Buyers' Clubs for the Open-Source Flow: Collective Provision in the FLOSS Value Network¶

Stefane Fermigier (Abilian) · sf@abilian.com

Draft working paper v0.8, 2026-09-07. Theory and empirical design. Every proposition carries a proof, a numeric cross-check, and a Monte-Carlo robustness report; the design is specified and no result is estimated. The instrument's operational half, the item wordings, the estimation shells and the power computation, is registered in ../notes/joining-forces/prereg-paper2.md and is not repeated here.

Download the PDF

Abstract¶

European supervisors ask firms to pool procurement and co-fund open-source alternatives to dominant providers. This paper models the institution that proposal implies: a buyers' club for the open-source flow. The code stock is free and inherited; funding is needed for the flow of maintenance, evolution, and compliance evidence, plus the excludable assurance, adaptation, and assistance services that hold a club together. Firms differ in capability (the cost of self-providing those services) and in need (the value of autonomy), the user typology of the FLOSS value network.

The model adds two open-source-specific margins to the collective-provision problem. First, membership can be paid in code: a firm can deliver fee-equivalent flow work at a private cost that falls with its capability, subject to a conversion friction. Riding (self-serving on the open flow without joining) then persists only when the friction is high relative to the fee. Since the fee falls as the club grows, scale converts riders into contributing members: the free-rider problem attenuates itself along the in-kind margin. Clubs sort into code-paying capable members and cash-paying incapable ones, the pattern observed in real consortia. With any compliance burden and an anchor of any positive size, near-costless code dues also make arbitrarily small clubs weakly viable, so the anchor buyer's role sharpens into jumping the critical mass for the cash population.

Second, the license regime enters: under effective reciprocity (copyleft reaching the deployment mode in use), riders' adaptations return upstream and defray the flow, so the club's fee and its under-participation wedge both fall with reciprocity strength, which makes free-riding partially self-financing. The generic collective-provision results hold under the extension. Participation is a game of strategic complements with a zero equilibrium and a critical mass; an anchor buyer of computable size eliminates the zero equilibrium of the cash benchmark; compliance burdens recruit the capable while pricing out marginal need; equilibrium membership is inefficiently low, with the same wedge on both margins under a permissive license. Under reciprocity the need-margin wedge shrinks; converting riders into members then pays only while the per-rider return stays below the wedge.

Two further results locate the club against the voluntary-provision reading of open source. When visible contribution carries signaling returns that rise with the club's audience, the conversion friction falls with club size and rider conversion arrives earlier. The club also weakly dominates voluntary provision firm by firm. Its strict contribution falls on the population priced out of self-provision and, once the alternative's value is read as depending on a funded flow, on every self-provider as well; below its critical mass it has nothing to add. The club is also a natural open-source steward in the Cyber Resilience Act's sense. A club large enough to fork disciplines its suppliers, which connects this model to the enclosure threshold of the supply-side companion project.

The second half of the paper specifies the measurement on the open-source poll. Five hypotheses cover the typology prior on who joins, firm-level regulatory recruitment, conditional cooperation, the binding constraint on cooperation, and a within-firm criticality contrast in the accepted sovereignty premium. They are elicited through commitment-style outcomes with behavioral anchors, a conditional-cooperation ladder that doubles as a provision-point elicitation, a randomized binding-constraint item, and a mirrored supply block. The mapping from elicited constraints to policy instruments is fixed before fieldwork. Identification is limited to contrasts within the stated-preference boundary; the design is specified and no result is estimated.

JEL: D71, H41, L17, H57, O34, L86.

About this paper¶

Cite this paper¶

Fermigier, S. (2026). Buyers' Clubs for the Open-Source Flow: Collective Provision in the FLOSS Value Network.
Draft working paper v0.8, Abilian Econ Lab.
https://econ.lab.abilian.com/papers/collective-provision/