Scientific Research Consortia

The Mechanism

The Structural Genomics Consortium was founded in 2004 to determine the three-dimensional structure of human proteins relevant to disease and make the results freely available without patent restriction. Its members include the Wellcome Trust, several major pharmaceutical companies (GSK, Pfizer, Merck, Bayer, Takeda, and AbbVie among them at various points), and academic groups at Toronto, Oxford, and Frankfurt. The pharmaceutical participants are direct commercial competitors in nearly every disease area the consortium investigates. They have spent the last twenty years collaborating productively at the bench level while remaining commercial rivals at the drug level, and the legal architecture that makes this work is a written consortium agreement of considerable sophistication.

The template the SGC and comparable consortia have refined over four decades has several recurring elements. Background intellectual property — what each party brings into the collaboration — remains the party's own, with defined licensing access granted to other consortium members for collaboration purposes only. Foreground intellectual property — what gets created during the collaboration — has explicit ownership and licensing rules negotiated upfront. Field-of-use carve-outs partition the IP space so that consortium outputs are licensable across all members for research purposes but each member retains exclusive rights to commercialize within their specific area of focus. Publication and disclosure sequencing coordinates the timing of academic publication and patent filing so that open-science norms and IP protection coexist. Governance architecture — steering committees, decision rules for new members, exit provisions, dispute resolution — handles the ongoing decisions no upfront agreement could fully anticipate. The supporting infrastructure (specialized attorneys, template documents refined across hundreds of agreements, university technology-transfer offices, organizations like the National Academies that publish best-practice frameworks) is substantial and mature. The same template has been deployed beyond pharmaceutical research — in semiconductor consortia like IMEC, in genomics, in pre-competitive AI safety work — with the shape adapting to context while the structural moves recur.

The Structural Principle

Multi-party relationships are almost never pure cooperation or pure competition. Real-world parties are typically allied on some dimensions and rivalrous on others, and the question that determines whether they can work together productively is whether they have machinery for keeping the alliance-on-some-dimensions from contaminating the competition-on-others. The underlying game is a stag hunt with retained side-stakes: cooperation produces a larger pie than any party could capture alone, but only when each party can credibly commit not to defect in pursuit of the smaller individual rabbit, and only when the cooperative gain doesn't compromise the side-stakes each party still needs to defend. Without machinery to handle this, three failure modes dominate — full collaboration that bleeds competitive position downstream, refusal to collaborate to protect future competition (so the valuable shared work never gets done), or informal collaboration that fractures the moment commercial interests diverge.

The consortium template is the technology for accessing the value of partial alignment without triggering the failure modes. It works by time-and-space partitioning the relationship: defining precisely which collaborations happen, on what terrain, with what flow of outputs back to shared use, and with what carve-outs that preserve each party's position outside the consortium's scope. The partitioning lets parties be allies at the consortium level and competitors at the commercial level simultaneously. The Hanseatic League — the confederation of Northern European trading cities that coordinated shared infrastructure from roughly 1200 to 1500 while each member city remained sovereign and competitive — is one historical example of the same structural move.

What makes the contemporary scientific-research form distinctive is the precision: written agreements that can specify carve-outs at the molecule level, governance rules that can adapt to changing membership, and IP-handling sophisticated enough that genuine competitors can collaborate at one layer without compromising their positions at another. The technology also generalizes beyond information goods. The same template that lets pharmaceutical companies share a research instrument lets a consortium of separately-owned projects share any expensive resource — equipment, infrastructure, expertise, time — as long as the partial-alignment-partial-divergence structure holds.

Where This Could Land

The operational diagnostic the consortium template offers is a single question that any project considering multi-party collaboration can ask of itself: where do we cooperate, where do we compete, and how do we keep those two from contaminating each other? The question decomposes into three sub-questions. What is the shared work — the activities, outputs, or resources we genuinely benefit from holding in common? What positions does each party retain outside the shared work — the territories, markets, customers, IP, or capabilities each must keep separate to preserve their independent identity? And how do we draw the boundary between the two so the shared work doesn't bleed into the retained positions over time?

A project that can answer those three questions cleanly has the structural shape consortium machinery was built for. The diagnostic travels because the question travels.

In multi-country creative co-productions involving three or more sovereign parties, the shared work is the project itself, the retained positions are each country's industry infrastructure and downstream rights, and the boundary handles which assets flow back to shared use versus which remain proprietary to each national co-producer. Current entertainment law structures these arrangements through stacks of bilateral agreements that nobody quite holds together, and the consortium template handles the underlying problem more cleanly than the bilateral-stack approach can. In cross-border infrastructure projects, the shared work is the infrastructure, the retained positions are each jurisdiction's regulatory authority and economic policy, and the boundary handles how investment, operation, and benefits flow without dissolving sovereignty. In joint research-and-development between commercially competitive firms — common in manufacturing and increasingly in AI safety — the shared work is the pre-competitive research, the retained positions are each firm's product market, and the boundary handles when pre-competitive becomes competitive. The pattern repeats across multi-foundation philanthropic initiatives, regional industry coordination, and multi-municipality cultural institutions. The application varies; the structural shape doesn't.

The activation gap is specific and addressable. The professional community of consortium attorneys — concentrated at a small set of specialized law firms, at the technology-transfer offices of major research universities, and at organizations that publish best-practice frameworks — has spent forty years refining a template that handles exactly the partial-alignment problem most fields are still solving badly through bilateral agreements or informal arrangements. They have rarely been asked to write a creative co-production agreement, a multi-municipality cultural-institution charter, or a multi-foundation philanthropic-initiative structure. The structural problems they routinely solve are closer to those use cases than current entertainment law, municipal law, or philanthropic law machinery is. The translation work involves articulating, for a project in any domain, what the equivalent of background IP, foreground IP, field-of-use carve-outs, and disclosure sequencing are in the project's context. The concepts don't map identically across fields, but they map closely enough that a consortium attorney can produce useful structure from the translation.

Rubedo's Interest

Producing IMAX-format documentary footage across the corridor network at the scale the catalogue's other entries point toward is capital-intensive in ways no single film's budget can carry alone, and the equipment and crew costs are dominated by mobilization rather than by per-shot expense. A summer sprint shoot across twelve EU treaty corridors with a single IMAX camera package and crew, with each participating co-production paying into a shared pool that funds the sprint, would amortize the cost across all twelve while producing footage that gets sorted in post for each individual film. Each film retains its own creative identity, its own narrative arc, its own downstream rights; the consortium machinery handles ownership, carve-outs, scheduling, and disposition of unused footage. The same template that lets pharmaceutical competitors share a research instrument lets twelve films share a camera package. The structural move is identical; the resource is different.