because non-passport inventory cannot be assumed deployable across borders.
Monitor: monitoring in RescPool would extend beyond market intelligence to cover operational readiness through two layers: readiness monitoring (fill-rate against target levels, shelf-life profile, quality audits, and “deployability checks” including time‑to-pick, time‑to-dispatch, cold-chain integrity and documentation completeness); and trigger monitoring (objective indicators of need, ideally drawing on or interoperating with other information systems – verified national usable stock levels, confirmed supply failure duration, validated demand surge indicators and logistics disruption evidence). A key feasibility point is that readiness monitoring should include legal readiness: periodic verification that passport arrangements remain valid across jurisdictions (serialisation rules, e‑leaflet acceptance, emergency dispensing protocols), so that the reserve does not become legally immobilised through regulatory drift.
Activate: RescPool activation would need to be fast enough to matter in a days-to-weeks bridging window. The mechanism therefore requires a pre‑defined authority chain and time‑bound decision standards. In practice, a technical secretariat executes release decisions under “red button” logic: pre‑specified thresholds trigger automatic authorisation for defined quantities and products, subject to prioritisation rules. Political oversight is retained but channelled to ex-ante rule‑setting and ex-post review, not case‑by-case release approvals vulnerable to paralysis under stress.
Replenish: replenishment should be treated as integral to feasibility: a viable RescPool would require a pre‑committed replenishment financing rule, contracting pathways that can replace released volumes without long delays, and an after-action loop updating basket composition, trigger calibration and readiness assumptions. For rolling components, replenishment includes performance enforcement of rotation partners (e.g. absorbing near-expiry stock) to keep wastage within political tolerance.
2.7.3. Third-country participation tiers: Eligibility, reciprocal commitments and access modalities
Because RescPool would be open to third countries, participation must be structured to remain operational under crisis time pressure and protect solidarity credibility among core participants. As a key feasibility principle: access rights should be pre‑committed and linked to reciprocal obligations – there is no realistic time to negotiate entitlement during a two‑week bridging crisis. A tiered model would therefore be required:
Tier 1 – Full participants: jurisdictions able to integrate into RescPool’s operational and legal architecture (including passport-ready dispensing conditions for relevant items). Commitments include financial contribution and/or hosting, acceptance of activation triggers and prioritisation rules, participation in exercises, and compliance with minimum national buffer requirements (deductible principle). These countries should have full access under the standard trigger and prioritisation framework, including predictable coverage for defined categories.
Tier 2 – “Associate” participants: third countries with high interdependence and demonstrated capability to meet operational requirements for specified product categories. Commitments include “skin in the game” via access bonds (financial deposits or in-kind equivalents), hosting and logistics pre‑commitments, and regulatory alignment for passport SKUs in the categories they seek access to. Access should be conditional and category-bounded, with ring-fencing protocols defined ex ante.
Tier 3 – Partner arrangements: RescPool should avoid discretionary exceptional access as an operational promise because it is too slow to execute. Instead, Tier 3 should be defined as pre‑authorised partner arrangements – swap lines, humanitarian corridors, logistics support – with explicit templates and conditions. Where pre‑authorisation is not feasible, co‑operation remains possible but should be treated as diplomatic support rather than deployable RescPool capacity.
Across tiers, feasibility would hinge on two cross-cutting design requirements: regulatory passport compatibility for any product expected to cross borders rapidly, and liability and pharmacovigilance clarity, including manufacturer indemnification provisions triggered upon custody transfer, without which suppliers will rationally resist participation.
2.7.4. Stress-testing on tracer archetypes
RescPool’s feasibility depends on whether it performs under distinct “product – shock” combinations. Three tracer archetypes are used here as stress tests: they illuminate where pooled reserves would deliver high marginal value, where they would risk operational failure (expiry, mismatch, legal immobility) and what design constraints would follow for basket selection, activation logic and replenishment.
The detailed stress test across the three tracer archetypes is presented in Annex 2.D, Section D.2. The assessment yields four operational imperatives for basket design.
First, RescPool should prioritise products where a short reserve release (days to weeks) materially prevents patient harm – favouring acute‑use hospital injectables over chronic treatments where buffer time is longer. Second, products should be eligible for pooling only if they possess a “regulatory passport” enabling cross-border dispensing without destination-market delays. Third, the reserve architecture should be matched to product characteristics: vendor-managed inventory with commercial FIFO rotation for high-volume generics; stockpiling of fungible inputs (reagents, consumables) rather than finished devices for fast-evolving technologies; and keep-warm capacity reservation rather than physical stock for products with high expiry risk. Fourth, RescPool’s success should be measured by operational speed (authorisation-to-patient delivery time), not inventory volume alone.
2.7.5. Operating model and design specifications
This section specifies how RescPool would operate in practice. The focus is on the design features that determine whether pooled reserves remain deployable under stress: i) matching stockpiling modalities to product characteristics, ii) embedding activation and allocation rules that function within a days-to-weeks bridging window, and iii) ensuring that products are not only available, but legally dispensable and logistically deliverable across borders.
Stockpile architecture options
RescPool would treat stockpile architecture as a portfolio allocation problem, matching medical products to reserve modalities based on their lifecycle risk profiles. Box 2.5 below provides the decision heuristic and the conditions for assigning products to modes.
The proposed RescPool’s design problem is not whether to stockpile, but where to stockpile: different product classes fail for different reasons (expiry, platform mismatch, cold-chain fragility, legal immobility). The reserve should therefore be built using a simple assignment rule that matches each item to the modality that minimises its dominant operational risk. A practical allocation heuristic (item-level) is:
Hold physically when availability certainty dominates and lifecycle risk is low: long shelf-life, stable SKU, predictable handling, and high bridge value (i.e. a short release meaningfully prevents harm).
Use rolling/vendor managed inventory when wastage risk dominates but routine turnover is high: predictable demand, feasible FIFO rotation through normal channels, and standardisable quality governance.
Use keep-warm when obsolescence/expiry or correlated-shock failure dominates: fast technology cycles (e.g. diagnostics), short shelf-life biologics or categories where warehousing concentrates waste and where commercial call-off may fail under systemic crises.
Two preconditions before any item enters the pooled layer are:
Deployability: can the item be legally dispensed cross-border under pre‑agreed emergency conditions? If not, it should remain nationally managed; and
Sustainability: is there a funded lifecycle path (rotation partner, keep-warm fee or replenishment rule) that prevents predictable expiry-driven waste?
A hybrid architecture would only be credible if RescPool can demonstrate high performance on passport coverage (legal deployability), wastage (rotation performance) and surge deliverability under stress (keep-warm enforceability).
RescPool’s architecture should comprise three elements:
1. Physical holding: A narrow set of passport-ready items would be held as physical inventory to maximise immediate availability and reduce dependence on crisis-time market liquidity. RescPool could use:
a. centralised hubs, where standardisation and Quality Assurance concentration are priorities; and/or
b. distributed hubs, where redundancy and last-mile speed are priorities.
2. Rolling reserves and vendor-managed inventory (VMI): For high-throughput medicines where expiry risk dominates, RescPool should use rolling holdings under contracts that preserve emergency call rights while rotating stock through routine channels. This modality is feasible only if RescPool has a contracted rotation engine (FIFO absorption at defined shelf-life thresholds) with enforceable Key Performance Indicators.
3. “Keep-warm” surge arrangements: Where warehousing concentrates waste or where product cycles are fast, RescPool should rely on keep-warm contracting rather than simple call-off. These arrangements pay for surge capability (idle readiness, critical input buffers, logistics commitments) and are designed to remain credible under correlated shocks.
In practice, each modality would need to be operationalised through measurable deliverability standards: readiness audits for physical holdings, rotation performance for rolling holdings and enforceable surge deliverability for keep-warm arrangements.
Allocation and activation governance
Allocation governance would be RescPool’s binding constraint. The mechanism would only work if participants trust that access will be granted within the bridging window – days, not weeks. This would require trading political discretion for speed, while embedding accountability in pre‑defined rules:
The trigger: activation cannot wait for committee deliberation. RescPool should rely on a compact set of objective thresholds that function as hard “tripwires”. When these thresholds are breached, the mechanism moves automatically:
Clinical continuity: Verified risk of interruption to essential services (ICU, emergency care).
Supply failure: Confirmed disruption duration exceeding the bridging window (>7 days with no replenishment).
Stock depletion: “Usable days of coverage” falling below critical thresholds (e.g. <3 days).
These triggers are designed to be binary and measurable, removing the ambiguity that breeds delay.
Prioritisation rules would need to be pre‑specified and grounded in clinical criticality and proportionality:
Clinical priority: Essential services receive first allocation (continuity-of-care criteria), not “first come, first served”;
Tiered entitlements: Access is strictly bounded by pre‑negotiated commitments – Tier 1 full access, Tier 2 category-bounded, Tier 3 pre‑authorised only;
Ring-fencing: Solidarity is protected by contractual access bonds – financial deposits or in-kind commitments that forfeit if participants impose export bans during crises.
To outrun the bridging window, RescPool should rely on a “red button” protocol: a standing technical secretariat, combining civil protection operations with medicines regulatory expertise, would hold delegated authority to execute releases when triggers are hit. Service standards would be rigid: authorisation within 24 hours for pre‑defined categories, with quantities determined by algorithmic rules (not bargaining). In sum, governance would be the binding institutional constraint on reserve viability: without credible pre‑commitment, established through clear triggers, prioritisation frameworks, and rapid decision timelines, the instrument would risk either paralysis (delayed deployment) or reputational failure (perceived unfairness).
Comparative governance benchmarks
RescPool’s governance architecture could draw on operational lessons from comparable stockpiling and emergency deployment mechanisms. The closest precedents – the EU rescEU medical reserve, the US Strategic National Stockpile and COVAX – show that rapid deployment is achievable within existing civil protection architecture (rescEU and the SNS both achieve sub‑12‑hour dispatch from a release decision), that pharmaceutical reserves face cross-border regulatory requirements these mechanisms do not, and that committee‑based crisis allocation, as in COVAX, predictably fails to keep pace with supply shocks. The full comparator profiles – including the IEA oil security system and the EU gas storage framework – are set out in Annex 2.D, Section D.1.
Across these benchmarks, a consistent governance principle emerges: rule‑making and rule‑application must be institutionally separated. The proposed “red-button” concept operationalises this separation: pre‑defined triggers and prioritisation rules enable technical release authority to act within hours, while political oversight focusses on whether the rules themselves remain appropriate.
Logistics and technical operability
Even with inventory and governance, RescPool would fail if products cannot be delivered and legally dispensed. Operational design must integrate i) storage and custody integrity, ii) distribution interfaces, and iii) regulatory/traceability operability that prevents the “white box” failure mode.
Storage, cold chain, custody and recall readiness: RescPool would require harmonised readiness standards: validated storage conditions (including cold chain), periodic audits, batch/lot control and recall-ready documentation. Custody transition rules would need to specify responsibility at each stage (host → mechanism → receiving country) including pharmacovigilance roles during the bridging window. A pre‑negotiated Product Liability Protocol, including manufacturer indemnification upon custody transfer, would be a prerequisite.
Distribution interfaces and last-mile delivery: RescPool should pre‑agree handover points and responsibilities for onward distribution (national depots, hospital networks, public health channels). It would require surge transport arrangements, including cold-chain logistics and, where third-country movements are relevant, customs facilitation protocols.
Interoperability, traceability and regulatory alignment: RescPool’s design would need to include a Regulatory Passport protocol to ensure cross-border dispensability: emergency SKUs (e.g. multilingual packs), operable serialisation/traceability handling under emergency conditions and clear pharmacovigilance responsibilities during deployment.
2.7.6. Expected effects
This section sets out what the proposed RescPool would be expected to achieve, how those effects would arise and how performance should be monitored. A central premise is that RescPool’s benefits would be conditional: the instrument would only shift outcomes and behaviour if the design prerequisites are in place – most importantly i) hybrid-by-default architecture with lifecycle sustainability, ii) rapid activation with credible prioritisation, and iii) regulatory passport deployability that prevents “white box” inventory.
Causal pathways: Bridge supply, panic dampening, logistics substitution and solidarity externalities
Bridge supply: RescPool’s primary effect would be to convert pooled inventory and surge capability into days-to-weeks of clinical continuity during disruptions. When essential medicines stock out, patient harm escalates in hours; procurement systems adjust in weeks. RescPool would convert pooled inventory into clinical continuity until slower mechanisms (re‑sourcing, manufacturing scale‑up, regulatory transfer) take effect. This is not merely “having stock” but preserving life‑sustaining care while slower mechanisms (re‑sourcing, manufacturing scale‑up, regulatory transfer) come online. The pathway succeeds only when two conditions align: the reserve must release within the bridging window (red-button speed) and the product must be legally dispensable upon arrival (regulatory passport). Without this operational precision, inventory becomes a warehouse, not a “lifeline”.
Panic dampening: Beyond clinical continuity, RescPool would reshape strategic behaviour under scarcity. In supply shocks, uncertainty triggers defensive over-ordering – hospitals hoard, countries ban exports and parallel procurement amplifies scarcity. A credible pooled backstop could interrupt this spiral by changing the “calculus of fear”: when actors believe a rules-based reserve will deliver within days, the incentive to over-stockpile diminishes. Credibility hinges on objective triggers, predictable allocation rules, and proof of deployability through exercises and past performance. However, this pathway requires empirical validation: pharmaceutical markets differ from commodity reserves (petroleum, grain) in regulatory complexity and information asymmetry, and no systematic evidence yet demonstrates that stockpile announcements reduce defensive ordering during medicine shortages. RescPool’s credibility effect should be monitored through procurement pattern analysis before and after deployment, credibility perception surveys, and stress-test exercises that assess whether known reserve availability modifies ordering behaviour. If panic-dampening effects do not materialise, RescPool would still deliver value through direct bridge supply and logistics substitution, but behavioural-spillover expectations should be adjusted accordingly.
Logistics substitution: RescPool’s value would include the ability to substitute for logistics failure – mobilising surge transport, cold chain and distribution co‑ordination when commercial channels are disrupted. This pathway is strongest for time‑sensitive goods and for countries or regions where domestic logistics capacity becomes a binding constraint under stress. It would depend on pre‑agreed interfaces with national distribution channels and exercise‑tested standard operating procedures for pick-pack-dispatch, custody transfer and last-mile handover.
Solidarity externalities (“spillover control”): By providing a credible pooled backstop, RescPool would reduce incentives for unilateral export restrictions and “beggar-thy-neighbour” competition during regional shocks. This effect would be strongest when third-country participation is pre‑committed (via access bonds and tiered entitlements) rather than crisis-negotiated, ensuring that solidarity is structured. Conversely, if the pool is perceived as vulnerable to free‑riding or political rationing, it would accelerate fragmentation rather than preventing it.
These pathways would form a cascade: bridge supply would prevent immediate harm, panic dampening would preserve market function, logistics substitution would overcome infrastructure failure and solidarity externalities would prevent regional contagion. But this cascade would only flow if the mechanism is operationally credible: a reserve that is slow to activate, legally immobile or politically contested can worsen behaviour by signalling insecurity rather than reassurance.
Target outcomes and monitoring framework
RescPool should be monitored as an operational capability, not as an inventory stocktake. The detailed performance monitoring framework – covering four domains (speed, product correctness, legal deployability, and lifecycle sustainability) with specific KPIs, measurement methods, and target ranges – is presented in Annex 2.D, Section D.3. The guiding assessment principle is that high nominal stock levels are not meaningful if authorisation and dispatch cannot meet the bridging window; speed and legal deployability are the binding metrics, not volume.
2.7.7. Risks: Moral hazard, crowd-out and political rationing
RescPool’s potential benefits would be counterbalanced by predictable political economy risks that, if unmanaged, could erode the mechanism’s viability. The most immediate danger would be moral hazard: if participants perceive the pool as substituting for national buffers, they may rationally underinvest in domestic preparedness, crowding out national capacity and degrading RescPool into a routine supply channel – a clinically and financially unsustainable outcome. To prevent this, access must be governed by a deductible principle, conditioning drawdown eligibility strictly on evidence that national minimum buffers were maintained prior to the shock.
Beyond incentive structures, the reserve would face acute operational threats related to inventory lifecycle and regulation. Without a rigorous rotation engine, medical stockpiles inevitably accumulate expired or obsolete inventory, inviting reputational scandals that undermine future funding. Addressing this requires hybrid architectures that use enforceable FIFO rotation agreements and “keep-warm” industrial reservations rather than static warehousing.
Even more critical is the risk of “white box” immobility, where visible stock cannot be dispensed due to divergent labelling, serialisation or language requirements. This failure mode – possessing the physical asset but lacking the legal right to use it – is particularly damaging to public trust. Consequently, a “regulatory passport” protocol should be a hard condition for feasibility.
Finally, the mechanism should withstand the pressure of political rationing during simultaneous shocks. If allocation relies on consensus-based negotiation in the heat of a crisis, resulting delays will exceed the bridging window, causing the reserve’s credibility to collapse. Avoiding such paralysis requires delegation of “red-button” activation authority based on pre‑agreed clinical criteria, alongside strict ring-fencing for third-country commitments to prevent free‑rider dynamics. These risks are manageable, provided RescPool is constructed not as symbolic stockholding, but as a rules-based operational capability with enforceable governance.
Operationalising the deductible principle
Access to RescPool would be conditioned on demonstrated national preparedness – a “deductible principle” that preserves incentives for domestic readiness and prevents the mechanism from crowding out national buffers. Participating states would be required to maintain minimum national buffers (calibrated to product criticality and national health system size) verified through periodic audits. Enforcement would operate through graduated consequences rather than binary exclusion, and a humanitarian override would apply for catastrophic events affecting countries with otherwise compliant records. The detailed design of the deductible framework – including product-specific requirements, audit mechanisms, enforcement escalation, and anti-gaming provisions – is presented in Annex 2.D, Section D.4.
2.7.8. Legal architecture and feasibility
RescPool would operate at the intersection of EU civil protection competence, pharmaceutical regulatory requirements and third-country co‑operation arrangements. Its policy rationale is consistent with the EU Stockpiling Strategy, which recognises the need to strengthen EU-level preparedness through better co‑ordinated reserves for critical goods, including medical countermeasures and essential supplies. The existing rescEU framework, established under Decision 2019/420, provides the closest institutional precedent: it already enables EU-financed strategic reserves with defined deployment procedures and operational readiness requirements. On this basis, a RescPool-type mechanism could potentially be developed within, or alongside, existing civil protection architecture.
However, the principal feasibility constraint would not be the creation of reserve capacity as such, but pharmaceutical law operability. Medicines differ from generic emergency goods because deployment depends on jurisdiction-specific requirements for emergency dispensing, labelling and leaflets, serialisation and traceability under the Falsified Medicines Directive, and pharmacovigilance responsibilities across jurisdictions. For third-country participation, bilateral or multilateral agreements would be required to define eligibility, contribution modalities, access rights, liability arrangements and customs facilitation protocols. These arrangements would need to be agreed ex ante, since negotiating them during a crisis would undermine the purpose of the mechanism.
While this report does not provide a full legal analysis of the applicable EU and national legal bases, Annex 2.D, Section D.5 provides an indicative mapping of the main legal and operational interfaces that would need to be resolved for RescPool to function, including civil protection competence, pharmaceutical regulatory requirements and third-country participation instruments.
Operational legal feasibility
Medicines and other regulated products must be deployable and dispensable across jurisdictions. RescPool’s functionality would depend on a pre‑negotiated “regulatory passport” ensuring that cross-border deployment is not obstructed by national pharmacy, labelling or dispending rules. This protocol would need to define emergency SKUs compatible with Tier 1 jurisdictions, incorporating multilingual patient information solutions (such as QR-coded e‑leaflets) and harmonised serialisation handling. Without such arrangements, pooled inventory could be physically available but administratively undispensable.
Because RescPool would be open to third countries, legal feasibility would also depend on standard participation instruments defining eligibility, commitments and access rights. Tier 2 participation should be governed through ex-ante agreements specifying contribution modalities, whether financial or in kind; applicable product categories; access entitlements; ring-fencing rules; audit rights; and dispute‑resolution arrangements. Tier 3 participation should be limited to pre‑authorised partner arrangements – such as corridors, swap lines or logistics support – rather than discretionary crisis-time access.
Similarly, industry participation would require a further layer of legal certainty. RescPool should therefore include a comprehensive Product Liability Protocol specifying when and how liability transfers following custody changes between the host, the mechanism and the receiving country; who holds pharmacovigilance responsibilities during the bridging window; how recalls and field safety corrective actions are executed across borders; and who bears recall costs and compensation risks where applicable.
Political feasibility: Member State buy-in, third-country interest and solidarity constraints
Political feasibility would depend on whether participants trust that RescPool can i) deliver access rapidly, ii) allocate fairly under stress, and iii) avoid free‑rider dynamics. Two design choices would be particularly decisive:
1. First, deductible access would be needed to protect national preparedness incentives. RescPool should function as a catastrophe layer, not as a substitute for national buffers. Access should therefore be conditional on demonstrating that minimum national preparedness obligations were met ex ante – for example, minimum buffer policies, routine stockholding standards or equivalent readiness requirements – and that a verified supply failure exists. This would preserve incentives for domestic preparedness and reduce the risk that pooled reserves become a routine supply channel.
2. Second, third-country credibility would require reciprocal and bounded access rights: participation by third countries would only be politically feasible if access entitlements are linked to defined commitments and governed by tiered rules. In practice, this could imply ex-ante access bonds, either financial deposits or in-kind equivalents, calibrated to the level of expected entitlement. Without such safeguards, Member States may be reluctant to pool meaningful quantities, reducing the mechanism to symbolic levels.
These arrangements would be designed in alignment with the EU Stockpiling Strategy, which sets out the broader co‑ordination framework for Member States’ material preparedness for crises (European Commission, 2025[71]). Above all, RescPool’s operational credibility would require that release decisions are not subject to crisis-time political bargaining. Political oversight would remain essential, but it should be concentrated on ex-ante rule‑setting (product basket, triggers, entitlements) and ex-post review (replenishment, audit, lessons learned).
Operational readiness (institutional roles, logistics capacity, quality assurance)
Operational feasibility would represent the principal binding constraint for the mechanism. To function effectively, RescPool should be designed not as a static repository but as a standing capability with clearly defined institutional roles and exercise‑tested procedures. A minimum viable division of labour would begin with a central Operational Secretariat acting as the executive function, responsible for maintaining readiness, executing “red-button” releases and managing performance monitoring. This must be paired with a specific Regulatory Interface function to maintain the Regulatory Passport, ensuring that serialisation and traceability remain operable while co‑ordinating pharmacovigilance and recall readiness.
At the physical level,host states and logistics operators would serve as the execution arm, responsible for storage maintenance, cold-chain integrity, and the immediate “pick-pack-dispatch” cycle. Their efforts should interface seamlessly with receiving jurisdictions, which would be required to define specific handover points, manage last-mile distribution, and participate actively in after-action reviews. To ensure integrity across this distributed network, RescPool would require harmonised quality assurance standards covering storage validation and batch control. Crucially, for “rolling” and “keep-warm” components, auditability should extend to physically verifying rotation performance and actual surge capacity.
Implementation pathway (minimum viable pilot, phased scale‑up, stress-testing exercises)
RescPool should be implemented through a gated pathway that would resolve hard legal-operational constraints before procurement and expand scope only when performance is evidenced through exercises and audits. The sequencing presented in Table 2.13 below is designed to avoid two predictable failure: i) building inventory that is legally immobile and ii) scaling a reserve whose activation and lifecycle governance have not been proven under realistic time pressure.
2.7.9. Impacts on consumers, industry and environment
Consumer impacts
RescPool’s primary consumer benefit would be continuity of access to essential medicines during supply disruptions. When critical products become unavailable, patients face immediate harm: treatment interruptions, therapeutic substitutions with inferior alternatives, delayed procedures and, in acute cases, preventable morbidity or mortality. RescPool’s bridging function would convert days-to-weeks of stockout into continued availability, directly protecting patient welfare. Secondary consumer effects are more nuanced:
Price stability: by dampening panic-driven demand spikes, RescPool could moderate crisis-period price escalation. This effect is strongest where shortages would otherwise create grey-market or parallel-import price inflation.
Equity of access: pooled reserves with pre‑defined allocation rules could improve equity compared to crisis-time competition where larger or better-connected health systems secure scarce supply at the expense of smaller systems. However, tiered participation would create differentiated access; Tier 2/3 participants with subordinate or conditional access rights may face allocation disadvantage during simultaneous multi-country shortages.
Industry impacts
RescPool’s potential effects on industry would vary by mechanism component:
Keep-warm capacity reservation: manufacturers receiving availability payments would benefit from revenue predictability and reduced demand volatility. For products where routine‑period demand does not justify maintained capacity (e.g. pandemic countermeasures, niche essential medicines), keep-warm contracts could be market-sustaining, preventing capacity exit that would otherwise occur. This would be particularly relevant for SME manufacturers who cannot cross-subsidise low-volume products.
Rolling inventory partnerships: FIFO rotation arrangements (where wholesalers or distributors absorb near-expiry reserve stock into commercial channels) would create operational burden but also commercial opportunity. Manufacturers would benefit from stable demand; distributors would gain access to predictable supply. However, rotation partnerships would require contractual discipline: unenforced rotation commitments would create expiry losses that would ultimately fall on the public purchaser.
Physical stockpile procurement: direct government procurement for reserves would represent additional demand, benefiting manufacturers. However, if reserves are perceived as displacing commercial sales rather than adding to them, the net demand effect would be neutral. The moral hazard mitigation (deductible access requiring national buffers) would be designed to prevent crowd-out, but industry perception of net demand effect would influence willingness to participate.
Regulatory passport requirements: emergency SKU development (multilingual packaging, e‑leaflets, serialisation adaptations) would impose compliance costs on manufacturers. However, standardised templates would reduce burden compared to bespoke country-by-country arrangements. For products already sold across multiple EU markets, the marginal regulatory cost would be modest; for single‑market products, it might be more significant.
Environmental impacts
RescPool would have countervailing environmental effects:
Storage and cold-chain energy consumption: physical stockpiles, particularly for cold-chain products (vaccines, biologicals), require energy-intensive storage. European benchmarks for refrigerated distribution warehouses indicate operational emissions of approximately 40‑45 kg CO2 per m² of floor area per year (use phase), equivalent to about 0.01‑0.02 kg CO2 per m³ per day at typical building heights. Ultra-low-temperature (ULT) storage at ‑70 to ‑80°C is substantially more energy-intensive: product-level benchmarks indicate electricity consumption on the order of 5‑20 kWh per unit per day, corresponding to several kg CO2e per m³ of internal volume per day under a representative European grid mix (PCAF European Buildings Database, 2022[72]; IIR, 2021[73]). Published lifecycle assessments of COVID‑19 mRNA vaccines suggest that cold-chain cooling contributes roughly 0.01‑0.2 kg CO2 per administered dose, depending on temperature regime and logistics configuration. A substantial reserve of temperature‑sensitive products would therefore generate meaningful ongoing emissions.
Wastage and disposal: despite rotation mechanisms, some expiry-driven wastage is inevitable. Disposal via incineration or controlled destruction (the standard methods for pharmaceutical waste) carries non-negligible environmental costs, including air pollutants, residual contamination and the release of active pharmaceutical ingredients into the environment. WHO guidance notes that pharmaceutical waste incineration can release particulate matter, dioxins and furans, and toxic metals where combustion controls are inadequate (WHO, 2017[74]). Recent reviews further identify improper pharmaceutical disposal as a contributor to water contamination, antimicrobial resistance and air pollution (Xu et al., 2025[75]). These impacts should be monitored and incorporated into lifecycle environmental assessments of RescPool’s stockpile architecture. The hybrid model’s emphasis on keep-warm manufacturing capacity over large physical reserves would partially limit wastage exposure but does not eliminate it.
Emergency logistics efficiency: RescPool’s pre‑positioned distribution capacity may actually reduce net emissions during crises relative to improvised emergency logistics. Air freight can be on the order of 25‑100 times more carbon‑intensive per tonne‑kilometre than maritime transport; emergency responses that depend heavily on airlift and redundant uncoordinated shipments are therefore substantially more carbon-intensive than pre‑planned, consolidated transport chains (IPCC, 2014[76]). Freight decarbonisation analyses identify optimised routing, shipment consolidation, and modal shift away from air as the principal levers for reducing logistics emissions. Surge transport arrangements designed with these criteria from the outset could embed such efficiencies structurally, rather than treating them as post-crisis improvements.
Procurement specifications: EU public procurement rules and Green Public Procurement (GPP) guidance explicitly permit contracting authorities to apply environmental criteria and lifecycle costing – through best price-quality ratio and cost-based award criteria rather than price alone (see Section 2.6.2) – in pharmaceutical and medical product tenders. Several European health systems, including Nordic countries and France, have already integrated sustainability criteria covering supplier environmental performance, production waste, and packaging into pharmaceutical procurement, demonstrating that environmental requirements can coexist with supply-security objectives. RescPool contracting could build on these precedents by incorporating criteria such as supplier emissions standards, sustainable packaging, and cold-chain efficiency requirements, provided they remain secondary to security of supply.
2.7.10. Overall assessment and recommendations
RescPool would be feasible and potentially high-impact as a health-sector-led instrument for bridging temporary supply disruptions, provided it is designed as an operational capability rather than a symbolic stockholding initiative. Its comparative advantage within the Objective 3 toolkit is distinctive: it is the only instrument that could deliver physical continuity within a days-to-weeks window, where monitoring (MedMIS) would anticipate shortages but not provide supply, and procurement re‑sourcing (MedPPA) would operate on timelines too slow to prevent immediate service disruption.
However, feasibility would be conditional. For regulated products, the critical risk is the “white box” failure mode: inventory that is physically present but legally immobile due to packaging, patient information, serialisation and dispensing constraints. Pooled reserves also predictably fail when activation becomes committee‑driven at the point of execution, or when lifecycle governance is underpowered. RescPool’s net value would therefore hinge on three pillars:
1. Regulatory passport deployability: emergency SKUs and traceability operability across participating jurisdictions;
2. “Red-button” activation and pre‑commitment allocation governance; and
3. Industrialised sustainability: hybrid architecture with enforceable rotation and keep-warm contracting where appropriate.
Where these pillars are established, RescPool could reduce patient-level harm, dampen destabilising behavioural responses (panic buying, export restrictions), and provide a structured platform for credible third-country engagement based on reciprocity. Where they are absent, the mechanism would risk becoming an expensive reserve that cannot deploy under stress.
Preconditions and “hinge points”
The “hinge points” listed in Table 2.14 below should be treated as pass/fail criteria for operational viability: they translate the feasibility assessment into testable conditions that can be verified through templates, audits and exercises.
Table 2.15 presents a comparative assessment of the three instruments proposed to manage supply and enhance the security of supply.
Note: Ratings consolidate the assessments set out in each instrument fiche (MedMIS scorecard, Table 2.6; MedPPA net assessment, Table 2.10 and conditions for proceeding; RescPool overall assessment and hinge‑point checklist, Table 2.14). They are not a new analysis. The three instruments are independently viable; the table compares them on common dimensions and does not imply a mandatory bundle.
References
[88] AfricaCDC (n.d.), African Pooled Procurement Mechanism (APPM).
[40] AMIS (2017), AMIS brochure, http://www.amis-outlook.org/amis-about/en/ (accessed on 16 February 2026).
[41] AMIS (2016), Five-year progress report of the Global Food Markey Information Group, AMIS: IG/10.
[46] Årdal, C. et al. (2021), “Supply chain transparency and the availability of essential medicines”, Bulletin of the World Health Organization, Vol. 99/4, pp. 319-320, https://doi.org/10.2471/blt.20.267724.
[24] Blot, E. (2024), Sourcing critical raw materials through trade and cooperation frameworks, Institute for European Environmental Policy, https://ieep.eu/wp-content/uploads/2024/03/Sourcing-critical-raw-materials-through-trade-and-cooperation-frameworks-IEEP-2024.pdf (accessed on 3 February 2026).
[44] Brockhaus, J. and M. Kalkuhl (2014), “Can the Agricultural Market Information System (AMIS) help to reduce food price volatility?”, https://doi.org/10.22004/AG.ECON.170391.
[10] Cernat, L. (2023), “The Art of the Mini-Deals: The Invisible Part of EU Trade Policy”, European Centre for International Political Economy (ECIPE), Brussels.
[82] Cernuschi, T. et al. (2011), “Advance market commitment for pneumococcal vaccines: putting theory into practice”, Bulletin of the World Health Organization, Vol. 89/12, pp. 913-918, https://doi.org/10.2471/BLT.11.087700.
[13] Cimini, F. et al. (2024), “Exploring New Interregional Opportunities for Pharmaceutical Supply Chains. The Potential of Mercosur Countries to Advance the EU’s Global Health Strategy”, Stiftung Wissenschaft und Politik 5, https://doi.org/10.18449/2024C05.
[9] Claussen, K. (2022), “Trade’s Mini-Deals 315 (2022), Available at SSRN: https://ssrn.com/abstract=3836909”, Virginia Journal of International Law, Vol. 62/2, pp. 315-382.
[67] Council of European Energy Regulators (2024), “ACER-CEER study on the impact of the measures included in the EU and National Gas Storage Regulations”, https://www.ceer.eu/wp-content/uploads/2024/04/ACER-CEER-Consultancy-study-on-the-impact-of-the-measures-included-in-the-EU-and-National-Gas-Storage-Regulations.pdf (accessed on 12 February 2026).
[16] Council of the European Union (2026), Proposal for a Regulation of the European Parliament and of the Council laying a framework for strengthening the availability and security of supply of critical medicinal products as well as the availability of, and accessibility of, medicinal products of common interest, and amending Regulation (EU) 2024/795 - Letter to the Chair of the European Parliament Committee on Public Health, https://data.consilium.europa.eu/doc/document/ST-11059-2026-INIT/en/pdf (accessed on 10 August 2026).
[5] Critical Medicines Alliance (2025), Strategic report of the critical Medicines Alliance, Critical Medicines Alliance, Brussels, https://health.ec.europa.eu/document/download/3da9dfc0-c5e0-4583-a0f1-1652c7c18c3c_en?filename=hera_cma_strat-report_en.pdf (accessed on 3 February 2026).
[53] Dayan, M. et al. (2024), “The future for health after Brexit”, Nuffield Trust.
[85] de Bengy Puyvallée, A. and K. Storeng (2022), “COVAX, vaccine donations and the politics of global vaccine inequity”, Globalization and health, Vol. 18/1, https://doi.org/10.1186/s12992-022-00801-z.
[61] Department of Health and Social Care (DHSC) (2022), “Annual Report and Accounts 2020-21”, UK Government / Department of Health and Social Care.
[39] Drechsler, D. (2023), “AMIS – enhancing food market transparency and policy coordination”, Rural 21 02, p. 13.
[3] Dufour, G. et al. (2025), “Targeted Trade Deals As a New Form of Global Value Chain Governance”, in Global Value Chain Development Report 2025 - Rewiring GVCs in a changing global economy, https://doi.org/10.30875/9789287085641c015.
[30] EFPIA (2025), Annual Regulatory GMP/GDP Inspection Survey 2024 Data, European Federation of Pharmaceutical Industry Associations, https://www.efpia.eu/media/fq4p0zxy/efpia-2024-reg-inspection-survey-v1-public-version.pdf (accessed on 25 March 2026).
[50] EMA (2023), “Methodology to identify critical medicines for the “Union List of critical medicines””.
[52] Espitia, A., N. Rocha and M. Ruta (2020), “A pandemic trade deal: Trade and policy cooperation on medical goods”, in Revitalising Multilateralism Pragmatic Ideas for the New WTO Director-General, Centre for Economic Policy Research.
[2] European Commission (2026), Critical medicines Act, https://health.ec.europa.eu/medicinal-products/critical-medicines-act_en (accessed on 24 March 2026).
[11] European Commission (2026), EU trade relationships by country/region, Negotiations and agreements, https://policy.trade.ec.europa.eu/eu-trade-relationships-country-and-region/negotiations-and-agreements_en (accessed on 24 March 2026).
[27] European Commission (2026), Opening procedures at EMA to non-EU authorities (OPEN) framework | European Medicines Agency (EMA), https://www.ema.europa.eu/en/partners-networks/international-activities/multilateral-coalitions-initiatives/opening-procedures-ema-non-eu-authorities-open-framework (accessed on 24 March 2026).
[65] European Commission (2026), Proposal for a regulation on public contracts and concessions, https://single-market-economy.ec.europa.eu/publications/proposal-regulation-public-contracts-and-concessions_en.
[22] European Commission (2026), Strategic projects on critical raw materials gain momentum in second selection round for potential funding and faster permitting, https://single-market-economy.ec.europa.eu/news/strategic-projects-critical-raw-materials-gain-momentum-second-selection-round-potential-funding-and-2026-01-19_en (accessed on 24 March 2026).
[21] European Commission (2025), Annex to the Commisison Decision recognising certain critical raw material projects located in third countries and in overseas countries or territories as Strategic Projects under Regulation (EU) 2024/1252 of the European Parliament and of the Council, European Commission, Brussels, https://single-market-economy.ec.europa.eu/document/download/60c576a5-435e-43e6-83de-c81f3652259b_en?filename=C_2025_3491_1_EN_annexe_acte_autonome_part1_v3.pdf (accessed on 24 March 2026).
[20] European Commission (2025), Annex to the Commission Decision recognising certain critical raw material projects as Strategic Projects under Regulation (EU) 2024/1252 of the European Parliament and of the Council.
[70] European Commission (2025), “COMMISSION STAFF WORKING DOCUMENT Summarising evidence supporting the legislative proposal laying a framework for strengthening the availability and security of supply of critical medicinal products as well as the availability of, and accessibility of, m…”, https://health.ec.europa.eu/document/download/a0d79c14-3702-459e-864b-67eb3bca6896_en?filename=mp_swd-2025-263_act_en.pdf (accessed on 18 February 2026).
[71] European Commission (2025), EU stockpiling strategy: Boosting the EU’s material preparedness for crises.
[15] European Commission (2025), Proposal for a Critical Medicines Act, https://health.ec.europa.eu/publications/proposal-critical-medicines-act_en (accessed on 24 March 2026).
[51] European Commission (2024), Assessment of the supply chain vulnerabilities for the first tranche of the Union list of critical medicines: technical report.
[56] European Commission (2023), Crisis Management - HERA, https://health.ec.europa.eu/health-emergency-preparedness-and-response-hera/crisis-management_en (accessed on 16 February 2026).
[17] European Commission (2023), Team Europe Initiative on manufacturing and access to vaccines, medicines and health technologies in Africa, https://international-partnerships.ec.europa.eu/policies/team-europe-initiatives/team-europe-initiative-manufacturing-and-access-vaccines-medicines-and-health-technologies-africa_en (accessed on 24 March 2026).
[89] European Commission (2023), Team Europe Initiative on manufacturing and access to vaccines, medicines and health technologies in Africa - International Partnerships, https://international-partnerships.ec.europa.eu/policies/team-europe-initiatives/team-europe-initiative-manufacturing-and-access-vaccines-medicines-and-health-technologies-africa_en (accessed on 17 February 2026).
[58] European Commission (2022), European Commission Union Registry - EU Emissions Trading System (EU ETS), https://climate.ec.europa.eu/eu-action/carbon-markets/eu-emissions-trading-system-eu-ets/union-registry_en (accessed on 17 February 2026).
[19] European Commission (2021), Partnership on manufacturing vaccines, medicines and health technologies and strengthening health systems in Latin America, https://international-partnerships.ec.europa.eu/policies/global-gateway/partnership-manufacturing-vaccines-medicines-and-health-technologies-and-strengthening-health_en (accessed on 24 March 2026).
[78] European Commission (2020), Comfort letter: coordination in the pharmaceutical industry to increase production and to improve supply of urgently needed critical hospital medicines to treat COVID-19 patients, https://competition-policy.ec.europa.eu/antitrust-and-cartels/legislation/coronavirus_en (accessed on 22 January 2026).
[93] European Commission (2019), rescEU - European Civil Protection and Humanitarian Aid Operations, https://civil-protection-humanitarian-aid.ec.europa.eu/what/civil-protection/resceu_en (accessed on 17 February 2026).
[6] European Commission (n.d.), Critical Medicines Alliance - Public Health - European Commission, https://health.ec.europa.eu/health-emergency-preparedness-and-response-hera/overview/critical-medicines-alliance_en (accessed on 29 April 2026).
[47] European Commission et al. (2021), Future-proofing pharmaceutical legislation: study on medicine shortages, https://data.europa.eu/doi/10.2875/211485 (accessed on 16 February 2026).
[62] European Court of Auditors (2025), Critical shortages of medicines EU measures were of added value, but structural problems remain, https://www.eca.europa.eu/ECAPublications/SR-2025-19/SR-2025-19_EN.pdf (accessed on 17 February 2026).
[91] European Parliament and Council of the European Union (2025), Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL Amending Regulation (EU)2017/1938 as regards the role of gas storage for securing gas supplies ahead of the winter season, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52025PC0099&qid=1750954213781 (accessed on 17 February 2026).
[66] European Parliament and Council of the European Union (2019), Regulation (EU) 2019/943 of the European Parliament and of the Council of 5 June 2019 on the internal market for electricity (recast).
[63] European Parliament and Council of the European Union (2014), Directive 2014/24/EU of the European Parliament and of the Council of 26 February 2014 on public procurement and repealing Directive 2004/18/EC.
[68] European Parliament and Council of the European Union (2009), Directive 2009/81/EC of the European Parliament and of the Council on the coordination of procedures for the award of certain works contracts, supply contracts and service contracts by contracting authorities or entities in the fields of defence and security.
[55] European Parliamentary Research Service (2023), The European public health response to the COVID-19 pandemic: lessons for future cross-border health threats, European Parliamentary Research Service (EPRS), https://www.europarl.europa.eu/RegData/etudes/STUD/2023/740218/EPRS_STU(2023)740218_EN.pdf (accessed on 16 February 2026).
[45] FAO (2012), “Proceedings of the Ministerial Meeting on Food Price Volatility”.
[42] FAO (2011), “Improving Food Market Transparency and Policy Response for Food Security: The Agricultural Market Information System A Scoping Report”, https://www.fao.org/fileadmin/user_upload/amis/docs/AMIS_scoping_report.pdf (accessed on 11 February 2026).
[23] Financial Times (2026), fDi Markets: the in-depth crossborder investment monitor from the Financial Times, https://www.fdimarkets.com/ (accessed on 25 March 2026).
[49] Frank, R., T. McGuire and I. Nason (2021), “The Evolution of Supply and Demand in Markets for Generic Drugs”, The Milbank Quarterly, Vol. 99/3, pp. 828-852, https://doi.org/10.1111/1468-0009.12517.
[18] G7 Italia (2024), Finance Track Enhancing pharmaceutical manufacturing capacity in Africa: current scenario and future perspectives for G7 Members, G7 Italia, https://www.g7italy.it/wp-content/uploads/Report_Pharmaceutical-Manufacturing-in-Africa_20241014.pdf (accessed on 17 February 2026).
[12] Gleeson, D. et al. (2019), “Analyzing the impact of trade and investment agreements on pharmaceutical policy: provisions, pathways and potential impacts”, Globalization and Health, Vol. 15/S1, https://doi.org/10.1186/s12992-019-0518-2.
[64] GmbH, G. (ed.) (2022), Study on Best Practices in the Public Procurement of Medicines, European Commission.
[59] HHS (2023), “Impact of Drug Shortages on Consumer Costs - REPORT TO CONGRESS”, U.S. Department of Health and Human Services Office of the Assistant Secretary for Planning and Evaluation, https://www.ncbi.nlm.nih.gov/books/NBK603205/pdf/Bookshelf_NBK603205.pdf (accessed on 17 February 2026).
[81] Hughes-McLure, S. and E. Mawdsley (2022), “Innovative Finance for Development? Vaccine Bonds and the Hidden Costs of Financialization”, Economic Geography, Vol. 98/2, pp. 145-169, https://doi.org/10.1080/00130095.2021.2020090.
[28] ICH (2026), The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) Official web site, https://www.ich.org/ (accessed on 25 March 2026).
[25] ICMRA (2026), International Coalition of Medicines Regulatory Authorities (ICMRA) | International Coalition of Medicines Regulatory Authorities (ICMRA), https://icmra.info/drupal/en/home (accessed on 24 March 2026).
[90] IEA (2022), Oil Security Policy, https://www.iea.org/reports/oil-security-policy (accessed on 17 February 2026).
[73] IIR (2021), The carbon footprint of the cold chain, International Institute of Refrigeration, https://doi.org/10.18462/iir.INfood07.04.2021.
[32] IMDRF (2026), International Medical Device Regulators Forum, https://www.imdrf.org/ (accessed on 25 March 2026).
[76] IPCC (2014), “Transport”, Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, https://www.ipcc.ch/site/assets/uploads/2018/02/ipcc_wg3_ar5_chapter8.pdf (accessed on 17 February 2026).
[26] IPRP (2025), International Pharmaceutical Regulators Programme, https://www.iprp.global/home (accessed on 24 March 2026).
[86] Kim, S. and J. Skordis-Worrall (2017), “Can voluntary pooled procurement reduce the price of antiretroviral drugs? a case study of Efavirenz”, Health policy and planning, Vol. 32/4, pp. 516-526, https://doi.org/10.1093/heapol/czw165.
[83] Kremer, M., J. Levin and C. Snyder (2020), “Advance Market Commitments: Insights from Theory and Experience”, AEA Papers and Proceedings, Vol. 110, pp. 269-273, https://doi.org/10.1257/pandp.20201017.
[34] Lee, S. and D. Prabhakar (2021), COVID-19 Non-Tariff Measures: The Good and the Bad, through a Sustainable Development Lens.
[8] Mattoo, A., N. Rocha and M. Ruta (2020), “Overview: The evolution of deep trade agreements”, in Mattoo, A., N. Rocha and M. Ruta (eds.), Handbook of Deep Trade Agreements, The World Bank, Washington, https://doi.org/10.1596/978-1-4648-1539-3.
[60] NAO (2020), “The supply of personal protective equipment (PPE) during the COVID-19 pandemic A picture of the National Audit Office logo”, https://www.nao.org.uk/wp-content/uploads/2020/11/The-supply-of-personal-protective-equipment-PPE-during-the-COVID-19-pandemic.pdf (accessed on 17 February 2026).
[54] Napier M., B. (2024), The cost of drug shortages, The Office of Health Economics (OHE), https://www.ohe.org/publications/the-cost-of-drug-shortages/. (accessed on 16 February 2026).
[79] NHS England Digital (2022), One giant step for secure data access, https://digital.nhs.uk/blog/data-points-blog/2023/one-giant-step-for-secure-data-access (accessed on 4 February 2026).
[1] OECD (2025), Government initiatives to enhance medical supply chains based on international cooperation, OECD, https://www.oecd.org/en/publications/securing-medical-supply-chains-in-a-post-pandemic-world_119c59d9-en/support-materials.html.
[36] OECD (2025), OECD Supply Chain Resilience Review: Navigating Risks, OECD Publishing, Paris, https://doi.org/10.1787/94e3a8ea-en.
[35] OECD (2025), OECD Trade Facilitation Indicators: Monitoring Policies up to 2025, OECD Publishing, Paris, https://doi.org/10.1787/fd6f27dc-en.
[43] OECD (2024), Export restrictions on staple crops since 2007 - an overview based on the OECD database on export restrictions on staple crops.
[38] OECD (2024), Securing Medical Supply Chains in a Post-Pandemic World, OECD Health Policy Studies, OECD Publishing, Paris, https://doi.org/10.1787/119c59d9-en.
[7] OECD (forthcoming), Trade facilitation beyond the border: The evolving role of mutual recognition instruments, OECD Publishing, Paris.
[77] Okeagu, C. et al. (2021), “Principles of supply chain management in the time of crisis”, Best Practice & Research Clinical Anaesthesiology, Vol. 35/3, pp. 369-376, https://doi.org/10.1016/j.bpa.2020.11.007.
[80] PAHO (2023), The Revolving Fund for Access to Vaccines, an Engine of Equity, https://www.paho.org/en/revolving-fund-access-vaccines-engine-equity (accessed on 16 February 2026).
[69] Panteli, D. et al. (2024), “Policy options for sustainable access to off-patent antibiotics in Europe”, npj Antimicrobials and Resistance, Vol. 2/1, https://doi.org/10.1038/s44259-024-00061-4.
[72] PCAF European Buildings Database (2022), Financing towards net-zero buildings, Partnership for Carbon Accounting Financials, https://carbonaccountingfinancials.com/financing-towards-net-zero-buildings (accessed on 17 February 2026).
[29] PIC/S (2025), Pharmaceutical Inspection Co-operation Scheme, https://picscheme.org/ (accessed on 25 March 2026).
[31] PIC/S (2025), PIC/S Accession requirements, https://picscheme.org/en/accessions-accession (accessed on 25 March 2026).
[4] SGMAP (2015), “Évaluation des politiques publiques (EPP) : principes, processus et méthode”, https://www.eval.fr/wp-content/uploads/2018/09/epp_map_memo.pdf (accessed on 21 August 2026).
[92] Socal, M., J. Sharfstein and J. Greene (2021), “The Pandemic and the Supply Chain: Gaps in Pharmaceutical Production and Distribution”, American Journal of Public Health, Vol. 111/4, pp. 635-639, https://doi.org/10.2105/ajph.2020.306138.
[14] Trade Impact B.V. (2023), Trade Sustainability Impact Assessment (SIA) in support of Free Trade Agreement and Investment Protection Agreement negotiations between the European Union and the Republic of India, European Commission, Brussels, https://circabc.europa.eu/ui/group/09242a36-a438-40fd-a7af-fe32e36cbd0e/library/809883ba-4c8d-42a8-b06e-fb988861e55f/details?download=true (accessed on 3 February 2026).
[57] Välikangas, L., M. Luistro-Jonsson and S. Jarvenpaa (2022), “Health crisis and the EU’s HERA: amplifying partial organizing with resourcing for stability, agility, and evolvability”, Journal of Organization Design, Vol. 11/4, pp. 169-187.
[84] WHO (2023), COVAX - Working for global equitable access to COVID-19 vaccines, https://www.who.int/initiatives/act-accelerator/covax (accessed on 17 February 2026).
[48] WHO (2018), Addressing the global shortage of, and access to, medicines and vaccines, https://apps.who.int/gb/ebwha/pdf_files/EB142/B142_13-en.pdf (accessed on 16 February 2026).
[74] WHO (2017), Safe management of wastes from health-care activities: a summary, https://web.archive.org/web/20260202094320/https://www.who.int/publications/i/item/WHO-FWC-WSH-17.05 (accessed on 17 February 2026).
[87] World Bank (2021), Africa announces the rollout of 400m vaccine doses to the African Union Member States and the Caribbean, https://www.worldbank.org/en/news/press-release/2021/08/05/africa-announces-the-rollout-of-400m-vaccine-doses-to-the-african-union-member-states-and-the-caribbean (accessed on 17 February 2026).
[33] World Trade Organization (n.d.), The WTO’s Pharma Agreement, https://www.wto.org/english/tratop_e/pharma_ag_e/pharma_agreement_e.htm (accessed on 25 March 2026).
[75] Xu, Y. et al. (2025), “Antibiotic resistance occurrence and ecological impact in landfill leachate: A review on compound effect of antibiotics and non-antibiotics”, Emerging Contaminants, Vol. 11/3, p. 100508, https://doi.org/10.1016/j.emcon.2025.100508.
[37] Zhao, N. et al. (2023), “Analysis of a maritime transport chain with information asymmetry and disruption risk”, Ocean & Coastal Management, Vol. 231, p. 106405, https://doi.org/10.1016/j.ocecoaman.2022.106405.
Note: ATMP = Advanced Therapies and Medicinal Products (mainly refer to cell and gene therapies). Detailed information on scope of agreement is listed on the EMA website. This table only provides a short summary. Amendments to MRAs to take into account new EU member countries are not considered.
1. Joint statement: Sustainable economic growth in the European Union and Canada through the Comprehensive Economic and Trade Agreement, https://www.international.gc.ca/trade-commerce/trade-agreements-accords-commerciaux/agr-acc/ceta-aecg/2024-02-09-joint-statement-declaration_conjointe.aspx?lang=eng.
This annex provides the detailed technical material underpinning the MedMIS evaluation in Section 2.5. Each subsection is referenced from the body text and is designed to be consulted independently. Tables, operational specifications and institutional precedents that were condensed in the body to maintain narrative focus are preserved here in full.
The annex is organised as follows:
Section B.1 describes the AMIS operational architecture;
Section B.2 presents the minimum viable indicator set;
Section B.3 provides the stress-test across tracer product categories;
Section B.4 elaborates the “transparency trust” function for biologicals;
Section B.5 details the tiered information governance model;
Section B.6 analyses the COVID‑19 Comfort Letter precedent; and
Section B.7 provides cost benchmarks and decomposition.
Cross-reference: Section 2.5.2 (Evidence and lessons from the AMIS)
The Agricultural Market Information System (AMIS) data collection follows harmonised crop calendars and marketing-season-based reporting cycles. Participating countries submit data at least monthly through national focal points (typically officials in agriculture‑related ministries) using common templates developed by the Secretariat. These templates cover required data and metadata, including information on forecasting methods. Secretariat staff perform quality and consistency checks on incoming submissions; where inconsistencies or gaps are found, files may be returned for clarification. Once validated, national data are entered into an access-restricted section of the AMIS Market Database accessible only to focal points and the Secretariat. Commodity and country specialists within the Secretariat then combine these national submissions with other public and private sources – including the Food and Agriculture Organization of the United Nations (FAO), International Grains Council (IGC), the United States Department of Agriculture (USDA) statistics, remote‑sensing information from the Group on Earth Observations Global Agricultural Monitoring Initiative (GEOGLAM), and commercial datasets – to construct global market balances and forecasts (AMIS, 2016[41]).
The main public output is the AMIS Market Monitor,7 published ten times per year. Each edition presents a collective assessment of market conditions and outlook by the ten Secretariat organisations, providing a curated narrative designed to improve transparency and signal to policymakers any developments that may warrant attention.
A notable feature of AMIS’s data architecture, which has implications for MedMIS design, is its differentiated approach to information access. National data submitted by focal points are stored in a dedicated, restricted section of the AMIS database visible only to focal points and the Secretariat. Internal deliberations and RRF meeting reports circulate among participants but are not published. Public products – the Market Monitor, harmonised databases and policy databases – are openly accessible regardless of a country’s reporting performance (AMIS, 2016[41]).
When global market conditions warrant policy action, AMIS shifts from routine monitoring to intensified dialogue among RRF participants. Under the agreed protocol, the AMIS Chair, supported by the Secretariat, assesses after each Market Monitor release whether prevailing conditions merit initiating dialogue. Even under stable conditions, written exchanges may share views on the market assessment. If disturbances emerge, the Secretariat intensifies its analysis and communication, and RRF participants consider possible policy responses. When conditions are judged abnormal or unstable, extraordinary RRF meetings can be convened at short notice so that participants can “liaise as soon as feasibly possible” to review the situation, discuss policy options that should be taken or avoided, and consider common communication strategies. The RRF does not have binding powers: it promotes early information exchange and policy discussion, assists in mobilising political support for appropriate responses, and briefs the Committee on World Food Security, but it cannot prescribe or enforce particular national measures.
Source: Based on AMIS (2017[40]), http://www.amis-outlook.org/amis-about/en/ and AMIS (2016[41]).
Cross-reference: Section 2.5.4 (Operational feasibility: Minimum viable dataset and architecture)
The table below specifies the illustrative minimum viable operational indicator set for MedMIS, organised across five indicator families. Each indicator is defined with its operational scope, primary data holders and intended analytical use.
Cross-reference: Section 2.5.5 (Testing the MedMIS proof-of-concept across tracer product categories)
The table below applies a structured stress test of the proposed MedMIS minimum viable configuration across three tracer product categories: injectable corticosteroids (small-molecule generics with concentrated API markets), PCR diagnostic tests (complex multi-sectoral assembly networks), and seasonal influenza vaccines (biologicals with inelastic short-term supply). These categories represent structural archetypes delineating the boundary conditions under which MedMIS can deliver added value and where its utility may be limited.
Cross-reference: Section 2.5.5 (Testing the MedMIS proof-of-concept across tracer product categories)
For biologicals with inelastic supply (e.g. seasonal influenza vaccines), MedMIS could not function as an early-warning system for production, as lead times are fixed. Instead, it would pivot to a “transparency-trust” mechanism designed to prevent defensive hoarding. During recent crises (H1N1, COVID‑19), the dominant failure mode was informational asymmetry: lacking visibility on global supply, governments engaged in defensive double‑ordering and bilateral deals, creating “phantom” scarcity that froze voluntary exchange and reallocation of stock (OECD, 2024[38]; Okeagu et al., 2021[77]). For these products, MedMIS would operate through three channels:
Demand signal validation. Rather than monitoring factory output, MedMIS would aggregate and anonymise national procurement volumes. By verifying that total global orders do not exceed total manufacturing capacity, the system would defuse the “panic premium” and signal to Ministries that aggressive over-procurement is unnecessary.
Ancillary component surveillance. While bulk antigen supply is fixed 6‑9 months in advance, the supply of fill-finish inputs (glass vials, lipid nanoparticles, stoppers) remains elastic. MedMIS would retain high utility here: detecting bottlenecks in these fungible components allows policymakers to intervene (e.g. fast-tracking customs for glass tubing) to ensure that available bulk antigen is not wasted due to packaging constraints.
Surplus reallocation. MedMIS could maintain a ledger of stock positions to enable accelerated donation or reallocation of surplus doses once national saturation points are reached. This adapts the institutional precedent of the European Commission’s COVID‑19 Clearing House for Medical Equipment (established April 2020), which validated the core governance logic of a central platform identifying available industrial capacity and matching it with Member State needs.
Cross-reference: Section 2.5.8 (Risks and negative externalities)
Poorly governed transparency in pharmaceutical markets can amplify the volatility it seeks to mitigate. In inelastic markets with lean inventories, shortage alerts can trigger defensive stockpiling, converting manageable “liquidity” constraints into systemic “solvency” crises. This transparency-induced “bullwhip effect” is particularly acute in pharmaceuticals because: i) demand is price‑inelastic and prone to panic; ii) inventories are small with long production and release lead times; and iii) revealing disruptions at single‑source facilities triggers immediate competitive procurement, disproportionately affecting jurisdictions with weaker purchasing power.
To address these externalities, MedMIS would adopt a tiered information governance model distinguishing between two access levels:
Tier A (restricted): granular intelligence (facility disruptions, inventory positions, upstream bottlenecks) accessible only to the Secretariat and designated authorities. This tier enables co‑ordinated mitigation (reserve releases, regulatory flexibilities) while keeping the supply shock invisible to speculative markets.
Tier B (public): aggregated assessments designed to stabilise expectations (e.g. ”global supply meets baseline demand”). These outputs provide directional reassurance without revealing facility-specific vulnerabilities.
The system creates value only if transitions between tiers follow pre‑agreed escalation triggers. Restricted information would move to public disclosure only when: i) disruption duration exceeds the bridging capacity of reserves; ii) clinical substitution is not feasible; or iii) demand management measures have been exhausted. Without disciplined thresholds, premature disclosure risks triggering the panic the mechanism was designed to forestall.
Cross-reference: Section 2.5.8 (Risks and negative externalities)
On 8 April 2020, the European Commission issued a Comfort Letter to Medicines for Europe, the European generic medicines association, enabling rapid co‑operation to prevent shortages of critical hospital medicines during the first COVID‑19 wave (European Commission, 2020[78]). Issued alongside the Commission’s Temporary Framework on antitrust guidance for crisis-related co‑operation, the Comfort Letter established that operational transparency can be made compatible with EU competition law, provided data exchange is shielded by robust architecture.
The Commission accepted that public-health objectives could require limited co‑ordination around production and supply, including exchanges related to capacity and stock levels. It constrained the mechanism to what was objectively necessary and proportionate for shortage mitigation, imposing governance conditions to prevent drift into competitively sensitive co‑ordination.
The core governance feature was a neutral intermediary: firms were prohibited from sharing sensitive business information directly with one another. Information flowed to a third party and was made available only in aggregated form, limiting the risk that participants could infer competitors’ strategic positions. The arrangement was open to all manufacturers, with strong emphasis on documentation: meetings were recorded and agreements documented, enabling ex post scrutiny. The Commission retained a steering role, ensuring co‑operation remained aligned with public-health objectives.
Three design features of the Comfort Letter arrangement are directly transferable to MedMIS architecture:
Secretariat as data intermediary: raw company data flows only to a neutral MedMIS Secretariat; participants receive access solely to aggregated outputs.
Hard perimeter on content: exchanges strictly limited to shortage mitigation, excluding price‑related discussions, tendering strategies, or forward-looking commercial planning.
Auditability and controlled communications: communications documented and controlled, with formal guidance obtained in advance from competition authorities providing a clear, enforceable legal foundation.
Note on applicability
The COVID‑19 Comfort Letter addressed a form of competitor co‑operation in which sensitive information could be exchanged only under tight safeguards, including Commission oversight, strict necessity, time limitation, documentation, and collection by a neutral intermediary with disclosure to participants in aggregated form only. Under MedMIS’s proposed trustee architecture, firms would submit data to the Secretariat and receive only aggregated outputs, without direct information-sharing among competitors.
This structural difference reduces one important category of competition risk, although the broader safeguards reflected in the Comfort Letter, including purpose limitation and governance controls, remain relevant. The Comfort Letter is therefore relevant less as a direct template for MedMIS than as a precedent for the proposition that carefully designed crisis-related co‑ordination can be made compatible with EU competition law. Because MedMIS would be a standing mechanism rather than a temporary emergency arrangement, it would likely benefit from clearer ex-ante legal grounding than the ad hoc COVID‑19 framework provided.
Cross-reference: Section 2.5.9 (Implementation burden and value‑for-money)
MedMIS cost estimates are grounded in three established international mechanisms that each illuminate a distinct cost driver: data collection scale, verification depth and institutional authority.
The Agricultural Market Information System (AMIS) (EUR 1 – 2m/year, ~10 – 12 FTE): routine market monitoring for four crops with heavy reliance on in-kind international organisation inputs. Anchors the cost floor. MedMIS would exceed AMIS costs because it requires trustee architecture and real-time verification for commercially sensitive inputs.
The Extractive Industries Transparency Initiative (EITI) (USD 11m/year, ~50 FTE): standing validation secretariat with structured verification cycles. Best structural analogue for MedMIS, though EITI validates retrospectively while MedMIS would require real-time triangulation.
The Financial Stability Board (FSB) (CHF 13 – 16m/year, ~40 FTE): high-authority co‑ordination body with continuous monitoring and de facto enforcement credibility. Upper-bound reference; MedMIS would not require FSB-level regulatory authority.
Source: Based on published budget data for AMIS (FAO), EITI (Annual Report), and FSB (Annual Report).
Structural observations
Scenario A’s EUR 6‑8M envelope reflects the fixed cost of establishing a standing secretariat with real-time verification capacity. This floor is binding: crisis-mode credibility requires permanent staff continuity, institutional memory, and familiarity with participant relationships – qualities that episodic or consultancy-based models cannot provide. Below this envelope, the mechanism would lack the verification depth to resolve contested data submissions within crisis timeframes.
Scenario B’s scaling to EUR 12‑16M reflects expanded institutional capacity for verification and surge readiness. The cost difference between scenarios is driven by: i) verification infrastructure for triangulating across commercial datasets, regulatory signals, customs flows and logistics intelligence; ii) staffing depth for 24/7 operational readiness during crises; and iii) security assurance and auditability requirements. Operational credibility and crisis-time decision support, rather than data volume, drive the cost curve.
Scenario A cost decomposition
Costs comprise a fixed institutional base (existing regardless of product count) and a variable per-product layer scaling with basket size.
Source: Staff costs based on international-organisation salary scales. Infrastructure benchmarks from comparable secure data platforms (NHS England Digital, 2022[79]).
The decomposition clarifies why the Scenario A envelope is estimated at EUR 6‑8m and why the product ceiling is binding. The fixed institutional base absorbs the bulk of the budget regardless of product scope. Each additional product costs relatively little, but cumulative variable costs narrow the margin for fixed-cost quality. Beyond approximately 25‑30 products, either per-product coverage thins below decision-grade quality or the budget must expand toward Scenario B.
Staff costs dominate at 65‑70% of budget across both scenarios. Year 1 costs will exceed the steady-state envelope by approximately EUR 1.5‑2.5m due to the non-amortised portion of the initial platform build and first-year security certification.
Sustainability: Hinge‑point criteria
To mitigate institutional drift and ensure that costs remain proportionate to demonstrated value, MedMIS should follow a pre‑committed conditional timeline with two strategic decision points (hinge points). These are substantive tests of whether the mechanism’s design preconditions have been met and whether operational performance justifies continued investment.
Month 24 (feasibility hinge). The project assesses whether essential preconditions have been met:
Has a credible competition-law safe harbour been secured (formal or documented ex-ante guidance from EC DG COMP)?
Has functional critical mass been achieved (e.g. at least 50% of OECD pharmaceutical procurement volume participating, plus top‑3 API manufacturing jurisdictions for tracer products)?
Have routine baseline‑setting and validation workflows demonstrated operational capability (e.g. through exercises or initial data runs)?
Failure to meet these preconditions triggers an immediate and permanent transition to a “light” co‑ordination forum: diplomatic convening and voluntary intelligence‑sharing continue, but systematic data collection ceases. This prevents cost-creep from a failing pilot and preserves the institution’s credibility for a future attempt under better conditions. If successful, the project proceeds to full Phase 1 operations (routine monitoring, escalation protocols, initial crisis-readiness exercises).
Month 48 (operational utility hinge). The project evaluates whether the pilot has demonstrated decision-utility in at least one simulated or actual crisis event, or near-miss triggering the mechanism. Evidence includes: Did MedMIS-provided intelligence materially improve the speed, coherence or proportionality of member-state response? If yes, the mechanism advances to Phase 2 (expanded participation, deepened upstream supply-chain monitoring, additional product categories). If no, MedMIS reverts to light co‑ordination.
For systemically important producer hubs (China, India) unlikely to fund the Secretariat directly, participation incentives could be framed as regulatory reciprocity rather than extracted contribution. This could include: prioritised regulatory pathways for post-approval variations relevant to continuity (site transfers, alternative suppliers); mutual reliance on inspection outputs (reduced duplicative inspections); and access to Secretariat intelligence on supply disruptions affecting their export markets.
Bounded conditions for value‑for-money
MedMIS would represent value‑for-money under bounded conditions: i) strict scope discipline; ii) a pre‑cleared competition-law “safe harbour” funded within the core budget; iii) credible participation by key demand-side jurisdictions and at least some major upstream hubs; and iv) pre‑committed decision gates that allow the mechanism to adjust or exit gracefully if preconditions are not met. Without these conditions, the instrument risks being either an expensive forum with insufficient visibility and participation to shift behaviour, or a low-credibility platform that fails to attract sustained investment and trust from industry participants.
Cost-sharing and differentiated contributions
Because MedMIS would be open to third countries, cost-sharing should reflect asymmetric interdependence and differentiated capacity. Assessed contributions should scale with objective proxies for exposure and benefit:
Core EU participants (Tier 1): proportionate contribution based on GDP or assessed share of pharmaceutical procurement volume (baseline model). These participants receive full access to Tier A (real-time operational intelligence) and voting rights on scope, triggers and allocation policy.
Associated third-country participants (Tier 2): assessed contributions scaled at 50‑75% of equivalent EU participants, reflecting lower institutional capacity to absorb participation costs. Alternatively, in-kind contributions (e.g. data provision, logistics support) can substitute for cash contributions, creating flexibility for capacity-constrained partners.
Producer hub participation: systemically important manufacturer jurisdictions that cannot fund the Secretariat should participate through regulatory reciprocity arrangements (outlined above), not cash contribution. This avoids the appearance of extractive conditionality while creating credible incentives for engagement.
Asymmetric access to Tier A intelligence: members contributing disproportionately (either through assessed contributions or through hosting, supply of critical data) receive first access to real-time alerts and early intelligence. This creates a “club good”: benefits improve with participation quality and scale.
Note that these cost estimates are illustrative – they are intended to establish order-of-magnitude benchmarks for institutional design decisions.
This annex provides the detailed technical material underpinning the MedPPA evaluation in Section 2.6. Each subsection is referenced from the body text and is designed to be consulted independently. Institutional profiles, empirical benchmarks, legal analysis and operational specifications that were condensed in the body to maintain narrative focus are preserved here in full.
The annex is organised as follows:
Section C.1 profiles six global procurement mechanisms relevant to MedPPA design;
Section C.2 presents empirical benchmarks on the cost and value of resilience premiums;
Section C.3 details the participation architecture, including the two‑layer opt-in model, procurement law compatibility, WTO Agreement on Government Procurement (GPA) analysis and allocation governance;
Section C.4 specifies the financing and liability architecture, including legal base options, the liability protocol and currency risk provisions; and
Section C.5 provides quality assurance, logistics and deployment specifications.
Cross-reference: Section 2.6.2 (Institutional models and global practice).
Six international procurement mechanisms provide operational lessons directly relevant to MedPPA feasibility. The body text (Section 2.6.2) synthesises the design lessons; the profiles below preserve the full institutional detail.
PAHO Revolving Fund
The Pan American Health Organization (PAHO) Revolving Fund has operated for over 40 years, enabling countries in the Americas to purchase vaccines and public health supplies at prices substantially below what they could achieve individually. For a basket of 13 routine vaccines (2018‑2022), PAHO reports that countries would have paid 75% more if they had bought outside the Fund, equivalent to Fund prices approximately 43% below non-pooled prices (PAHO, 2023[80]).
The mechanism’s longevity reflects more than volume consolidation. Its working-capital (revolving) fund and payment/collection functions allow PAHO to pay suppliers promptly and extend 60‑day interest-free credit to participating states, reducing supplier credit risk. The Fund charges a 4.25% surcharge on the net procurement price, which capitalises the revolving fund and finances this liquidity infrastructure.
This feature is directly relevant to MedPPA design: a financial architecture capable of centralising payment liability, rather than relying solely on administrative co‑ordination, provides the supplier assurance needed to maintain supply security during volatile conditions. The PAHO experience indicates that a low-single‑digit operational premium can finance the liquidity backbone that enables substantially larger savings through volume consolidation.
Gavi, the Vaccine Alliance: IFFIM and AMC
Gavi’s relevance lies in its use of standing instruments that couple pooled procurement with durable financing to shape market behaviour. Two mechanisms are instructive:
The International Finance Facility for Immunisation (IFFIM), established in 2006, converts long-term legally binding donor pledges into immediate cash through “vaccine bonds”, improving liquidity and smoothing annual appropriation constraints. This frontloading mechanism guarantees suppliers payment while allowing donors to spread contributions over 20 years, providing Gavi with predictable liquidity independent of annual appropriations (Hughes-McLure and Mawdsley, 2022[81]). For MedPPA, this demonstrates how a pooled procurement mechanism can manage cash-flow gaps and provide manufacturers with payment certainty without requiring immediate full budget availability.
The Advance Market Commitment (AMC), first piloted for pneumococcal vaccine in 2009, legally committed donors to purchase qualifying vaccines at a pre‑agreed price once criteria were met, incentivising manufacturers to invest in capacity and supply for low-income markets. The AMC utilised a USD 1.5 billion donor subsidy fund that effectively doubled per-dose revenue for manufacturers on the initial tranche (raising the price from a USD 3.50 tail price to USD 7.00), creating capacity that did not previously exist (Cernuschi et al., 2011[82]; Kremer, Levin and Snyder, 2020[83]). Once capacity was established, the long-term tail price ceiling was locked in, and suppliers honoured commitments during subsequent shortages. The AMC demonstrates that upfront capacity payments function as capital investments to ensure physical availability, not as ongoing inefficiencies: the time‑limited premium secured both manufacturing capacity and long-term price stability.
COVAX facility
COVAX was launched in 2020 under the ACT-Accelerator as a global effort to support equitable access to COVID‑19 vaccines, with procurement and delivery functions administered by Gavi. By end‑2023, WHO reported that COVAX had delivered nearly 2 billion doses to 146 economies and helped avert an estimated 2.7 million deaths in lower-income economies (WHO, 2023[84]).
COVAX’s early-phase constraints illustrate that aggregation does not defeat scarcity: pooled demand cannot prevent shortfalls when global supply is inelastic or when manufacturing under-delivers. Pooled procurement proved most effective when paired with at-risk financing alongside clear, legitimate allocation rules (de Bengy Puyvallée and Storeng, 2022[85]). For MedPPA design, this implies that the feasibility case should rest on contractual instruments that purchase priority and capacity (not only price), and governance arrangements that maintain coalition discipline when scarcity peaks.
Global Fund Pooled Procurement Mechanism
The Global Fund’s Pooled Procurement Mechanism (PPM) aggregates orders for HIV, TB, and malaria commodities from over 80 countries. Over two decades, it has driven significant price reductions for generic medicines (Kim and Skordis-Worrall, 2017[86]). Its key success factor is the integration of purchasing with grant financing: as the Global Fund also controls the funding stream, it provides certainty to suppliers. This feature suggests that MedPPA would be most effective if linked to existing health financing streams for participating third countries rather than relying on ad hoc national contributions, particularly in emergencies when fiscal space, approvals, and payment timing become binding constraints.
AVAT (African Vaccine Acquisition Task Team)
During COVID‑19, the African Union established AVAT as a pooled procurement channel to complement COVAX and secure additional doses for AU Member States. Through the Africa Medical Supplies Platform, AVAT negotiated a deal with Johnson & Johnson providing access to up to 400 million doses, underpinned by a USD 2 billion Afreximbank financing facility that supported payment and guarantee functions (World Bank, 2021[87]).
The operational lesson is that regional pooled procurement becomes credible when it couples demand aggregation with a financing/guarantee backbone and a centralised purchasing platform. AVAT also demonstrates the value of a bloc mechanism for smaller markets that cannot secure comparable deals individually, while highlighting that political sponsorship and financial engineering are decisive feasibility conditions.
African Pooled Procurement Mechanism (APPM)
Building on the AVAT experience, the African Union Summit in February 2024 decided to establish a continent-wide African Pooled Procurement Mechanism (APPM) under the leadership of Africa CDC, with support from Afreximbank and the UN Economic Commission for Africa (UNECA). In October 2025, Africa CDC formally launched the APPM at the African Pharmaceutical Suppliers Meeting in Addis Ababa, covering all 55 AU Member States. The mechanism is anchored to the African Continental Free Trade Area (AfCFTA) Pharmaceutical Initiative and is designed to consolidate demand for medicines, vaccines, diagnostics, and other essential health products, while prioritising African manufacturers (AfricaCDC, n.d.[88]). While the APPM has not yet generated the published cost and performance data needed for systematic benchmarking, the AVAT-to-APPM trajectory validates the feasibility of transitioning from crisis-specific bloc procurement to a standing pooled mechanism.
MAV+ (Manufacturing and Access to Vaccines, Medicines and Health Technologies)
The MAV+ initiative illustrates a strategic approach where demand aggregation is explicitly used to support industrial policy goals. Launched as a Team Europe Initiative in 2021, MAV+ adopts a three‑pillar approach: strengthening local manufacturing capacity (supply side), enhancing access and demand aggregation (demand side), and supporting the regulatory environment (European Commission, 2023[89]).
For many third countries, the attraction of such a mechanism is not price minimisation but the “resilience premium”, using the contract to support local manufacturing or guarantee off-take for nascent industries. This dynamic is relevant to MedPPA design: the mechanism’s value for non-EU partners differs from its value for EU countries: the governance architecture would need to accommodate both.
Summary: Key design features across comparator mechanisms
The table below summarises the principal design features of each comparator mechanism and their relevance to MedPPA design.
Cross-reference: Section 2.6.2 (Cost estimates, benchmarks, and value‑for-money justification)
This section presents the empirical evidence underlying the resilience premium estimates in Section 2.6. The body text summarises the findings; the detailed benchmarks and source data are preserved below.
PAHO Revolving Fund
The Fund charges a 4.25 per cent surcharge on the net procurement price. This fee capitalises a revolving fund that provides 60‑day interest-free credit to member states and guarantees prompt payment to suppliers. For a basket of 13 routine vaccines (2018‑2022), countries would have paid 75% more if they had bought outside the Fund, equivalent to Fund prices approximately 43% below what countries would pay individually (PAHO, 2023[80]). A low-single‑digit operational premium thus finances the liquidity and payment certainty infrastructure that enables substantially larger savings through volume consolidation.
Gavi Advance Market Commitment (AMC)
To secure pneumococcal vaccines for low-income markets, the AMC utilised a USD 1.5 billion donor subsidy fund. This subsidy effectively doubled per-dose revenue for manufacturers on the initial tranche (raising the price from a USD 3.50 tail price to a USD 7.00 AMC price). This time‑limited premium incentivised manufacturers to build production capacity that did not previously exist. Once capacity was established, the long-term tail price ceiling of USD 3.50 was locked in, and suppliers honoured supply commitments during subsequent shortages (Cernuschi et al., 2011[82]). The AMC demonstrates that upfront capacity payments function as capital investments: they secure both physical availability and long-term price stability that individual countries could not negotiate.
AVAT (COVID‑19)
The African Union’s AVAT mechanism relied on a USD 2 billion guarantee facility from Afreximbank. While the guarantee carried a financing cost, typically in the low single‑digit percentage range of covered exposure in comparable structures, it replaced the sovereign payment risk of 55 individual nations with a single, highly creditworthy counterparty (World Bank, 2021[87]). This allowed the AU to secure up to 400 million doses of Johnson & Johnson vaccine, a deal commercially unachievable for individual African states given their sovereign credit profiles and payment capacity constraints. The AVAT institutional architecture has since evolved into the African Pooled Procurement Mechanism (APPM), a standing continent-wide mechanism launched in 2025 (see Annex 2.C, Section C.1).
The cost of inaction
Mechanisms that relied solely on lowest-price procurement without resilience clauses failed to protect buyers during the COVID‑19 crisis. Three sets of data illustrate the scale of cost escalation:
Medicines: A US analysis found average price increases of over 16 per cent following shortages, with substitute products experiencing higher multiples (HHS, 2023[59]).
PPE: UK procurement data recorded price increases of 166 per cent for respiratory masks and over 1 000 per cent for body bags (NAO, 2020[60]).
Total fiscal cost: The UK Department of Health and Social Care wrote off GBP 8.7 billion worth of PPE as unusable, overpriced, or undelivered, of which GBP 4.7 billion reflected the gap between purchase price and subsequent market value (Department of Health and Social Care (DHSC), 2022[61]).
These data suggest that resilience premia in the range assessed for MedPPA (low-to-mid single digits to approximately 20% above minimum-cost procurement, depending on product type and supply risk) would be lower than documented crisis-period cost escalation. The comparison is necessarily indicative rather than controlled: no study has directly measured the avoided cost of a resilience premium in a European pharmaceutical context. The premium should therefore be treated as a design parameter to be refined through pilot implementation, with a prospective evaluation framework tracking avoided crisis-period costs against cumulative premia paid during routine periods.
Cross-reference: Section 2.6.2 (Governance, funding, legal and operational arrangements)
This section provides the detailed legal analysis of the two‑layer opt-in architecture, procurement law compatibility, WTO GPA implications, allocation authority, and dispute resolution provisions that would underpin MedPPA’s participation model.
Two-layer opt-in architecture
Participation would operate through two distinct layers. Separating these layers is a design priority, because many procurement co‑operation failures arise when institutional membership and product-specific commitment are conflated.
Membership opt-in: defines eligibility tier, baseline governance rights, reporting and data-sharing obligations, dispute resolution, minimum readiness requirements (logistics capability, designated focal points, regulatory pathways), and any standing financial obligations (e.g. administrative fees, option premia).
Product-specific opt-in: Binds participants for a defined product in a defined period, triggering: volume commitment rules (firm versus indicative demand); financing terms (pay-as-you-go, pooled fund drawdown, reservation fees); allocation rules under scarcity; quality and regulatory release requirements; and enforcement conditions, including exclusivity or functional equivalents.
This separation ensures that institutional commitment does not automatically create obligations for every product, while product-specific binding creates the enforceable demand signals that suppliers require to allocate capacity and prioritise deliveries.
EU procurement law compatibility
The tiered participation model must be assessed against applicable procurement law at two levels. Under EU public procurement law (Directive 2014/24/EU, sector-specific instruments, and the single regulation the Commission proposed in September 2026 to replace them; see Section 2.6.2), structuring a pooled mechanism that reserves tranches or confers preferential access on specific countries raises questions of equal treatment and transparency, particularly where the procurement agent is an EU institution or body subject to the Financial Regulation. The legal architecture would need to demonstrate either that tier-specific access rights fall within established frameworks for third-country participation (such as the existing JPA provisions) or that separate procurement exercises are conducted for distinct participant groups.
WTO Agreement on Government Procurement (GPA) analysis
Where participating governments are parties to the WTO Agreement on Government Procurement (GPA) or bilateral agreements containing procurement chapters, differential access (reserved capacity, priority delivery, surge options) may raise non-discrimination concerns unless justified under applicable public health or national security exceptions. Tier-specific access rights should therefore be either: i) justified under recognised exceptions with clear documentation; ii) structured as separate procurement exercises rather than discriminatory conditions within a single tender; or iii) limited to procurement categories not covered by applicable trade commitments. Early engagement with both EU legal services and trade law authorities would be essential.
Allocation authority and the fiduciary dilemma
The fiduciary tension becomes acute once third countries participate. If the procurement agent is an EU-based body, it is politically and legally accountable to EU institutions and taxpayers. Under scarcity, this creates the question: are third-country claims equal, subordinate or separately defined?
A workable design requires an explicit choice among allocation architectures:
Equal claim rights: conceptually clean but politically and fiscally demanding, typically requiring strong burden-sharing and a clear legal mandate.
Subordinate access: more compatible with an EU fiduciary perimeter but reduces third-country willingness to commit and weakens the consolidated bargaining position.
Separate tranches: often most feasible, with capacity or volumes earmarked ex ante by tier or participant group, creating predictable claim rights and reducing contestation.
To preserve crisis velocity while maintaining legitimacy, the governance model would need to operationalise: delegated execution authority (a clear legal mandate for the procurement agent to act once triggers are met); pre‑approved contractual templates; decision thresholds (consensus for constitutional rules, qualified majority for emergency activation, delegated authority for execution); fast dispute escalation with time limits; and ex-post accountability through structured reporting and independent review after activation events.
Dispute resolution
The framework agreement should include binding dispute resolution compatible with emergency execution: time‑limited procedures and interim measures allowing procurement to proceed while disputes are resolved. Orderly exit clauses must not undermine ongoing framework contracts or reserved-capacity options. Safeguards should support lawful pooled procurement: clear public-interest purpose, limits on information exchanged, secure handling of commercially sensitive data, and pre‑specified rules to prevent inappropriate co‑ordination spillovers.
Cross-reference: Section 2.6.2 (Governance, funding, legal and operational arrangements)
This section provides the detailed analysis of security-of-supply financing layers, treaty options, liability protocols, currency risk, and affordability channels that would underpin MedPPA’s financing model. The financing and legal design would determine whether suppliers treat pooled demand as bankable and executable when supply tightens.
From bulk purchasing to security-of-supply contracting
Where the policy objective is assured access, pooled procurement must be able to purchase resilience, priority production slots, surge options, minimum fill rates, and enforceable delivery schedules. This typically requires front-loaded commitment instruments that compensate suppliers for maintaining surge readiness rather than relying exclusively on pay-on-delivery purchasing. A practical architecture would distinguish three complementary financing layers:
Framework contracting with national call-offs: for products with elastic supply, participants could purchase through framework agreements with national payments at call-off. This preserves administrative simplicity while establishing common specifications, qualified supplier rosters, and standardised contracting terms.
Central guarantee capacity: to support advance commitments and reduce payment-risk pricing, the mechanism should include a pooled guarantee function, either a dedicated facility or centrally backed payment guarantee, that can underpin down payments, staged milestone payments, and performance‑linked contracts. This does not require full fiscal pooling for all purchases, but provides the credibility needed for suppliers to allocate capacity and prioritise deliveries under uncertainty.
Surge option layer: for products where scarcity is the defining risk (e.g. vaccines), the mechanism should contract for surge optionality: paid rights to call off volumes within a defined activation window. To be operationally meaningful, these commitments should be supported by multi-year appropriations authority or ring-fenced budget envelopes, so suppliers can treat the option as durable rather than contingent on annual budget cycles. The option layer should translate into explicit entitlements (e.g. tiered tranches or pre‑allocated shares) so that pre‑payment corresponds to defined claim rights under scarcity.
Delivery performance instruments
Security-of-supply contracting requires that financing be paired with delivery performance instruments, including: minimum service levels (e.g. fill-rate commitments); escalation clauses; transparent delivery schedules; and remedies for non-performance (e.g. liquidated damages, reallocation rights, or step-in options to secondary suppliers). These instruments convert legal remedies into prepaid economic leverage.
Legal base options and institutional pathways
The establishment of a binding international framework would require identification of an appropriate legal base. Four alternative pathways exist:
1. Article 168(5) TFEU (public health incentive measures): permits adoption of incentive measures to protect and improve human health, excluding harmonisation of national laws. A Council Decision could establish participation rules, financing mechanisms, and procurement procedures without requiring Treaty amendment. Achievable within a single MFF cycle (3‑5 years).
2. Article 122 TFEU (economic emergency measures): permits Council measures “appropriate to the economic situation” in a “spirit of solidarity.” This base supported the Emergency-Support-Instrument (ESI)-funded COVID‑19 vaccine procurement. Permits rapid Council action. Constraint: inherently temporary and situation-specific; using this for a standing mechanism would stretch its intended scope.
3. International agreement under Article 216/218 TFEU: enables multilateral framework agreements with third countries, potentially hosted by an existing international organisation. Provides legal certainty for third-country participation. Constraint: requires Council authorisation, Commission negotiation, and potentially national parliament ratification. Timeline: 3‑7 years.
4. Intergovernmental agreement outside EU Treaty framework: maximum flexibility but excludes EU institutions from governance and creates parallel legal architecture.
5. The most feasible pathway would likely be a hybrid approach: an internal EU framework under Article 168(5) TFEU establishing the governance architecture for EU participants, combined with an international agreement under Article 216 TFEU enabling structured third-country participation.
Liability protocol
The framework should pre‑establish a liability architecture covering: manufacturer indemnification terms; adverse event compensation arrangements and their financing sources (either harmonised national indemnities aligned to a common template or a pooled/no-fault facility with ring-fenced funding); and jurisdictional handling of claims and arbitration rules. Without this, novel product procurement (notably vaccines) can stall when suppliers require liability shields that participants cannot credibly provide through ad hoc arrangements.
Because pooled procurement delivers value only if products are legally deployable, the agreement should incorporate regulatory reliance conditions, including unilateral reliance arrangements for Tier 2 participants, as a participation prerequisite for product-specific opt-in, not a best-effort aspiration.
Currency risk and affordability
If third-country participants remit in local currency while framework contracts are denominated in euros, the mechanism would require either a currency hedging facility or a requirement that participants pre‑fund in the contract currency, each carrying distinct cost and accessibility implications.
If participation extends beyond high-income partners, the financing design should include an explicit affordability channel for eligible Tier 2 participants (e.g. donor window, concessional co-financing, or blended finance aligned with co‑operation instruments), structured as product- and tranche‑specific support linked to product-specific opt-in commitments and readiness conditions.
Cross-reference: Section 2.6.3 (Quality assurance, logistics and the “single buyer / single approver” logic)
This section provides the detailed supplier qualification, chain-of-custody, cold-chain, and customs specifications that would underpin MedPPA’s quality assurance framework.
Supplier and site qualification
A credible pooled mechanism requires standing quality capacity beyond paper compliance:
Verification of GMP/GDP status and quality systems, relying on recognised inspections and mutual recognition where available, and standardising the evidence package required for all participating tiers.
Harmonised requirements for certificates of analysis, batch documentation, stability data where relevant, and traceability identifiers, so the same evidence package supports release across participating jurisdictions.
Predefined escalation procedures for deviations, quality signals, and recalls, including decision authority and information-sharing protocols across participants.
Chain-of-custody requirements and minimum traceability expectations for diversion-prone or high-value products.
In crisis mode, the quality model should preserve safety while enabling speed through reliance, standardised evidence, and rapid signal-sharing rather than duplicative pre‑distribution checks.
Logistics operability: Customs, cold chain, and last-mile readiness
Procurement centralisation cannot substitute for logistics readiness. The mechanism should embed operational prerequisites and service standards into product-specific opt-in and contracting:
Expedited clearance protocols. Standardised customs documentation packs, pre‑agreed routing and designated entry points, and expedited clearance procedures aligned to national emergency frameworks. Where full pre‑clearance is not legally available ex ante, the mechanism should rely on pre‑positioned documentation, predefined emergency procedures, and rapid activation steps that can be triggered under national law.
Cold chain and handling standards. Verified minimum handling capacity for the products opted into (storage, monitoring, validated transport), with clear hand-off points (incoterms/custody transfer) and accountability for excursions.
Distribution model clarity. Whether delivery is centralised through an implementing partner or delivered to national depots with national distribution thereafter; the model should be aligned to participant capacity and to the product’s handling requirements.
Readiness-linked product choice and support. Product selection and contracting should reflect verified cold-chain realities. Where Tier 2 participants lack required infrastructure for certain ultra-cold products, the mechanism should either procure formulations compatible with existing capacity, or pair participation with readiness support linked to the financing architecture (Section C.4), so that cold-chain and handling requirements are not purely theoretical.
This keeps the instrument operationally grounded: a mechanism cannot be “open” in practice if the products it procures are systematically incompatible with the verified delivery capacity of participating partners. Quality assurance and regulatory operability are the mechanisms through which pooled procurement becomes deliverable access. A MedPPA model that standardises contracting but leaves approval, liability readiness, packaging compatibility, and logistics to ad hoc national solutions risks being fast at procurement and slow at deployment.
This annex provides the detailed technical material underpinning the RescPool evaluation in Section 2.7. Each subsection is referenced from the body text and is designed to be consulted independently. Comparator mechanism profiles, stress-test results, performance specifications and legal analysis that were condensed in the body to maintain narrative focus are preserved here in full.
The annex is organised as follows:
Section D.1 profiles four strategic reserve mechanisms and one crisis-time allocation mechanism relevant to RescPool design;
Section D.2 presents the stress-test across three tracer product archetypes;
Section D.3 specifies the performance monitoring framework, including KPIs, measurement methodology, reporting cadence and the exercise protocol;
Section D.4 details the deductible principle, covering calibration, verification, enforcement and override architecture; and
Section D.5 presents legal base options and institutional pathways.
Cross-reference: Section 2.7.2 (Lessons from comparator reserve systems)
This subsection profiles four strategic reserve mechanisms whose governance features are relevant to pooled pharmaceutical reserve design. Each profile describes the mechanism’s institutional structure, activation model, and operational track record. The body text (Section 2.7.2) presents the cross-cutting governance principles derived from these comparators.
International Energy Agency (IEA) oil security system
The IEA oil security system, established in 1974, requires each member country to hold emergency oil stocks equivalent to at least 90 days of net oil imports. The Co‑ordinated Emergency Response Mechanism enables collective releases during severe supply disruptions, with stocks released into the market rather than physically allocated among members, thereby limiting crisis-time distribution disputes (IEA, 2022[90]).
The system’s operational credibility now rests on six co‑ordinated actions: the 1991 Gulf War, 2005 Hurricane Katrina, 2011 Libya disruption, two releases in 2022, and the March 2026 response to the Middle East conflict and Strait of Hormuz disruption. On 11 March 2026, IEA member countries agreed to make 400 million barrels available to the market, the largest co‑ordinated release in the Agency’s history. Initial volumes had already begun to reach the market by 19 March, with Asia-Oceania releases starting immediately and stocks from the Americas and Europe beginning from the end of March.
The speed of this response reflects two design features: pre‑agreed burden-sharing arrangements that reduce the need for ad hoc negotiation, and an institutional framework that allows political decisions to focus on activating a known mechanism rather than designing one during the crisis.
The IEA model is directly relevant to the activation-speed and allocation-rule design elements of pharmaceutical reserves, but two structural differences limit transferability. First, crude oil is comparatively fungible, whereas medicines are regulated products requiring jurisdiction-specific authorisation. Second, IEA releases operate through market mechanisms, whereas pharmaceutical reserve deployment would generally require administrative allocation to health systems and providers.
EU Gas Storage Regulation
Adopted as emergency legislation following Russia’s invasion of Ukraine, the amended EU Gas Storage Regulation mandates that Member States fill underground gas storage to 90% capacity annually, with binding intermediate trajectories monitored by the Commission. Member States lacking storage facilities must store 15% of their annual consumption in facilities located in other Member States, thereby establishing a concrete solidarity obligation (European Parliament and Council of the European Union, 2025[91]). Default solidarity rules apply where bilateral agreements are absent, preventing crisis-time negotiation paralysis.
Three design features are relevant to RescPool’s governance:
1. Binding fill targets with intermediate milestone: the Regulation imposes graduated trajectory targets throughout the filling season, with Commission monitoring at each stage. This architecture closes the gap between nominal reserve commitments and verified physical readiness, the same gap that pharmaceutical reserves would face if participation were measured by pledged volumes rather than verified inventory.
2. Automatic default rules: where Member States have not concluded bilateral solidarity agreements, the Regulation’s default rules apply automatically, ensuring that cross-border access does not depend on ad hoc political bargaining at the moment of greatest urgency.
3. Operator certification: the Regulation requires certification of storage operators to prevent undue foreign influence over critical infrastructure. Pharmaceutical reserves would face an analogous risk: reserves hosted or managed by entities without assured operational independence could prove inaccessible under geopolitical stress.
US Strategic National Stockpile (SNS)
The SNS, managed by HHS’s Assistant Secretary for Preparedness and Response, holds pharmaceuticals, vaccines, antidotes, and medical supplies. It is designed primarily for mass-casualty emergency response (bioterrorism, CBRN events, pandemic surge) rather than routine pharmaceutical shortage management, and its product scope and demand profile differ from those of a cross-border supply-continuity reserve. The lessons drawn here are therefore limited to activation governance and lifecycle management.
The SNS operates with delegated activation authority, enabling deployment without a presidential emergency declaration. Pre‑configured “12‑hour push packages” can be dispatched to any US location within 12 hours of a deployment decision (Socal, Sharfstein and Greene, 2021[92]). The SNS demonstrates that pre‑delegated technical release authority and a hybrid operational model (combining federal stockpiles, vendor-managed inventory, and capacity contracts) can achieve rapid activation. However, the COVID‑19 pandemic exposed several documented vulnerabilities: expired inventory due to insufficient formulary review cycles, inadequate surge quantities, and federal – state co‑ordination failures in last-mile distribution. These failures illustrate the gap between nominal stockpile levels and actual deployable capacity, and are notably the most transferable evidence, since the pandemic-era shortfalls (PPE, ventilators, supply-demand mismatches) were shortage‑like in character rather than classic CBRN emergencies.
Two further structural limitations affect transferability. The SNS operates within a single regulatory jurisdiction (one regulator, one legal framework, unified command), whereas a European pooled reserve would face cross-border regulatory requirements at every stage of the deployment chain. Additionally, the SNS experience confirms that lifecycle governance, regular formulary reviews, rotation discipline, and co‑ordination with receiving jurisdictions through documented handoff protocols, is essential to prevent reserves from degrading between crises.
RescEU
Established in 2019 as part of the EU Civil Protection Mechanism, rescEU provides strategic reserves fully financed by the EU budget, hosted across 22 Member States and co‑ordinated through the Emergency Response Co‑ordination Centre (ERCC) within DG ECHO. Aid must be ready for dispatch within 12 hours of acceptance of an offer (European Commission, 2019[93]). The EUR 540 million CBRN (chemical, biological, radiological, and nuclear) strategic stockpile represents a significant scale‑up from COVID-era reserves.
The rescEU framework is relevant to pharmaceutical reserve design in four respects:
1. 12‑hour operational readiness: rescEU has demonstrated that EU-level co‑ordination can move medical supplies from host-state warehouses to requesting Member States within hours, including cross-border delivery of equipment and generators to Ukraine. This operational track record establishes that rapid multilateral deployment is achievable within existing civil protection architecture.
2. Institutional infrastructure: rescEU’s hosting arrangements, ERCC activation protocols, and HERA co‑ordination on medical countermeasures provide tested institutional pathways for procurement, storage, and deployment decisions that could serve as a foundation rather than requiring parallel structures.
3. Regulatory gap for pharmaceuticals: current rescEU medical reserves focus on emergency equipment, PPE and CBRN countermeasures, products that do not require jurisdiction-specific marketing authorisation, multilingual patient information, or serialisation verification for cross-border dispensing. Pharmaceuticals introduce each of these requirements. The gap between rescEU’s current product scope and what pharmaceutical reserves would demand defines the operational architecture that would need to be developed.
4. Fiscal precedent: the EUR 540 million CBRN stockpile (including the EUR 242 million Finland-hosted component) demonstrates that EU budgetary authorities have accepted strategic stockpiling expenditure in this order of magnitude for health security purposes. Pharmaceutical lifecycle costs (rotation, keep-warm arrangements, regulatory passport maintenance) would introduce recurrent expenditure absent from static CBRN stockpiles, but the CBRN precedent establishes political feasibility for the fiscal envelope.
COVAX: A cautionary benchmark on crisis-time allocation governance
COVAX provides additional evidence on crisis-time allocation governance, though its scope (vaccine access for low- and middle‑income countries) differs from pooled reserves for supply disruptions. Its allocation framework relied on multiple advisory and allocation bodies, and distribution decisions became politically contested during supply-constrained periods, contributing to documented delays. The institutional lesson is that crisis-time release decisions requiring committee deliberation predictably fail to keep pace with supply shocks. Political deliberation functions well ex ante (defining basket composition, triggers, and participation tiers) and ex post (replenishment, accountability, rule revision), but case‑by-case activation during acute supply shortages requires delegated technical authority.
Across these mechanisms, four principles recur:
1. Pre‑commitment. Reserves that depend on crisis-time negotiation for activation, allocation, or burden-sharing predictably fail under stress.
2. Objective triggers and technical release authority. Sub‑24‑hour activation is achievable when political oversight is concentrated on ex-post accountability and replenishment rather than case‑by-case approvals.
3. Binding readiness standards. Audit mechanisms must confirm not merely that inventory exists but that it remains deployable under realistic stress scenarios.
4. Pre‑planned co‑ordination with receiving jurisdictions. Handoff protocols and capacity assessments must be institutionalised, not improvised during deployment.
The convergent operational benchmark is 12‑hour deployment from activation decision to dispatch readiness (achieved by both rescEU and the US SNS). For RescPool, this benchmark would need to be paired with regulatory passport arrangements ensuring that deployment speed is not degraded by last-mile regulatory or logistical barriers, a constraint largely absent from energy, gas or general emergency commodities.
Cross-reference: Section 2.7.4 (Stress-testing on tracer archetypes)
This section presents the full stress-test assessment for each tracer archetype. The body text (Section 2.7.4) presents the basket design implications derived from these results.
Tracer A: Corticosteroids
Corticosteroids represent high-criticality, broadly used medicines where continuity failures quickly translate into avoidable harm. They exemplify products where shortages often stem from quality events, concentrated upstream inputs and thin market economics.
RescPool value: the mechanism would offer high “bridge value” for finished doses: short-to-medium outages (quality remediation, plant disruptions, logistics failures) can be bridged with deployable finished-dose stock, avoiding disruption to acute care pathways. Credible pooled availability would also reduce incentives for defensive over-ordering during early shortage signals, provided activation rules are predictable and fast.
The “bridge‑to-nowhere” risk: for essential generics, apparent temporary disruptions may mask structural fragility – supplier exit, API consolidation or repeated GMP failures. In such cases, a reserve release preserves continuity only if the bridge has a destination: parallel re‑sourcing and supplier qualification must be activated, not merely “wait-and-see” buffering. Activation should therefore automatically initiate a MedPPA workstream (e.g. alternative supplier qualification) so that RescPool would not be repeatedly used as a substitute for a failing market.
Operational implications: Corticosteroids are best suited to rolling or vendor-managed inventory, minimising expiry and wastage, backed by a smaller physical safety layer for immediate releases. The product class requires first-in-first-out (FIFO) rotation partners (i.e. wholesalers) that can absorb near-expiry lots; without rotation partnerships, this product class becomes a predictable wastage risk. Cross-border deployability requires passport-ready packs or rapid conversion capacity to prevent jurisdiction-locked inventory.
Tracer B: PCR diagnostics
PCR diagnostics highlight a different failure mode: shortages often involve not one product but multi-component systems with cross-sector inputs, plastics, enzymes, reagents, primers, probes, and RNA extraction kits, combined with rapid technology cycles and platform heterogeneity.
RescPool value: A pooled reserve would add value only if it stocks the right thing: platform-agnostic bottlenecks that unlock diagnostics capacity across different testing systems. The comparative advantage would be strongest for inputs that are both cross-platform and not easily substituted during demand spikes.
kit stockpiling is operationally weak: Diagnostic platforms and kit designs evolve rapidly, meaning finished kits held in reserve bear a non-negligible risk of obsolescence. Even if a kit remains technically valid, it may not match the equipment installed in the receiving system, a “wrong SKU” failure where inventory exists but is unusable.
Operational implications: A reserve focussed on diagnostics would be more resilient if oriented toward reagents and consumables that are genuinely cross-platform (e.g. extraction consumables, selected reagents) rather than platform-specific test kits. Pairing pooled inputs with pre‑arranged regional assembly or fill-finish partnerships, public or contracted capacity that can convert pooled inputs into deployable kits, addresses the conversion step. Where diagnostics are regulated products, cross-border movement still requires traceability and quality documentation, though platform-agnostic inputs carry lower regulatory specificity burdens than finished kits.
Tracer C: Influenza vaccines
Seasonal influenza vaccines represent biologics with long lead times, strain specificity, and demanding storage and handling requirements. They test the limits of what a stockpile can plausibly achieve when supply is production-cycle constrained.
Reserve value: The highest marginal value lies not in holding large quantities of finished vaccines but in preventing ancillary bottlenecks that can derail vaccination campaigns during surge periods. Ancillaries and delivery system components, syringes, needles, diluents (where relevant), vials and stoppers, cold-chain consumables, often have longer shelf lives and can become binding constraints during rapid scale‑up. Surge cold-chain capacity and deployable logistics can prevent last-mile failures even when vaccine supply exists.
The limits of finished-dose vaccine stockpiling: mismatches between stockpiled vaccines and circulating strains reduce clinical value. Finished-dose holdings concentrate wastage risk and complicate liability and pharmacovigilance responsibilities across jurisdictions. Stockpiles cannot create antigen capacity on demand; they can only buffer within limited windows.
Operational implications: Influenza vaccination continuity is primarily a supply-chain operability problem (ancillaries and cold chain), not a finished-dose inventory problem. Where surge capability is needed, keep-warm arrangements (capacity reservation for packaging and fill-finish inputs) are typically more resilient than warehousing finished vaccine stocks.
Implications for basket design and operational readiness
The stress test highlights four operational imperatives that would need to serve as gatekeeping criteria for the RescPool basket:
“Bridge value” with exit strategy: RescPool should prioritise products where a short release materially prevents harm and where parallel re‑sourcing pathways can be activated. For structurally fragile generics, activation should trigger supplier qualification and procurement escalation to prevent the stockpile from becoming a permanent subsidy for chronic shortages.
Regulatory passport as eligibility condition: products should be eligible for pooling only if they possess a regulatory passport for immediate cross-border use: pre‑agreed emergency SKUs, multilingual e‑leaflets, and harmonised serialisation protocols. Without these, inventory risks being physically available but legally stranded.
Product-appropriate “zero-waste” architecture: high-volume generics: vendor-managed inventory with commercial FIFO rotation. Fast-evolving technologies (diagnostics): stockpile fungible inputs to avoid obsolescence. High-expiry-risk items: keep-warm capacity reservation fees rather than physical warehousing.
Operational speed as the primary metric: RescPool’s success should be measured by deployment speed and correctness, not inventory volume. Key indicators include passport coverage (share of stock legally deployable across all Tier 1 jurisdictions), activation velocity (<24h decision-to-dispatch), and rotation efficiency (>80% commercial absorption of near-expiry stock).
Cross-reference: Section 2.7.6 (Expected effects)
RescPool would be monitored as an operational capability, not as an inventory stock count. The performance question is whether pooled reserves are deployable, usable, and legally dispensable within the bridging window, while remaining financially and politically sustainable over time. The KPI framework below is structured around four domains that jointly determine whether the reserve performs under stress: speed, product correctness, legal deployability, and lifecycle sustainability.
The framework serves two functions: i) routine readiness audits conducted on a standing cycle, and ii) after-action review following activations or exercises. Each KPI is defined in auditable operational terms – authorisation, dispatch, handover, SKU match, custody documentation, passport status – allowing performance to be assessed consistently across products and events.
Note: Benchmarks are indicative: they would be calibrated over time using observed data from exercises and actual deployments.
Reporting cadence and responsibilities
The monitoring framework would operate on three cycles, each serving a distinct governance function:
Standing readiness reporting (quarterly): host-state operators and the Operational Secretariat would report on inventory status (fill rates, shelf-life distribution, passport coverage), rotation activity, and any quality or storage incidents. This reporting feeds routine readiness assessments and supports early identification of degradation before it reaches critical levels. The Operational Secretariat would compile; host-state operators and rotation partners would provide source data.
Annual readiness review: a comprehensive assessment covering all four KPI domains, incorporating results from the annual exercise, audit findings, and trend analysis across quarterly reports. The annual review would produce a consolidated readiness score per product category and per participating jurisdiction, identify systemic weaknesses, and trigger corrective action where indicators fall below target ranges. The Operational Secretariat would prepare the assessment; the governing body would review and decide on corrective measures.
After-action review (event-triggered): following any activation or significant exercise, a structured review would assess actual performance against KPI targets across the full deployment chain, from trigger validation through authorisation, dispatch, handover, and (where applicable) dispensing. After-action reviews are the primary mechanism for calibrating target ranges: observed performance distributions from real events replace indicative thresholds with empirically grounded benchmarks.
Exercise protocol
Exercises serve two functions: validating operational readiness before a real activation occurs, and generating the performance data needed to calibrate KPI benchmarks. Without a regular exercise cycle, the monitoring framework would measure only paper compliance rather than demonstrated capability.
Frequency. A full-chain exercise (trigger validation through dispensing-node handover) would be conducted annually. Tabletop exercises testing specific elements (activation decision making, allocation rule application, passport validation) would be conducted semi‑annually.
Scenario design. Exercises would be structured around the tracer archetypes assessed in Section D.2, ensuring that each major product type (stable generics, fast-evolving diagnostics inputs, biologics ancillaries) is tested over a rolling multi-year cycle. Scenarios would include both single‑product shortages (testing speed and product correctness) and multi-product concurrent shocks (testing allocation governance and prioritisation under competing demands). At least one scenario per cycle would involve a cross-border deployment to a Tier 1 partner, testing passport operability and receiving-jurisdiction handoff protocols under realistic conditions.
Pass/fail criteria. Exercise performance would be assessed against the indicative target ranges in Table D.2. A “pass” requires meeting threshold targets on all gating indicators (activation SLA, dispatch speed, passport coverage for the exercised product). Non-gating indicators (e.g. rotation performance, keep-warm deliverability) are assessed for trend compliance rather than binary pass/fail. Exercise failures trigger corrective action plans with defined remediation timelines, reviewed at the next quarterly reporting cycle.
Calibration function. Performance data from exercises and actual activations would be aggregated to produce observed distributions for each KPI. Once sufficient data points exist (indicatively, after three to five exercise cycles or activations), the governing body would formally adopt calibrated target ranges replacing the indicative thresholds, with tightening trajectories defined for subsequent periods.
Principles for interpreting metrics
Four principles govern how the KPIs above are read in assessments, to ensure they drive genuine operational readiness rather than paper compliance:
1. Speed as a gating indicator: high nominal stock is not meaningful if authorisation and dispatch times cannot fit inside the disruption window. Activation and dispatch metrics function as pass/fail criteria: a reserve that consistently misses the 24‑hour authorisation SLA is operationally non-functional regardless of inventory volume.
2. Correctness as a first-order indicator: delivering the wrong SKU, platform mismatch for diagnostics inputs, incorrect formulation or strength for medicines, constitutes operational failure even if delivery is fast. SKU correctness and documentation completeness are tracked alongside volume, not subordinated to it.
3. Legal deployability as a hard constraint: passport coverage and serialisation operability determine whether inventory is legally dispensable across borders. A pooled reserve with low passport coverage risks becoming legally immobile, undermining both continuity effects and behavioural stabilisation. Passport coverage is therefore a gating indicator alongside speed.
4. Sustainability as part of capability: wastage/expiry, rotation performance and keep-warm deliverability determine whether the reserve can remain credible over time. High wastage erodes political support; weak rotation translates directly into lower effective readiness.
Cross-reference: Section 2.7.7 (Risks: Moral hazard, crowd-out and political rationing)
This section specifies the operational design of the deductible principle, conditioning access to pooled reserves on demonstrated national preparedness. The body text (Section 2.7.7) presents the rationale; this section details the calibration, verification, enforcement, and override architecture.
Design parameters
The deductible would function only if minimum preparedness obligations are defined in operational and auditable terms. A purely declaratory requirement invites self-certification and ex-post disputes; overly rigid access-denial rules risk being politically unsustainable in genuine emergencies. The framework below specifies preparedness expectations that are measurable ex ante and enforceable ex post while preserving a narrow humanitarian override.
Minimum buffer obligations would be calibrated along three axes:
1. Product-specificity. Requirements vary by clinical criticality and supply-risk profile. Higher buffer obligations apply to products with concentrated manufacturing, fragile cold chains, or recurrent shortage histories; lower requirements apply to diversified, stable supply chains.
2. Capacity adjustment. Obligations scale with national health system size and existing stockholding infrastructure, avoiding one‑size‑fits-all burdens that would deter participation or induce strategic under-reporting.
3. Auditability. Compliance is evidenced through verifiable indicators: documented stock levels for designated “buffer SKUs”, binding rotation contracts with suppliers or distributors, nominated buffer facilities subject to periodic inspection, and proof of surge distribution access (pre‑positioned logistics capacity or validated standby agreements).
Indicative buffer requirements
A workable framework would require participating states to maintain national buffers covering a specified number of days of normal consumption for basket products, differentiated by product category.
Note: Buffer obligations are illustrative and would require calibration against observed shortage durations and national stockholding baselines.
Verification mechanism
Compliance verification would operate through periodic reporting supplemented by risk-based spot audits. Audit rights must be explicitly established in participation agreements. Reporting would cover documented stock levels, rotation performance, and buffer facility readiness. The verification model follows the precedent of the EU Gas Storage Regulation, where the Commission monitors compliance against binding intermediate trajectory targets rather than relying on end-state self-reporting alone.
Enforcement architecture
Enforcement would operate through graduated consequences rather than binary exclusion:
De‑prioritisation (first breach): non-compliant participants remain eligible for allocation only after compliant participants are served. This preserves access while creating a material incentive for compliance.
Temporary suspension with financial penalties (repeated or egregious failures): contribution adjustments or access restrictions apply; restoration pathways exist once compliance is demonstrated through verified corrective action.
Outright permanent exclusion would be avoided to preserve the political viability of the mechanism and maintain incentives for re‑compliance.
Humanitarian override
The deductible framework includes a tightly bounded exception for extraordinary circumstances: cases where non-compliance resulted from unforeseeable supply shocks beyond national control (e.g. natural disaster destroying national stockpile facilities, simultaneous multi-product supply failures exceeding buffer capacity). The override would operate through:
Independent verification of force majeure circumstances
Ministerial-level attestation
Decisions delegated to a standing technical committee
The override is designed to be narrow and procedurally demanding. Without these constraints, the exception would predictably expand to accommodate political convenience, undermining the deductible’s incentive function.
Anti-gaming provisions
The framework addresses the risk of strategic gaming, with participants maintaining “paper” compliance while allowing buffers to degrade (e.g. documenting stock levels but failing to rotate, leading to expired inventory masquerading as readiness).
Protection mechanisms include:
Unannounced spot audits with physical verification of usable stock (not merely documented stock)
Rotation performance indicators tracked alongside volume metrics
Independent third-party storage audits for designated buffer facilities
Whistleblower provisions allowing industry actors (manufacturers, distributors) to flag non-compliance without commercial penalty
The deductible principle converts moral hazard from an abstract risk into a concrete governance requirement. Its effectiveness depends not on elaborate penalty structures, but on making preparedness obligations specific, measurable, and routinely monitored, preserving RescPool as a genuine catastrophe‑layer buffer rather than a substitute for national responsibility.
Cross-reference: Section 2.7.8 (Legal architecture and feasibility)
RescPool would operate at the intersection of EU civil protection competence, pharmaceutical regulatory requirements, and third-country co‑operation arrangements. The legal architecture would need to address each of these dimensions. This section presents the menu of legal base options for establishing a pooled pharmaceutical reserve within existing EU competences. The body text (Section 2.7.8) presents the feasibility assessment; the legal detail is preserved below.
EU competence (Article 196 TFEU)
Article 196 TFEU provides EU competence to “encourage co‑operation between Member States in order to improve the effectiveness of systems for preventing and protecting against natural or man-made disasters” and to “promote consistency in international civil protection work.” This base supports:
Co‑ordination of national stockpiling efforts
Establishment of pooled reserves (as demonstrated by rescEU)
Third-country co‑operation arrangements
The rescEU framework (Decision 2019/420 amending Decision 1313/2013/EU) already establishes precedent for EU-owned or co-financed emergency reserves, including medical countermeasures. A pharmaceutical extension could operate within this existing legal architecture, potentially requiring implementing decisions rather than new primary legislation.
Pharmaceutical regulatory interface
The novel legal challenge is pharmaceutical law operability. For medicines and regulated medical devices, cross-border deployment engages three regulatory domains:
1. Marketing authorisation requirements: emergency dispensing of products authorised in one Member State but not another requires derogation from normal requirements. The regulatory passport concept operationalises existing emergency provisions, specifically Article 5(2) of Directive 2001/83/EC for named-patient use and Article 126a for mutual recognition in emergencies, but requires pre‑negotiated templates and operational protocols to function at speed.
2. Serialisation and traceability: the Falsified Medicines Directive (2011/62/EU) and Delegated Regulation (EU) 2016/161 create serialisation requirements that can block cross-border movement if not pre‑addressed. Emergency derogation procedures exist but require activation by competent authorities; a regulatory passport would need to include pre‑agreed handling protocols.
3. Pharmacovigilance: responsibility for adverse event monitoring and recall execution during cross-border emergency deployment must be clearly allocated. A product liability protocol addressing custody transitions, indemnification triggers, pharmacovigilance responsibilities, and recall costs during the bridging window is a precondition for supplier participation.
Third-country participation instruments
For structured participation beyond EU Member States, the legal architecture would require:
Bilateral or multilateral agreements establishing eligibility, contribution modalities, access entitlements, and ring-fencing rules. These could be concluded under Article 216 TFEU (international agreements) or as administrative arrangements with lower institutional formality.
Customs facilitation protocols for emergency cross-border movements, potentially requiring pre‑agreed arrangements under existing customs co‑operation frameworks.
Liability and indemnification clarity for products crossing into non-EU jurisdictions where EU product liability rules do not apply.
Hybrid approach
The mechanism’s hybrid nature (health-sector governance combined with civil protection delivery) suggests a dual legal architecture:
Health-sector governance (product selection, regulatory passport design, quality assurance, pharmaceutical lifecycle management) grounded in Article 168(5) TFEU and the pharmaceutical regulatory framework, with HERA and EMA providing the technical and regulatory interface.
Physical reserve and deployment infrastructure building on Article 196 TFEU civil protection competence, extending the rescEU framework (Decision 2019/420) to pharmaceutical-specific requirements.
The primary legal work required would be: i) establishing the health-sector governance mandate (product basket, regulatory passport, quality assurance); ii) extending rescEU implementing provisions to accommodate pharmaceutical reserve management; iii) developing regulatory passport templates validated by national competent authorities; and iv) concluding third-country participation arrangements under Article 216 TFEU.
The assessment indicates that the legal base would be achievable within existing EU competences, provided pharmaceutical regulatory operability is addressed through pre‑negotiated templates and protocols rather than requiring new primary legislation.
← 1. According to the Herfindahl-Hirschman Index (HHI) measuring import concentration across suppliers to the EU. The HHI is calculated as the square of the respective market shares of countries and regions the EU imports from (only focussing on extra-EU trade).
← 2. The European Commission does not use the terminology “targeted trade deals”. This paragraph keeps the terminology of the paper quoted.
← 3. A Qualified Person (QP) is a licensed pharmacist, biologist, chemist (or equivalent qualified professional) authorised under EU Directive 2001/20/EC to certify batches of medicinal products for commercial sale or clinical supply.
← 5. Substantive equivalence/comparability determinations (e.g. new API sources or major manufacturing changes) remain product-specific and cannot be pre‑committed at a generic design level; the mechanism can, however, pre‑agree the process, prioritisation and information package required to accelerate assessments when needed.
← 6. The Union list of critical medicines is formally established under the pharmaceutical legislation on which political agreement was reached in December 2025. It should be distinguished from product lists used for crisis preparedness purposes (e.g. the HERA Medical Countermeasures List), which serve different policy objectives.